Handbook of Research on Global Diffusion of Broadband Data Transmission Yogesh K. Dwivedi Swansea University, UK Anastasia Papazafeiropoulou Brunel University, UK Jyoti Choudrie University of Hertfordshire, UK
Volume I
INFORMATION SCIENCE REFERENCE Hershey • New York
Acquisitions Editor: Development Editor: Senior Managing Editor: Managing Editor: Copy Editor: Typesetter: Cover Design: Printed at:
Kristin Klinger Kristin Roth Jennifer Neidig Sara Reed Jeannie Porter, Maria Boyer Sean Woznicki Lisa Tosheff Yurchak Printing Inc.
Published in the United States of America by Information Science Reference (an imprint of IGI Global) 701 E. Chocolate Avenue, Suite 200 Hershey PA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail:
[email protected] Web site: http://www.igi-global.com and in the United Kingdom by Information Science Reference (an imprint of IGI Global) 3 Henrietta Street Covent Garden London WC2E 8LU Tel: 44 20 7240 0856 Fax: 44 20 7379 0609 Web site: http://www.eurospanonline.com Copyright © 2008 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. 3URGXFWRUFRPSDQ\QDPHVXVHGLQWKLVVHWDUHIRULGHQWL¿FDWLRQSXUSRVHVRQO\,QFOXVLRQRIWKHQDPHVRIWKHSURGXFWVRUFRPSDQLHVGRHV not indicate a claim of ownership by IGI Global of the trademark or registered trademark. Library of Congress Cataloging-in-Publication Data Handbook of research on global diffusion of broadband data transmission / Yogesh K. Dwivedi, Anastasia Papazafeiropoulou, and Jyoti Choudrie, editors. p. cm. Summary: “This book explores broadband adoption and the digital divide through a global perspective, it provides research on constructs such as relative advantage, utilitarian outcomes, hedonic outcomes, and service quality. From over 100 noted experts in nearly 30 countries, WKLVZRUNDOORZVSROLF\PDNHUV,QWHUQHWVHUYLFHSURYLGHUVDQGRWKHUVWRJDLQPXOWLFXOWXUDOLQVLJKWLQWRZKDWIDFWRUVLQÀXHQFHFRQVXPHUV¶ decisions to adopt broadband”--Provided by publisher. ISBN 978-1-59904-851-2 (hardcover) -- ISBN 978-1-59904-852-9 (e-book) 1. Internet users--Attitudes. 2. Digital divide. 3. Internet service providers. 4. Consumer behavior. 5. Broadband communication systems. 6. Globalization--Social aspects. I. Dwivedi, Yogesh Kumar. II. Papazafeiropoulou, Anastasia. III. Choudrie, Jyoti.
TK5103.4.H36 2008 ¶GF 2007052995 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. All work contributed to this book set is original material. The views expressed in this book are those of the authors, but not necessarily of the publisher. If a library purchased a print copy of this publication, please go to http://www.igi-global.com/reference/assets/IGR-eAccess-agreement.pdf for information on activating the library's complimentary electronic access to this publication.
To my Mother and Father for their love and blessings…And to my lovingly and much awaited ‘Shagun’
To Athina
To my dearest, wonderful, encouraging and supportive Mum, Dad and Bobby
Editorial Advisory Board
Nikhilesh Dholakia University of Rhode Island, USA Guy Fitzgerald Brunel University, UK Sergio Godoy Universidad Catolica de Chile, Chile Heejin Lee University of Melbourne, Australia Catherine Middleton Ryerson University, Canada Challa Radhakumari Sri Sathya Sai University, India Aradhana Srivastava PRIA (Participatory Research in Asia), India Viswanath Venkatesh University of Arkansas, USA Michael D. Williams Swansea University, UK Vishanth Weerakkody Brunel University, UK
Table of Contents
Foreword .......................................................................................................................................... xxxi Preface ............................................................................................................................................ xxxiv Acknowledgment............................................................................................................................... xlvi
Volume I Chapter I Broadband Adoption and Diffusion (BAD): A Framework / Yogesh K. Dwivedi and Anastasia Papazafeiropoulou ......................................................................................................... 1
Section I National Policies Division I Africa Chapter II South Africa: The Long Walk to Broadband Freedom / Justin Henley Beneke.................................... 13 Division II Asia Chapter III Bridging the Digital Divide Through Broadband Deployment / Challa Radhakumari........................ 30 Chapter IV Broadband Policy, Market Competition, and User Adoption in Taiwan / Yu-li Liu.............................. 47 Chapter V ICT Competency of Bangladesh to Face Broadband Diffusion / Anwarul Islam and K.C. Panda..................................................................................................................................... 60
Chapter VI Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society / Dal Yong Jin............................................................................................................... 78 Chapter VII Structural Changes and Regulatory Challenges in the Japanese Telecommunications Industry / Hidenori Fuke ....................................................................................................................................... 90 Division III Australia and New Zealand Chapter VIII $QDO\VLVRIWKH,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG,QWHUQHW$FFHVV Qiuyan Fan ......................................................................................................................................... 109 Chapter IX Broadband for the Mass Market / Roger Saunders............................................................................. 126 Chapter X Competition, Regulation, and Broadband Diffusion: The Case of New Zealand / Bronwyn Howell.................................................................................................................................. 139 Division IV Europe Chapter XI Digital Divide and Broadband Access: The Case of an Italian Region / Enrico Ferro, J. Ramon Gil-Garcia and Natalie Helbig ........................................................................................... 160 Chapter XII Improving Broadband Access in Rural Areas / Ingjerd Skogseid....................................................... 177 Chapter XIII Metropolitan Broadband Networks: Design and Implementation Aspects, and Business Models / Antonios Alexiou, Christos Bouras, John Papagiannopoulos and Dimitris Primpas ........................ 196 Chapter XIV Small World: The Irish Broadband Experience / Diana Wilson, Kevin O’Reilly, and Dave Murray................................................................................................................................ 211 Chapter XV Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion: $*HUPDQ&DVH6WXG\Axel Schulz, Bernd Carsten Stahl, and Simon Rogerson ............................. 227
Division V North America Chapter XVI Competition in Broadband Provision and the Digital Divide / Wei-Min Hu and James E. Prieger ..................................................................................................... 241 Chapter XVII *RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ Elizabeth Fife, Laura Hosman,and Francis Pereira .......................................................................... 260 Chapter XVIII Regulation and the Deployment of Broadband / James E. Prieger and Sunhwa Lee ........................ 278
Section II Consumer-User Behaviors Division I Australia Chapter XIX Factors Affecting Broadband Adoption for Mainstream Consumers / Peter Adams.......................... 306 Division II Europe Chapter XX Developing a Dynamic View of Broadband Adoption / Herbert Daly, Adrina Ortiz, Yogesh K. Dwivedi, Ray J. Paul, J. Santos, and J.M. Sarriegi ........................................................... 322 Chapter XXI Employing the Content Validity Approach for Improving the Content of the Broadband Adoption Survey Instrument / Yogesh K. Dwivedi, Banita Lal, and Khalil Khoumbati..................... 337 Chapter XXII Inside the Microcosm: A Case Study of a Wireless Internet Hotspot / Pierre Vialle, Olivier Epinette, and Olivier Segard .............................................................................................................................. 349 Chapter XXIII 7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW Karianne Vermaas and Lidwien van de Wijngaert ............................................................................. 366
Division III Middle East Chapter XXIV Factors Affecting Attitudes towards Broadband Adoption in the Kingdom of Saudi Arabia / Vishanth Weerakkody .......................................................................................................................... 380 Division IV North America Chapter XXV Characteristics of Farm and Rural Internet Use in the United States / Peter L. Stenberg and Mitchell Morehart ........................................................................................... 395 Division V South America Chapter XXVI Broadband User Behavior Characterization / Humberto T. Marques Neto, Leonardo C.D. Rocha, Pedro H.C. Guerra, Jussara M. Almeida, Wagner Meira Jr., and Virgilio A.F. Almeida ................... 408
Volume II Chapter XXVII Precisions about the Broadband Divide in Chile / Sergio Godoy and M. Soledad Herrera............... 427
Section III Small and Medium-Sized Enterprises (SMEs) Division I Asia Chapter XXVIII A Survey on the Adoption and Usage of Broadband Internet / Roya Gholami, John Lim, and Sang-Yong Tom Lee...................................................................................................................... 448 Division II Europe Chapter XXIX Broadband Access and Broadband-Based Applications: An Empirical Study of the Determinants of Adoption Among Italian SMEs / Massimo G. Colombo, Luca Grilli, and Cinzia Verga ................... 466
Chapter XXX Broadband Diffusion to SMEs in the UK / Oluwasola Oni and Anastasia Papazafeiropoulou ........ 481 Chapter XXXI Environmental Drivers of E-Business Strategies Among SMEs / Alessandro Arbore and Andrea Ordanini ........................................................................................... 493 Chapter XXXII ([SORULQJ60(V¶$GRSWLRQRI%URDGEDQGLQWKH1RUWKZHVWRI(QJODQGBoumediene Ramdani and Peter Kawalek.............................................................................................................................. 504 Chapter XXXIII External Pressures for Adoption of ICT Services among SMEs / Andrea Ordanini and Alessandro Arbore ........................................................................................... 524
Section IV Impact on Emerging Applications Division I Entertainment Industry Chapter XXXIV IPTV Business Model Analysis / Kate Carney Landow, Michelle Fandre, Raghu Nambiath, Ninad Shringarpure, Harvey Gates, Artur Lugmayr, and Scott Barker.............................................. 538 Chapter XXXV The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry / Stanford L. Levin, John B. Meisel, and Timothy S. Sullivan .................................. 563 Division II Health Industry Chapter XXXVI Broadband for Health in Developing Countries / Aradhana Srivastava ............................................ 581 Chapter XXXVII Improving Health Services via Advanced ICT Networks / Peter Farr, Isabelle Ellis, and John Royle.................................................................................................................................... 593 Chapter XXXVIII Remote Patient Monitoring in Residential Care Homes: Using Wireless and Broadband Networks / Tanja Bratan, Malcolm Clarke, Joanna Fursse, and Russell Jones ................................................... 604
Division III Social Impact Chapter XXXIX Social Consequences of Broadband Access in Japan / Kenichi Ishii.................................................. 619 Division IV Communication Chapter XL Internet-Based Changes in Organizational Communication / Erik Lundmark and Alf Westelius........................................................................................................ 637 Chapter XLI 8ELTXLWRXV&RPPXQLFDWLRQYLD5HVLGHQWLDO*DWHZD\VAlex De Smedt............................................ 655
Section V Cross-Country Analysis Chapter XLII Adoption of Broadband Services: The Role of National Policies / Morten Falch ............................. 671 Chapter XLIII Broadband Diffusion and its Driving Forces / Banani Nandi and Chandana Chakraborty............... 689 Chapter XLIV Diffusion of Broadband Access in Latin America / Arturo Robles Rovalo, Claudio Feijóo González, and José Luis Gómez-Barroso ............................. 711 Chapter XLV Diffusion Forecasting and Price Evolution of Broadband Telecommunication Services in Europe / Dimitris Varoutas, Christos Michalakelis, Alexander Vavoulas, and Konstantina Deligiorgi ................................................................................................................ 729 Chapter XLVI Explaining Patterns of Broadband Deployment and Adoption in OECD Countries / Inmaculada Cava Ferreruela.............................................................................................................. 756 Chapter XLVII ICT Statistics for Broadband Promoting Regulatory Policy / Diana Korsakaite and Tomas Lamanauskas....................................................................................... 776
Chapter XLVIII Impact of Broadband VoIP on Telecoms: A Cross-Country Analysis / Bardo Fraunholz and Chandana Unnithan ........................................................................................ 796 Chapter XLIX The Adoption of Broadband Internet in Australia and Canada / Catherine Middleton and Shanton Chang .......................................................................................... 818 Chapter L The Evolution of Broadband Industry in the Developing World: A Comparison of China and India / Nir Kshetri, and Nikhilesh Dholakia ......................................... 841
Detailed Table of Contents
Preface ............................................................................................................................................ xxxiv Acknowledgment............................................................................................................................... xlvi
Volume I Chapter I Broadband Adoption and Diffusion (BAD): A Framework / Yogesh K. Dwivedi and Anastasia Papazafeiropoulou ......................................................................................................... 1 The aim of this chapter is to outline various factors reported in the 49 chapters of this handbook of research. 7KHDLPLVDOVRWRRUJDQL]HLGHQWL¿HGIDFWRUVLQDPHDQLQJIXOPDQQHULQRUGHUWRSURSRVHDIUDPHZRUNRI broadband adoption and diffusion. This chapter illustrates the fact that research on the adoption, diffusion, usage, and impact of broadband is clearly a global issue which requires a multidisciplinary approach. The proposed framework includes three levels of factors—macro factors, individual micro factors, and SME-level micro factors. These three levels of factors are relevant at different levels of development, deployment, and diffusion of broadband which persist in various developed and developing countries. The chapter concludes by suggesting that the proposed framework is based on a comprehensive set of factors observed in various countries, and future studies may use this framework to identify gaps and then bridge those gaps by conducting new studies.
Section I National Policies This section examines the macro or supply-side factors affecting broadband deployment and diffusion in various countries including Australia, Bangladesh, Germany, Greece, India, Ireland, Italy, Japan, New Zealand, Norway, South Africa, South Korea, Taiwan, and the United States of America. A number of pertinent factors and issues including national policy, market competition, ICT competency, structural changes, regulatory challenges, rural access, socio-political and ethical responsibility, and digital divide DUHGLVFXVVHGLQWKLVVHFWLRQ7KHVHFWLRQLVIXUWKHURUJDQL]HGLQWR¿YHGLYLVLRQVDFFRUGLQJWRWKHJHRgraphical areas where the studies included in this section were conducted.
Division I Africa Chapter II South Africa: The Long Walk to Broadband Freedom / Justin Henley Beneke .................................... 13 South Africa has fallen behind its international peersboth developing and developed marketsin the race to rollout broadband services. In fact, even within the African continent, it is neither the broadband leader nor progressive in comparison to its Northern African counterparts. This chapter explores the development of broadband services in South Africa, as well as touching on the challenges faced in bringing this phenomenon into the mainstream. Reasons for the lack of diffusion and adoption of such VHUYLFHVSRLQWWRKLJKHQGXVHUFRVWVRIWKHVHUYLFHDYHU\OLPLWHGJHRJUDSKLFDOIRRWSULQWRIERWK¿[HG line and mobile broadband infrastructure, as well as a lack of computer literacy and an understanding of what broadband is able to offer. The chapter looks at possible solutions, including introducing a greater degree of competition into the market to facilitate downward pressure on prices, as well as providing FRVWEDVHGDFFHVVWRLQWHUQDWLRQDOVXEPDULQH¿EHUFDEOHVDQGWKHXQEXQGOLQJRIWKHORFDOORRSWRIXUWKHU this objective. Division II Asia Chapter III Bridging the Digital Divide through Broadband Deployment / Challa Radhakumari ......................... 30 This chapter provides a summary relating to the functioning of two projects in the two Southern States of India, Andhra Pradesh and Kerala, to show how through broadband deployment in rural areas the digital divide can be bridged. By focusing on the implementation of the two projects, the chapter illustrates their contribution in practically using the broadband technologies in overcoming the hurdles to bridging WKHGLJLWDOGLYLGHDQGKLJKOLJKWVWKHFULWLFDOVXFFHVVIDFWRUVDVLGHQWL¿HGGXULQJWKHIXQFWLRQLQJRIWKH projects which helped the states in achieving their goals. The chapter also reveals through its analysis that the accessibility of services through broadband technology has brought an opportunity to the citizens to become a part of the current knowledge revolution, besides bringing about a great technological transformation to the areas where it is implemented and thus contributed to bridging the digital divide. The chapter is concluded by proving that moving from a manual to electronic process with broadband technology as an enabler; the two Southern States of India set an example, which will serve as a set of guidelines for application of similar projects in other geographical settings. Chapter IV Broadband Policy, Market Competition, and User Adoption in Taiwan / Yu-li Liu.............................. 47 7DLZDQKDVEHHQPDNLQJJUHDWHIIRUWVWRSURPRWHWKHIXO¿OOPHQWRILWV1DWLRQDO,QIRUPDWLRQ,QIUDVWUXFWXUH (NII), with broadband network technology and its application having rapidly developed in recent years. This chapter analyzes the broadband adoption, the competition among providers of broadband, and relevant policies in Taiwan. The research methods adopted include a literature review, in-depth interviews,
DQGVHFRQGDU\DQDO\VLVRISUHYLRXVVXUYH\VFRQGXFWHGE\7:1,&6LQFH7DLZDQ¶VEURDGEDQGSHQHWUDWLRQ LVWKH¿IWKKLJKHVWLQWKHZRUOGWKHGLVFXVVLRQVRIWKHPDMRUIDFWRUVFRQWULEXWLQJWREURDGEDQGGHSOR\ment in this chapter might provide some experiences from which other countries can learn. Chapter V ICT Competency of Bangladesh to Face Broadband Diffusion / Anwarul Islam and K.C. Panda ............................................................................................................ 60 As a developing country, Bangladesh has taken keen initiatives to develop its sustainable information infrastructure. Teledensity and overall IT infrastructure is now in a growing stage. Recently, Bangladesh KDVEHHQFRQQHFWHGZLWK6($0(:(VXEPDULQHFDEOHHVWDEOLVKLQJDQRSWLFDO¿EHUEDFNERQH LWV teledensity is changing in rapid pace. But, the broadband diffusion in Bangladesh is not on par with other Asian countries, since it is in an embryonic stage in broadband diffusion. This chapter tries to show the initiatives taken and the existing condition of Bangladesh to fetch the countrywide broadband diffusion. Efforts have been made to unmask the overall development of ICT infrastructure in Bangladesh to judge the environment of broadband diffusion in the country. Chapter VI Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society / Dal Yong Jin............................................................................................................... 78 This chapter attempts to ascertain the causes of the rapid growth of broadband services in the context of the broader socio-cultural elements. It recognizes technology as a socio-cultural product which has historically been constituted by certain forms of knowledge and social practice, so this chapter explores cultural elements contributing to the diffusion of broadband services in the context of the cultural enYLURQPHQWLQ.RUHD)XUWKHUWKLVFKDSWHUGLVFXVVHVWKHVLJQL¿FDQWUROHRIWKHSHRSOHDVXVHUVLQWKH process of the rapid diffusion and growth of broadband services. In particular, it emphasizes the way in ZKLFKWKH HFRQRPLFFULVLVDVRQHRIWKHPRVWVLJQL¿FDQWVRFLRFXOWXUDOWXUQLQJSRLQWVLQPRGHUQ .RUHDQKLVWRU\KDVLQÀXHQFHGWKHGHSOR\PHQWRIEURDGEDQGVHUYLFHVDVKLJKVSHHG,QWHUQHWFRQQHFWLRQV have developed since 1997. Chapter VII Structural Changes and Regulatory Challenges in the Japanese Telecommunications Industry / Hidenori Fuke ....................................................................................................................................... 90 The structure of the telecommunications industry in Japan has been changing revolutionarily. The FKDQJHVDUHREVHUYHGLQ¿YHSKDVHV GHYHORSPHQWRIFRPSHWLWLRQLQWRWKHORFDOFDOOPDUNHWGLIIXVLRQ of broadband Internet and development of inter-platform competition, rapid growth of cellular services and Internet access via cellular, decline of POTS (plain old telephone service), and structural changes from vertical integration to layered structure and development of media convergence. These changes require total review of the regulatory framework that was formed in the POTS era. This chapter reviews: D HVVHQWLDOIDFLOLWLHVUHJXODWLRQ E DXQLYHUVDOVHUYLFHV\VWHPDQG F DÀDWUDWHSULFLQJV\VWHPRI the Internet to solve problems that are likely to distort the new industry structure and would stress the importance of a regulatory system that is competition, technology, and content neutral.
Division III Australia and New Zealand Chapter VIII $QDO\VLVRIWKH,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG,QWHUQHW$FFHVV Qiuyan Fan ......................................................................................................................................... 109 Like many other governments in the world, the Australian government has taken a multi-faceted approach to promoting broadband Internet access. This chapter provides an in-depth analysis of the impact of policy issues on broadband Internet access in Australia. The primary goal of this chapter is to develop a holistic XQGHUVWDQGLQJRI$XVWUDOLD¶VQDWLRQDODSSURDFKHVSHUWDLQLQJWREURDGEDQG,QWHUQHWDFFHVV7KLVUHVHDUFK KDVFOHDUO\LQGLFDWHGWKDWWKHVWDWHRIEURDGEDQG,QWHUQHWDFFHVVLVFORVHO\UHODWHGWRWKHJRYHUQPHQW¶V policy and regulatory framework. The government has been basing its actions on market forces as a principal driver for broadband Internet connectivity. However, market forces only play roles in improving broadband Internet access in the major cities and have little effect in regional and rural Australia. The research has indicated that the regulatory competition regime, by and large, has failed to address concerns of market dominance and market power in the telecommunications sector as is evidenced by a relatively lower price-performance ratio of broadband services in Australia. Chapter IX Broadband for the Mass Market / Roger Saunders ............................................................................. 126 This chapter suggests that there is no new application to stimulate adoption of broadband by the mass market. Many new applications have been introduced but have not created the desired growth. One application that could have mass market attraction is voice over Internet protocol (VoIP), and it is the most likely killer application. But failure by major communications carriers to develop VoIP is slowing broadband penetration to this larger market segment. It is postulated that this resistance results from the risk to current carrier call revenue from VoIP and that infrastructure to support high-speed broadband PD\QRWJHQHUDWHDGHTXDWHUHYHQXHLQWKHVKRUWWHUPWRVDWLVI\VKDUHKROGHUVRU¿QDQFLDOPDUNHWV1R international standards have yet been set, and VoIP between the Internet platforms is not fully integrated. Also the multitude of broadband packages offered by the various competing carriers creates confusion in the mass market which, as a result, defers purchase decisions. Chapter X Competition, Regulation, and Broadband Diffusion: The Case of New Zealand / Bronwyn Howell .................................................................................................................................. 139 New Zealand offers a thought-provoking case study of the effects of different competition and regulatory policies on broadband diffusion rates. Despite having one of the highest rates of Internet connection and usage in the OECD, widely available broadband infrastructure, and low broadband prices, broadband uptake per capita languishes in the bottom third of the OECD. While low uptake has typically been atWULEXWHGWRFRPSHWLWLRQDQGUHJXODWRU\IDFWRUVDVVRFLDWHGZLWK1HZ=HDODQG¶VµOLJKWKDQGHG¶UHJXODWRU\ UHJLPHWKLVFKDSWHUSURSRVHVWKDWDPRUHFUHGLEOHH[SODQDWLRQOLHVLQDFRPELQDWLRQRI1HZ=HDODQG¶V legacy of demand-side regulations, in particular the retail tariff options for voice telephony, and the
limited value being derived by New Zealand residential consumers from the small range and narrow adoption of applications currently used that necessitate broadband connections. The New Zealand case illustrates the effect that legacy regulations can have on both the diffusion of new technologies per se and the choices made by consumers between different generational variants within that technology. The case indicates a need for more research on the effect of telecommunications industry regulations on demand-side uptake factors. Division IV Europe Chapter XI Digital Divide and Broadband Access: The Case of an Italian Region / Enrico Ferro, J. Ramon Gil-Garcia and Natalie Helbig ................................................................... 160 Reducing digital divide in order to build an information society for all is one of the top priorities for European policymakers. A better understanding of the determinants of broadband access at the individual level represent a key starting point for any e-inclusion policy. Based on a review of the literature on digital divide and broadband access, the authors document different approaches to understanding the digital divide and argue that these perspectives can also help to understand broadband access. Combining the digital divide and broadband literature provides a systematic and theory-based approach to the selection and inclusion of variables in different models. This chapter presents a case study conducted LQDQ,WDOLDQUHJLRQ7KHDXWKRUVSURYLGHVRPHLPSOLFDWLRQVRIWKH¿QGLQJVDQGDUJXHWKDWSROLF\PDNHUV should explore the relationship between IT skills acquisition, broadband access, and Internet use in order to develop more effective policies and programs. Chapter XII Improving Broadband Access in Rural Areas / Ingjerd Skogseid ....................................................... 177 The chapter explores the characteristics of rural broadband infrastructure development. Taking the existing installed base into consideration, small rural communities can initiate bottom-up cultivation of broadband infrastructure. Such initiatives are important contributions to overcoming the disparity in broadband access. In effect, they aggregate demand by creating a larger total market for suppliers than the individual needs of the actors. The proposal is to use descriptive clusters as a way to reveal the LQVWDOOHGEDVH$VHWRITXHVWLRQVDQGDQVZHUVZLOOEHGH¿QHGWRUHYHDOWKHLQVWDOOHGEDVH7KHUHVXOWLQJ data can be used to acquire an overview of the types of resources available and the choices that need to be made. The use of descriptive clusters places emphasis on the local context and culture. With a bottom-up strategy, questions must be answered in relation to the local context. The responses and lessons learned may vary from one location to the next, making blueprint implementations impossible. Chapter XIII Metropolitan Broadband Networks: Design and Implementation Aspects, and Business Models / Antonios Alexiou, Christos Bouras, Dimitris Primpas, and John Papagiannopoulos ...................................... 196 This chapter presents the design principles that cover the implementation of broadband infrastructure in the region of Western Greece, by examining all the necessary parameters that arise while implementing
such a critical developmental project. The broadband infrastructure that is deployed is either based on RSWLFDO¿EHU RQELJPXQLFLSDOLWLHV RURQZLUHOHVVV\VWHPV 2)'0EDVHGDQG:L)LFHOOV )XUWKHUPRUH the authors present as two case studies all issues of the designing of the Metropolitan Area Network of Patras, the third largest city of Greece and the Wireless Access Network of Messatida. The major target of the broadband networks is to interconnect the buildings of the public sector in the city and also deploy LQIUDVWUXFWXUH ¿EHURUZLUHOHVVV\VWHPV WKDWZLOOFUHDWHFRQGLWLRQVRIFRPSHWLWLRQLQSURYLGLQJERWKDFcess and content services to the advantage of the end consumer. The usage of the broadband infrastructure by service providers will be based on the open availability of the infrastructure in a cost-effective way. 7KLVFKDSWHUDOVRSUHVHQWVWKHPDLQFKDUDFWHULVWLFVRIDSURSRVHGEXVLQHVVSODQWKDWHQVXUHV¿QDQFLDO viability of the broadband infrastructure and guarantees the administration, growth, and exploitation of infrastructure. Chapter XIV Small World: The Irish Broadband Experience / Diana Wilson, Kevin O’Reilly, and Dave Murray ................................................................................................................................ 211 In this chapter the authors consider from a marketing perspective the political, cultural/social, and economic factors, both micro and macro, affecting the supply/demand nexus of broadband services for the Irish consumer. This chapter charts the development of broadband and its current situation of rollout and uptake, examines the reasons for its continuing poor performance, and offers recommendations on how Ireland may close the gap and perhaps even move ahead. Utilized data was collated from a variety of resources, journals and press and trade publications. The authors attended a ministerial conference on the state of broadband to which many representatives of the telecommunications industry had been LQYLWHG7KH\VXUYH\HGSHRSOHZLWKDQGZLWKRXWWKHIDFLOLW\DQGLQWHUYLHZHGNH\SOD\HUVLQWKH¿HOG7KH chapter concludes that, although the market is beginning to grow strongly, it is from a low base, and as a result Ireland still lags behind many of its European counterparts. There is still a lack of competition which is having an adverse effect on both supply and demand of broadband for the Irish domestic conVXPHU,UHODQGLVDVPDOOFRXQWU\DQGWKHLQFXPEHQWVWLOOKROGVFRQWURORIWKHPRVWSUR¿WDEOHDUHDVRIWKH technology, particularly wholesale and selling to other operators. Also, the Irish are still not convinced WKDWEURDGEDQGLVWKHµNLOOHUDSS¶WKH\QHHG7KLVPD\FKDQJHLQWKHIXWXUHDVWKHWHFKQRORJ\GHOLYHUV more of the content-rich multimedia fare that the Irish already enjoy in other formats. Chapter XV Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion: A German Case Study / Axel Schulz, Bernd Carsten Stahl, and Simon Rogerson ............................. 227 There is considerable interest worldwide in broadband diffusion, with research focusing on aspects such as the provision of broadband in remote areas and the socio-economic factors that determine the OLNHOLKRRGRIDGRSWLRQ7KLVFKDSWHULGHQWL¿HVWKHSROLFLHVDQGLQLWLDWLYHVXVHGWRHQFRXUDJHEURDGEDQG awareness, availability, and adoption. Using the case study of a local broadband initiative in remote and rural Germany, the chapter asks the question of who can and should be responsible for broadband provision, and how such responsibility ascriptions are realized.
Division V North America Chapter XVI Competition in Broadband Provision and the Digital Divide / Wei-Min Hu and James E. Prieger .......................................................................................................................... 241 This chapter examines the supply of DSL broadband by the incumbent local exchange company (LEC) LQ¿YH86VWDWHVLQWKHHDUOLHU\HDUVRIGHSOR\PHQW(PSLULFDODQDO\VLVVKRZVWKDWLQFRPHRWKHUGHmographics, and cost factors are important determinants of entry and availability. After controlling for other factors, the racial characteristics of the area do not affect DSL provision. Active competition in broadband from competitive LECs reduces deployment of DSL by the incumbent, but potential competition from competitive LECs has the opposite effect. Competition from cable companies also negatively LQÀXHQFHVWKHLQFXPEHQW¶VGHFLVLRQWRVXSSO\'6/7KHDXWKRUV¶REMHFWLYHLQJDXJLQJWKHLPSRUWDQFHRI the various factors is to highlight the important drivers of broadband provision for policymakers. Chapter XVII Governmental and Cultural Factors in Broadband Adoption / Elizabeth Fife, Laura Hosman, and Francis Pereira ............................................................................................................................ 260 7KRXJKWKHSRWHQWLDOEHQH¿WVRIEURDGEDQG,QWHUQHWDGRSWLRQDUHJUHDWWKHOHYHOVRIWDNHXSYDU\JUHDWO\ around the world. Some governments have adopted aggressive policies to deploy broadband networks and to encourage the use of these applications, while others have not. In the former cases, governments DUHPRWLYDWHGWRSURPRWHEURDGEDQGDGRSWLRQLQRUGHUWRUHDOL]HERWKHFRQRPLFDQGVRFLDOEHQH¿WV7KLV chapter argues that the high level of broadband adoption rates witnessed in certain Asian economies is attributable in part to the aggressive policies pursued by these governments. Independent of these policies however, social factors can also have an impact on whether broadband-related technology will be DGRSWHG(YHQLIHFRQRPLFDQGVRFLDOEHQH¿WVH[LVWWKHUHIRUHDVLQWKHFDVHRIWHOHPHGLFLQHLQWKH8QLWHG States, cultural and social factors may in fact hinder the deployment of such applications and retard the growth rate of broadband access. Chapter XVIII Regulation and the Deployment of Broadband / James E. Prieger and Sunhwa Lee ........................ 278 This study examines the impact of telecommunications regulatory policy on broadband service deployPHQW8VLQJ86GDWDFRYHULQJDOOIRUPVRIDFFHVVWHFKQRORJ\ FKLHÀ\'6/DQGFDEOHPRGHP DQGDOO areas served by major carriers, authors investigate the impact of state and federal regulation on broadband availability. Alternative regulation increases the probability of broadband availability, particularly for price caps. Unbundled network element (UNE) rates, the prices incumbent carriers charge to competitors for access to the local exchange network, also matter. Areas with lower UNE rates have a slightly higher probability of broadband availability. The effects of UNE rates on broadband deployment are ODUJHVWZKHUHLQFHQWLYHUHJXODWLRQLVLQSODFH7KHDXWKRUV¶REMHFWLYHLQH[DPLQLQJUHJXODWRU\IDFWRUVLV to highlight the role of incentive regulation and local telecommunications competition policypolicies used or available around the worldin stimulating broadband service deployment.
Section II Consumer-User Behaviors This section examines the micro or demand-side factors affecting broadband adoption and usage in various countries including Australia, Brazil, Chile, France, The Kingdom of Saudi Arabia, The Netherlands, the United Kingdom, and the United States of America. Chapters included in this section provides in-depth discussion on socio-behavioral, attitudinal, and demographic factors affecting the adoption of broadband and the digital divide at the consumer level. This section also focuses on broadband user behavior and characterization. Similar to Section I, according to geographical area where studies (included within WKLVVHFWLRQ ZHUHFRQGXFWHGWKLVVHFWLRQLVIXUWKHURUJDQL]HGLQWR¿YHGLYLVLRQV Division I Australia Chapter XIX Factors Affecting Broadband Adoption for Mainstream Consumers / Peter Adams .......................... 306 7KLVFKDSWHUH[SORUHVZKHWKHUSDVWH[SHULHQFHVZLWKWHOHFRPPXQLFDWLRQVSURYLGHUVDQGFXUUHQWµSODQV¶ RQRIIHUVHUYHDVEDUULHUVEHWZHHQDQLQGLYLGXDOFRQVXPHU¶VSHUVXDVLRQSKDVHRIWKHLQQRYDWLRQGHFLsion process and the decision phase. With broadband in approximately one-third of Australian homes, it is important that telecommunications providers understand why the future mainstream segment of consumers will want to adopt broadband, and any barriers to this. This analysis suggests studies are needed to investigate whether the telecommunications providers are collectively confusing potential broadband consumers in their attempts to differentiate a generic product in the market. It argues that future technology adoption studies need to consider including the complexity of the actual purchase decision when developing constructs for quantitative models. The author argues that if we are to build a picture of why mainstream consumers adopt broadband, more than just the perceptions of using the technology itself need to be investigated. Division II Europe Chapter XX Developing a Dynamic View of Broadband Adoption / Herbert Daly, Adrina Ortiz, Yogesh K. Dwivedi, Ray J. Paul, J. Santos, and J.M. Sarriegi ........................................................... 322 The widespread domestic use of broadband Internet technology has been recognized to have a positive LQÀXHQFHRQQDWLRQDOHFRQRPLHVDQGLPSURYHWKHOLIHRIFLWL]HQV'HVSLWHVXEVWDQWLDOLQYHVWPHQWWRGHvelop the infrastructure, many of the countries have experienced slow adoption rates for broadband. This chapter develops a view of UK broadband adoption using dynamic modeling techniques based on an existing statistical study. The contrasting approaches to modeling are compared. Principles of a dynamic modeling system are introduced, and an appropriate form for broadband adoption chosen. The process of building a dynamic model based on an existing static model of broadband adoption is presented. Finally, the new perspective of the dynamic model is explored using the causal loop analysis technique.
Chapter XXI Employing the Content Validity Approach for Improving the Content of the Broadband Adoption Survey Instrument / Yogesh K. Dwivedi, Banita Lal, and Khalil Khoumbati ..................... 337 The overall aim of this chapter is to validate the content of the broadband adoption survey instrument IURPWKHKRXVHKROGFRQVXPHU¶VSHUVSHFWLYH7KHREMHFWLYHVRIWKLVFKDSWHUDUH ¿UVWWRHQVXUHWKDWWKH LGHQWL¿HGFRQVWUXFWVDQGWKHLUUHVSHFWLYHLWHPVDGHTXDWHO\FRYHUUHOHYDQWGLPHQVLRQVRIIDFWRUVWKDWDIIHFW FRQVXPHUVLQWKHGRPDLQRIEURDGEDQGDGRSWLRQ VHFRQGWRGHWHUPLQHZKHWKHUWKHLGHQWL¿HGFRQVWUXFWV and their respective items adequately cover relevant dimensions; and third, to conduct a pre-test and pilot WHVWRQWKHUHVXOWLQJVXUYH\LQVWUXPHQWVLQRUGHUWRREWDLQIHHGEDFNOHDGLQJWRLPSURYHPHQWVLQWKH¿QDO TXHVWLRQQDLUH7KHLQLWLDOLWHPVIRUHDFKFRQVWUXFWZHUHLGHQWL¿HGIURPERWKWKHWHFKQRORJ\DGRSWLRQ OLWHUDWXUHDQGH[SORUDWRU\VWXGLHVRQEURDGEDQGDGRSWLRQ9DOLGDWLRQRIWKHLGHQWL¿HGLWHPVZDVWKHQ SHUIRUPHGHPSOR\LQJDYDULDWLRQRIDTXDQWLWDWLYHDSSURDFKWRFRQWHQWYDOLGLW\7KH¿QGLQJVREWDLQHG from the content validation are then presented and subsequently discussed. Chapter XXII Inside the Microcosm: A Case Study of a Wireless Internet Hotspot / Pierre Vialle, Olivier Epinette, and Olivier Segard .................................................................................................. 349 The objective of this chapter is to highlight critical elements affecting the diffusion of broadband wireless Internet at a hotspot location, through a case study. The research deals with a wireless Internet services project in the main Paris airport, and comprises two components. First, this chapter analyzes WKHHPHUJHQFHRIDYDOXHFKDLQIRUDQHZVHUYLFHEDVHGRQWKHFRRSHUDWLRQRIVHYHUDO¿UPV,QSDUticular, the authors show how different actors can or cannot position themselves on this value chain, according to their resources and capabilities. Second, the authors explore the perceptions and attitudes of business passengers in order to better understand the potential adoption and use of hotspot services, and provide a preliminary framework of analysis. The research is drawn from a qualitative survey via in-depth interviews of potential suppliers (airline companies, service providers, airport managers) and business passengers. Chapter XXIII 7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHWKarianne Vermaas and Lidwien van de Wijngaert ............................................................................................................ 366 7KLVFKDSWHULGHQWL¿HVDVPDOOQXPEHURIUHODWLYHO\KRPRJHQHRXVJURXSVRI'XWFK,QWHUQHWXVHUV ERWK broadband and narrowband), based on their usage pattern. Using individual and behavioral characterLVWLFVWKHDXWKRUVIXUWKHULQYHVWLJDWHWKHQDWXUHRIWKHGLIIHUHQWJURXSV7KH8VHVDQG*UDWL¿FDWLRQV approach is employed as a starting point for an online survey. Cluster analysis and logistic regression ZHUHXVHGIRUGDWDDQDO\VLV7KLVFKDSWHULGHQWL¿HVDQGSUHVHQWV¿YHFOXVWHUVRIGLIIHUHQW,QWHUQHWXVHUV based on patterns of behavior. Results show that the Internet users in the sample consist of a large group that more or less conducts the same online activities. Results also show that narrowband and broadband users differ in their Internet behavior.
Division III Middle East Chapter XXIV Factors Affecting Attitudes towards Broadband Adoption in the Kingdom of Saudi Arabia / Vishanth Weerakkody .......................................................................................................................... 380 Utilizing a survey approach, this research set out to explore the reasons for the slow progress in broadband adoption and investigates the factors that may be affecting the adoption of broadband by KSA consumers. 3DUWLFXODUHPSKDVLVZDVSODFHGRQLQGLYLGXDOOHYHOIDFWRUVVXFKDVVRFLDODQGFXOWXUDOLQÀXHQFHV7KH NH\¿QGLQJVZHUHWKDWWKHIDFWRUVZLWKWKHPDLQLQÀXHQFHRQDWWLWXGHWRZDUGVDGRSWLRQRIEURDGEDQG were usefulness, service quality, age, usage, type of connection, and type of accommodation. Contrary WRSUHGLFWLRQDOWKRXJKVRFLRFXOWXUDOIDFWRUVVXFKDVUHJXODWLRQWKURXJK¿OWUDWLRQRIEURDGEDQGZHUH IRXQGWRKDYHQRVLJQL¿FDQWLQÀXHQFHRQWKHDGRSWLRQRIEURDGEDQGFRQVXPHUVZHUHDZDUHDQGODUJHO\ did not like the regulation. The chapter also provides a discussion on research implications, limitations, and future directions. Division IV North America Chapter XXV Characteristics of Farm and Rural Internet Use in the United States / Peter L. Stenberg and Mitchell Morehart ........................................................................................................................ 395 The Internet became enmeshed in U.S. businesses management practices over the last decade. During this period access and use of the Internet increased for all regions of the United States, most types of work places, and all income groups. In this study, the authors examine Internet use by farm and rural workers, and proprietors using descriptive statistics and market demand analysis. In their market demand analysis approach, the primary methodology the authors used is categorical dependent variable analysis. The results indicate income is a critical element, though other factors such as age of proprietor DQGUXUDOXUEDQORFDWLRQDUHDOVRVLJQL¿FDQWLQPDUNHWGHPDQGGHWHUPLQDWLRQ Division V South America Chapter XXVI Broadband User Behavior Characterization / Humberto T. Marques Neto, Leonardo C.D. Rocha, Pedro H.C. Guerra, Jussara M. Almeida, Wagner Meira Jr., and Virgilio A.F. Almeida ................... 408 This chapter presents a broadband user behavior characterization from an Internet service provider standpoint. Understanding these user behavior patterns is important to the development of more ef¿FLHQW DSSOLFDWLRQV IRU EURDGEDQG XVHUV 7KH FKDUDFWHUL]DWLRQ GLYLGHV WKH XVHUV LQWR WZR FDWHJRULHV UHVLGHQWLDODQGVPDOORI¿FHKRPHRI¿FH 62+2 DQGHPSOR\VIRXUFKDUDFWHUL]DWLRQFULWHULD VHVVLRQ arrival process, session duration, number of bytes transferred within a session, and user request patterns.
The results show that both residential and SOHO session inter-arrival times are exponentially distributed, and point out that a typical SOHO user session is longer and transfers a larger volume of data. Analysis also uncovers two main groups of session request patterns within each user category: (1) sessions that comprise traditional Internet services, such as WWW services, e-mail, and instant messenger; and (2) VHVVLRQVWKDWFRPSULVHSHHUWRSHHU¿OHVKDULQJDSSOLFDWLRQVEDVLFDOO\7KLVFKDSWHUDOVRDQDO\]HVDQG FODVVL¿HVWKHHEXVLQHVVVHUYLFHVPRVWFRPPRQO\DFFHVVHGE\XVHUVZKLFKGLGQRWYDU\VLJQL¿FDQWO\ across the user categories.
Volume II Chapter XXVII Precisions about the Broadband Divide in Chile / Sergio Godoy and M. Soledad Herrera ............... 427 This chapter aims to quantify more exactly the adoption of broadband at the household level in Chile, by assessing its impact on three types of digital divide: between users and non-users of the Internet, between usage at home and elsewhere, and between home broadband users and modem home users. This was done by a statistical analysis of WIP-Chile surveys of 2003 and 2006. At least in Chile, the main digital gap is still between users and non-users of the Internet, both in terms of age and education level. Income mainly affects the probability of having broadband access at home. Since broadband has rapidly expanded among all socioeconomic segments, it is becoming less relevant as a predictor of access and Web usage. Other factors are also weak predictors of both residential use of the Internet and broadband connections at home.
Section III Small and Medium-Sized Enterprises (SMEs) This section examines the factors affecting broadband deployment, diffusion, and use by SMEs in various countries including Italy, Singapore, and the United Kingdom. According to geographical area where the studies (included within this section) were conducted, this section is further organized in two divisions. Division I Asia Chapter XXVIII A Survey on the Adoption and Usage of Broadband Internet / Roya Gholami, John Lim, and Sang-Yong Tom Lee ...................................................................................................................... 448 7KLVFKDSWHUDUJXHVWKDWLQVSLWHRILQFUHDVLQJVLJQL¿FDQFHRIEURDGEDQG,QWHUQHWWKHUHDUHQRWPDQ\UHsearch papers explicitly addressing issues pertaining to its adoption and post-adoption. Previous research on broadband has mainly focused on the supply-side aspect at the national level, ignoring the importance of the demand side, which may involve looking more deeply into the use as well as the factors impacting
RUJDQL]DWLRQDODQGLQGLYLGXDOXSWDNH,QDQDWWHPSWWR¿OOWKLVJDSWKLVFKDSWHUHPSLULFDOO\YHUL¿HVDQ integrated theoretical model, comprising the theory of planned behavior and the IS continuance model, WRH[DPLQHIDFWRUVLQÀXHQFLQJEURDGEDQG,QWHUQHWDGRSWLRQDQGSRVWDGRSWLRQEHKDYLRURIVRPH organizations in Singapore. Overall, strong support for the integrated model has been manifested by the UHVXOWVREWDLQHGSURYLGLQJLQVLJKWLQWRLQÀXHQWLDOIDFWRUV$WWKHDGRSWLRQVWDJHSHUFHLYHGEHKDYLRUDO control has the greatest impact on behavioral intention. Findings also suggest that, as compared to attiWXGHVXEMHFWLYHQRUPVDQGSHUFHLYHGEHKDYLRUDOFRQWUROPRUHVLJQL¿FDQWO\DIIHFWWKHEURDGEDQG,QWHUQHW adoption decision. At the post-adoption stage, intention is no longer the only determinant of broadband ,QWHUQHWFRQWLQXDQFH UDWKHULQLWLDOXVDJHVLJQL¿FDQWO\DIIHFWHGEURDGEDQG,QWHUQHWFRQWLQXDQFH Division II Europe Chapter XXIX Broadband Access and Broadband-Based Applications: An Empirical Study of the Determinants of Adoption Among Italian SMEs / Massimo G. Colombo, Luca Grilli, and Cinzia Verga ........................................................................ 466 Why do some small and medium enterprises (SMEs) adopt Internet broadband technologies (high-speed connection and complementary applications), and others do not? This chapter aims at analyzing the issue through an econometric investigation. Relying on the (thin) previous empirical literature on the topic and focusing on a large and representative sample of Italian SMEs, we analyze the determinants of broadband connection and adoption of complementary applications. Results of the econometric analysis UHYHDOWKDW DPRQJ¿UPVSHFL¿FFKDUDFWHULVWLFVVL]HDQGWKH¿UP¶VQHHGWRFRPPXQLFDWHDUHPDMRU determinants both of broadband connection and use of complementary applications, while indicators of ¿UPHI¿FLHQF\DQGRIWKHFRPSHWLWLYHSUHVVXUHVXIIHUHGIURPWKH60(SRVLWLYHO\DIIHFWRQO\WKHQXPEHU of used applications, while not exerting any impact on the decision to connect; (2) among location-speFL¿FFKDUDFWHULVWLFVWKHOHYHORIWHOHFRPPXQLFDWLRQVLQIUDVWUXFWXUHSRVLWLYHO\LQÀXHQFHVERWKFRQQHFWLRQ and applications use, while the presence within the local labor market of a young and skilled workforce PDNHV60(VDGRSWPRUHDSSOLFDWLRQV DQG WLPHVSHFL¿FYDULDEOHVOLNHWKRVHUHODWHGWRWKHDFWXDODQG IXWXUHSULFHRIWKHKLJKVSHHG,QWHUQHWFRQQHFWLRQDIIHFW60(V¶GHFLVLRQVWRDGRSWEURDGEDQG7KHVH ¿QGLQJVKDYHLPSRUWDQWLPSOLFDWLRQVIRUVXSSOLHUVDQGSROLF\PDNHUV Chapter XXX Broadband Diffusion to SMEs in the UK / Oluwasola Oni and Anastasia Papazafeiropoulou ........ 481 Broadband is a relatively new technology, and its adoption in the United Kingdom has been an issue GXHWRLWVSHUFHLYHGEHQH¿WVIRUEXVLQHVVHVDQGPRUHVRIRU60(V,QWKLVFKDSWHUWKHDXWKRUVDUJXHWKDW previous research focuses on home uses of broadband, particularly for educational purposes with little attention to its adoption by SMEs. The authors argue that the existing diffusion of innovation theories are inadequate for the study of broadband diffusion, and they propose a more socio-technical approach for that purpose. This study can be useful for SMEs considering adoption of new technologies such as broadband, as well as policymakers that seek to apply effective technological adoption policies.
Chapter XXXI Environmental Drivers of E-Business Strategies among SMEs / Alessandro Arbore and Andrea Ordanini .......................................................................................................................... 493 ,QIURQWRIWUDGLWLRQDOLQWHUSUHWDWLRQVRIWKHGLJLWDOJDSEDVHGRQHQGRJHQRXVFRQGLWLRQVRIWKH¿UPVWKH authors emphasize the importance that some external pressures may have on the e-business strategy of small and medium enterprises (SMEs). The environmental factors analyzed in this chapter are market position, competitive intensity, and institutional pressures. SMEs have been grouped according to their level of e-business involvement, in relation to the number of e-business solutions adopted so far. Three layers are proposed: excluded, tentative, and integrated e-business SMEs. A general conclusion of this FKDSWHULVWKDWGLIIHUHQWPRGHOVVHHPWRH[SODLQH[FOXVLRQDQGLQYROYHPHQW6SHFL¿FDOO\WZRIDFWRUV among those analyzed are revealed to be more suitable in explaining e-business exclusion. They are the size of an SME and a lack of institutional pressures to adopt. On the opposite side, this chapter suggests that e-business involvement seems to be primarily prompted by a selective competitive environment and not by imitative behaviors, as for the previous case. Chapter XXXII ([SORULQJ60(V¶$GRSWLRQRI%URDGEDQGLQWKH1RUWKZHVWRI(QJODQG Boumediene Ramdani and Peter Kawalek .......................................................................................... 504 7KLVFKDSWHUH[SORUHVWKHIDFWRUVLPSDFWLQJ60(V¶ VPDOOWRPHGLXPVL]HGHQWHUSULVHV¶ DGRSWLRQRI broadband. It argues that ICTs are highly differentiated technologies for which there is not necessarily DVLQJOHDGRSWLRQPRGHO:KLOHPRVWODUJH(XURSHDQFRPSDQLHVDUHFRQQHFWHGWREURDGEDQG60(V¶ connectivity is lagging behind. The question of why one SME adopts broadband while another does QRWLVVWLOOXQGHUVWXGLHG7KHUHIRUHWKHSXUSRVHRIWKLVFKDSWHULVWR¿OOWKLVJDSE\LQYHVWLJDWLQJWKH WHFKQRORJLFDORUJDQL]DWLRQDODQGHQYLURQPHQWDOIDFWRUVLPSDFWLQJ60(V¶DGRSWLRQRIEURDGEDQG7KLV chapter starts by highlighting the importance of ICT innovations adoption in general and broadband in particular. Based on the ICT innovations adoption literature, an SMEs broadband adoption framework ZDVGHYHORSHGDQGHPSLULFDOO\YDOLGDWHGLQYROYLQJQLQH60(V¶NH\GHFLVLRQPDNHUVLQWKH1RUWKZHVW of England. Chapter XXXIII External Pressures for Adoption of ICT Services among SMEs / Andrea Ordanini and Alessandro Arbore ........................................................................................... 524 This chapter emphasizes the importance that external sources of pressure may have on the level of ICT involvement among small and medium enterprises (SMEs) in Italy. While past research tends to prioritize the role of endogenous conditions for the adoption of information and communication technologiesthat LV¿QDQFLDOUHVRXUFHVRUJDQL]DWLRQFRQGLWLRQVPDQDJHPHQWFXOWXUHthe high dependence of SMEs on their environment requires paying special attention to external pressures as well. Both competitive and institutional pressures are proposed and tested through an ordinal regression model on a sample of 285 SMEs. The results provides both policy and management implications.
Section IV Impact on Emerging Applications This section examines the impact of broadband on emerging ICT applications and business models. Chapters included in this section provide in-depth discussion on the impact of broadband on communication, society, and delivery of entertainment and health services to consumers and citizens. According to type of impact, this section is further organized in four divisions. Division I Entertainment Industry Chapter XXXIV IPTV Business Model Analysis / Kate Carney Landow, Michelle Fandre, Raghu Nambiath, Ninad Shringarpure, Harvey Gates, Artur Lugmayr, and Scott Barker.............................................. 538 This chapter focuses on evaluating Internet protocol television (IPTV) business models from different VHUYLFHSURYLGHUVWKURXJKWKHSULVPRIWKHVWDUPRGHO7KHWRROLVEDVHGRQ3RUWHU¶V¿YHIRUFHVDVGHveloped by M.E. Porter. The star model extends the Five Forces into a set of metrics to evaluate current and future business offerings. The star model is a simple tool used to identify the strengths and weaknesses of different business models in an appealing geometric shape. To highlight how to use this tool, sample partnership models are analyzed to evaluate the strength of a combined service. This tool will help IPTV service providers, and all potential investors, to build or identify a sound business model for their target market. The star model is explored through multiple case studies in this chapter including CBS, AOL, Google, Sling Media, and YouTube. Chapter XXXV The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry / Stanford L. Levin, John B. Meisel, and Timothy S. Sullivan .................................. 563 This chapter describes the far-reaching effects of broadband Internet access on the motion picture indusWU\,W¿UVWSURYLGHVDVXPPDU\RIWKHHIIHFWVRQWKHLQGXVWU\¶VEXVLQHVVPRGHO WKH,QWHUQHW SDUWLFXODUO\ when combined with broadband connections) provides a new window for the movie studios to utilize in releasing their products. It next examines the ways that legal, political, and cultural environments are DOUHDG\LQÀXHQFLQJWKHLQGXVWU\¶VVHDUFKIRUDQHZEXVLQHVVPRGHOWRUHSODFHWKHROG)LQDOO\LWGUDZV on lessons from the music industry to predict how the industry will ultimately incorporate broadband technology into a new business model. The authors believe that the motion picture industry provides an H[FHOOHQWFDVHVWXG\RIEURDGEDQG¶VHIIHFWVRQDPDWXUHLQGXVWU\ Division II Health Industry Chapter XXXVI Broadband for Health in Developing Countries / Aradhana Srivastava ............................................ 581 This chapter highlights the major issues in the use of broadband technologies in healthcare in developing countries. The use of Internet technologies in the health sector has immense potential in developing
countries, especially in the context of public health programs. Some of the main uses of ICT in health include remote consultations and diagnosis, information dissemination and networking between health SURYLGHUVXVHUJURXSVDQGIRUXPV,QWHUQHWEDVHGGLVHDVHVXUYHLOODQFHDQGLGHQWL¿FDWLRQRIWDUJHWJURXSV for health interventions, facilitation of health research, and support to healthcare delivery and administration. The technology has immense potential, but is also constrained by lack of policy direction, and problems with access to technology and lack of suitable infrastructure in developing nations. However, given its crucial role in public health, comprehensive efforts are required from all concerned stakeholders if universal e-health is to become a reality. Chapter XXXVII Improving Health Services via Advanced ICT Networks / Peter Farr, Isabelle Ellis, and John Royle .................................................................................................................................... 593 This chapter describes an innovative broadband initiative that connects a group of general practices, medical specialists, hospitals, and other health providers in rural areas of Australia through a managed virtual private network. It provides secure connectivity for a variety of mission-critical healthcare delivery applicationsfor example, transmission of pathology and radiology test results direct to clinicians. The medical practices involved are small to medium enterprises (SMEs), and the key aspects of ICTs for them are the impact on costs, productivity, and customer service. The formal evaluation process examined the GRPDLQVRIDSSURSULDWHQHVVHI¿FLHQF\DQGFRVWHIIHFWLYHQHVV%HLQJWKH¿UVWVXFKKHDOWKQHWZRUNRILWV kind in Australia, the project encountered challenges, and by overcoming these has been guiding government policy in respect to e-health. Initially funded from March 2005 via a Commonwealth Government grant, the GoldHealth network moved into a sustainable mode from mid-2006. The chapter provides insights into GoldHealth and should be a useful guide to any similar broadband network initiatives for the health sector elsewhere in the world. Chapter XXXVIII Remote Patient Monitoring in Residential Care Homes: Using Wireless and Broadband Networks / Tanja Bratan, Malcolm Clarke, Joanna Fursse, and Russell Jones ...................................................................................................... 604 7KH8.¶V1DWLRQDO+HDOWK6HUYLFH 1+6 LVXQGHUJRLQJJUHDWUHIRUP'ULYHQE\DGHPDQGIRUKLJKHU quality healthcare provision, ICTs are increasingly being used as tools to realize this change. The authors investigated the use of remote patient monitoring (RPM) using wireless and broadband networks in three community care homes between 2003 and 2006. The aim of the project was to determine for what conditions and in which setting the RPM was most useful, and to establish an organizational and FOLQLFDOLQIUDVWUXFWXUHWRVXSSRUWLW(YDOXDWLRQRIWKHSURMHFWGHPRQVWUDWHGFOLQLFDOEHQH¿WVVXFKDVWKH early detection of cardiac events, allowing prompt intervention, and routine monitoring of other conditions. A change in work practices resulted in a more collaborative approach to patient management and led to an increase in communication between healthcare professionals from different sectors, as well DVWKHHVWDEOLVKPHQWRISURWRFROVIRUVHHNLQJDGYLFH7HFKQLFDOO\WKHHTXLSPHQWODUJHO\PHWWKHXVHUV¶ needs. In conclusion, the monitoring proved a useful tool for the management of chronic diseases and has great potential to contribute to the reform of the NHS.
Division III Social Impact Chapter XXXIX Social Consequences of Broadband Access in Japan / Kenichi Ishii .................................................. 619 In Japan, both the cheapest wired broadband services and the most advanced 3G mobile phone services are widely available. Because of recent pro-competitive policy drives such as the “e-Japan policy,” the Japanese broadband market has become very competitive. While the digital divide has narrowed in recent years in terms of Internet access, a divide still exists with regard to Internet usage. Comparison between narrowband and broadband users demonstrates that broadband services currently are used mainly for entertainment. Unlike wired Internet use, mobile Internet is not used for information-gathering activities. Results do not support the media substitution effect of the Internet. Mobile Internet use VLJQL¿FDQWO\DQGSRVLWLYHO\FRUUHODWHVZLWKVRFLDOL]LQJZLWKIULHQGVZKHUHDVZLUHG,QWHUQHWXVHGRHVQRW correlate with socializing. Experience of past policies suggests that customer orientation will be a key factor in the success of the “U-Japan” policy. Division IV Communication Chapter XL Internet-Based Changes in Organizational Communication / Erik Lundmark and Alf Westelius....... 637 This chapter presents a descriptive study of the use of ICT and the change in communication patterns in Swedish sport associations over the period from 1994 to 2003. The change is discussed in light of Internet and broadband diffusion. Results show that new channels for communication have been adopted, primarily Web sites and e-mail, but few established channels have been dropped. While there are associations that save time, money, and increase the spirit of community using ICT, many organizations experience the increased number of communication channels as a burden, since maintaining them takes H[WUDUHVRXUFHVEXWWKHEHQH¿WVDUHQRWDOZD\VHDV\WRGHWHFWRUPHDVXUH&HUWDLQFKDUDFWHULVWLFVFRPPRQDPRQJQRQSUR¿WRUJDQL]DWLRQV 132V DVZHOODV,QWHUQHWDQGEURDGEDQGDFFHVVKDYHLQÀXHQFHG the development of ICT use. Chapter XLI Ubiquitous Communication via Residential Gateways / Alex De Smedt ............................................ 655 This chapter focuses on Residential Gateway, a type of home equipment providing broadband access to users in their homes. The chapter shows how such a device evolves from a simple modem to an advanced gateway system that contributes to the access inside the home of any data on any compatible device. The objective of this chapter is to give the reader technical insight into the enabling mechanisms and technologies dealing with such functionalities. The text indicates particular technical solutions, but explanations are kept high-level in order to allow non-technical readers to understand the basics and concepts of the solutions. A number of references show that the technical exposé is becoming a reality.
Section V Cross-Country Analysis This section examines both macro or supply-side factors and micro or demand-side factors affecting broadband deployment, diffusion, adoption, usage, and impact in various countries including Australia, Canada, China, Denmark, Germany, Greece, India, Latin America, Sweden, and the United States of America. This section is further organized into nine chapters. Chapter XLII Adoption of Broadband Services: The Role of National Policies / Morten Falch ............................. 671 Broadband is seen as a key infrastructure for developing the information society. For this reason many governments are actively engaged in stimulating investments in broadband infrastructures and use of broadband services. This chapter compares a wide range of broadband strategies in the most successful markets for broadband. This is done through analysis of national policies in three European countries (Denmark, Sweden, and Germany) and the United States, Japan, and South Korea. It is concluded that successful implementation of broadband depends on the kind of policy measures to be taken at the national level. Many countries have provided active support for stimulating diffusion of broadband, and national variants of this type of policy in different countries are important for an explanation of national differences in the adoption of broadband. Chapter XLIII Broadband Diffusion and its Driving Forces / Banani Nandi and Chandana Chakraborty............... 689 In light of the emerging consensus on potential impact of broadband technology on economic growth and development, this chapter analyzes the cross-country differences in growth of broadband technology by examining the key demand and supply factors driving diffusion in the observed countries. In addition, utilizing empirical evidence and country case analyses, the chapter offers tentative policy suggestions for accelerating broadband diffusion under alternative circumstances. Chapter XLIV Diffusion of Broadband Access in Latin America / Arturo Robles Rovalo, Claudio Feijóo González, and José Luis Gómez-Barroso .................................................................. 711 7KH³JHRJUDSKLF´GLJLWDOGLYLGHLVREYLRXVZKHQFRPSDULQJPRUHGHYHORSHGFRXQWULHVWRWKHUHVW,WV¿UVW and most obvious sign is the difference in the diffusion of broadband access. However, it is clear that there are also lines of separation in smaller geographic ranges: between countries in the same geographic DUHDLQVLGHHDFKFRXQWU\DQGVRPHWLPHVLQHDFKVSHFL¿FUHJLRQ7KLVFKDSWHUVKRZVWKLVVLWXDWLRQE\ studying the broadband access diffusion in Latin America on a three-level basis (regional, national, and local). At the national level, a few explanatory variables of the different situations presented by the countries chosen for the study are researched. Additionally, a description of the environment (market and public action) where this diffusion is occurring is also included.
Chapter XLV Diffusion Forecasting and Price Evolution of Broadband Telecommunication Services in Europe / Dimitris Varoutas, Christos Michalakelis, Alexander Vavoulas, and Konstantina Deligiorgi............................................................................... 729 This chapter is concerned with the methodologies for the study of the diffusion patterns and demand estimation, as well the pricing schemas for broadband telecommunication services in Europe. Along with the introduction of diffusion models and price indexes which can represent broadband convergence and diversity, a description of the theoretical models and methodologies are given, and application of these models in the European telecommunication market is performed. Evidence from Europe outlines telecom market behavior and contributes to better understanding of broadband diffusion worldwide. To this direction, a price index is constructed regarding the ADSL technology. Chapter XLVI Explaining Patterns of Broadband Deployment and Adoption in OECD Countries / Inmaculada Cava Ferreruela.............................................................................................................. 756 The aim of this chapter is to provide some insights about the explaining patterns of broadband deployment and adoption. This problem is addressed by examining these insights in light of the results of an exhaustive cross-national empirical analysis that uses a comprehensive panel data set from the 30 OEDC countries with more than 40 features. The results suggest that technological competition and the low cost of deploying infrastructures on one side, and the predisposition to use new technologies as well as some social indicators on the other, appear to be the key drivers for broadband deployment and adoption, respectively. Chapter XLVII ICT Statistics for Broadband Promoting Regulatory Policy / Diana Korsakaite and Tomas Lamanauskas ....................................................................................... 776 This chapter introduces the statistical analysis of a number of ICT market indicators as a means to develop sound regulatory policies aiming to promote broadband take-up. The chapter provides analysis of the concept of broadband, statistical analysis of ICT indicators time series and cross-country series DJDLQVW EURDGEDQG SHQHWUDWLRQ GHWHUPLQDWLRQ RI FRQWHUPLQRXV IDFWRUV ZLWK VWDWLVWLFDOO\ VLJQL¿FDQW LQÀXHQFHRYHUEURDGEDQGSHQHWUDWLRQLQ(XURSHDQVWDWHVDQGGHULYDWLRQRIFDOOVIRUUHJXODWRU\SROLF\ SDUWLFXODUDFWLRQVRXWRIVWDWLVWLFDOO\VLJQL¿FDQWUHODWLRQV7KHRYHUDOODLPRIWKHFKDSWHULVWRRIIHUD way as to how ICT statistical data could be employed to suggest preconditions for possible practical solutions in the domain of broadband promotion, and to use this to bring the rhetoric of statistics down to the operational level. Chapter XLVIII Impact of Broadband VoIP on Telecoms: A Cross-Country Analysis / Bardo Fraunholz and Chandana Unnithan ........................................................................................ 796 VoIP is a technology that has received much attention over the past few years. Speculations are rampant WKDWLWZLOOEHµWKH¶WHFKQRORJ\IRUWHOHFRPPXQLFDWLRQVRIWKHIXWXUHDVEURDGEDQGJDLQVPDVVPDUNHW
penetration in every nation. It holds the promise of ubiquity and eliminates the need for a separate infrastructure for telecommunications. In this chapter, the authors have undertaken a cross-country analysis of two economies, Germany and India, at varied levels of broadband VoIP diffusion, to examine the future potential of this technology in the respective nations and their telecommunications industries. A brief analysis presented in this chapter reveals some valuable insights regarding the impact of VoIP in both economies which may prove to be useful for other economies and telecommunication industries. Chapter XLIX The Adoption of Broadband Internet in Australia and Canada / Catherine Middleton and Shanton Chang ............................................................................................................................. 818 %URDGEDQG,QWHUQHWFRQQHFWLYLW\LVVHHQDVDPHDQVWRLQFUHDVHWKHHI¿FLHQF\DQGFRPSHWLWLYHQHVVRIDQ economy. But despite ongoing efforts to promote broadband in Australia, uptake has been much slower than expected. This chapter aims to identify areas that have been holding up broadband development in Australia. In examining multiple areas for attention (competition, user characteristics and behaviors, applications, network characteristics, and pricing), the authors refer to the experience of Canada, a leader in broadband deployment, to show the differences in each area. The chapter outlines objectives for the development of a more user-friendly broadband environment in Australia which would encourage broadband adoption. Although both countries discussed here have their own policy agendas and some unique circumstances related to broadband deployment, the chapter provides valuable insights for policymakers and industry leaders in Australia and in other countries which are struggling to develop widespread broadband deployment. Chapter L The Evolution of Broadband Industry in the Developing World: A Comparison of China and India / Nir Kshetri and Nikhilesh Dholakia .......................................... 841 7HOHFRPPXQLFDWLRQVQHWZRUNVRI,QGLDDQGWKH3HRSOH¶V5HSXEOLFRI&KLQDDUHDPRQJWKHODUJHVWLQWKH world. The two economies have a number of areas for broadband use ripe for exploration. Broadband networks in some regions in these two economies are even more developed than in some parts of the industrialized world. There are, however, a number of reasons to believe that these two countries may exhibit distinct and varied patterns of broadband diffusion. This chapter compares and contrasts the diffusion patterns of broadband technology in the two economies. The authors examine factors driving broadband diffusion in the two economies in three major categories: demand and cost conditions, industry structure, and export conditions.
xxxi
Foreword
I have been interested in the development and evolution of broadband for many years, and the one thing that has struck me in that time is the dynamic nature of the subject. Just when we think that we have it captured and perhaps have established a mental model for understanding it, at least in our own domain, broadband confounds us by reinventing itself and bringing up new issues and challenges in our own area and around the world. This handbook is testimony to the diverse nature of the subject, covering a wide range of issues and experiences in many sectors, industries, governments, regulatory frameworks, and areas of the world. The book is truly international and provides examples and experiences that will surprise and engage even the most knowledgeable. For example, its chapters range from the affects of broadband on the motion picture industry in the United States, and thus the world, through to actual and potential healthcare EHQH¿WVLQWKHGHYHORSLQJZRUOGIURPFDSLWDOLVWHFRQRPLFVWRVRFLDODQGFRPPXQLW\EHQH¿WVDQGEDFN As well as the experiences in a global context, the book also covers the critical role of governments and regulation and the economic development aspects of broadband. 7KHUHFDQEHIHZWHFKQRORJLHVWKDWKDYHEHHQPRUHLQÀXHQWLDODQGWKDWKDYHPRUHVHULRXVDQGSRWHQWLDOO\EHQH¿FLDOVRFLDOLPSOLFDWLRQVWKDQEURDGEDQGDQGLWVHHPVZHDUHRQO\DWWKHEHJLQQLQJRIWKHVH affects. This handbook provides an essential guide to that diversity and the issues to be addressed for successful broadband implementation.
Professor Guy Fitzgerald Brunel University, UK
Guy Fitzgerald is professor of information systems at Brunel University and is head of the Department of Information Systems and Computing. Prior to this he was the cable and wireless professor of business information systems at Birkbeck College, University of London, and before that he was at Templeton College, Oxford University. As well as being an academic, he has also worked in the computer industry with companies such as British Telecom, Mitsubishi, and CACI Inc., International. His research concerns the effective management and development of information systems and he has published widely in these areas. He is probably best known for his work in relation to development techniques and methodologies and is the author of a major text in this area entitled Information Systems Development: Methodologies, Techniques and Tools, now in its fourth edition. He is also well known for his research in the areas of strategy, outsourcing, and executive information systems. His PRVWUHFHQWUHVHDUFKLVFRQFHUQHGZLWKWKHGHYHORSPHQWRIÀH[LEOHLQIRUPDWLRQV\VWHPVWRHQKDQFHRUJDQL]DWLRQDODJLOLW\+H is founder and co-editor of the Information Systems Journal (ISJ), an international journal from Blackwell Publishing, and he has been a member of many international Program Committees, including the International Conference on Information Systems (ICIS) and the European Conference on Information Systems (ECIS).
xxxii
Foreword
The vital role played in modern society by information and communication technologies (ICTs) is nowadays recognized by both businesses and governments alike. The UN openly acknowledges the importance of the digital environment as a tool for economic and business development,1 and in Europe numerous EU initiatives2 actively promote and monitor the adoption of ICT-supported practices in order to raise productivity and growth. The changes brought about by the corresponding increasingly networked society have been many, with the last dozen years or so witnessing countless profound changes at societal, industry-sector, organizational, and individual levels. In this day and age, many of us now routinely make use of ICT for both work and leisure purposes, with speed and convenience of information exchange encouraging us to conduct transactions and communicate electronically with employers, work colleagues, friends, family members, businesses, and government agencies. As a result of the emergence and regular use of a wide variety of ICT-supported ways of doing things, we have witnessed terms such as e-business, e-government, e-health, e-learning, message boards, chat rooms, and blogging become part of our recognized vocabulary. However, the continued adoption and widespread use of ICT in daily life depends heavily upon the availability of reliable high-speed networks, and there is no doubt that broadband is a key enabling technology that allows such activity to occur reliably and at acceptable speeds. 7KLVREYLRXVVLJQL¿FDQFHRIEURDGEDQGKDVEHHQUHFRJQL]HGE\JRYHUQPHQWVDURXQGWKHZRUGDQG despite the massive investments required in terms of the provision of new network infrastructures, many have introduced policies to promote broadband availability and uptake. Indeed, the availability of affordable broadband was a key objective of the e-Europe action plan, and the current strategic framework for the European Information Society (i2010) places particular emphasis on broadband coverage and reducing the digital divide. The social and economic importance of broadband availability3 is such that broadband diffusion is often viewed as a means by which international competitiveness and economic development may be benchmarked. ,WLVZLWKLQWKLVFRQWH[WRIZLGHVSUHDGUHFRJQLWLRQRIWKHVLJQL¿FDQFHRIEURDGEDQGDQGLWVFRUUHVSRQGing promotion and take-up that this publication has been produced, and I am delighted to have been provided with the opportunity to write the Foreword to the Handbook of Research on Global Diffusion of Broadband Data Transmission. The handbook is clearly a valuable resource, providing a timely and relevant collection of chapters addressing a variety of issues pertaining to the adoption and use of broadband and the reduction of the GLJLWDOGLYLGHDWERWKPDFURDQGPLFUROHYHOV,WLVGLYLGHGLQWR¿YHVHFWLRQVHDFKFRPSULVLQJDQXPEHU of chapters and each addressing a different theme of broadband adoption. The international nature of WKHLVVXHDQGLWVWUHDWPHQWZLWKLQWKHKDQGERRNLVFOHDUO\LOOXVWUDWHGLQWKH¿UVWVHFWLRQZKLFKH[DPLQHV national policies affecting broadband deployment and diffusion in a range of countries including Aus-
xxxiii
tralia, Bangladesh, Germany, Greece, India, Ireland, Italy, Japan, New Zealand, Norway, South Africa, South Korea, Taiwan, and the United States of America. The second section focuses upon demand-side LHFRQVXPHUUHODWHG IDFWRUVLQÀXHQFLQJEURDGEDQGDGRSWLRQDQGXVHLQFRXQWULHVVXFKDV$XVWUDOLD Brazil, Chile, France, the Kingdom of Saudi Arabia, The Netherlands, the United Kingdom, and the United States of America. The third section is dedicated to examining issues of broadband use in small and medium-sized enterprises (SMEs) in Italy, Singapore, and the United Kingdom, while the fourth DQG¿IWKVHFWLRQVDUHGHYRWHGWRH[DPLQLQJWKHLPSDFWRIEURDGEDQGRQHPHUJLQJ,&7DSSOLFDWLRQVDQG business models (including those in the entertainment and health industries) and macro and micro supply and demand factors in Australia, Canada, China, Denmark, Germany, Greece, India, Latin America, Sweden, and the United States of America respectively. I am pleased to be able to recommend the handbook. It will prove highly useful to readers who are looking for substantive material on broadband promotion and adoption, and I should particularly draw attention to the variety of international perspectives presented. Overall the handbook provides an appealing treatment of the area, and I am sure it will be viewed as a valuable information resource on what is a highly topical and relevant subject.
Professor Michael D. Williams Swansea University, UK September 2007 ENDNOTES 1
2
3
General Assembly Resolution 56/183 endorsed the need to develop the so-called information society. See for instance, the eEurope 2005 Action Plan, e-Business W@tch, and the e-Business Support Network. Broadband Access: The New Highways to Prosperity—speech delivered by Vivian Reding, the member of the European Commission responsible for information society and media, at the “Bridging the Broadband Gap Through EU Spectrum Policy” event, Brussels, March 2006.
Michael Williams is a professor in the School of Business and Economics at Swansea University in the UK. He holds a BSc IURPWKH&1$$DQ0(GIURPWKH8QLYHUVLW\RI&DPEULGJHDQGD3K'IURPWKH8QLYHUVLW\RI6KHI¿HOG+HKDVLPSOHPHQWHG DQG HYDOXDWHG LQIRUPDWLRQ V\VWHPV LQ GRPDLQV LQFOXGLQJ ¿QDQFH WHOHFRPPXQLFDWLRQV PDQXIDFWXULQJ DQG JRYHUQPHQW LV the author of numerous refereed and invited papers, and has obtained external research funding from sources including the (XURSHDQ8QLRQWKH1XI¿HOG)RXQGDWLRQDQGWKH:HOVK$VVHPEO\*RYHUQPHQW+HFXUUHQWO\VHUYHVDVDQLQYLWHGPHPEHURI the project expert group for a European Union funded project examining transformative use of ICT.
xxxiv
Preface
As the Internet has become a part of everyday life, broadband has been considered as the necessary evolutionary step as a technology that offers fast, always-on Internet connections with access to services, DSSOLFDWLRQVDQGFRQWHQWZLWKUHDOOLIHVW\OHDQGSURGXFWLYLW\EHQH¿WV,QWHUQDWLRQDORUJDQL]DWLRQVVXFK as the International Telecommunication Union (ITU) and the Organization for Economic Cooperation and Development (OECD) forecast broadband to be a vital means of enhancing competitiveness in an economy and also of sustaining economic growth. Examples of delivering economic value include the potential of improving the productivity and competitiveness of small- and medium-size enterprises 60(V DQGODUJHUFRPSDQLHVDVEURDGEDQGSURYLGHVDQHI¿FLHQWFKDQQHOIRUVXSSO\FKDLQPDQDJHPHQWLPSOHPHQWDWLRQDQGVDYLQJFRVWVRIRI¿FHVSDFHE\VXSSRUWLQJHIIHFWLYHKRPHZRUNLQJ6LPLODU to commercial organizations, broadband also offers the potential to governments of creating electronic services and delivering them to citizens in a cost-effective and transparent manner. Electronic services in the public sector have the potential to reduce the cost of delivery and increase the quality of healthcare, WKHUHE\LQFUHDVLQJWKHFLWL]HQV¶WUXVWDQGFRQ¿GHQFHLQSXEOLFVHUYLFHV %URDGEDQGFDQDOVRLPSURYHFLWL]HQV¶OLYHVLQVHYHUDOZD\V,WFDQKHOSHTXLSFKLOGUHQZLWK,&7V skills for employment purposes and improve the way they obtain education. Similarly, since broadband facilitates working at home, it can help people to obtain a better work/life balance that is characterized E\PRUHHPSRZHUPHQWPRUHSURGXFWLYLW\DQGOHVVVWUHVV%URDGEDQGDOVRRIIHUVGLUHFWEHQH¿WVWRHOderly people, as it can be utilized to provide personalized care at homehence, removing the need to live in hospitals or care homes. Since broadband Internet has the potential to profoundly impact science, business, and societyand transform almost every aspect of everyday lifeit is appropriate and timely to understand the deployment and adoption of broadband technologies. Numerous researchers around the world have realized the importance of studying this research area and have focused upon accumulating knowledge in this area. %URDGEDQGUHVHDUFKKDVEHHQSUROL¿FIRUDSKHQRPHQRQWKDWLVTXLWH\RXQJ+RZHYHUDQDQDO\VLVRIWKH current literature on broadband suggests that the available body of knowledge is fragmented with some studies looking at adoption or usage patterns and some at the impact of broadband to existing or new Internet applications. This handbook is an effort to collect and group existing research results in order to offer an overall picture and comprehensive understanding of exploratory issues related to the deployment, diffusion, adoption, usage, and impact of broadband technology from a global perspective. 7RDFFHVVWKHODWHVWUHVHDUFKDQGSURYLGHDQRXWOHWWRUHVHDUFKHUVLQWKH¿HOGRIEURDGEDQGWKHHGLWRUV decided to launch this handbook where researchers from all over the world would assist in providing the necessary coverage of possible research issues within the area. The primary objective of this project was to assemble as much research coverage as possible related to the deployment, diffusion, adoption, use, and impact on emerging applications from studies conducted in various geographical settings. Our mission through this handbook is to provide an understanding of the global diffusion of broadband, examining
xxxv
factors affecting its deployment, diffusion adoption, usage, and impact on consumers and businesses from a global perspective. Additionally, the handbook helps to understand differences in the adoption of broadband in different countries and examine policy issues at national and international levels. In order to provide the most balanced coverage of concepts and issues related to the selected topics of this handbook, researchers from around the world were asked to submit proposals describing their proposed coverage and the contribution of such coverage to the handbook. All proposals were carefully UHYLHZHGE\WKHHGLWRUVLQOLJKWRIWKHLUVXLWDELOLW\UHVHDUFKHUV¶UHFRUGVRIVLPLODUZRUNLQWKHDUHDRIWKH proposed topics, and the best proposal for topics with multiple proposals. The goal was to assemble the preeminent research in broadband from all over the world to contribute entries to the handbook. Upon the receipt of full entry submissions, each submission was forwarded to at least two expert external reviewers on a double-blind, peer review basis. Only submissions with strong and favorable reviews were chosen as entries for this handbook. In many cases, submissions were sent back for several revisions prior WR¿QDODFFHSWDQFH$VDUHVXOWWKLVKDQGERRNLQFOXGHV HQWULHVKLJKOLJKWLQJDVSHFWVRIGHSOR\PHQW diffusion, adoption, and use of broadband in various geographical settings. All entries are written by knowledgeable, distinguished scholars from many prominent research institutions around the world. The extended and comprehensive coverage of broadband research in this distinctive book will contribute towards theory, practice, and policy. The theoretical contribution of this collection of studies is that it synthesizes the appropriate literature in order to enhance knowledge of broadband deployment, diffusion, adoption, usage, and impact from the global perspective. This handbook contributes to various theories and models from information systems, management, marketing, economics, and other social sciences disciplines. Some of the theories that this handbook contributes includes diffusion of innovations, technology acceptance model, theory of planned behavior, decomposed theory of planned behavior, model of adoption of technology in households, socio-technical approach, studies on developing countries, policymaking for telecommunications, and consumer behavior. Considering the relatively slow and heterogeneous adoption of broadband today, it can be learned that the policymakers and providers of the innovationin this case the telecommunications industryKROGDVSHFL¿FLQWHUHVWLQWKH¿QGLQJV of this handbook. Policymakers in various countries, particularly in the developing world, are currently investigating how to increase the diffusion of broadband within their own countries, and so information RQRWKHUFRXQWULHV¶H[SHULHQFHVFDQSURYHXVHIXO$GGLWLRQDOO\WKHWHOHFRPPXQLFDWLRQVLQGXVWU\LVLQterested in determining how to improve their current strategies. Therefore, for both policy and practice, this handbook will offer an understanding of the broadband diffusion strategies at both the macro and micro levels. This is particularly useful as there is little research published in the area of deployment, consumer adoption, usage, and impact of broadband. Understanding the usage and impact of broadband will be helpful for content developing organizations to integrate compelling content with new generation broadband, as well as to broadband service providers seeking to improve their services. In order to cater to the information needs of a diverse spectrum of readers and at the same time efIHFWLYHO\SUHVHQWWKLVJOREDOEXWFRPSOH[WRSLFWKLVKDQGERRNLVVWUXFWXUHGLQWR¿YHVHFWLRQVZLWKHDFK section including a number of divisions and consequent chapters. A brief description of each section, division, and chapter is provided below. Section I: National Policies examines macro and supply-side factors affecting broadband deployment DQGGLIIXVLRQLQYDULRXVJHRJUDSKLFDOUHJLRQVZKLFKDUHJURXSHGLQWR¿YHGLYLVLRQV $IULFD$VLD$XVtralia and New Zealand, Europe, and North America). A number of important factors including national policy, market competition, ICT competency, and digital divide are discussed within this section. 1LQHFKDSWHUVDUHLQFOXGHGLQWKLVVHFWLRQZKLFKDUHIXUWKHURUJDQL]HGLQWR¿YHGLYLVLRQVDVQRWHG above, according to their geographical orientation as follows: Division I. Africa (Chapter II: South Africa: The Long Walk to Broadband Freedom); Division II. Asia (Chapter III: Bridging the Digital
xxxvi
Divide Through Broadband Deployment, Chapter IV: Broadband Policy, Market Competition, and User Adoption in Taiwan, Chapter V: ICT Competency of Bangladesh to Face Broadband Diffusion, Chapter VI: Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society, Chapter VII: Structural Changes and Regulatory Challenges in Japanese Telecommunications Industry); Division III. Australia and New Zealand &KDSWHU9,,, $QDO\VLVRIWKH,QÀXHQFH RI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG,QWHUQHW$FFHVV&KDSWHU,; %URDGEDQGIRUWKH0DVV 0DUNHW&KDSWHU; &RPSHWLWLRQ5HJXODWLRQDQG%URDGEDQG'LIIXVLRQ 7KH&DVHRI1HZ=HDODQG Division IV. Europe &KDSWHU;, 'LJLWDO'LYLGHDQG%URDGEDQG$FFHVV 7KH&DVHRIDQ,WDOLDQ5HJLRQ &KDSWHU;,, ,PSURYLQJ%URDGEDQG$FFHVVLQ5XUDO$UHDV&KDSWHU;,,, 0HWURSROLWDQ%URDGEDQG1HWZRUNV 'HVLJQDQG,PSOHPHQWDWLRQ$VSHFWVDQG%XVLQHVV0RGHOV&KDSWHU;,9 6PDOO:RUOG 7KH,ULVK %URDGEDQG([SHULHQFH&KDSWHU;9 6RFLDO3ROLWLFDODQG(WKLFDO5HVSRQVLELOLW\LQ%URDGEDQG$GRSWLRQ and Diffusion: A German Case Study); and Division V. North America &KDSWHU;9, &RPSHWLWLRQ LQ%URDGEDQG3URYLVLRQDQGWKH'LJLWDO'LYLGH&KDSWHU;9,, *RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ %URDGEDQG$GRSWLRQ&KDSWHU;9,,, 5HJXODWLRQDQGWKH'HSOR\PHQWRI%URDGEDQG $EULHIDFFRXQW of these chapters is provided below. Chapter II, “South Africa: The Long Walk to Broadband Freedom” by Justin Henley Beneke, aims to explore the development of broadband services in South Africa, as well as touching on the challenges faced in bringing this phenomenon into the mainstream. Reasons for the lack of diffusion and adoption of such services point to high end user costs of the service, a very limited geographical footprint RIERWK¿[HGOLQHDQGPRELOHEURDGEDQGLQIUDVWUXFWXUHDVZHOODVDODFNRIFRPSXWHUOLWHUDF\DQGDQ understanding of what broadband is able to offer. The author of the chapter concludes with possible solutions to these challenges. Chapter III, “Bridging the Digital Divide through Broadband Deployment” by Challa Radhakumari, provides a summary relating to the functioning of two projects in the two Southern States of India, Andhra Pradesh and Kerala, to show how through broadband deployment in rural areas the digital divide can be bridged. By focusing on the implementation of the two projects, the chapter illustrates their contribution in practically using broadband technologies in overcoming the hurdles to bridging the digital divide, and KLJKOLJKWVWKHFULWLFDOVXFFHVVIDFWRUVDVLGHQWL¿HGGXULQJWKHIXQFWLRQLQJRIWKHSURMHFWVZKLFKKHOSHG the two states in achieving their goals. The chapter concludes by providing recommendations for application of similar projects in other geographical settings. Chapter IV, “Broadband Policy, Market Competition, and User Adoption in Taiwan” by Yu-li Liu, analyzes broadband adoption, competition among providers of broadband, and relevant policies in Taiwan. The research methods adopted include a literature review, in-depth interviews, and secondary DQDO\VLVRISUHYLRXVVXUYH\VFRQGXFWHGE\7:1,&6LQFH7DLZDQ¶VEURDGEDQGSHQHWUDWLRQLVWKH¿IWK highest in the world, the discussions of the major factors contributing to broadband deployment in this chapter can provide some useful experiences from which other countries may learn. Chapter V, “ICT Competency of Bangladesh to Face Broadband Diffusion” by Anwarul Islam and K.C. Panda, examines the initiatives taken by Bangladesh to develop its sustainable information infrastructure, reporting that teledensity and overall IT infrastructure is now in the growing stage. Nevertheless, the broadband diffusion in Bangladesh is not on par with other Asian countries, as it is still in an embryonic stage in terms of broadband diffusion. This chapter, therefore, tries to show the initiatives taken and the existing condition of Bangladesh to achieve countrywide broadband diffusion. Chapter VI, “Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society” by Dal Yong Jin, attempts to ascertain the causes of the rapid growth of broadband services in the context of the broader socio-cultural elements. It recognizes technology as a socio-cultural product which has historically been constituted by certain forms of knowledge and social
xxxvii
practice, so this chapter explores cultural elements contributing to the diffusion of broadband services in WKHFRQWH[WRIWKHFXOWXUDOHQYLURQPHQWLQ.RUHD)XUWKHULWGLVFXVVHVWKHVLJQL¿FDQWUROHRIWKHSHRSOH as users, in the process of the rapid diffusion and growth of broadband services. Chapter VII, “Structural Changes and Regulatory Challenges in the Japanese Telecommunications Industry” by Hidenori Fuke, examines the structure of the telecommunications industry in Japan which KDVEHHQXQGHUJRLQJLPPHQVHFKDQJH7KHFKDQJHVDUHREVHUYHGLQ¿YHSKDVHV GHYHORSPHQWRIFRPpetition into the local call market, diffusion of broadband Internet and development of inter-platform competition, rapid growth of cellular services and Internet access via cellular, decline of POTS (Plain Old Telephone Service), and structural changes from vertical integration to layered structure and development of media convergence. These changes require total review of the regulatory framework that was formed in the POTS era. This chapter proposes a review of essential facilities regulation, a universal VHUYLFHV\VWHPDQGDÀDWUDWHSULFLQJV\VWHPRIWKH,QWHUQHWLQRUGHUWRVROYHSUREOHPVWKDWDUHOLNHO\WR distort the new industry structure. Chapter VIII,³$QDO\VLVRIWKH,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG,QWHUQHW Access” by Qiuyan Fan, provides an in-depth analysis of the impact of policy issues on broadband Internet access in Australia. The primary goal of this chapter is to develop a holistic understanding RI$XVWUDOLD¶VQDWLRQDODSSURDFKHVSHUWDLQLQJWREURDGEDQG,QWHUQHWDFFHVV7KHJRYHUQPHQWKDVEHHQ basing its actions on market forces which help in improving broadband Internet access in the major cities, but which, however, have little effect in regional and rural Australia. The research has indicated that the regulatory competition regime has failed to address concerns of market dominance and market power in the telecommunications sector as is evidenced by a relatively lower price-performance ratio of broadband services in Australia. Chapter IX, “Broadband for the Mass Market” by Roger Saunders, suggests there is no new application to stimulate adoption of broadband by the mass market. Many new applications have been introduced but KDYHQRWFUHDWHGWKHGHVLUHGJURZWK2QHDSSOLFDWLRQWKDWLVLGHQWL¿HGDVSRWHQWLDOO\KDYLQJPDVVPDUNHW attraction is Voice over Internet Protocol (VoIP), and it is the most likely killer application. However, failure by major communications carriers to develop VoIP is slowing broadband penetration to this larger market segment. Finally, the multitude of broadband packages offered by the various competing carriers creates confusion in the mass market which, as a result, defers purchase decisions. Chapter X, “Competition, Regulation, and Broadband Diffusion: The Case of New Zealand” by Bronwyn Howell, offers a thought-provoking case study of the effects of different competition and regulatory policies on broadband diffusion rates. Despite New Zealand having one of the highest rates of Internet connection and usage in the OECD, widely available broadband infrastructure, and low broadband prices, broadband uptake per capita languishes in the bottom third of the OECD. The New Zealand case illustrates the effect that legacy regulations can have on both the diffusion of new technologies per se, and the choices made by consumers between different generational variants within that technology. Chapter XI, “Digital Divide and Broadband Access: The Case of an Italian Region” by Enrico Ferro, J. Ramon Gil-Garcia, and Natalie Helbig, looks at the issue of digital divide, and based on a review of the literature on digital divide and broadband access, the authors document different approaches to understanding the phenomenon and argue that these perspectives can also help to understand broadband access. This chapter presents a case study conducted in an Italian region. The authors provide some LPSOLFDWLRQVRIWKH¿QGLQJVDQGDUJXHWKDWSROLF\PDNHUVVKRXOGH[SORUHWKHUHODWLRQVKLSEHWZHHQ,7 skills acquisition, broadband access, and Internet use in order to develop more effective policies and programs. Chapter XII, “Improving Broadband Access in Rural Areas” by Ingjerd Skogseid, explores the characteristics of rural broadband infrastructure development. Taking the existing installed base into
xxxviii
consideration, small rural communities can initiate bottom-up cultivation of broadband infrastructure. Such initiatives are important contributions to overcoming the disparity in broadband access. The proposal is to use descriptive clusters as a way to reveal the installed base. This can be used to acquire an overview of the types of resources available and the choices that need to be made. Chapter XIII, “Metropolitan Broadband Networks: Design and Implementation Aspects, and Business Models” by Antonios Alexiou, Christos Bouras, Dimitris Primpas, and John Papagiannopoulos, presents the design principles that cover the implementation of broadband infrastructure in the region of Western Greece, by examining all the necessary parameters that arise while implementing such a critical GHYHORSPHQWDOSURMHFW7KHEURDGEDQGLQIUDVWUXFWXUHWKDWLVGHSOR\HGLVHLWKHUEDVHGRQRSWLFDO¿EHURU on wireless systems. The usage of the broadband infrastructure by service providers will be based upon the open availability of the infrastructure in a cost-effective way. This chapter also presents the main FKDUDFWHULVWLFVRIDSURSRVHGEXVLQHVVSODQWKDWHQVXUHV¿QDQFLDOYLDELOLW\RIWKHEURDGEDQGLQIUDVWUXFWXUH and guarantees the administration, growth, and exploitation of the infrastructure. Chapter XIV,³6PDOO:RUOG 7KH,ULVK%URDGEDQG([SHULHQFH´E\'LDQD:LOVRQ.HYLQ2¶5HLOO\ and Dave Murray, considers the political, cultural/social, and economic factors, both micro and macro, affecting the supply/demand nexus of broadband services for the Irish consumer. This chapter suggests that although the market is beginning to grow strongly, it is from a low base, and as a result the country still lags behind many of its European counterparts. There is still a lack of competition which is having an adverse effect on both the supply and demand of broadband. Also, the Irish consumers are still not FRQYLQFHGWKDWEURDGEDQGLVWKHµNLOOHUDSS¶WKH\QHHG7KHDXWKRUVFRQFOXGHWKDWWKLVPD\FKDQJHLQWKH future as the technology delivers more of the content-rich multimedia fare that the Irish already enjoy in other formats. Chapter XV, “Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion: A German Case Study” by Axel Schulz, Bernd Carsten Stahl, and Simon Rogerson, suggests that there is considerable interest worldwide in broadband diffusion, with research focusing on aspects such as the provision of broadband in remote areas and the socio-economic factors that determine the likelihood RIDGRSWLRQ7KLVFKDSWHULGHQWL¿HVWKHSROLFLHVDQGLQLWLDWLYHVXVHGWRHQFRXUDJHEURDGEDQGDZDUHQHVV availability, and adoption. Using the case study of a local broadband initiative in remote and rural Germany, the chapter asks the question of who can and should be responsible for broadband provision and how such responsibility ascriptions are realized. Chapter XVI, “Competition in Broadband Provision and the Digital Divide” by Wei-Min Hu and James E. Prieger, examines the supply of DSL broadband by the incumbent local exchange company /(& LQ¿YH86VWDWHVLQWKHHDUOLHU\HDUVRIGHSOR\PHQW(PSLULFDODQDO\VLVVKRZVWKDWLQFRPHRWKHU demographics, and cost factors are important determinants of entry and availability. After controlling for other factors, the racial characteristics of the area do not affect DSL provision. Active competition in broadband from competitive LECs reduces deployment of DSL by the incumbent, but potential competition from competitive LECs has the opposite effect. Competition from cable companies also negatively LQÀXHQFHVWKHLQFXPEHQW¶VGHFLVLRQWRVXSSO\'6/7KHDXWKRUV¶REMHFWLYHLQJDXJLQJWKHLPSRUWDQFHRI the various factors is to highlight the important drivers of broadband provision for policymakers. Chapter XVII, “Governmental and Cultural Factors in Broadband Adoption” by Elizabeth Fife, /DXUD+RVPDQDQG)UDQFLV3HUHLUDVXJJHVWVWKDWDOWKRXJKWKHSRWHQWLDOEHQH¿WVRIEURDGEDQG,QWHUQHW adoption are great, the levels of take-up vary greatly around the world. This chapter argues that the high level of broadband adoption rates witnessed in certain Asian economies is attributable in part to the aggressive policies pursued by the respective governments. The chapter concludes by suggesting that HYHQLIHFRQRPLFDQGVRFLDOEHQH¿WVH[LVWDVLQWKHFDVHRIWHOHPHGLFLQHLQWKH8QLWHG6WDWHVFXOWXUDO and social factors may in fact hinder the deployment of such applications and retard the growth rate of broadband access.
xxxix
Chapter XVIII, “Regulation and the Deployment of Broadband” by James E. Prieger and Sunhwa Lee, examines the impact of telecommunications regulatory policy on broadband service deployment. Using U.S. data covering all forms of access technology and all areas served by major carriers, the authors LQYHVWLJDWHWKHLPSDFWRIVWDWHDQGIHGHUDOUHJXODWLRQVRQEURDGEDQGDYDLODELOLW\7KHDXWKRUV¶REMHFWLYH in examining regulatory factors is to highlight the role of incentive regulation and local telecommunications competition policy in stimulating broadband service deployment. Section II: Consumer-User Behavior examines the micro and demand-side factors affecting broadband DGRSWLRQDQGXVDJHLQYDULRXVQDWLRQDOVHWWLQJVJURXSHGLQWR¿YHGLYLVLRQV $XVWUDOLD(XURSH0LGGOH East, North America, South America). Chapters included in this section provide in-depth discussion on socio-behavioral, attitudinal, and demographic factors affecting adoption of broadband and the digital divide at the consumer level. This section also focuses on broadband user behavior and characterizaWLRQ1LQHFKDSWHUVDUHLQFOXGHGLQWKLVVHFWLRQZKLFKDUHJURXSHGLQWR¿YHGLYLVLRQVDFFRUGLQJWRWKHLU geographical orientation as follows: Division I. Australia &KDSWHU;,; )DFWRUV$IIHFWLQJ%URDGEDQG Adoption for Mainstream Consumers); Division II. Europe &KDSWHU;; 'HYHORSLQJD'\QDPLF9LHZ RI%URDGEDQG$GRSWLRQ&KDSWHU;;, (PSOR\LQJWKH&RQWHQW9DOLGLW\$SSURDFKIRU,PSURYLQJWKH &RQWHQWRIWKH%URDGEDQG$GRSWLRQ6XUYH\,QVWUXPHQW&KDSWHU;;,, ,QVLGHWKH0LFURFRVP $&DVH 6WXG\RID:LUHOHVV,QWHUQHW+RWVSRW&KDSWHU;;,,, 7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW Division III. Middle East &KDSWHU;;,9 )DFWRUV$IIHFWLQJ$WWLWXGHVWRZDUGV%URDGEDQG$GRSWLRQLQ the Kingdom of Saudi Arabia); Division IV. North America &KDSWHU;;9 &KDUDFWHULVWLFVRI)DUPDQG Rural Internet Use in the USA); Division V. South America &KDSWHU;;9, %URDGEDQG8VHU%HKDYLRU &KDUDFWHUL]DWLRQ&KDSWHU;;9,, 3UHFLVLRQVDERXWWKH%URDGEDQG'LYLGHLQ&KLOH Chapter XIX, “Factors Affecting Broadband Adoption for Mainstream Consumers” by Peter Adams, H[SORUHVZKHWKHUSDVWH[SHULHQFHVZLWKWHOHFRPPXQLFDWLRQVSURYLGHUVDQGFXUUHQWµSODQV¶RQRIIHUVHUYH DVEDUULHUVEHWZHHQDQLQGLYLGXDOFRQVXPHU¶VSHUVXDVLRQSKDVHRIWKHLQQRYDWLRQGHFLVLRQSURFHVVDQG the decision phase. The author argues that future technology adoption studies need to consider including the complexity of the actual purchase decision when developing constructs for quantitative models. The chapter concludes by supporting that if we are to build a picture of why mainstream consumers adopt broadband, more than just the perceptions of using the technology itself need to be investigated. Chapter XX, “Developing a Dynamic View of Broadband Adoption” by Herbert Daly, Adrina Ortiz, Yogesh K. Dwivedi, Ray J. Paul, J. Santos, and J.M. Sarriegi, develops a view of UK broadband adoption using dynamic modeling techniques based on an existing statistical study. The contrasting approaches to modeling are compared. Principles of a dynamic modeling system are introduced, and an appropriate form for broadband adoption chosen. The process of building a dynamic model based on an existing static model of broadband adoption is presented. Finally, the new perspective of the dynamic model is explored using the causal loop analysis technique. Chapter XXI, “Employing the Content Validity Approach for Improving the Content of the Broadband Adoption Survey Instrument” by Yogesh K. Dwivedi, Banita Lal, and Khalil Khoumbati, illustrates WKHYDOLGDWLRQRIFRQWHQWRIWKHEURDGEDQGDGRSWLRQVXUYH\LQVWUXPHQWIURPWKHKRXVHKROGFRQVXPHU¶V SHUVSHFWLYH7KLVFKDSWHU¿UVWO\LGHQWL¿HVFRQVWUXFWVDQGWKHLUUHVSHFWLYHLWHPVWKDWDGHTXDWHO\FRYHU relevant dimensions of factors that affect consumers in the domain of broadband adoption, and second, LWGHWHUPLQHVZKHWKHUWKHLGHQWL¿HGFRQVWUXFWVDQGWKHLUUHVSHFWLYHLWHPVDGHTXDWHO\FRYHUHGWKHUHOHYDQW GLPHQVLRQV9DOLGDWLRQRIWKHLGHQWL¿HGLWHPVZDVSHUIRUPHGHPSOR\LQJDYDULDWLRQRIDTXDQWLWDWLYH DSSURDFKWRFRQWHQWYDOLGLW\7KH¿QGLQJVREWDLQHGIURPWKHFRQWHQWYDOLGDWLRQDUHSUHVHQWHGDQGGLVcussed. Chapter XXII, “Inside the Microcosm: A Case Study of a Wireless Internet Hotspot” by Pierre Vialle, Olivier Epinette, and Olivier Segard, highlights critical elements affecting the diffusion of broadband
xl
wireless Internet at a hotspot location, through a case study. The research deals with a wireless Internet services project in the main Paris airport, and comprises two components. First, this chapter analyzes the HPHUJHQFHRIDYDOXHFKDLQIRUDQHZVHUYLFHEDVHGRQWKHFRRSHUDWLRQRIVHYHUDO¿UPV,QSDUWLFXODU the authors show how different actors can or cannot position themselves on this value chain, according to their resources and capabilities. Second, the authors explore the perceptions and attitudes of business passengers in order to better understand the potential adoption and use of hotspot services, and provide a preliminary framework of analysis. Chapter XXIII,³7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW´E\.DULDQQH9HUPDDVDQG /LGZLHQYDQGH:LMQJDHUWLGHQWL¿HGDVPDOOQXPEHURIUHODWLYHO\KRPRJHQHRXVJURXSVRI'XWFK,QWHUQHW users (both broadband and narrowband), based on their usage patterns. Using individual and behavioral characteristics, the authors further investigated the nature of the different groups. This chapter presents ¿YHFOXVWHUVRIGLIIHUHQW,QWHUQHWXVHUVEDVHGRQSDWWHUQVRIEHKDYLRU7KHUHVXOWVVKRZWKDWWKH,QWHUQHW users in the sample consist of a large group that more or less conduct the same online activities. The results also show that narrowband and broadband users differ in their Internet behavior. Chapter XXIV, “Factors Affecting Attitudes towards Broadband Adoption in the Kingdom of Saudi Arabia” by Vishanth Weerakkody, sets out to explore the reasons for the slow progress in broadband adoption and investigates the factors that may be affecting the adoption of broadband by Kingdom of 6DXGL$UDELD .6$ FRQVXPHUV7KHNH\¿QGLQJV DUHWKDWWKHIDFWRUVZLWKWKHPDLQLQÀXHQFHXSRQ FRQVXPHUV¶DWWLWXGHVWRZDUGVDGRSWLRQRIEURDGEDQGDUH XVHIXOQHVV VHUYLFHTXDOLW\ DJH usage, (5) type of connection, and (6) type of accommodation. Contrary to prediction, socio-cultural IDFWRUVVXFKDVUHJXODWLRQWKURXJK¿OWUDWLRQRIEURDGEDQGZHUHIRXQGWRKDYHQRVLJQL¿FDQWLQÀXHQFH on the adoption of broadband. Chapter XXV, “Characteristics of Farm and Rural Internet Use in the United States” by Peter L. Stenberg and Mitchell Morehart, examines Internet use by farm and rural workers and proprietors using descriptive statistics and market demand analysis. The primary methodology used is categorical dependent variable analysis. The results indicate that income is a critical element, though other factors such as age RISURSULHWRUDQGUXUDOXUEDQORFDWLRQDUHDOVRVLJQL¿FDQWLQPDUNHWGHPDQGGHWHUPLQDWLRQ Chapter XXVI, “Broadband User Behavior Characterization” by Humberto T. Marques Neto, Leonardo C.D. Rocha, Pedro H.C. Guerra, Jussara M. Almeida, Wagner Meira Jr., and Virgilio A.F. Almeida, presents a broadband user behavior characterization from an Internet service provider standpoint. Analysis uncovers two main groups of session request patterns within each user category: (1) sessions that comprise traditional Internet services such as WWW services, e-mail, and instant messenger; and VHVVLRQVWKDWFRPSULVHSHHUWRSHHU¿OHVKDULQJDSSOLFDWLRQV7KLVFKDSWHUDOVRDQDO\]HGDQGFODVVL¿HGWKHHEXVLQHVVVHUYLFHVPRVWFRPPRQO\DFFHVVHGE\XVHUV Chapter XXVII, “Precisions about the Broadband Divide in Chile” by Sergio Godoy E. and M. 6ROHGDG+HUUHUD3TXDQWL¿HVWKHDGRSWLRQRIEURDGEDQGDWWKHKRXVHKROGOHYHOLQ&KLOHE\DVVHVVLQJ its impact on three types of digital divide: between users and non-users of the Internet, between usage at home and elsewhere, and between home broadband users and modem home users. In Chile, the main digital gap is still between users and non-users of the Internet, both in terms of age and education level. Income mainly affects the probability of having broadband access at home. Since broadband has rapidly expanded among all socio-economic segments, it is becoming less relevant as a predictor of access and Web usage. Section III: Small and Medium-Size Enterprises (SMEs) examines the factors affecting broadband deployment, diffusion, and use by SMEs in various countries. Six chapters are included in this section which are grouped into two divisions according to their geographical orientation as follows: Division I. Asia &KDSWHU;;9,,, $6XUYH\RQWKH$GRSWLRQDQG8VDJHRI%URDGEDQG,QWHUQHW DQGDivision II.
xli
Europe &KDSWHU;;,; %URDGEDQG$FFHVVDQG%URDGEDQG%DVHG$SSOLFDWLRQV $Q(PSLULFDO6WXG\ RIWKH'HWHUPLQDQWVRI$GRSWLRQ$PRQJ,WDOLDQ60(V&KDSWHU;;; Broadband Diffusion to SMEs in WKH8.&KDSWHU;;;, (QYLURQPHQWDO'ULYHUVRI(%XVLQHVV6WUDWHJLHVDPRQJ60(V&KDSWHU;;;,, ([SORULQJ60(V¶$GRSWLRQRI%URDGEDQGLQWKH1RUWKZHVWRI(QJODQG&KDSWHU;;;,,, ([WHUQDO3UHVsures for Adoption of ICT Services among SMEs). Chapter XXVIII, “A Survey on the Adoption and Usage of Broadband Internet” by Roya Gholami, -RKQ/LPDQG6DQJ
xlii
munication technologiesWKDWLV¿QDQFLDOUHVRXUFHVRUJDQL]DWLRQFRQGLWLRQVPDQDJHPHQWFXOWXUHthe high dependence of SMEs on their environment requires paying special attention to external pressures as well. Both competitive and institutional pressures are proposed and tested through an ordinal regression model on a sample of 285 SMEs. Section IV: Impact on Emerging Applications examines the impact of broadband on emerging ICT applications and business models. The chapters included in this section provide in-depth discussion of the impact of broadband on communication, society, and delivery of entertainment and health services to consumers and citizens. Eight chapters are included in this section, which is grouped into four divisions according to type of impact or industry sector to which they belong: Division I. Entertainment Industry &KDSWHU;;;,9 ,379%XVLQHVV0RGHO$QDO\VLV&KDSWHU;;;9 7KH,PSDFWRIWKH,QWHUQHW on the Law and Economics of the United States Motion Picture Industry); Division II. Health Industry &KDSWHU;;;9, %URDGEDQGIRU+HDOWKLQ'HYHORSLQJ&RXQWULHV&KDSWHU;;;9,, ,PSURYLQJ+HDOWK 6HUYLFHVYLD$GYDQFHG,&71HWZRUNV&KDSWHU;;;9,,, 5HPRWH3DWLHQW0RQLWRULQJLQ5HVLGHQWLDO Care Homes (Using Wireless and Broadband Networks); Division III. Social Impact &KDSWHU;;;,; Social Consequences of Broadband Access in Japan); and Division IV. Communication &KDSWHU;/ ,QWHUQHW%DVHG&KDQJHVLQ2UJDQL]DWLRQDO&RPPXQLFDWLRQ&KDSWHU;/, 8ELTXLWRXV&RPPXQLFDWLRQ via Residential Gateways). Chapter XXXIV, “IPTV Business Model Analysis” by Kate Carney Landow, Michelle Fandre, Raghu Nambiath, Ninad Shringarpure, Harvey Gates, Artur Lugmayr, and Scott Barker, focuses on evaluating Internet protocol television (IPTV) business models from different service providers through the prism of the star model. The star model is a simple tool used to identify the strengths and weaknesses of different business models in an appealing geometric shape. To highlight how to use this tool, sample partnership models are analyzed to evaluate the strength of a combined service. This tool will help IPTV service providers, and all potential investors, to build or identify a sound business model for their target market. The star model is explored through multiple case studies in this chapter including CBS, AOL, Google, Sling Media, and YouTube. Chapter XXXV, “The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry” by Stanford L. Levin, John B. Meisel, and Timothy S. Sullivan, describes the far-reachLQJHIIHFWVRIEURDGEDQG,QWHUQHWDFFHVVRQWKHPRWLRQSLFWXUHLQGXVWU\,W¿UVWO\SURYLGHVDVXPPDU\RI WKHHIIHFWVRQWKHLQGXVWU\¶VEXVLQHVVPRGHO WKH,QWHUQHW SDUWLFXODUO\ZKHQFRPELQHGZLWKEURDGEDQG connections) provides a new window for the movie studios to utilize in releasing their product. It next H[DPLQHVWKHZD\VWKDWOHJDOSROLWLFDODQGFXOWXUDOHQYLURQPHQWVDUHDOUHDG\LQÀXHQFLQJWKHLQGXVWU\¶V search for a new business model to replace the old. Finally, it draws on lessons from the music industry to predict how the industry will ultimately incorporate broadband technology into a new business model. Chapter XXXVI, “Broadband for Health in Developing Countries” by Aradhana Srivastava, highlights the major issues in the use of broadband technologies in healthcare in developing countries. The technology has immense potential, but is also constrained by lack of policy direction and problems with access to technology and lack of suitable infrastructure in developing nations. However, given its crucial role in public health, comprehensive efforts are required from all concerned stakeholders if universal e-health is to become a reality. Chapter XXXVII, “Improving Health Services via Advanced ICT Networks” by Peter Farr, Isabelle Ellis, and John Royle, describes an innovative broadband initiative that connects a group of general practices, medical specialists, hospitals, and other health providers in rural areas of Australia through a PDQDJHGYLUWXDOSULYDWHQHWZRUN 931 %HLQJWKH¿UVWVXFKKHDOWKQHWZRUNRILWVNLQGLQ$XVWUDOLDWKH project encountered challenges, and by overcoming these has been guiding government policy in respect to e-health. The chapter provides insights into the case study and should be a useful guide to any similar broadband network initiatives for the health sector elsewhere in the world.
xliii
Chapter XXXVIII, “Remote Patient Monitoring in Residential Care Homes (Using Wireless and Broadband Networks)” by Tanja Bratan, Malcolm Clarke, Joanna Fursse, and Russell Jones, investigates the use of Remote Patient Monitoring (RPM), using wireless and broadband networks, in three community care homes in the UK between July 2003 and January 2006. The aim of the project was to determine for what conditions, and in which setting, the RPM was most useful, and to establish an organizational and FOLQLFDOLQIUDVWUXFWXUHWRVXSSRUWLW(YDOXDWLRQRIWKHSURMHFWGHPRQVWUDWHGFOLQLFDOEHQH¿WVVXFKDVWKH early detection of cardiac events, allowing prompt intervention and routine monitoring of other conditions. A change in work practices resulted in a more collaborative approach to patient management and led to an increase in communication between healthcare professionals from different sectors, as well as the establishment of protocols for seeking advice. Chapter XXXIX, “Social Consequences of Broadband Access in Japan” by Kenichi Ishii, suggests that in Japan, while the digital divide has narrowed in recent years in terms of Internet access, a divide still exists with regards to Internet usage. A comparison between narrowband and broadband users demonstrates that broadband services currently are used mainly for entertainment. Unlike wired Internet use, mobile Internet is not used for information-gathering activities. The results do not support the PHGLDVXEVWLWXWLRQHIIHFWRIWKH,QWHUQHW0RELOH,QWHUQHWXVHVLJQL¿FDQWO\DQGSRVLWLYHO\FRUUHODWHGZLWK socializing with friends, whereas the wired Internet use did not correlate with socializing. Chapter XL, “Internet-Based Changes in Organizational Communication” by Erik Lundmark and Alf Westelius, presents a descriptive study of the use of information and communication technology (ICT) and the change in communication patterns in Swedish sport associations over the period from 1994 to 2003. The change is discussed in light of Internet and broadband diffusion. The results show that new channels for communication have been adopted, primarily Web sites and e-mail, but few established FKDQQHOVKDYHEHHQGURSSHG&HUWDLQFKDUDFWHULVWLFVFRPPRQDPRQJQRQSUR¿WRUJDQL]DWLRQV 132V DVZHOODV,QWHUQHWDQGEURDGEDQGDFFHVVKDYHLQÀXHQFHGWKHGHYHORSPHQWRI,&7XVH Chapter XLI, “Ubiquitous Communication via Residential Gateways” by Alex De Smedt, focuses on Residential Gateway, a type of home equipment providing broadband access to users in their homes. The chapter shows how such a device evolves from a simple modem to an advanced gateway system that contributes to the access inside the home of any data on any compatible device. The chapter gives technical insight into the enabling mechanisms and technologies dealing with such functionalities. Section V: Cross-Country Analysis examines both macro and supply-side factors and micro and demand-side factors affecting broadband deployment, diffusion, adoption, usage, and impact in various countries, including Australia, Canada, China, Denmark, Germany, Greece, India, Latin America, Sweden, and the United States of America. This section is further organized into nine chapters. Chapter XLII, “Adoption of Broadband Services: The Role of National Policies” by Morten Falch, compares a wide range of broadband strategies in the most successful markets for broadband. This is done through an analysis of national policies in three European countriesDenmark, Sweden, and Germanyand the United States, Japan, and South Korea. It is concluded that the successful implementation of broadband depends on the kind of policy measures to be taken at the national level. Many countries have provided active support for stimulating the diffusion of broadband, and national variants of this type of policy in different countries are important for an explanation of national differences in the adoption of broadband. Chapter XLIII, “Broadband Diffusion and its Driving Forces” by Banani Nandi and Chandana Chakraborty, analyzes the cross-country differences in the growth of broadband technology by examining the key demand and supply factors driving diffusion in the observed countries. In addition, utilizing empirical evidence and country case analyses, the chapter offers tentative policy suggestions for accelerating broadband diffusion under alternative circumstances.
xliv
Chapter XLIV, “Diffusion of Broadband Access in Latin America” by Arturo Robles Rovalo, Claudio Feijóo González, and José Luis Gómez-Barroso, studies the broadband access diffusion in Latin America on a three-level basis (regional, national, and local). At the national level, a few explanatory variables of the different situations presented by the countries chosen for the study are researched. Additionally, a description of the environment (market and public action) where this diffusion is occurring is also included. Chapter XLV, “Diffusion Forecasting and Price Evolution of Broadband Telecommunication Services in Europe” by Dimitris Varoutas, Christos Michalakelis, Alexander Vavoulas, and Konstantina Deligiorgi, is concerned with the methodologies for the study of the diffusion patterns and demand estimation, as well the pricing schemas for broadband telecommunication services in Europe. Along with the introduction of diffusion models and price indexes which can represent broadband convergence and diversity, a description of the theoretical models and methodologies are given and application of these models in the European telecommunication market is performed. Evidence from Europe outlines telecom market behavior and contributes to better understanding of broadband diffusion worldwide. Chapter XLVI, “Explaining Patterns of Broadband Deployment and Adoption in OECD Countries” by Inmaculada Cava Ferreruela, provides some insights about the explanatory patterns of broadband deployment and adoption. This problem is addressed by examining patterns in light of the results of an exhaustive cross-national empirical analysis that uses a comprehensive panel data set from the 30 OEDC countries with more than 40 features. The results suggest that technological competition and the low cost of deploying infrastructures on one side, and the predisposition to use new technologies as well as some social indicators on the other, appear to be the key drivers for broadband deployment and adoption respectively. Chapter XLVII, “ICT Statistics for Broadband Promoting Regulatory Policy” by Diana Korsakaite and Tomas Lamanauskas, introduces the statistical analysis of a number of information and communication technology (ICT) market indicators as a means to develop sound regulatory policies aiming to promote broadband take-up. The overall aim of the chapter is to offer a way of how ICT statistical data could be employed to suggest preconditions for possible practical solutions in the domain of broadband promotion, and in this way to bring the rhetoric of statistics down to the operational level. Chapter XLVIII, “Impact of Broadband VoIP on Telecoms: A Cross-Country Analysis” by Bardo Fraunholz and Chandana Unnithan, undertakes a cross-country analysis of two economiesGermany and Indiaat varied levels of broadband VoIP diffusion to examine the future potential of this technology in the respective nations and their telecommunications industries. A brief analysis presented in this chapter reveals some valuable insights regarding the impact of VoIP in both economies which may prove to be useful for other economies and telecommunication industries. Chapter XLIX, “The Adoption of Broadband Internet in Australia and Canada” by Catherine 0LGGOHWRQDQG6KDQWRQ&KDQJLGHQWL¿HVDUHDVWKDWKDYHEHHQKROGLQJXSEURDGEDQGGHYHORSPHQWLQ Australia. In examining multiple areas for attention (competition, user characteristics and behaviors, applications, network characteristics, and pricing), the authors refer to the experience of Canada, a leader in broadband deployment, to show the differences in each area. The chapter outlines objectives for the development of a more user-friendly broadband environment in Australia which would encourage broadband adoption. Finally, Chapter L, “The Evolution of Broadband Industry in the Developing World: A Comparison of China and India” by Nir Kshetri and Nikhilesh Dholakia, compares and contrasts the diffusion patterns of broadband technology in the two economies. The authors examine factors driving broadband diffusion in the two economies in three major categories: demand and cost conditions, industry structure, and export conditions.
1
Chapter I
Broadband Adoption and Diffusion (BAD): A Framework Yogesh K. Dwivedi Swansea University, UK Anastasia Papazafeiropoulou Brunel University, UK
ABSTRACT The aim of this chapter is to outline various factors reported in the 49 chapters of this handbook of UHVHDUFK7KHDLPLVDOVRWRRUJDQL]HLGHQWL¿HGIDFWRUVLQDPHDQLQJIXOPDQQHULQRUGHUWRSURSRVHD framework of broadband adoption and diffusion. This chapter illustrates the fact that research on the adoption, diffusion, usage, and impact of broadband is clearly a global issue which requires a multidisciplinary approach. The proposed framework includes three levels of factors—macro factors, individual micro factors, and SME-level micro factors. These three levels of factors are relevant at different levels of development, deployment, and diffusion of broadband which persist in various developed and developing countries. The chapter concludes by suggesting that the proposed framework is based on a comprehensive set of factors observed in various countries, and future studies may use this framework to identify gaps and then bridge those gaps by conducting new studies.
INTRODUCTION A recently published book on the issue of broadband adoption from the consumer perspective suggested that “the problem of the slow rate of broadband adoption and usage in many countries includes factors that may not be covered only by examining consumer perspectives” (Dwivedi, 2007). Since the deployment and diffusion of broadband is in different stages in different countries, developed and developing worlds require different types of
approaches for managing the deployment, diffusion, and adoption of broadband. For example, in terms of a developing country, there is a need to look at regulation and infrastructure development issues at the national level. However, in terms of developed countries where progress has been made towards infrastructure development in urban areas, there is a need to determine the factors affecting broadband adoption and use in the rural and farm economic sectors and the development of broadband infrastructure in rural communities.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband Adoption and Diffusion (BAD)
Researchers within the developed countries context should systematically examine the broadband issues at various levels, for example, heterogeneous coverage, adoption, and usage of broadband. This is also linked with the issue of the digital divide, so future research efforts should be focused on examining the varying levels of adoption in different sections of society and formulating strategies that policymakers may apply to create more homogeneity in the digital society (Dwivedi, 2007). Similar to individual consumers, small and medium-size enterprises (SMEs) have also been slow in adopting broadband. Factors that may affect individual consumers are likely to differ with those affecting the adoption of broadband by SMEs. Dwivedi (2007) provides an in-depth understanding of the factors affecting broadband adoption by consumers, but the factors affecting broadband adoption by SMEs remain largely untouched and therefore require the immediate attention of researchers (Dwivedi, 2007). New electronic services such as e-government services are currently being implemented in many countries. The diffusion and adoption of highspeed Internet is a prerequisite for the successful adoption of such emerging electronic services by citizens (Dwivedi, Papazafeiropoulou, Gharavi, & Khoumbati, 2006). This suggests that studying the impact of broadband on consumers, particularly in areas such as consumer adoption of new communication methods, music and software downloads, entertainment, retail, travel, and tourism on an LQGLYLGXDOEDVLVFDQEHEHQH¿FLDOLQGHWHUPLQLQJ the real impact of broadband (Dwivedi, 2007). Widespread adoption of broadband is also likely to change the way many businesses undertake their business processes by affecting the value chain. Therefore it is important to examine the business model of many sectors, particularly television, telecommunications, publishing, and the picture/photo industry (Dwivedi, 2007). For example, future research needs to examine the ZD\WKHDGRSWLRQRIEURDGEDQGZLOOLQÀXHQFHWKH economics of the motion picture industry. Also, with broadband facilitating the implementation of IPTV, the issue arises as to how IPTV will affect the current business model of the media
2
and broadcasting/television industry. Similarly, broadband is an enabling technology for PC-to-PC communication such as VoIP, so a further issue which requires attention is how broadband VoIP ZLOO LQÀXHQFH WKH WHOHFRPPXQLFDWLRQV LQGXVWU\ (Dwivedi, 2007). The issues discussed above are certainly critical to the success of the adoption and diffusion of broadband at global, national, and local levels. The aim of this chapter is to outline the issues and factors that are explored and discussed within this handbook of research by various authors from different JHRJUDSKLFDOUHJLRQV7KLVFKDSWHU¿UVWLOOXVWUDWHV DQGGHVFULEHVWKHSUR¿OHDQGEDFNJURXQGRIWKH contributors. Then it proceeds to list and outline various factors that affect broadband deployment, diffusion, adoption, usage, and the impact of broadband. Finally, before concluding the chapter, a brief discussion on a proposed framework on broadband adoption and diffusion is provided.
PROFILE OF AUTHORS Research Discipline of Authors Table 1 illustrates the fact that the authors of this handbook represent various research areas, as the expertise of their home departments varies widely. A total of 24 research disciplines are represented by authors including business, management, economics, IT, computer science, health science, and information systems (see Table 1). This clearly suggests that research on broadband deployment, diffusion, adoption, usage, and impact is the concern of many disciplines and, hence, is truly multidisciplinary in nature.
Geographical Location of Authors Table 2 illustrates the global nature of research on broadband-related issues. Chapters from more than 25 countries including both developed and developing countries from all continents were included in this handbook. However, the predominance of the chapter contributions is from Australia, European Countries, India, Japan, the UK, and the United States.
Broadband Adoption and Diffusion (BAD)
Table 1. Research discipline of authors SN
Discipline
Frequency
1
Business, Management & Economics
18
2
Center for Technology in Government
01
3
Computer Engineering Department
01
4
Center for Computing and Social Responsibility
01
5
Consultancy/Industry
05
6
Department of Communication
03
7
Department of Radio & TV
01
8
Department of Library & Information Science
01
9
Department of Global Media
01
10
Department of Information Systems
12
11
Department of Public Affairs and Public Policy
02
12
Health and Medical Sciences
02
13
Information and Communication Technology
01
14
Information & Communication
01
15
Industrial Management Engineering
01
16
Information Technology Management
02
17
International Business
01
18
Informatics and Communication
01
19
Management & Engineering
01
20
Public Sector
02
21
Systems Information Engineering
01
22
Telecommunications Department
01
23
Volunteer Research Organization
01
24
Others
11
Number of Authors Contributing in Each Chapter Table 3 shows that the majority of chapters were written by more than one author: 16 chapters were single authored while the remaining 34 chapters were co-authored by two or more authors. This again highlights the fact that the area of research is complex and often requires varied skills and knowledge.
FACTORS AFFECTING BROADBAND DEPLOYMENT AT NATIONAL LEVEL: NATIONAL POLICIES AND OTHER MACRO FACTORS This section examines the macro or supply-side factors affecting broadband deployment and diffusion in various countries. A total of 24 pertinent factors and issues including national policy, market competition, ICT competency, structural changes,
3
Broadband Adoption and Diffusion (BAD)
Table 2. Geographical location of authors SN
Discipline
Frequency
1
South Africa
01
2
India
04
3
Taiwan
01
4
Bangladesh
01
5
Canada
02
6
Japan
02
7
Australia
05
8
New Zealand
01
9
Italy
04
10
Norway
01
11
Greece
02
12
Ireland
01
13
Germany
03
14
USA
06
15
UK
06
16
France
01
17
The Netherlands
01
18
The Kingdom of Saudi Arabia
01
19
Brazil
01
20
Singapore
01
21
Sweden
02
22
Belgium
01
23
Denmark
01
24
Latin America
01
25
OECD Countries–Spain
01
26
EU–Lithuania
01
27
South Korea
01
28
China
01
29
Mexico
01
regulatory policy/challenges, socio-political, and HWKLFDO UHVSRQVLELOLW\ ZHUH LGHQWL¿HG IURP YDULous chapters (see Table 4 for a list of the factors). However, the dominant factors that were reported most frequently include competition, campaigns/ promotion, infrastructure-related issues, regulation, development and access cost, and government intervention/policy. Table 4 illustrates all macro
4
factors and the corresponding chapters in which they are reported and discussed.
Consumer-User Behavior This section of the book examines the micro or demand-side factors affecting broadband adoption and usage in various countries including Australia,
Broadband Adoption and Diffusion (BAD)
Table 3. Authors per chapter SN
Number of Authors Per Chapter
Frequency
1
1
16
2
2
17
3
3
11
4
4
03
5
5
00
6
6
02
7
7
01
Brazil, Chile, France, The Kingdom of Saudi Arabia, The Netherlands, the United Kingdom, and the United States of America. The chapters included in this section provide in-depth discussions on 34 factors covering socio-behavioral, attitudinal, and demographic factors affecting adoption of broadband and the digital divide at a consumer level. This section also focuses on broadband user behavior and characterization. Table 5 further illustrates all factors and the corresponding chapters in which they are reported and discussed.
SMALL AND MEDIUM-SIZE ENTERPRISES This section of the book examines the factors affecting broadband deployment, diffusion, and use by SMEs in various countries including Italy, Singapore, and the United Kingdom. Table 6 further illustrates 34 different factors and corresponding chapters in which they are reported and discussed.
IMPACT OF BROADBAND ON EMERGING APPLICATIONS This section of the book examines the impact of broadband on emerging ICT applications and business models. The chapters included in this section provide in-depth discussions on the impact of broadband on communication, society, and delivery of entertainment and health services to consumers
and citizens. Table 7 further illustrates various types of impact and the corresponding chapters in which they are reported and discussed
BROADBAND ADOPTION AND DIFFUSION (BAD): A FRAMEWORK Within this section we provide a very brief description of the proposed framework on broadband adoption and diffusion. The proposed framework is illustrated in Figure 1. The framework is based RQIDFWRUVLGHQWL¿HGDQGGLVFXVVHGE\YDULRXVDXthors of this handbook. This framework suggests that countries, mainly developing countries that are at an early stage of broadband development, should give major emphasis on macro-factors that affect deployment at a national level by creating an appropriate regulatory, technological, and socioeconomic environment. Once an appropriate and conducible environment is established, the next step is to focus on addressing the factors affecting adoption and use by households and SMEs. This is the focus of attention for the majority of developed countries. The majority of developed nations, for example Australia, have already deployed a broadband infrastructure; however, consumers are slow in adopting it due to various socio-economic reasons. Therefore, the emphasis for such countries should be removing the barriers of individual adoption and use of broadband. Failing to implement such strategies may lead to a digital divide at various levels including global, national, regional, and consumer-user levels. Broadband adoption and use is likely to affect both industry and society. Some of the socio-economic effects are discussed within this handbook. However, to avoid a harmful effect on society and to exploit the useful effects of broadband on various industries, integrated and continuous efforts from various stakeholders are needed.
CONCLUSION This chapter has outlined the various factors reported in the 49 chapters that follow in this handbook of research. Furthermore, the chapter 5
Broadband Adoption and Diffusion (BAD)
Table 4. Macro factors SN
National Policies/ Macro Factors
Chapters
1
Broadband Availability & Coverage
C2, C5, C12, C18
2
Competence
C2
3
Public Awareness Campaigns/ Promotion and Communication
C3, C14, C44
4
Training Policy
C3, C5
5
Broadband Policy
C4
6
Telecommunication Liberalization
C4, C14, C45
7
Competition
8
Infrastructure Development
C4, C5, C7, C8, C9, C12, C13, C17, C20, C46
9
Broadband Strategies
C4
C4, C5, C6, C7, C8, C10, C14, C16, C18, C20, C43, C45, C46, C49
10
Socio-Cultural Factors
C6, C17
11
Regulation/Regulatory Policies
C7, C8, C10, C14, C18, C42, C49
12
Development Cost/Access Pricing
C7, C8, C9, C12, C16, C18, C45, C49, C50
13
Government Support–Intervention/Government Policy
C8, C9, C20, C42, C43, C45, C47
14
Location
C11, C16, C43, C44
15
Responsibility
C15
16
Authority and Norms
C15
17
Market Size & Composition
C16
18
Telephone Penetration
C16
19
Investment/Foreign Direct Investment (FDI)
C17, C20
20
Leadership
C21
21
Network Provisioning
C21
22
Authentication and Security
C21
23
Industry Structure
C50
24
Export Conditions
C50
LGHQWL¿HG DQG RUJDQL]HG IDFWRUV LQ D PHDQLQJful manner in order to propose a framework of broadband adoption and diffusion. This chapter has illustrated that research on adoption, diffusion, usage, and the impact of broadband is clearly a global issue which requires a multidisciplinary approach. The proposed framework includes three levels of factorsmacro factors, individual micro factors, and SME-level micro factors. These three levels of factors are relevant at different levels of development, deployment, and diffusion of broadband, and persist in various developed and developing countries. Both the industry and policymakers should carefully choose and address the issues
6
relevant to their country and regions for encouraging homogenous adoption and usage of broadband. This framework is based on a comprehensive set of factors observed in various countries. However, as this is a complex and evolving area, future studies may use this framework to identify gaps and then bridge those gaps by conducting new studies.
REFERENCES BSG Report. (2001). Report and strategic recommendations. London.
Broadband Adoption and Diffusion (BAD)
Table 5. Micro factors SN
Micro Factors
Chapters C2, C4, C5, C6, C9, C14, C19, C20,
1
Cost/Price/Subscription Fee C21, C43, C49
2
Knowledge/Awareness/Information Sources
C3, C19, C20, C21, C43
3
6RFLDO,QÀXHQFH5HFRPPHQGDWLRQV
C3, C4, C14, C20, C21
Demographics/Users Characteristics (e.g., age, gender, education, income,
C2, C3, C4, C5, C6, C11, C16, C22,
occupation, and ethnicity)
C21, C23, C25, C27, C46, C49
4 5
Speed
C4
6
Citizen/Consumer/User Training
C3, C5, C11, C20
7
Stability
C4
8
Content/Applications
C4, C5, C6, C9, C22, C49
9
Reliability
C4
10
Convenience
C4, C14
11
Supplier Reputation
C4
12
Individualism
C6
13
PC at Home
C11
14
Household Size
C11
15
Customer Needs
C14, C20, C23
16
Perceived Usefulness/Utilitarian Outcomes
C19, C20, C21, C26
17
Image
C19
18
Perceived Complexity
C19
19
Technology Orientation
C19
20
Life Style Motivations (career, family, entertainment)
C19
21
Purchase Complexity
C19
22
Commitment to Technology
C19
23
Relative Advantage
C20, 21
24
Service Quality
C20, C21
25
Usage Pattern/User Behavior
C20, C23, C26, C49
26
Hedonic Outcomes
C20, C21
27
Personality
C22
28
Experience
C22, C23
29
Attitude towards Computing and Wireless Internet
C22
30
Affordability
C43
31
Broadband UseInformation
C23
32
Broadband UseCommunication
C23
33
Broadband UseEntertainment
C23
34
Broadband UseTransactions
C23
7
Broadband Adoption and Diffusion (BAD)
Table 6. SMEs
8
SN
Factors Affecting BB Adoption by SMEs
Chapters
1
Attitude
C28
2
Perceived Behavioral Control
C28
3
Subjective Norms
C28
4
Behavioral Intention
C28
5
Perceived Usefulness
C28
6
&RQ¿UPDWLRQ
C28
7
Satisfaction
C28
8
Continuance Intention
C28
9
Initial Usage/Continuous Usage
C28
10
Firm Size
C29, C31, C33
11
)LUP¶V1HHGWR&RPPXQLFDWH
C29
12
Existing Level of Telecommunication Infrastructure
C29
13
Price
C29
14
Compatibility
C30, C32
15
Complexity
C30, C32
16
Relative Advantage
C30, C32
17
Social Groups Involved in the Production and Use of the Innovation
C30
18
Lack of Institutional Pressure to Adopt
C31
19
Trailability
C32
20
Observability
C32
21
Top Management Support
C32
22
ICT Experience
C32
23
Location
C32
24
Competitive Pressure
C32, C33
25
External ICT Support
C32
26
Government Pressure
C32
27
Financial Resources
C33
28
Organization Conditions
C33
29
Management Culture
C33
30
Leading Market Position
C33
31
Business Group Membership
C33
33
Technology Legitimacy
C33
34
Manufacturing
C33
Broadband Adoption and Diffusion (BAD)
Table 7. Emerging applications SN
Industry/Sector Affected by BB Adoption and Usage
Chapters
1
IPTV Business Model
C34
2
Motion Picture Industry
C35
3
Health Sector
C36, 37, 38
4
Social Consequences
C39
5
Organizational Communication
C40
6
VoIP
C48
7
Evolution of Facilitating/Supporting Technologies
C41
Figure 1. Broadband adoption and diffusion (BAD): A framework National Policies/Macro Factors (See Table 4)
Deployment of Broadband
Micro Factors Consumer Level Factors (See Table 5)
Diffusion, Adoption, and Use of Broadband at Various Levels
Broadband Digital Divide
Impact of Broadband (See Table 7)
Micro FactorsSMELevel Factors (See Table 6)
Dwivedi, Y.K. (2007). Consumer adoption and usage of broadband. Hershey, PA: IGI Global. Dwivedi, Y.K., Papazafeiropoulou, A., Gharavi, H., & Khoumbati, K. (2006). Examining the socio-economic determinants of adoption of an HJRYHUQPHQW LQLWLDWLYH µ*RYHUQPHQW *DWHZD\¶ The Electronic GovernmentAn International Journal, 3(4), 404-419. Rice, C. (1997). Understanding customers. Oxford: Butterworth-Heinemann Rogers, E.M. (1995). Diffusion of innovations. New York: The Free Press.
KEY TERMS Broadband: Always-on access, at work, at home, or on the move, provided by a range of ¿[HGOLQHZLUHOHVVDQGVDWHOOLWHWHFKQRORJLHVWR progressively higher bandwidths capable of supporting genuinely new and innovative interactive content, applications, and services, and the delivery of enhanced public services (BSG Report, 2001). Consumers-Users: Those who pay for services DQGJRRGVZKLOHµXVHUV¶DUHLQGLYLGXDOVZKRDUH affected by or who affect the products or services. In other words, users are those who use the prod-
9
Broadband Adoption and Diffusion (BAD)
ucts and services but do not pay for them (Rice, 1997). For example, a child can be categorized as a user since he or she uses broadband for online gaming and to undertake homework; however, the child does not pay for the service. Contrastingly, WKHSDUHQWFDQEHLGHQWL¿HGDVDFRQVXPHUDVZHOO as a user since he or she pays and uses the service (Rice, 1997). Diffusion: The process by which an innovation is communicated through certain channels
10
over time among the members of a social system (Rogers, 1995). IPTV: Digital television content delivered through IP-based networks. VoIP: Refers to voice over Internet protocol. This is the family of technologies that allow IP networks to be used for voice applications, such as telephony, voice instant messaging, teleconferencing, and so forth.
Section I
National Policies
Division I
Africa
13
Chapter II
South Africa:
The Long Walk to Broadband Freedom Justin Henley Beneke University of Cape Town, South Africa
ABSTRACT South Africa has fallen behind its international peers—both developing and developed markets—in the race to rollout broadband services. In fact, even within the African continent, it is neither the broadband leader nor progressive in comparison to its Northern African counterparts. This chapter aims to explore the development of broadband services in South Africa, as well as to touch upon the challenges faced in bringing this phenomenon into the mainstream. Reasons for the lack of diffusion and adoption of such VHUYLFHVSRLQWWRKLJKHQGXVHUFRVWVRIWKHVHUYLFHDYHU\OLPLWHGJHRJUDSKLFDOIRRWSULQWRIERWK¿[HG line and mobile broadband infrastructure, as well as a lack of computer literacy and an understanding of what broadband is able to offer. The chapter continues to look at possible solutions, including introducing a greater degree of competition into the market to facilitate downward pressure on prices, as ZHOODVSURYLGLQJFRVWEDVHGDFFHVVWRLQWHUQDWLRQDOVXEPDULQH¿EHUFDEOHVDQGWKHXQEXQGOLQJRIWKH local loop to further this objective.
INTRODUCTION South Africa, situated at the southern-most tip of Africa, has an abundance of good fortunesnatural beauty, mineral wealth, a warm climate, and the like. Yet, most South Africans are considerably
more likely to know about the happenings in the lives of American movie stars or British musicians WKDQWRKDYHDWUXHFRQFHSWLRQRIµEURDGEDQG¶ This chapter aims to explore the development of broadband services in South Africa, as well as touching on the challenges faced in bringing this phenomenon into the mainstream.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
South Africa
A BRIEF HISTORY OF BROADBAND DEPLOYMENT IN SOUTH AFRICA 7HONRP6RXWK$IULFD¶VVWDWHFRQWUROOHGWHOHFRPmunications company, was awarded a monopoly RQ¿[HGOLQHWHOHFRPPXQLFDWLRQVVHUYLFHVEHWZHHQ 1997 and 2002 (Bridges.org, 2001). As part of its obligations, the company was mandated to implement PSTN services in rural and semi-urban areas around the country, many of which would otherwise not have proved commercially viable to operators in a competitive environment. Although initially successful in its endeavors to connect the masses, the gravity of the situation soon came to the fore as the company was forced to disconnect over two million telephone lines owing to non-payment. Telecommunications prices remained high due to the need for cross-subsidization (i.e., the wealthy QHHGLQJWR¿QDQFHQHWZRUNUROORXWWRWKHSRRU DV well as a lack of competition in the sector. Valueadded telecommunications services were also rather limited in the late nineties and early part of the current decade. Due to the fact that Telkom had LQYHVWHGVLJQL¿FDQWFDSLWDOLQLWVintegrated services digital network (ISDN) platform, it was reluctant to cannibalize on this and install ADSL lines. The incumbent was also very wary of introducing a new service to market which would have the effect of negatively impacting on its metered-driven call revenues. Nonetheless, fearing competition shortly after the expiration of its monopoly, the company
announced the provisioning of broadband services in 2002. ,QWKLVUHVSHFW7HONRPODXQFKHGLWV¿UVW¿[HG line consumer broadband offering in August 2002. This was an always-on connection to the Internet that operated at a downlink speed of 512 kilobits per second and an uplink speed of 256 kilobits SHUVHFRQG7KLVPHDQWWKDWDRQHPHJDE\WH¿OH could be downloaded in less than 20 seconds and DQLGHQWLFDO¿OHXSORDGHGLQVHFRQGV Previously, consumers were restricted to dialup connections operating at 56 kilobits per second (analogue), 64 kilobits per second (singe-band ISDN), or 128 kilobits per second (dual-band ISDN). Apart from being inherently slower, this service was billed on a per-second basis, with the result that users accessing the Internet through dial-up were typically watching the clock to ensure that their online sessions were kept as brief as possible. Their exposure to the Internet was thus minimized as lengthy online sessions equated to a hefty telephone account at the end of the month. The table below lists the cost of spending 10, 20, and 40 hours online per month during each year between the period 1993 and 2003. The Telkom costs are allocated for 10, 20, or 40 hours of local call charges during peak call times (i.e., business hours). The Telkom price constitutes the line rental cost plus the timed call charges. Figure 1 reveals the escalating cost of spending time online. It is particularly worthwhile
Table 1. Internet access costs over the decade spanning 1993 and 2003 (Internet.org.za, 2003)
14
South Africa
noting the sharp increase since the year 2001. However, while ADSL promised users a greater degree of freedom, the service was launched with DWKUHHJLJDE\WHPRQWKO\GDWDµVRIWFDS¶PHDQLQJ that consumers could only use the service at its maximum capability until this amount of data had EHHQ VHQW DQGRU UHFHLYHG $IWHU WKLV µFDS¶ ZDV reached, the service would be severely restricted in terms of international transmission speeds. This resulted in the accessing of international Web sites being throttled, while local Web sites continued to remain fast to access. Telkom has subsequently reversed this policy and applied a µKDUGFDS¶PHDQLQJWKDWDOO,QWHUQHWFRQQHFWLYLW\ is terminated once the cap has been reached. Local bandwidth is no longer unlimited. That pointed out, this restriction was seen as being largely unproblematic for Internet users merely wishing to browse the World Wide Web and check e-mail. However, for those power userswhose online activities included a greater variety of tasksthis was less than desirable. Nonetheless, despite the limitations, the deployment of broadband heralded an era where it was thought that consumers need not be as time orientated in their online behavior as was previously the case, and hence the South
African Internet community deemed it a step in the right direction. Table 2 illustrates the migration from dial-up to broadband. It is evident that dial-up subscribers have increased year-on-year until 2004, but have started to decline since this time. Early in 2004, Sentech (the government-owned national television signal carrier) launched its MyWireless service. This allowed Internet users the ability to experience broadband connectivity through a wireless medium. The service used advanced radio telephony (in particular, a form of 3G communications technology) to send/receive data between the PC and Internet. However, this was not a true mobile solution in that the computer needed WREHLQD¿[HGORFDWLRQ LHUHPDLQVWDWLRQDU\ while an Internet session was in progress. FurtherPRUH6HQWHFK¶VOLPLWHGQHWZRUNSUHVHQFHPHDQW that the service was only available in a handful of metropolitan areas. The service was launched without a data cap, instead relying on fair use by its customers. This attracted many would-be ADSL customers who were not enthusiastic about using a service which essentially limited them to only a few hours of usage per day. Unfortunately, the MyWireless service attracted a considerable amount of negative press due to teething network problems
Figure 1. A graphical representation of Internet access costs between 1993 and 2003 (Internet.org.za, 2003)
15
South Africa
Table 2. Growth and decline of dial-up subscribers (Goldstuck, 2005) Year
Subscribers (growth/decline)
1994
15,000
1995
33,600 (155%)
1996
79,700 (137%)
1997
196,620 (146%)
1998
366,235 (86%)
1999
560,000 (53%)
2000
782,000 (40%)
2001
960,000 (22%)
2002
1,008,000 (5%)
2003
1,048,320 (4%)
2004
1,088,000 (4%)
2005
1,080,000 (-1%)
2006
1,050,000 (-3%)
and their lack of support in dealing with customHUV¶IUXVWUDWLRQ7KLVUHVXOWHGLQDVPDOOJURXSRI disgruntled customers becoming incredibly vocal about their experiences (even going so far as to establish a Web site, www.SentechSucks.co.za) and creating a negative sentiment in the market which has hindered the development of MyWireless and opened the door for a competitor to gain market share. The company was also forced to reconsider its “all you can eat” philosophy in terms of usage and has subsequently imposed absolute restrictions on uploads and downloads. Wireless Business Solutions (WBS), a private operator tasked with the responsibility of running the network of National Lottery Point of Sale terminals, commercially launched its iBurst service in early 2005. This was very similar in nature to 6HQWHFK¶VRIIHULQJDOWKRXJKLWIHDWXUHGDKLJKHU theoretical download transmission rate of 1 Mbps. It did, however, include a monthly three gigabyte data cap, similar to that of Telkom. The iBurst service capitalized on the mistakes committed by 6HQWHFK¶VMyWireless service and, combined with improved funding for network expansion, now
16
Figure 2. A MyWireless modem
boasts a subscriber base which is rumored to be four times the size of its direct competitor. While all of the above broadband services UHTXLUHXVHUVWRUHPDLQLQD¿[HGORFDWLRQZKLOH XVLQJWKH,QWHUQHWWKHFHOOXODUQHWZRUNV¶RIIHULQJV (i.e., those of MTN and Vodacom) were designed WREHPRUHÀH[LEOH7KLVDOORZHGDXVHUWREHLQ motion, for example traveling on a train or bus, while online. 3ULRUWR&KULVWPDV9RGDFRPIXO¿OOHGLWV promise of launching 3G services (operating at a speed of 384 kilobits per second) in the local market. MTN followed suit in mid-2005. Due to the celOXODUQHWZRUNV¶H[WHQVLYHFRYHUDJHWKURXJKRXWWKH country, the area in which wireless broadband could EHREWDLQHGZDVVLJQL¿FDQWO\H[SDQGHG+RZHYHU the high-speed service was initially launched on a limited scalewithin metropolitan areasand is still largely restricted in this respect. Furthermore, data charges (i.e., cost per megabyte transferred) were considerably higher than those offered by ¿[HGORFDWLRQEURDGEDQGSURYLGHUV7KLVUHVXOWHG in potential 3G users being offered a one gigabyte package for a similar price to that being charged for three gigabyte (or larger) accounts on the other
South Africa
networks. This immediately signaled that users would be made to pay for true mobility and extended network coverage. In 2006, MTN and Vodacom optimized their 3G services with the introduction of High-Speed Downlink Packet Access (HSDPA). This increased download speeds to a maximum theoretical limit of 1.8 Mbps, providing the fastest wireless form of broadband. This development was seen as a major step forward for customers who required multimedia services (particularly Web 2.0 applications such as YouTube, MySpace, etc.) that are bandwidth intensive by nature. Figure 3 illustrates the split of broadband users EHWZHHQ¿[HGOLQH 7HONRP ¿[HGZLUHOHVV 6HQ-
tech and iBurst), and mobile (MTN and Vodacom) providers. Table 3 provides a year-on-year tally of the number of broadband users in South Africa. While the market is experiencing a promising growth rate in percentage terms at the time of writing, it should be acknowledged that the growth is based on a very small pool of users. The penetration rate of broadband in the local market remains under one percentage point of the population of South Africa (46 million). In contrast, Denmark, The Netherlands, Iceland, Korea, Switzerland, and Finlandthe world leaders in terms of broadband penetrationare reported to have a penetration rate
Table 3. Growth of broadband subscribers (Smit, 2005) Year
Subscribers (growth/decline)
2003
15,667
2004
51,124 (226%)
2005
147,532 (189%)
2006*
277,000 (88%)
2007*
405,000 (46%)
2008*
533,000 (32%)
* = projected totals
Figure 3. Broadband market share by access technology
25% Fixed-line Fixed-w ireless 15%
60%
Mobile
17
South Africa
of over a quarter of the population. Even the average penetration rate of the 30 countries surveyed by the Organization of Economic Cooperation and Development15.5%illustrates the gulf in existence (OECD, 2006).
TOWARDS A PROFILE OF THE BROADBAND USER BASE ([SORULQJWKHSUR¿OHRIWKHW\SLFDOEURDGEDQGXVHU LVGLI¿FXOWIRUDQXPEHURIUHDVRQV Firstly, a single broadband connection may be utilized by more than one person. Unlike a cellular phone which is very personal in nature, ,QWHUQHWFRQQHFWLRQV LQFOXGLQJWKRVHFODVVL¿HGDV broadband) may easily be shared by a multitude of individuals. This is often practiced for cost purposes where it makes little sense for a family, for example, to pay multiple times over for essentially the same service. In this respect, it may be useful WRGLVWLQJXLVKEHWZHHQWKHEURDGEDQGµFXVWRPHU¶ RUµVXEVFULEHU¶ LHWKHSXUFKDVHURUDFFRXQWSD\HU DQGWKHEURDGEDQGµXVHU¶ LHDQLQGLYLGXDOZKR makes use of the connection without directly paying for this privilege). 6HFRQGO\ LW LV UDWKHU GLI¿FXOW WR GLVWLQJXLVK between consumer broadband subscribers and business broadband subscribers. Broadband services may be marketed at either the business or home market, but there is no guarantee that individuals will register in this manner. In fact, not all broadband services are stipulated for one form of use or the otherthis is often left to the discretion of the user. To this end, many broadband users may simply opt for the cheapest offering, even if this is not aligned to the manner in which the broadband provider anticipates that the service will be utilized. Furthermore, in a 6PDOO 2I¿FH +RPH 2I¿FH 62+2 HQYLURQPHQWDVLQJOHEURDGEDQG account may be used for business purposes during RI¿FH KRXUV DQG KRXVHKROG SXUSRVHV WKHUHDIWHU In this respect, it may be useful to distinguish between consumer broadband services and business broadband services, although it is acknowledged that this is not always possible.
18
Thirdly, with the rapid deployment of broadband technology around the world, the face of typical broadband is likely to be constantly changing. As the service increases in popularity, it seems inevitable that the constitution of the broadband user base will tend more towards the mainstream and further away from early adopters. In South Africa, industry experts appear to DJUHH WKDW HDUO\ DGRSWHUV DUH µWHFKLHV¶ LH LQdividuals from a technical background who are either professionally employed within the ICT industry or who are computer enthusiasts with high levels of computer literacy). Furthermore these individuals are thought to be mostly male, based in metropolitan areas, well educated (i.e., possess VRPH IRUP RI WHUWLDU\ HGXFDWLRQ DIÀXHQW LH earning professional salaries, being able to afford a cell phone, car, house, etc.), and relatively young (i.e., tending towards 30 years of age). However, it seems that the broadband user base may be getting younger as entire families are starting to make use of the household broadband Internet connection. Furthermore, as the technology becomes more user friendly and access to LQIRUPDWLRQLQUHDOWLPHEHFRPHVµPLVVLRQFULWLFDO¶LWVHHPVOLNHO\WKDWFRUSRUDWHXVHUV DOWKRXJK not necessarily technologically minded) may adopt these services for access to their corporate networks while traveling. This facility would allow business users to check their e-mail, view FRQ¿GHQWLDO FRUSRUDWH UHSRUWV DQG VR IRUWK DW any time, from any location in which a broadband service is offered. It is thought that these developments will serve WRVNHZWKHSUR¿OHRIWKHFXUUHQWEDVHRIEURDGband users. Two local studies, conducted by Webchek and the Online Publishers Association, shed further light on this issue. The Project SA Web User study, conducted in 2005 by Webchek, is a follow-up study on the SUR¿OHRIWKHDYHUDJH6RXWK$IULFDQ,QWHUQHWXVHU Initially conducted in 1999/2000, the latest study points to the movement of the average user from “techno whizzkid” to “the man in the street.” The survey was conducted by means of a computer-aided telephonic interviewing process.
South Africa
Table 4. Webchek’s overview of the South African Internet user base Age
Almost half of South African Web users are between 25 and 44 years of age. Yet, an increase in users over the age of 55 was noted since the 1999/2000 survey.
Language
The majority of South African Web users are English speaking.
Gender
The ratio of females vs. males is approximately 60/40.
Monthly Household Income
Stage in Lifecycle
Level of Education
The average monthly household income earned is just over R13,000 (approximately US$2,000), which marks a decrease since the previous study. The number of respondents with children in the home has increased from 41% in 2001 to 53% in 2006. On average, respondents had two children. ,QOLQHZLWKSUHYLRXV¿QGLQJVWZR¿IWKVRI6RXWK$IULFDQ:HEXVHUV\HDUVDQGROGHUKDYHHLWKHU FRPSOHWHGXQLYHUVLW\RUKDYHVRPHRWKHUIRUPRIWHUWLDU\TXDOL¿FDWLRQ The ratio of respondents working in the IT industry has decreased from 1999/2000. It is also interesting to
Employment Category
note that 27% of Web users have their own business. Of these respondents, 40% had owner-run enterprises. Seven out of ten businesses appear to be run from home.
7HFKQRJUDSKLF3UR¿OH
Only 20% of respondents see themselves as technologically advanced. )RUW\¿YHSHUFHQWRIWKHVDPSOHKDYH:HEDFFHVVDWKRPHRQO\DQGDWZRUNRQO\ZLWKKDYLQJ access at both home and work. The number of days the Web is accessed per month has shown a slight upward curve from 2002 among home users. The Web was accessed 16.8 days a month on average at home
Usage and Behavioral
in April 2006 (compared to 14.4 days on average in 2002). Thirty-seven percent of home users access the
Patterns
Web everyday at home. At work, the Web is accessed 23 days a month on average (basically every working day). Home users spend an average of 5.8 hours online per week and work users 7.4 hours per week. Most LQWHUHVWLQJO\6RXWK$IULFDQ:HEXVHUVVWLOOWHQGWRXVHWKH,QWHUQHWPRVWO\IRUVHDUFKLQJIRUVSHFL¿F information and sending and receiving e-mail.
A random sample of individuals was contacted from the telephone directory. Potential respondents had to be at least 18 years of age and had to access the World Wide Web (that is, not e-mail only) at least once a month, either at home and/or at work in order to be included in the sample. No quota controls were exercised on gender and on race, and the survey was limited to the major metropolitan areas of South Africa. The Project SA Web User ¿QGLQJVDUHSUHVHQWHGLQ7DEOH A second study, commissioned by the Online Publishers Association and executed by Nielsen/ Netratings, also found that online media casts a wider net than previously thought. Demographic details are based on online surveys completed by over 37,000 individuals in October 2006. Web usage statistics were measured between July and September 2006. The Nielsen/Netratings¿QGLQJV are presented in Table 5.
7KXVERWKVWXGLHVFRQ¿UPWKHVXVSLFLRQWKDW,Qternet usage is entering an era where it is becoming a personal productivity and entertainment tool for the population at large. Where access is available, the service appears to be utilized by individuals across many walks of life. It would appear that broadband technology is spearheading growth of ,QWHUQHWVHUYLFHVLQWKHKRPHHQYLURQPHQWVSHFL¿cally, a phenomenon that will be discussed in the following section.
DRIVERS AND INHIBITORS OF BROADBAND DIFFUSION AND ADOPTION *ROGVWXFN LGHQWL¿HVWKUHHGULYHUVRIEURDGband adoption in South Africa.
19
South Africa
Table 5. The Online Publishers Association overview of the South African Internet user base Eighty-one percent of all respondents were in full-time employment. Sixty-two percent of visitors to Access in the Workplace
the Internet did so from their workplace compared to only 27% from home, 6% from an educational establishment, and 3% from an Internet café. Forty-one percent of respondents, however, did not have access to the Internet from home. Online media consumption is strongest in urban areas: 27% of access takes place from Johannesburg
Geographical Footprint
(Gauteng), 12% from Pretoria (Gauteng), 16% from Cape Town (Western Cape), and 6% from Durban (KwaZulu-Natal), with other larger cities also showing higher usage than other areas. Fifty-nine percent of Web users are younger than 35: 18-24 year olds constitute the largest age group (32%), followed by 25-29 year olds (17%) and 30-34 year olds (15%). Education levels are high,
Education and Income
with 79% having some further education and 34% having a degree or post-graduate degree. Income levels are also high, with 39% of Web users having an annual household income of over R150,000 (approximately US$20,000) and of these, 18% having an annual household income over R400,000 (approximately US$55,000). (QJOLVKZDVWKHPRVWFRPPRQO\UHSRUWHG¿UVWODQJXDJHRI,QWHUQHWXVHUV RIUHVSRQGHQWV
Language Spoken
IROORZHGE\$IULNDDQV 7KHWUDGLWLRQDO$IULFDQODQJXDJHVRI,VL=XOX DQG,VL;KRVD were the next most reported mother tongues. However, they are both slowly on the increaseup 2% and 1%, respectivelysince the October 2004 survey.
•
•
20
Lifestyle-Oriented Demand: The new JHQHUDWLRQRIRQOLQHJDPHUVDQG¿OHVZDSping enthusiasts demand speed and quality of access. For the gamers, speed and quality are paramounteven more so than for the EXVLQHVVPDUNHW)RUWKH¿OHVKDULQJPDUNHW which uses applications that accelerate data usage through multiple upload requests from DW\SLFDOXVHU¶VPDFKLQHDGDWDFDSRIDQ\ size is highly resented. SME Demand: ADSL has found a ready and enthusiastic market among small businesses. Such users are not as price-sensitive as home users, nor are they as focused on capacity issues like contention ratios and data limits. Here, the solution represents DVLJQL¿FDQWFRVWVDYLQJE\SUHYHQWLQJWKH business having to dial-up on a regular basis to send/retrieve e-mail and has also had the positive spin-off of allowing the enterprise to EHFRPHPRUHHI¿FLHQWLQWHUPVRILWVRQOLQH communications. Thus, the SOHO market, micro-enterprises, and small business alike KDYHUHDOL]HGLQVWDQWFRVWEHQH¿WVDVDUHVXOW and their experience has become an excellent word-of-mouth marketing exercise for the LQFXPEHQW¿[HGOLQHRSHUDWRU
•
Pent-Up Consumer Demand: Home users who have been on the Internet since the midor late-1990s have become frustrated with the limitations and cost of dial-up, and have been ready to spend a higher upfront monthly amount on a better form of connectivity. Many such individuals have been exposed to the merits of broadband on their travels RYHUVHDVRUKDYHKHDUGDERXWWKHEHQH¿WVRI the medium through communication with their friends, colleagues, and families settled overseas. While it is true that consumers have resisted the price levels set by Telkom, the need for speed is ever present. The World Wide Web of the current decade is most different from that of the nineties, whereby Web designers usually took account of slow (typically dial-up) connections by building sites that could download quickly. However, in a broadband-dominated developed world, where most Web sites reside, this priority is fading into the background. Bandwidthintensive sites therefore demand high-speed connectivity, and price is often a secondary factor here. Similarly, large e-mail attachments are becoming commonplace, rather
South Africa
than the exception as they were in the 1990s, and dial-up connections are ill equipped to KDQGOHWKHÀRZRIWKLVGDWD7KHUHIRUHOHQJWK\ downloads result in higher cost to dial-up users, and timeouts mean wasted time and energy…over and above simple frustration. The 3CC framework is an alternative view of the forces at play. It considers both the drivers and inhibitors of growth in the broadband market. This is represented in Table 6.
IS BROADBAND MERELY ENHANCED DIAL-UP CONNECTIVITY? 7KH GH¿QLWLRQ RI EURDGEDQG LQ 6RXWK $IULFD LV somewhat unclear as there appear to be disparities between the characteristics of local broadband services and foreign broadband services, reinforcing the notion that South Africans are obtaining a raw deal in contrast to their international counterparts. In this respect, many broadband providers in South Africa restrict the usage of their service in a number of noteworthy respects. These are discussed below. Firstly, while most standard broadband services in developed countries are not capped, ADSL accounts in South Africa typically feature a two-, three-, or four-gigabyte data cap. The 3G networks, strictly speaking, do not cap individuals, but charge users in a manner which results in the service being prohibitively expensive for anything other than basic online tasks. This means that using a broadband connection in South Africa is not feasible if individuals wish to download music albums, full-length movies, or even stream live television. In this respect, either the data cap will rapidly be exhausted, or if the user is subscribed on a pay-per-usage basis, the end-of-month account is likely to be exorbitant. Telkom, the largest provider of broadband connectivity in South Africa, considers downloading large amounts of data (usually required for obtaining multimedia content) through an $'6/ OLQH DV µDEXVH¶ RI WKH VHUYLFH 7KLV YLHZ
is strongly counteracted by the local broadband community who argue that the facility is supposed to be used for more than just e-mail and Web access purposes. However, at the present time, it appears that broadband users, based on the current service offerings, are only able to use WKHLUEURDGEDQGFRQQHFWLRQVDVDJORUL¿HGGLDOXS connection. This has caused much resentment and frustration. In fact, the Internet community felt so strongly about the issue that they have reported Telkom to the telecommunications regulator, the Independent Communications Authority of South Africa, as well as to the Competition Commission on the basis of anti-competitive behavior. Secondly, many broadband service providers XWLOL]HµSURWRFROVKDSLQJ¶7KLVWHFKQRORJ\JLYHV network preference to certain applications over other applications. In order to understand protocol shaping, it should be understood that most consumer broadEDQGDFFRXQWVDUHµFRQWHQGHG¶7KLVPHDQVWKDWD 512 kbps connection, for example, is not guaranteed the full 512 kbps data stream at all times. Instead, a number of users essentially share the same Internet pipeline, meaning that if all users are concurrently online, each user is likely to achieve a mere fraction of the theoretical performance. Protocol shaping takes this a step further. The principle here is that certain applications should be given more than their fair share of bandwidth, while others should be given less than their fair share. In this manner, certain applications can be given low priority and thus their use discouraged. File swappers are often cited as an example of parties ZKRDUHµKHDY\XVHUV¶DQGDJDLQVWZKRPQHWZRUN protection should be implemented. For example, if the network detects that the XVHU LV XSORDGLQJGRZQORDGLQJ ¿OHV WKURXJK D peer-to-peer application (e.g., BitTorrent), his or her performance is likely to be intentionally degraded and thus bandwidth removed from the FRQQHFWLRQDQGGRQDWHGWRRWKHUµQHWZRUNIULHQGO\¶ applications such as Web browsing. To this end, it is argued that potential abusers of the network are throttled so that performance is not degraded for the entire broadband user base. Unsurprisingly, the individuals concerned complain that they are being unfairly penalized. 21
South Africa
Table 6. The 3CC framework Factors
Costs
Drivers of Growth
Inhibitors of Growth
Since introducing its ADSL service, Telkom has
A lack of competition, underscored by a prolonged
reduced subscription rates twiceonce in 2005 and
telecoms monopoly, has failed to drive down costs to the
again in 2006. The incumbent has also upgraded
extent witnessed in foreign markets (Mzolo, 2005). As
the speeds of its 192 kbps lines to 384 kbps and its 1
calculated by Christian (2005), in certain circumstances,
Mbps lines to 4 Mbps free of charge, meaning that
7HONRP¶V$'6/VHUYLFHLVRYHUPRUHH[SHQVLYH
consumers are achieving better value. Prices are
than other countries. In fact, Winston Smith, MyWireless
expected to fall as the market grows (thereby bringing
portfolio manager at Sentech, even determined that
DERXWHFRQRPLHVRIVFDOH DQGFRPSHWLWLRQLQWHQVL¿HV
the cost of a dedicated symmetrical 8 Mbps Internet
It has been calculated that the original price of a 512
connection in South Africa is approximately 1,000 times
kbps line has almost halvedfrom R680 to R362in
more expensive than a comparative service in Sweden
nominal terms since inception. Mobile data rates were
(MyBroadband, 2006a).
slashed from R25.00 per megabyte in 2004 to R 2.00 per megabyte in 2005a reduction of over 90% (du Plessis, 2005). These mobile data rates now compare YHU\IDYRUDEO\ZLWK6RXWK$IULFD¶VLQWHUQDWLRQDOSHHUV Telkom has realized the potential of data revenue and
Fixed-line voice and data services are only available to
is rapidly equipping its exchanges to accommodate
a tenth of the population, whereas 3G services (mobile
ADSL lines. The phenomenal growth of subscribers
broadband) coverage is very limited in terms of its
between 2004 and 2006 bears testimony to this policy.
geographic footprint. Broadband coverage outside
The two leading cellular networks are competing,
metropolitan areas in South Africa is virtually non-
head-to-head, to acquire the more lucrative business
existent.
customers who require advanced services such as mobile broadband access. For this reason, Vodacom has already expanded its 3G service (enhanced by Coverage
HSDPA) into CBD and residential areas, whereas MTN is currently aggressively rolling out 3G services into the suburbs. The third cellular operator, Cell C, is currently investing in EDGE technology. Additionally, iBurst,D¿[HGZLUHOHVVEURDGEDQG provider, is engaging in extensive network expansion which presently includes 15 cities and large towns (MyBroadband, 2006b).
Competence
As highlighted early in this chapter, Internet services
Broadband appears to be an unknown quantity for
are starting to reach the “man in the street.” In part,
many South Africans, the vast majority of which
this is due to computer education in the workplace,
are not computer literate and are unlikely to have
as well as in the primary, secondary, and/or tertiary
ever accessed the Internet before (Monteiro, 2006).
education environment. A number of companies and
2QO\¿YHPLOOLRQ3HUVRQDO&RPSXWHUVDUHSUHVHQWLQ
QRQSUR¿WRUJDQL]DWLRQVLQFOXGLQJWKH,QWHUQHW6HUYLFH
the country (Laschinger & Goldstuck, 2006); and
Providers Association and Telkom Foundation, have
four million individuals have access to the Internet
been proactive in donating hardware and software
through their school, college, home, or place of work
to educational institutions and providing training to
(Goldstuck, 2005). This equates to a penetration rate of
educators so that they can impart these skills to their
approximately 10% for each of the above, meaning that
students.
the vast majority of South Africans have no exposure to computers, let alone the Internet.
22
South Africa
Herein lies a fundamental problem: broadband promises a multimedia-rich environment whereby XVHUVFDQH[SHULHQFHWKHWUXHEHQH¿WVRIWKHLQformation age. However, by labeling multimediaVHHNLQJXVHUVDVµKHDY\XVHUV¶RUHYHQµDEXVHUV¶ of the network, the potential for broadband to revolutionize the way in which the Internet is used LVVHYHUHO\VWLÀHG)XUWKHUXQKDSSLQHVVLVFDXVHGE\ the fact that this is a relatively unique stance foisted on consumers in the South African market.
RATIONALE PROVIDED FOR HIGHER COST STRUCTURE IN THE COUNTRY There has been much published and ushered in private conversations about the high costs of Internet, particularly broadband, access in South Africa. Telkom, feeling the need to defend its pricing structure, has contributed a number of key reasons to explain its seemingly high tariffs. Yet, WKHWHOHFRPPXQLFDWLRQVLQFXPEHQWLVGH¿DQWWKDW South African consumers should not expect the same services, at the same price, that are available to their overseas counterparts. This is discussed below. Firstly, Telkom claims that the South African broadband subscriber base is still relatively small LQVWDWXUH7KXVLWKDVEHHQXQDEOHWREHQH¿WIURP the economies-of-scale as experienced elsewhere in the world. It has argued that as the number of broadband subscribers continues to grow, subscription rates should naturally decrease. There appears to be some merit in this argument as there have already been two price decreases brought about VLQFH$'6/ZDV¿UVWODXQFKHG Secondly, Telkom claims that international bandwidth is exceptionally expensive in South Africa due to the geographical distance from the main Internet centers (i.e., United States of America, Western Europe, South East Asia, etc.). While this LVWUXH6$7WKHSULPDU\¿EHURSWLFVXEPDULQH cable linking South Africa with the rest of the world, is exclusively controlled by Telkom. It is argued that 7HONRP¶VPRQRSRO\RYHUWKLVFULWLFDOUHVRXUFHKDV driven access charges far above those that would
have been tolerated if competition over submarine international bandwidth had been in place. In fact, it is claimed that a very small percentage of the SAT-3 cable capacity is currently being utilized by the incumbent. Due to the fact that the cable is only intended to last for use over the next two decades, critics argued that this essential national UHVRXUFHLVVLPSO\µURWWLQJDZD\¶ Thirdly, Telkom claims that the cost to establish its ADSL network has been extremely capital intensive. The incumbent argues that it has had to convert previously analogue exchanges into digital IRUPDWDQGWKHQHTXLSWKHVHWRKDQGOH¿[HGOLQH broadband. It therefore maintains that it needs to recoup these costs and that further investment in LQIUDVWUXFWXUHZLWKRXWWKHUHTXLUHG¿QDQFLDOUHturn, does not make business sense. In this respect, the company has even threatened to pull the plug on its broadband network if it were to be forced to slash the price tag of its broadband offerings. )RXUWKO\ 6RXWK $IULFD¶V WHOHGHQVLW\ LV ORZHU than that of many broadband leaders. In the centre of London, for example, residents live close together. Less cable is therefore required to reach the residents. In South Africa, due to high levels of dispersion of the population, more cable is required to build the comprehensive network. Another key point in this regard is that there exists a negative correlation between the length of the local loop and the theoretical limit of the ADSL data throughput rate. Currently, the local loop is EHWZHHQIRXUWR¿YHNLORPHWHUVLQOHQJWK7HONRP feels this should be closer to two kilometers. In order for ADSL2+ speeds to be possible, it will be necessary to redevelop parts of the network in order to implement shorter local loops. Again, the infrastructure costs may not prove economically viable unless favorable returns can be generated on the service.
SOLUTIONS FOR THE ROAD AHEAD A prominent author of Internet issues in South Africa, Arthur Goldstuck, has developed a hierarchy
23
South Africa
of Internet needs. The model is closely based on 0DVORZ¶VKLHUDUFK\JURXQGHGLQKXPDQEHKDYLRU theory. The model encompasses levels for access, quality, utility, community, and strategy (in ascending order). This is depicted in Figure 4. Unfortunately, it appears that while their international counterparts have been able to ascend to the top of the pyramid (i.e., the community or strategy level), problems at the quality level have
prevented South African broadband users from REWDLQLQJPD[LPXPEHQH¿WIURPWKH,QWHUQHW In terms of remedies, the ADSL Hearings, conducted throughout 2005 and 2006, challenged the stranglehold on the broadband market by the dominant players. A complaint was laid with the Independent Communications Authority of South Africa (ICASA) by MyADSL.co.za founder, 5XGROSK 0XOOHU LQWR WKH H[FHVV SUR¿WVHHNLQJ
Figure 4. Goldstuck’s hierarchy of Internet needs (Goldstuck, 2005)
ValueAdded Services
Infrastructure
Access
This infrastructure level equates to connectivity and merely entails establishing a point-to-point connection into the Internet.
Quality
This infrastructure level equates to the nature of the connectionfor example, narrowband (dial-up), broadband (ADSL/3G), and so forth. In essence, the more mature and affordable the technology (that is, increased throughput rates and greater reliability at lower costs), the better.
Utility
This value-added services layer equates to the spread of applications that the consumer is able to use on the Internet. In essence, the more online activities the user can pursue (e.g., downloading music, reading news, etc.), the better.
Community
This value-added services layer equates to a social dimension and entails the consumer reaching out and experiencing the world through human interaction. Examples include using instant messaging and joining online discussion forums.
Strategy
7KLVYDOXHDGGHGVHUYLFHVOD\HUHTXDWHVWRWKHXVHU¶VVHFRQGOLIHZKHUHE\KHRUVKHIHHOVWKDWWKHRQOLQHZRUOGLV an extension of his or her life in the physical world. Hence, almost everything achieved off-line may actually be completed online as well. This may entail operating a business online, pursuing online relationships, shopping online, and the like. Therefore, this layer encompasses all of the above by taking the consumer into a seamless virtual realm.
24
South Africa
behavior of the incumbent. A number of remedies were proposed at the hearings. )LUVWO\7HONRP¶VFXUUHQWSULFLQJVWUXFWXUHDOlows for them to charge a line rental fee, as well as an ADSL access component fee. Muller argued that this amounted to double charging for essentially the same piece of copper line and that the internal costs to the incumbent did not merit an ADSL access fee of between US$50 to US$100. Instead, he argued that a once-off fee to cover the installation of the port at the local exchange should be levied. This would have the effect of vastly reducing broadband subscription fees and hence making the service more affordable to the µPDQLQWKHVWUHHW¶ 6HFRQGO\VXEPDULQH¿EHUFRQQHFWLYLW\VKRXOG be expanded or access charges to the existing pipeline reduced. A study conducted by BalancingactAfrica (Southwood, 2006) revealed that some 29 of 55 African countries obtain 80% of their total international Internet bandwidth by satellite (an inherently more expensive medium than submarine ¿EHURSWLFFDEOH 0XFKRIWKHUHPDLQLQJEDQGZLGWK LV SURYLGHG WKURXJK WKH 6$7 VXEPDULQH ¿EHU cable, of which Telkom is a principal stakeholder. As Telkom has exclusive rights to sell bandwidth on this cable within South Africa, prices have UHPDLQHG DUWL¿FLDOO\ KLJK 7KUHH VROXWLRQV KDYH EHHQSURSRVHG7KH¿UVWLVWRQDWLRQDOL]HWKHDVVHW However, this is fraught with danger as it sends the wrong signal to the international community about the respect of property rights within the country. The second is to force Telkom, through legislation, to offer wholesale bandwidth at cost or near-cost prices. The third is to build a rival cable, a decision currently being debated within the Ministry of Public Enterprises (McLeod, 2007). There are in fact a number of alternative cable systems in the pipeline to connect South Africa with the Internet backbone. EASSY (the Eastern Africa Submarine SYstem) is the foremost cable development. This was initially proposed in 2003 DV D 1HZ 3DUWQHUVKLS IRU $IULFD¶V 'HYHORSPHQW (NEPAD) initiative. The cable system is designed to feature landing points in a multitude of countries along the East coast of Africa including Kenya, Tanzania, Mozambique, and South Africa. Those
land-locked countries wanting access to EASSY (e.g., Botswana, Lesotho, Zambia, etc.) will be accommodated through a terrestrial network designed to bring them into the fold and alleviate their reliance on satellite bandwidth. The cable is planned to extend from Mtunzini, located just north of Durban, South Africa, to Port Sudan, Sudan, a distance of about 9,900 kilometers. The cable is expected to be launched into operation by the end of 2008. Thirdly, local loop unbundling (LLU) was proposed as a solution to give other players in the market direct access to households and businesses. During its period of exclusivity, Telkom was given the sole right to provide telecommunications infrastructure connecting commercial and private establishments in the country. This meant that no two parties could establish a connection between themselves, even if a company wished to connect WZR UHJLRQDO RI¿FHV WRJHWKHU 7KLV DOVR KDG WKH unfortunate effect of cellular operators, Internet service providers, and the like needing to use the incumbent to connect their inter-company nodes if not located within the same premises. This dramatically increased telecommunications costs
Figure 5. The EASSY cable system (Kenduiywo, 2005)
25
South Africa
and kept Internet penetration to modest levels (as highlighted earlier in the chapter). Local loop unbundling would allow the second network operator, launched in late 2005, to provide direct telecommunications services to consumers DQG EXVLQHVVHV ZLWKRXW µSLJJ\ EDFNLQJ¶ RQ WKH Telkom network. The current legislation allows for additional players in the market to apply for licenses to pursue this approach as well.
THE BOTTOM LINE Political will is required to liberate the market (i.e., introduce greater competition) or to drive down prices through legislation. Unfortunately, the Department of Communicationsresponsible for overseeing the sectorand the Independent Communications Authority of South Africathe body tasked with regulating the sectorhave a checkered past in this respect. The ADSL Resolutions, the outcome of the hearings described above, were released in mid-2006, but to date have been side-stepped by the various operators in the market. ICASA is hemorrhaging from a shortage of staff, and Minister of Communications Ivy MatsepeCassabury has by her own admission been slow to bring about positive reform in the sector. The frustrations of local users are evident in their thousands and are vented on a daily basis on activist sites such as www.MyADSL.co.za and www. Hellkom.co.za (a pun intended on combining the ZRUGVµKHOODQGµ7HONRP¶
26
ers in this respect, it is thought that the heightened state of competition will drive prices downwards, in addition to expanding broadband coverage into both urban and semi-urban areas. As the power engine of the sub-Saharan economy in Africa, broadband may prove a key ingredient to economic growth and prosperity within the country. Yet, there are other national prioritiesQDPHO\¿JKWLQJ+,9$,'6FRQVWUXFWing housing, implementing social reforms, and the liketo contend with. Broadband is still perceived to be an elitist service and therefore not within the domain of the average South African. Until pricing, coverage, and computer literacy concerns are addressed, it is unlikely that the broadband revolutionwhich has swept many a society into the information agewill have the same affect on South African shores. The long walk to broadband freedom continues.
ACKNOWLEDGMENT The 3CC framework was developed in consultation with Rudolph Muller, founder of MyADSL and Information Systems lecturer at the University of Johannesburg. Also, many of the views detailed in this chapter echo the sentiments and comments expressed on the MyADSL/MyBroadband discusVLRQIRUXPV7KDQNVWRWKHµIRUXPLWHV¶IRUWKHLU indirect contributions.
REFERENCES BizCommunity.com. (2006, November 17). Online SA consumers fuelling ad revenue growth. Retrieved from http://www.bizcommunity.com/ Article/196/23/12516.html Boer, S. (2006, July 27). Latest SA Web user survey results released. Retrieved from http://www.bizcommunity.com/Article.aspx?c=11&l=196&ai=11040 Bridges.org. (2001, May 2). South Africa telecommunications overview, commentary, and statistics (policy brief). Retrieved from http://www.bridges. org/publications/123
South Africa
Christian, C. (2005, November 11). Telkom’s ADSL QRZRI¿FLDOO\RYHUPRUHH[SHQVLYHWKDQ other countries. Retrieved from http://www.mybroadband.co.za/nephp/?m=show&id=1167 du Plessis, M. (2005, September 15). Tit for tat. Retrieved from http://www.iweek.co.za/ViewStory.asp?StoryID=155165 Franz, L. (2005, November 1). Pushing the city limits. Retrieved from http://www.iweek.co.za/ ViewStory.asp?StoryID=157791 Goldstuck, A. (2005). Internet access in South Africa 2005. Johannesburg: World Wide Worx. Internet.org.za. (2003). Cost of Internet access in South Africa (1993-2003). Retrieved October 2, 2006, from http://www.internet.org.za/costs. html Kenduiywo, P. (2005, January 30). Eastern Africa submarine cable system. Proceedings of the Workshop on Information Society and Regulation: Access and Infrastructure, Ghana. Laschinger, K., & Goldstuck, A. (2006). PC users in South Africa 2006. Johannesburg: World Wide Worx. Mawson, N. (2006, May 19). Telkom moves to shorten local loop in high-demand areas. Retrieved from http://www.engineeringnews.co.za/eng/ news/today/?show=86554 McLeod, D. (2006, June 2). Ivy’s extra mile. Retrieved IURPKWWS IUHH¿QDQFLDOPDLOFR]DWHFKnology/atech.htm McLeod, D. (2007, January 26). Undersea wrangling. Retrieved from KWWS VHFXUH¿QDQFLDOPDLO co.za/07/0126/technology/dfeat.htm Monteiro, A. (2006, March 29). Education, lack of engineers hamper SA IT. Retrieved from http:// www.moneyweb.co.za/shares/ict_sector/995758. htm Muller, R. (2006, April 20). SA slipping further behind in broadband rankings. Retrieved from http://moneyweb.iac.iafrica.com/shares/ict_sector/252408.htm
MyBroadband.co.za. (2006a, May 8). Telkom 1000 times more expensive. Retrieved from http://www. mybroadband.co.za/nephp/?m=show&id=2680 MyBroadband.co.za. (2006b, May 29). iBurst coverage expansion moving forward. Retrieved from http://www.mybroadband.co.za/nephp/ ?m=show&id=2943 Mzolo, S. (2005, December 20). Lower telecom costs ‘only talk’. Retrieved from http:// ZZZ¿QFR]DDUWLFOHVGHIDXOWGLVSOD\BDUWLFOH asp?ArticleID=1518-1786_1853631 OECD. (2006, October 13). OECD broadband statistics: June 2006. Retrieved from http://www. oecd.org/document/9/0,2340,en_2649_34223_ 37529673_1_1_1_1,00.html Simons, D. (2002). Telkom lights ADSL, but is the price right? Retrieved September 3, 2006, from http://www.iec.org/events/2002/dslwfeurope/ newsletter/feature.html Southwood, R. (2006, November 6). Satellite dominates cellular backhaul. Retrieved from http://www. mybroadband.co.za/nephp/?m=show&id=4799 Smit, T. (2005). SA wireless access and broadband market. Johannesburg: BMI-TechKnowledge. 9DSL; -XQH Telkom drops ADSL and data prices further. Retrieved from http://www. telkom.co.za/common/aboutus/mediacentre/pressrelease/articles/article_789b.html
KEY TERMS 3G: The abbreviation for third-generation cellular technology. The services associated with 3G provide the ability to transfer simultaneously both voice data and non-voice data (e.g., downloading movies or sending/receiving e-mail). A number of applications are associated with 3G including videoconferencing, high-speed Web access, and even Internet telephony. EASSY (Eastern Africa Submarine System): $ SURSRVHG VXEPDULQH ¿EHU RSWLF FDEOH WKDW LV designed to connect a multitude of countries along
27
South Africa
the east coast of Africa including Kenya, Tanzania, Mozambique, and South Africa. Landlocked countries (e.g., Botswana, Lesotho, Zambia, etc.) will also have indirect access to the cable. The proposed termination pointfor onward connection to Europeis Djibouti. Integrated Services Digital Network (ISDN): The set of protocols which transform a regular copper telephone line into a digital platform that is inherently faster and more reliable than its analogue counterpart. Local Loop Unbundling (LLU): A regulatory process of allowing multiple locally and nationally based telecommunications operators to make XVHRIFRQQHFWLRQVIURPWKHWHOHSKRQHH[FKDQJH¶V FHQWUDO RI¿FH WR WKH FXVWRPHU¶V SUHPLVHV 7KH physical wire connection between customer and company is often referred to as the “local loop” and was historically owned by the incumbent local exchange carrier.
28
Public-Switched Telephone Network (PSTN): $QHQJLQHHULQJVWDQGDUGZKLFKGH¿QHVWKHPDMRULW\ RIWKHZRUOG¶VWUDGLWLRQDOWHOHSKRQHV\VWHPV7KH technology operates by automatically connecting customers according to the number dialed. SAT-3: Refers to South Atlantic 3. This is DQ H[LVWLQJ VXEPDULQH ¿EHU RSWLF FDEOH OLQNLQJ Portugal and Spain to South Africa, with connections to several West African countries along the route. It forms part of the SAT-3/WASC/SAFE cable system, where the SAFE cable links South Africa to Asia. Small and Medium-Sized Enterprise (SME): A company whose headcount or turnover falls below certain limits. In South Africa, SMEs typically have fewer than 100 employees. Teledensity: Refers to telephone density. This is the number of telephone connections within a particular area, typically a square kilometer. Manhattan Island would have an extremely high teledensity whereas the outback regions in Australia would have an extremely low teledensity.
Division II
Asia
30
Chapter III
Bridging the Digital Divide Through Broadband Deployment Challa Radhakumari Sri Sathya Sai University, India
ABSTRACT This chapter provides a summary relating to the functioning of two projects in the two Southern States of India, Andhra Pradesh and Kerala, to show how through broadband deployment in rural areas the digital divide can be bridged. By focusing on the implementation of the two projects, the chapter illustrates their contribution in practically using the broadband technologies in overcoming the hurdles to EULGJLQJWKHGLJLWDOGLYLGHDQGKLJKOLJKWVWKHFULWLFDOVXFFHVVIDFWRUVDVLGHQWL¿HGGXULQJWKHIXQFWLRQing of the projects which helped the states in achieving their goals. The chapter also reveals through its analysis that the accessibility of services through broadband technology have brought an opportunity to the citizens to become a part of the current knowledge revolution, besides bringing about a great technological transformation to the areas where it is implemented, and thus contributed to bridging WKHGLJLWDOGLYLGH7KHFKDSWHULV¿QDOO\FRQFOXGHGE\SURYLQJWKDWPRYLQJIURPDPDQXDOWRHOHFWURQLF process with broadband technology as an enabler; the States set an example, which will serve as a set of guidelines for application of similar projects in other geographical settings.
INTRODUCTION Information and communication technology (ICT) has the potential to contribute to fostering empowerment and participation, and making the governPHQWSURFHVVHVPRUHHI¿FLHQWDQGWUDQVSDUHQWE\ encouraging communication and information sharing among people and organizations, and within the government itself. Information technology essentially refers to the digital processing, storage,
and communication of information of all kinds (Agarwal, 2001). Therefore, IT can potentially be used in every sector of the economy. The true impact of IT on growth and productivity continues to be a matter of debate, even in the United States, which has been the leader and largest adopter of IT (Arora, Arunachalam, Asundi, & Fernandes, 2000). However, there is no doubt that the IT sector has been a dynamic one in many developed countries, and India has stood out as a developing
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Bridging the Digital Divide Through Broadband Deployment
country where IT, in the guise of software exports, KDV JURZQ GUDPDWLFDOO\ GHVSLWH WKH FRXQWU\¶V relatively low level of income and development. $QH[DPSOHRI,7¶VEURDGHULPSDFWFRPHVIURP the case of IT-enabled services, a broad category covering many different kinds of data processing and voice interactions that use IT infrastructure as inputs, but do not necessarily involve the production of IT outputs (Lenhart et al., 2003). The Internet boosts immeasurably our collective capacity to archive information, search through large quantities of it quickly, and retrieve it rapidly. It is said that the Internet will expand access to education, good jobs, and better health; and that it will provide citizens with direct access to government (Helen, 2004). In so far as such claims are plausible, Internet access is an important resource, DQGLQHTXDOLW\LQ,QWHUQHWDFFHVVLVDVLJQL¿FDQW concern for all social scientists. With each passing day, there are more and more reports and studies on the opportunities and challenges associated with the global extension of information and communication technologies. Many aspects of the linkages between ICT and the resulting improvements in the commonwealth of citizens, country by country, are explored in these reports and studies (Olorunda, 2006). But, while ICT applications in rural and remote areas of both developed and developing countries are acknowledged in such reports and studies, it is usually only in a cursory fashion. This is resulting in an increasing digital divide (SocialText. com, 2006). With that in mind, this chapter focuses on matters relating to the efforts being made for bridging the digital divide in some of the developing countries like India, and how these efforts are inducing the other areas to launch similar and other innovative measures for minimizing the digital divide problem. The chapter outlines the functioning of two broadband projects in two different States of India and their role in contributing to the development of the areas, and highlights how broadband technology is practically used for bridging the digital divide by overcoming various hurdles in the implementation process. The critical success factors that created the necessary environment for
achieving the goals of the projects and contributing to their sustained development are also outlined in the chapter.
REVIEW OF LITERATURE 0DUN:DUVFKDXHU¶VERRNUHSUHVHQWVDQLPSRUWDQW and necessary stage in the evolution of scholarly WKLQNLQJ DERXW WKH GLJLWDO GLYLGH D VLJQL¿FDQW concept that has entered the lexicon of contemporary literature about technology transfer (OECD, 2005). The concept of the digital divide is perhaps itself a case study in the transfer or diffusion of NQRZOHGJH:KHQ¿UVWDSSHDULQJDVDEUHDNWKURXJK conceptualization capable of intellectually organizing and fueling an area of study, concepts and their implications tend to be presented starkly and forcefully (Modovix, 2006). Some reports also seem to suggest that developing economies such DV1LJHULDDUHWU\LQJWRµOHDSIURJ¶LQWRQHZWHFKnologies, eliminating the need for building a new infrastructure for telecommunications (Olarunda & Olarunda, 2006). The digital divide is widely regarded as a uniWDU\SKHQRPHQRQ$QGDVD¿UVWDSSUR[LPDWLRQLW is indeed useful to distinguish, in a general way, between the rich and powerful who are part of the Information Age and the poor and powerless who are not. Viewed analytically, there is not one, but three digital divides; and a fourth one is seen emerging in many nations (Keniston & Deepak Kumar, 2003).
Wealth and Power 7KH¿UVWGLYLGHLVWKDWZKLFKH[LVWVZLWKLQHYHU\ nation, industrialized or developing, between those who are rich, educated, and powerful, and those who are not. For example, income and education in the United States distinguish dramatically between those who own computers and those who do not, as between those who can access the Internet and those who cannot. In the United States, where household telephone penetration is about 95%, in 1999 households with incomes over $75,000 (roughly, the top 10%) were twenty times more
31
Bridging the Digital Divide Through Broadband Deployment
likely to have Internet access than those in the lowest income brackets: 80% of the rich and 5% of the poor had access to the Internet. If we analyze home ownership of computers, rich households were nine times more likely to own one. If we compare Americans with four years or more of college with those who have six years or less education, FRPSXWHURZQHUVKLS¿JXUHVDUH YVDQG the Internet access percentages are 49% vs. 3%. Similar results were found in a survey in Australia (Keniston & Deepak Kumar, 2003). As of mid-2002, no comparable studies have been conducted in India, where telephone connectivity is extremely low (about 3%) and the installed base of computers and Internet connections are even lower. But the overall pattern is clearly similar to that in America. As of early 2002, there were approximately six million computers in India, of which perhaps two-thirds were in businesses, VFKRROV JRYHUQPHQW RI¿FHV DQG VR IRUWK OHDYing, at a high estimate, two million computers in households. In mid-2002, there were probably about a million Internet connections in India, again most of them in institutional settings rather than LQGLYLGXDOKRXVHKROGV$¿JXUHRIPLOOLRQ,QGLDQ Internet-connected households (out of about 200 million households) in 2002 would be on the high side (Keniston & Deepak Kumar, 2003). :KRDUHWKHµFRQQHFWHG¶LQ,QGLD"2EYLRXVO\ as a group, they are a small, rich, successful, and English-speaking minority. For all of its ancient cultural wealth, despite the persistence of old elites and the emergence of new elites, India remains one RIWKHZRUOG¶VSRRUHVWVRFLHWLHV0RUHWKDQ of the population is illiterate; tens of millions of children are not in school. Telephone connectivity in India is about 3% and will not rise much above that level unless the cost of connections (the soFDOOHGµODVWPLOH¶FRVW FDQEHORZHUHG,QVKRUW there is a massive digital divide in India based on income, related to education and urban residence, and correlated with economic, political, and cultural power. (Keniston, 2003).
Linguistic and Cultural A second digital divide, less often noted, is linguistic and cultural. In many nations this divide 32
separates those who speak English or another West European language from those who do not. But even in the United States, where well over RIDOOLQKDELWDQWVVSHDNÀXHQW(QJOLVKWKHUH are large differences in access to ICTs among different ethnic and cultural groups. For example, in 1998, Asian American households (largely of 6RXWK $VLDQ RU 6RXWK 3DFL¿F $VLDQ H[WUDFWLRQ had 55% computer ownership, white Americans had 52%, while Americans of Hispanic origin had 25% and blacks 23% respectively. An even larger gap separated Asian Americans and whites from blacks and Hispanics with regard to Internet access (Pierce, 2003). These cultural disparities, dramatic in the U.S., are far more notable in India, where they are compounded by linguistic issues. An estimated 60-80% of all Web sites in the world are in English, while DOPRVWDOOWKHUHVWDUHLQRQHRIWKHPDMRUµ1RUWKHUQ¶ ODQJXDJHV OLNH -DSDQHVH *HUPDQ )UHQFK Spanish, Portuguese, and increasingly Chinese. But in India, like the rest of South Asia, only an HVWLPDWHGRIWKHSRSXODWLRQVSHDNVÀXHQW English while the rest (more than 900 million Indians and about 1.2 billion South Asians) speak other languages. For Indians who speak no (or little) English, the barriers to the Information Age are almost inseparable. All widely used operating systems require some knowledge of English or one of the µ1RUWKHUQ¶ ODQJXDJHV 7KXV LQ SUDFWLFH XQOHVV Indians know English, which most Indians do not, no matter how wealthy, brilliant, educated, prosperous, or motivated they may be, computer use and Internet access are effectively out of the TXHVWLRQ7KHUHVXOWLVDVHOIFRQ¿UPLQJSURSKHF\ since there is so little software in any language other than English, virtually everyone in South Asia who uses computers knows English. Of course the 50 or so million Indians who VSHDN ÀXHQW (QJOLVK E\ QR PHDQV FRQVWLWXWH D representative sample of the Indian population: they again tend to be prosperous, urban, highly HGXFDWHGFRQFHQWUDWHGLQWHFKQLFDO¿HOGV7KH\ are, in a word, members of the Indian elite, where English is the lingua franca. For the great majority of Indians, however, computers are linguistically inaccessible.
Bridging the Digital Divide Through Broadband Deployment
To linguistic inaccessibility in India is added the absence of culturally relevant content. The number of Web sites in 2000 in India is small in any case, but the number of sites in Indian languages is miniscule. To be sure, a few gifted programmers are attempting to change this, and sites are beginning to appear in languages with vast populations of mother tongue speakers like Hindi, Bengali, or Tamil. But to all intents and purposes, the many ancient, rich, and sophisticated cultures that make up India remain almost invisible on the Web (Keniston, 2003).
Wide Gap Between the Rich and the Poor Nations The third digital divide follows inevitably from the ¿UVWWZRthat is, the growing digital gap between the rich and the poor nations. At one extreme are WKH8QLWHG6WDWHVDQGWKHµ1RUGLF¶FRXQWULHVOLNH Sweden, Germany, Finland, and Iceland, where household telephone connectivity is well over 90%, computer saturation is over 50%, and home-based Internet connectivity averages over 50%. At the other extreme lies most of Africa, most of South America, South Asia, China, Indonesia, and so onthe 80% of the world where telephone connectivity is 3% or less (less than 30 million/1 billion in India), home computer ownership is 1-2%, and Internet connectivity is less than half of that. The reason why the digital divide between nations is increasing seems clear. If widespread access to ICTs gives a nation an advantage, and lack of access leaves it at a disadvantage, then the maxim, “To those who have shall be given” applies with special force to the international digital divide. The international disparity in access to ICTs is of FRXUVH DQ DVSHFW RI LQGHHG D UHÀHFWLRQ RI RWKHU disparities between rich and poor nations. But insofar as ICTs are themselves enabling, facilitating, and wealth-creating, the international divide in information technology widens the already great gulf between North and South.
Digerati The fourth digital divide, which is emerging in countries like India and America, can be called the µGLJHUDWL¶'LJHUDWLLPSOLHVHPHUJHQFHRIDVSHFLDO HOLWHJURXSZKLFKUHSUHVHQWVWKHEHQH¿FLDULHVRI the enormous successful information technology industry and the other knowledge-based sectors of the economy such as biotechnology and pharmacology. Time and again in India, for example, brilliant graduates of Indian Institutes of Technology or major engineering colleges and universities who chose to concentrate in the natural sciences, mechanical engineering, or chemical engineering comment that their equally gifted classmates who entered computer science or biotechnology are now earning many times their incomes and living in an altogether different way. Unlike older Indian elites, the privileges of the new digerati are based not on caste, inherited wealth, family connections, or access to traditional rulers, but on a combination of education, brainpower, special entrepreneurial skills, and ability to stay on the “cutting edge” of knowledge. The critical question about the fourth digital divide, however, is whether the prosperity of this new digital elite spreads to the rest of society, especially to urban poor and to rural villagers, or whether it creates an increasingly separate, cosmopolitan, knowledge-based enclave. The study of the implementation of Akshaya Project and how it is contributing to bridging the digital divide, and conferring various socio-ecoQRPLFEHQH¿WVKDVLWVUHOHYDQFHQRZSDUWLFXODUO\ when the policymakers world over are considering broadband as the instrument for accelerating the HFRQRPLFGHYHORSPHQW7KHVWXG\LVDOVRVLJQL¿FDQW as it highlights the technological transformation the districts have undergone, besides examining YDULRXVHFRQRPLFDQGVRFLDOEHQH¿WVWKHEURDGEDQG deployment could confer on the citizens (Radhakumari, 2006). The Akshya Project of Kerala demonstrates how in developing countries local residents are aware of WKHEHQH¿WVRI,&7LQXSOLIWLQJWKHLUOLYHVEXWLWLV the lack of infrastructure and other facilities that DUHGHSULYLQJWKHPRIWKHDFFHVVWRWKHEHQH¿WV
33
Bridging the Digital Divide Through Broadband Deployment
Using the awareness levels the local residents are encouraging the local governments to play an active role in bridging the digital divide. The Akshaya Project, which is an ambitious e-literacy program of the Kerala State, has proved that information technology implementation in rural areas can also promote entrepreneurship development (Radhakumari, 2006).
BACKGROUND FOR INTRODUCTION OF THE PROJECTS A.P Broadband Network $QGKUD3UDGHVKDQG.HUDODDUHWKH¿UVWWZRVWDWHV in India to have initiated the efforts to deploy broadband to rural areas for bridging the digital divide so as to achieve the goal of balanced spread of e-services availability.
A.P Broadband Project Andhra Pradesh has emerged as one of the leaders in e-governance in India. A number of pioneering initiatives have been taken up by the state to transform itself into a knowledge society. Innovative e-initiatives such as e-Seva, which provides over 120 services to citizens; the Computerized Administration of Registration Department (CARD) project for computerization of registration of immovable property; e-procurement for bringing about transparency and reduction in government procurement procedures; and so on, have been launched by the state government for providing better quality of services to the people of Andhra Pradesh. The state has witnessed a rapid expansion of telecom networks both in the public and private VHFWRUGRPDLQVLQUHFHQW\HDUV
34
be in danger of being left behind in the emerging digital economy. Therefore the state has decided to take action on three fronts to address the problem RIGLJLWDOGLYLGH ¿UVWO\WRSXWLQSODFHWHOHFRPmunications infrastructure in order to provide affordable bandwidth for large sections of the community; secondly, to make low-cost information access appliances which are easy to learn and use by rural population, available in large numbers, so as to increase IT penetration rates; and thirdly, to develop content relevant to the lives of people and making it available over networks. :KLOHWKHVWDWHKDVPDGHVLJQL¿FDQWJDLQVLQ WKH¿HOGRI,7DSSOLFDWLRQVRYHUWKHODVWWHQ\HDUV in urban areas, especially in the state capital of Hyderabad, it is observed that the situation in rural areas continues to be grim. Under such scenario YLVXDOL]LQJFHUWDLQGLUHFWEHQH¿WVRIPDNLQJ,&7 access a universal phenomenon such as provision of access to markets, healthcare, entertainment, and MREV IUHHÀRZRILQIRUPDWLRQ LQFUHDVHGWUDQVSDUency in government operations; easy management RIGLVDVWHUVOLNHF\FORQHVDQGÀRRGVWKURXJKORJLVWLF support; procurement of information in matters like hygiene, infant care, crop disease management, which in turn would improve the overall quality of life of citizens in rural areasthe State of Andhra Pradesh has proceeded with the plan of making ICT a universally accessible and affordable phenomenon. Aspiring to become a leader in the information HFRQRP\DQGZLVKLQJWRSURYLGHWKHEHQH¿WVRI information technology revolution to broad sections of its population, the A.P government has taken initiative to facilitate and accelerate the implementation of a broadband network in the state that will provide low-cost bandwidth to people residing in both urban and rural areas. Therefore a proposed action under the title “A.P Broadband Network” is conceived with the aim of creating ubiquitous Internet access through broadband deployment to all the rural areas, and accelerate the economic, social, and personal development of citizens; and help to leapfrog the state while providing a demonstration of how information technology can improve the quality of life in rural areas.
Bridging the Digital Divide Through Broadband Deployment
Akshaya Project of Kerala Realizing that diffusion of broadband services at DIIRUGDEOHSULFHOHYHOVZLOOKDYHVLJQL¿FDQWLPSDFW on GDP: attract new investment; create additional MREVGHYHORSDTXDOL¿HGDQGHI¿FLHQWODERUSRRO increase productivity, and access new and improved services. The state government of Kerala has made VLJQL¿FDQW HIIRUWV WR FUHDWH WKH HQYLURQPHQW IRU stimulating the explosive growth of infrastructure for making the broadband access a universally affordable phenomenon. While the other countries, world over, are still speaking about delivering µXQLYHUVDOO\DIIRUGDEOHDFFHVVWREURDGEDQG¶IRU all of their citizens, India has galloped towards a revolution in information technology (IT) through broadband deployment. Perceiving better, speedier, and more transparent administration; development of entrepreneurs of all sizes, big and small, all over who would in turn cater to the demands at national and international levels; growth of IT professionals of international standards through specialized IT education; and all of which in turn would augment the quality of life of citizens, the state government of Kerala in India has taken a big leap by leveraging broadband services for taking e-governance to rural areas DQGODXQFKHGWKHµ$NVKD\D3URMHFW¶7KHSURMHFW LVDQH[DPSOHIRUGLIIXVLRQRI,7EHQH¿WVWKURXJK broadband deployment in rural areas. The efforts made by the villagers in initiating the project and contributing to its sustained functioning demonstrates that it is the lack of infrastructure and other facilities that are responsible for digital divide in the rural areas and not the talent or awareness levels (Radhakumari, 2006). In this backdrop, the above two integrated IT broadband projects have been conceived by the WZRVWDWHVIRUDFKLHYLQJWKHREMHFWLYHRIµSURYLGLQJ XUEDQDPHQLWLHVLQUXUDODUHDV¶ 385$ DQGWKXV µEULGJLQJWKHGLJLWDOGLYLGH¶7KHWZRSURMHFWVDUH chosen as the focal point for this chapter for two VLJQL¿FDQWUHDVRQV a.
The broadband project of Andhra Pradesh VHWVDQH[DPSOHIRUGLIIXVLRQRI,7EHQH¿WV through broadband deployment in rural
areas. The project lays a strong foundation for broadband application in the entire state which will accommodate many more services in the future besides making the current einitiatives available at the doorstep of each and every household in rural areas. E 7KH6WDWHRI.HUDODLVWKH¿UVWLQ,QGLDWRDWtempt to deploy broadband under the name µ$NVKD\D 3URMHFW¶ IRU EULGJLQJ WKH GLJLWDO GLYLGH WKURXJK D µFHQW SHUFHQW HOLWHUDF\¶ program; thus it is preparing the people of the entire state to be a part of the knowledge revolution, which is the result of the IT revolution. The project serves as a role model for taking IT to the grassroots level in the villages. The initiative for implementation of the project came from local bodies and has ÀRZQXSZDUGVWRWRSOHYHODGPLQLVWUDWLRQ
Functioning of the Projects The following discussion is devoted to describing the functioning of the chosen projects in terms of WKHLU REMHFWLYHV WHFKQRORJ\ DGRSWLRQ ¿QDQFLDO LPSOLFDWLRQVUROORXWRIWKHSURMHFWVEHQH¿WVUHQdered, and the critical success factors relating to the implementation of the projects.
A.P Broadband Project According to the Telecom Regulatory Authority RI,QGLDEURDGEDQGLVGH¿QHGDV³DQDOZD\VRQ data connection that is able to support various interactive services, and has the capability of a minimum download speed of 256 kbps” (TRAI, 2006a). However, in countries like South Korea, the understanding of broadband is bandwidth in excess of 2 Mbps. In view of the rapid changes in WHFKQRORJLHVLWPD\QRWEHDSSURSULDWHWRGH¿QH broadband in terms of information transmission UDWHV,WZRXOGEHXVHIXOWRGH¿QHEURDGEDQGDVD high-capacity two-way link between an end user and access network suppliers capable of supporting full motion, interactive video applications (TRAI, 2006b). This is an approach that has also been adopted, for example, by Canada, following the recommendations of its National Broadband
35
Bridging the Digital Divide Through Broadband Deployment
Task Force. For the implementation of the A.P Broadband Project in the state, the capacity to deliver a symmetrical bandwidth of at least 512 kbps with a guaranteed latency of less than 100 milliseconds and less than 1% packet loss, to the end user, with the currently available technologies, is considered as broadband.
Objectives of the A.P Broadband Project The project has set for itself double-direction goals: (a) customer centered and (b) business centered. •
•
36
Customer-centered objectives: The principal objective of the A.P Broadband Project is to accelerate the deployment of a broadband network in the state by determining the technology solutions and selecting credible partners to provide high-quality, affordable, and equitable broadband access for citizens, businesses, and public institutions. To provide seamless connectivity for citizens and businesses is the additional objective of the project which is accomplished by connecting the Andhra Pradesh Broadband Network with public sector electronic services through virtual links and peering relationships. Business-centered objectives: The A.P Broadband Project has the following key business-centered objectives: { Offering delivery of anytime, anywhere citizen-friendly electronic services and Internet access for citizens, businesses, and public institutions at affordable rates { Rolling out of broadband communications especially to rural areas that do not have access to high-quality communications { Aggregation of government demand for bandwidth and providing such EDQGZLGWKIRUD¿[HGDQQXDOIHHWREH determined from year to year { Providing solutions for voice, video, and data access for residential connections with suitable information appliances
{
Making available applications and content solutions for enabling the A.P Broadband Project to deliver expected EHQH¿WV
Techniques Adopted for Achievement of the Objectives To achieve the key objective of widespread availability and deployment of broadband connectivity for rural and urban households at an affordable tariff, a base tariff structure was proposed by the bidder that meets this objective. A separate tariff structure was also proposed for new applications and value-added services that will be provided. In addition, it is also arranged to aggregate, to the extent feasible, the bandwidth requirements of JRYHUQPHQWRI¿FHVDJHQFLHVDQGLQVWLWXWLRQVDQG source this from the proposed broadband network LQH[FKDQJHIRUD¿[HGDQQXDOIHH 6SHFL¿FDOO\WKHVWDWHKDVDGRSWHGWKHIROORZLQJ four strategies for achievement of the objectives: 1. 2. 3. 4.
Make an overall project implementation model that is acceptable Work for outcomes in the set timeframe Establish an open IP-based network, capable of application service provision Make available the broadband network access at affordable prices within the state
Rollout of the Broadband Network Telecommunication technologies have also been witnessing rapid advancement in recent years. :KLOH¿EHURSWLFQHWZRUNVFDQWKHRUHWLFDOO\VXSSRUWLQ¿QLWHEDQGZLGWKZLUHOHVVWHFKQRORJLHVKDYH matured to the point that they can today easily cross the broadband threshold. New wireless technologies make it possible to provide reliable and high-quality bandwidth to people in remote areas. Technologies like 802.11g and 802.16 are being used for catering to broadband requirements. The FRVWVRIKLJKVSHHGQHWZRUNVXVLQJ¿EHUDQGZLUHless have been coming down dramatically. Gigabit Ethernet has made it much cheaper to establish
Bridging the Digital Divide Through Broadband Deployment
broadband communications infrastructure and is opening up possibilities of leapfrog development. The broadband infrastructure established in the state is capable of reaching all parts of the state and able to deliver the following bandwidths to government and public institutions: • •
•
•
10Gb to Class I locations, which includes each of the district headquarters 1Gb to Class II locations, which includes each Mandal headquarters and other selected locations (about 1,200 locations) 1 Gb to Class III locations, which includes educational institutions, hospitals, and other designated locations, provided the customer LQVWDOOVRUSD\VIRULQVWDOODWLRQRIDGLUHFW¿EHU connection, otherwise a 50Mb connection should be provided at a tariff not exceeding WKHWDULIIIRUD0%GLUHFW¿EHUFRQQHFtion 100/50 Mb to Class IV locations, which includes each village (about 25,000 locations) (50 Mb connectivity to villages and governPHQWRI¿FHVVKDOOQRWH[FHHGRIWKHWRWDO YLOODJHVJRYHUQPHQWRI¿FHVZKHUHYHUGLUHFW ¿EHUFRQQHFWLRQLVQRWIHDVLEOH
The different levels of the network centers are: •
•
•
State Network Center (SNC): The SNC is WKHKXE ZKHUHDOOVWDWHQHWZRUN WUDI¿FDJgregates. The SNC offers various services like data services, voice services on VoIP, and videoconferencing for all the state government departments. Additionally, it offers access to Internet and state data centers, and provides connectivity to content servers. The SNC will also host the helpdesk. District Network Center (DNC): The DNC LVWKHKXEZKHUHDOOGLVWULFWQHWZRUNWUDI¿F aggregates. All DNCs shall connect to the SNC. Mandal Aggregation Center (MAC): The MAC is the hub where all Mandal network WUDI¿FDJJUHJDWHV$OO0$&V KLJKXVDJHDQG low usage) are connected to their respective DNCs.
•
Village Aggregation Center/PoP (VAC): All VACs are connected to their respective MACs.
%DVHGRQWKHIXQFWLRQDOLW\DQGQHHGIRUHI¿FLHQW management, the project is segregated into two distinct areas: (a) broadband backbone network and (b) logical aggregation network. The network is designed as layered, hybrid topology to meet the government architecture and topology requirements. The scope covers the following number RI*R$3RI¿FHVWREHFRQQHFWHGIRUHDFKFODVV of location: • •
At Hyderabad200 locations 'LVWULFW2I¿FHV40 locations 6XE'LYLVLRQDO2I¿FHV25 locations 0DQGDO2I¿FHV15 locations Villages21,150 locations
Services on Broadband Network The various services offered on broadband network in the state are categorized as services to the education department, services to all users/citizens, Internet access services, and other services to be supported on the network. •
•
Services to education department on A.P Broadband Network: { Free services: All CBSE curriculum from Class 1 to 12 that is available with the consortium partners provided online on the education portal can be accessed by the students at no extra charge except the connectivity charge { Chargeable services: The network and software have provision for interactive coaching classes. This is however a chargeable service. The fees will depend on the interactive course being taken Services offered to all users/citizens: The project will entail the deployment of infrastructure providing a broadband access service consisting of the network connection and broadband access service, meeting WKHPLQLPXPFKDUDFWHULVWLFVGH¿QHGLQWRD
37
Bridging the Digital Divide Through Broadband Deployment
broadband network providing the delivery of IP services. { Simultaneous services: The broadband access service will provide QoS functionality to support simultaneous Internet access, videoconferencing, and voice services { ISP services: The broadband access VHUYLFHZLOOEHDEOHWRÀH[LEO\SURYLGH user access to a wide, unrestricted number of ISPs and ASPs providing general DQGVSHFL¿FVHUYLFHVHOHFWURQLFDOO\ { Minimum bandwidth: The broadband access service and customer network equipment ID supporting the requirement for delivery of symmetrical 512kbps service. The broadband access service will be capable of delivering the H[SDQVLRQFDSDELOLW\GH¿QHGEHORZ Bandwidth kbps Primary Initial Minimum 2 Mbps
•
•
38
Secondary 100 Mbps
Internet access services: The broadband access service is supporting a generic IP Internet service, utilizing standard Web browser and e-mail client software Other services to be supported on the network: Broadband network will support the following additional services: easy set up and networking; expansion of capacity and capability; system management and call processing tools; network security; H.323 DQG6,3 HGXFDWLRQVSHFL¿FDSSOLFDWLRQVDQG hardware devices for distance learning; voice service (VoIP); simultaneous and interactive exchange of video, audio, and data point to point, and point to multi-point conferencing; interaction and collaboration applications; Web server and presentation systems; and FXVWRPL]HGPHHWLQJURRPRI¿FHRUGHVNWRS systems { Voice services (VoIP): The broadband access service will be capable of the carULDJHRI9R,3WUDI¿FXVLQJDQDSSURSULDWH level of QoS. Suitable customer network equipment will be made available to sup-
{
port the delivery of VoIP, however the provision of customer premises equipment is not to be an integral part of the scope of the project. PSTN interface/connectivity: The VoIP network will interconnect to the PSTN and other telecom networks. The various services considered are: ¾ Data services ¾ VOIP voice services ¾ Video services ¾ Internet connectivity ¾ Connectivity to the state data center
Functioning of the Akshaya Project of Kerala: Initiation of the Project 7KH µ$NVKD\D¶ PHDQLQJ µSHUSHWXDWLQJ SURVSHULW\¶PDUNVWKHEHJLQQLQJRIDGULYHWRµEULGJHWKH GLJLWDOGLYLGH¶LQWKH6WDWHRI.HUDOD7KHLQLWLDWLRQ for the project, unlike any other IT projects, has not come from the government as an imposition, but came as a result of the initiation taken by the three-tier Panchayatsthat is, the local/block panchayat, village/grama panchat, and the district panchayatafter decentralization of planning in India, making the project a unique one among the other IT projects in Kerala in particular and all the other IT projects of other states in general. Understanding the situation that those who do not move along with the technical progress will be marginalized and will not be able to grow along with the world, if the leap in the IT is missed, the grama Panchayat of Malappuram proposed a plan for spreading e-literacy to every family and WKXVWDNLQJWKHEHQH¿WVRIWKH,7WRWKHJUDVVURRWV levels. With the support of block panchayat and the district panchayat, the Malappuram local panchayat approached different agencies to undertake this project by setting aside Rs. 60 lakhs during 2001-2002 plan periods. When the state ,7PLVVLRQZDV¿QDOO\DSSURDFKHGLWWRRNXSWKLV project proposal as a novel idea and extended full cooperation and the needed support for the materialization of the plan. Thus the Akshaya Project was launched in the Malappuram district by none
Bridging the Digital Divide Through Broadband Deployment
other than the President of India, Sri A.P.J Kalam, on November 28, 2002, on a pilot basis. The project is being implemented by the local bodies, with support and monitoring from the district and state administration.
Objectives of the Akshaya Project $NVKD\D¶VPDLQDLPLVWRWUDQVIRUPWKHOLYHVRI 6.4 million families in Kerala, cutting across the social, cultural, and linguistic barriers, to bring KRPHWKHEHQH¿WVRILQIRUPDWLRQWHFKQRORJ\WRWKH common man, by familiarizing at least one person from each family with computers and empowering him or her to access relevant e-content in the regional language.
Mode of Achievement of the Objective Envisaging a 15-hour training for 10 days of oneand-a-half-hour duration each to train the citizens, the Government of Kerala sought to create 6,000 modern community information centers called µ$NVKD\DFHQWHUV¶DQGSURYLGHFRQYHQLHQWDFFHVV to information services within a radius of two kilometers from the each family residence.
Spread of the Project’s Objective to the State The government conceived a three-stage implementation plan for taking the Akshaya Project to WKHHQWLUHVWDWH'XULQJWKH¿UVWVWDJHWKHSURMHFW is implemented in one district, Malappuram, on DSLORWEDVLVZKLFKKDVEHQH¿WHGSHRSOH spread over 78 panchayats and municipalities in the district, with at least 1,000 persons from each panchayat converted into e-literates who are ready to use this knowledge to take full advantage in becoming familiar with the use of the Internet. The project is in its second phase of implementation in the district. Based on the successful implementation of the project during the pilot scheme, the project was rolled out to another six districts, which conVWLWXWHVWKHVHFRQGSKDVHRIWKH$NVKD\D3URMHFW¶V implementation.
Presently, the project is completing its second stage of implementation and is all set to begin the third stage of its implementation by covering the remaining six districts in the state. With the completion of the second stage of implementation, the project is able to bring credit to the state RI DFKLHYLQJ µFHQW SHUFHQW HOLWHUDF\¶ LQ VHYHQ districts.
Mode of Functioning of the Project The functioning mode of the project is the publicprivate partnership (PPP) model, where the three tier panchayats take the lead role in planning for the project, and the execution is being taken care of by the selected private entrepreneurs. The project has set for itself a three-phase LPSOHPHQWDWLRQ VWUDWHJ\ 7KH ¿UVW SKDVH GHDOV ZLWKWKHDFKLHYHPHQWRIµFHQWSHUFHQWHOLWHUDF\¶ in the state. The second phase covers the “e-vidya” program, under which all those citizens who have undergone WUDLQLQJLQWKH¿UVWSKDVHDQGZKRDUHDPELWLRXV in furthering their knowledge of computers, are given an opportunity to take part in the 30-hour computer training program that offers knowledge relating to the entire basic computer operation. This FRXUVHRIIHUVDFHUWL¿FDWHWRDOOWKRVHFDQGLGDWHV who complete the e-vidya course successfully. 7KHFHUWL¿FDWHLVUHFRJQL]HGE\WKH*RYHUQPHQW of Kerala. The e-vidya program is open to all and is not restricted to the trainees of the e-literacy plan. 7KHWKLUGSKDVHFRPSULVHVDQµ,QWHUQHWSDFNDJH¶ which envisages 10 hours of comprehensive training to each person through e-mail, chat, and other Internet facilities, so that they become familiar with the use of the Internet, including browsing. After completion of this training, each trainee is given a free e-mail account.
Financial Implication of the Project to the Citizen The learner pays only Rs. 40 of the total fee of Rs. 120the remaining being paid by the grama
39
Bridging the Digital Divide Through Broadband Deployment
panchayat, block panchayat, and the district panchayats in the ratio of Rs. 60 and Rs. 10 each UHVSHFWLYHO\GXULQJWKH¿UVWVWDJHRIWKHSURMHFW¶V implementation. Learners belonging to SC/ST communities are exempted from paying even this nominal amount. During the second phase of Akshaya implementation, the cost of the total e-vidya course comes to Rs. 450/-, which is paid in full by the trainees. The total cost of the Internet package program of the project during the third phase is estimated to be Rs. 140 per person, of which the government proposes to contribute Rs. 100, leaving the remaining Rs. 40 as the share of the trainee.
Financial Implication of the Project to the Government The government is required to make an investment in providing the connectivity to each Akshaya center. The establishment cost of the coordinating RI¿FHthat is, the salaries of the Akshaya staff, HOHFWULFLW\ FKDUJHV UHQW RI WKH RI¿FH DUHD DQG so onwill constitute the annual recurring cost required to be met by the government.
Private Entrepreneur Selection, Training, and Motivation; and the Role All the educated unemployed youth who have secured a minimum 12th standard degree with basic computer knowledge and with the minimum entrepreneurial abilities like capacity to take a certain amount of risk, managing the setting up DQGUXQQLQJRIDQ$NVKD\DFHQWHUDQG¿QDOO\ZLWK minimum zeal to serve citizens, are all eligible to become the private entrepreneurs for running the $NVKD\DFHQWHUV7KH¿QDOVHOHFWLRQKRZHYHULV made on the basis of interviews and the opinion of the local bodies. Every selected entrepreneur should be ready to make an initial investment of Rs. 3 lakhs to establish an Akshaya center, requiring a geographical DUHDRIVTXDUHIHHWDQG¿YHFRPSXWHUV,Q order to promote entrepreneurship development, the government also assists those entrepreneurs
40
ZKRDUHIXO¿OOLQJDOOWKHRWKHUIRUPDOLWLHVEXWDUH XQDEOHWRVHFXUH¿QDQFLDOVXSSRUWLQJHWWLQJWKH ORDQVIURPEDQNVE\VWDQGLQJDVWKHLUµVXUHWLHV¶ Entrepreneurs who appear for interviews are DOVR JLYHQ DQ RSSRUWXQLW\ WR ¿OH FRPSODLQWV LI any, with the Appeal Committee. The selected HQWUHSUHQHXUV PXVW ¿OH DQ XQGHUWDNLQJ DW WKH $NVKD\D3URMHFWRI¿FHVWDWLQJRQD5VVWDPS paper that they are ready to serve the project for a minimum of three years. The selected entrepreneurs are trained in management skills to run the Akshaya centers in different batches in two-day and three-day workshops and seminars. Experts from C-DIT (Center for Development of Imaging Technology) become the trainers for this program. The entrepreneur is expected to recover his or her investment by conducting the e-literacy course and providing various utility services, besides using the centers for running other computer courses for the public as part of the business.
Selection of Trainers/Instructors (YHU\$NVKD\DFHQWHUKDVPLQLPXP¿YHLQVWUXFtors, one of whom is required to possess either a PGDCA or MCA or BSc degree in computer science. The others can be persons with computer knowledge and a minimum of Plus Two education background. All the instructors are given training by C-DIT experts who introduce them to the contents of four CDs developed by C-DIT for the e-literacy campaign. These CDs, meant for e-literacy learners, contain self-study material, some computer games, and certain simple computer programs. Three instructors are selected, one for every 500 learners; they are paid by the entrepreneur.
6HOHFWLRQRI%HQH¿FLDULHVRU7UDLQHHV Trainees are selected through household surveys. Entrepreneurs, local body representatives, volunteers of Kudumbasree (a poverty alleviation program of the Local Government), and Nehur Yuva Kendra and the Literacy Mission are under-
Bridging the Digital Divide Through Broadband Deployment
taking the e-literacy campaign to create awareness among the public about the entire Akshaya 3URMHFW¶VPRWLYH
•
%HQH¿WVRIWKH%URDGEDQG3URMHFWVWR Different Sections of the Society
Akshaya centers are acting as common service centers: Akshaya centers are functioning as the information centers and the common service providers offering all e-services under one roof
,QFLGHQWDO%HQH¿WV 7KH EHQH¿WV RI LPSOHPHQWLQJ WKH SURMHFWV DUH experienced by all the citizens from two angles, direct and incidental:
•
'LUHFW%HQH¿WV •
•
•
•
Achievement of cent percent e-literacy: The project implementation has brought an added credit to the State of Kerala of achieving the VWDWXVRIWKH¿UVWVWDWHWREULQJDERXWQRWRQO\ cent percent literacy but also e-literacy Low-cost e-governance: Adoption of the PPP model for implementation of the project has enabled the state government to bring about e-governance at a very low cost. Since the private party is contributing to the infraVWUXFWXUH GHYHORSPHQW WKH JRYHUQPHQW¶V ¿QDQFLDO EXUGHQ LV UHGXFHG WR VXSSO\LQJ of only connectivity to different Akshaya centers (Radhakumari, 2006). Equal access: All age groups and all sexes have equal preferences in accessing the facilities offered by the Akshaya Project of Kerala. Irrespective of age and sex, all family members in villages, barring school children who have the facility of acquiring computer knowledge as part of their course, have equal HOLJLELOLW\WREHFRPHWKHEHQH¿FLDU\RIWKH project. Active involvement of local people: Since the initiation came from the local bodies (Akshaya Project) and private parties (A.P), there is active involvement of both parties in ¿QDQFLQJDQGPRQLWRULQJWKHHQWLUHIXQFWLRQing of the projects, which is promoting the sustained development of the projects; this thus eases ongoing burden on the government
•
•
Entrepreneurship development: Through involvement of locally available educated unemployed youth as private partners for implementation of the project, the Akshaya Project has incidentally contributed to the promotion of self-employment strategy and thus has proved that ICT adoption in rural areas can promote entrepreneurship development Promotion of backward communities: These communities are given the opportunity to enjoy the satisfaction of acquiring computer knowledge and participation in the knowledge revolution of the current times Enthusiasm of the senior trainers is inspiration: The enthusiasm and zeal of the elderly citizens in enrolling and attending the Akshaya classes are serving as a great inspiration to all the youngsters in the State of Kerala, who fail to make use of the opportunities knocking at their doors
Uniqueness of the Akshaya Project • • •
• •
•
Initiation and monitoring of the entire functioning of the project by the local bodies Implementation with the help of Kerala IT mission Active role of the local educated unemployed youth, which is one of the important parties to the functioning of the project Contribution to entrepreneurship development Generation of self-employment for the educated unemployed youth, thus preventing the brain drain Minimization of the chances of frustration and suicidal tendencies among the youth on account of severe unemployment problems
41
Bridging the Digital Divide Through Broadband Deployment
•
•
•
•
Increase in the number of jobs and creates a source of regular income, especially for the entrepreneurs and the instructors, and assured source of income for the trainees in the future Sense of achievement and satisfaction of participating in the government programs to the entrepreneurs and trainers, which otherwise could never be dreamed of by this category of people Practically setting an example for “broadband deployment for bridging the digital divide,” and thus taking the state to achieving cyber ÀXHQF\ Contribution to breaking the age-old social and cultural barriers in the traditional bound areas
•
Akshaya Project of Kerala •
&ULWLFDO6XFFHVV)DFWRUV,GHQWL¿HG from the Implementation of the Projects These factors represent the most important variables, which contributed to the successful implementation of both the broadband projects in the two chosen states of A.P and Kerala:
•
Andhra Pradesh Broadband Project •
•
•
42
Rapid roll out of broadband access services across the state, especially to rural areas that do not have access to high-quality communications. Adoption of a strategy with which infrastructure laying, and providing application and content solutions were done simultaneously. This had made the project a great success and made the state overcome one of the important and common barriers to broadband deployment; that is, the long-time gap between infrastructure development and applications. Offering anytime, anywhere citizen-friendly delivery of electronic services and Internet access for citizens, businesses, and public institutions at affordable rates
(I¿FLHQWLGHQWL¿FDWLRQRIULVNVWKDWDUHOLNHO\ to impact the deployment capability at the initial stage, and adoption of appropriate instantaneous risk management strategies Giving higher scores for bidders whose proposals include adoption of strategies for higher penetration of access devices and better availability of content
•
•
Active involvement of the local bodies and local people of the district: Realization by the local bodies themselves that leveraging information technology for the overall development of the district, along with the most current technological requirements of the world, made the entire project functional, and thus automatically avoided the resistance problem from the local people. This has also contributed to the adoption of technological change, with much ease, by the people of the district. Involvement of educated unemployed as private entrepreneurs: At the pre-emptive stage of moving the district forward, the policymakers created partnerships with the talented, educated unemployed youth of the district and the Tulip Pvt. Ltd., the private service provider, for the adoption of proven knowledge and technology; this constituted the most important sustainability factor for the perpetual functioning of the project. Connectivity: Without the broadband network, the district would not have been able to HVWDEOLVKLQWUDQHWDQG,QWHUQHWµFRQQHFWLYLW\¶ with its various Akshaya e-centers, nor fully leverage information technology applications for its operations. Neither the organizational restructuring nor the strategic technology plan would have been feasible. Organizational review: Decentralization of plan process has given a spurt to the local ERG\¶VDFWLYHUROHLQWKHGHYHORSPHQWRIWKH GLVWULFW$VDUHVXOWWKHORFDOSDQFKD\DW¶VLQvolvement in setting up of Akshaya e-centers, selection of private entrepreneurs, monitoring
Bridging the Digital Divide Through Broadband Deployment
•
•
•
•
of the functioning of the project, and so on, have made the project a highly successful venture. Adoption of strategic technology plan: Based on the geographical scenario of Malappuram district, provision of connectivity became a major hurdle. Many private and public sector enterprises including BSNL could not come forward to help the state cross the hurdle. Finally, the Tulip Pvt. service provider ventured and provided the wireless broadband network connectivity. Speed of plan implementation: With the close involvement, supervision, and scrutiny of the local bodiescomprising members from three tier panchayats (local, grama, and district)the decision to bring broadband network to the district, by crossing all the hurdles such as technical, geographical, political, social, cultural, and other non-technical factors, and investing in IT, was made possible. Simultaneous program of selection, training and motivation of trainees: Once the SURMHFW SURSRVDO ZDV ¿QDOL]HG DQG LPSOHmented, private entrepreneurs and the trainers were selected, trained, and motivated, simultaneously, without loss of time for designing, planning, and implementation of the programs, individually. This was only possible because the same enthusiastic team RIRI¿FLDOVDQGQRQRI¿FLDOVZHUHDOORZHGWR be involved in the project from the concepWLRQ RI WKH LGHD XQWLO WKH ¿QDO VWDJH RI LWV implementation. Training: The organizers of the project made DVLJQL¿FDQWDQGVXVWDLQLQJFRPPLWPHQWWR initial and ongoing staff training. The adoption of new software required for training the private entrepreneurs and the trainers, and the initial intensive training designed and delivered through workshops and seminars, GHYHORSHGWKHVWDIINQRZOHGJHWRDSUR¿FLHQW level by promoting individual skill sets. This has also made the entrepreneurs and the trainers embrace the sudden technological change in the district with willingness and enthusiasm.
•
•
Government conceding the plea of local bodies and revoking the transfer orders: When the Akshaya pilot project in MalapSXUDP GLVWULFW UHDFKHG WKH ¿QDO NLFNRII stage, transfer orders were issued to one of WKHLPSRUWDQWRI¿FLDOVZLWKRXWZKRVHSUHVence the project would have collapsed. The local bodies, then in unison, made a plea to the government to revoke the transfer orders at this crucial stage, which was immediately conceded. This constituted another critical success factor for the implementation of the project. Leadership: Senior administrators and HOHFWHGRI¿FLDOVVHWWKHSUHFHGHQWIRUWKHWHFKQRORJLFDO FKDQJH WKURXJK WKHLU µOHDGHUVKLS E\H[DPSOH¶7KH\ZHUHWKHHDUO\WHFKQRORJ\ adopters. This cultural change was one of the most important building blocks for success of the Akshaya Project.
CONCLUSION The study of implementation of the Akshaya Project LQ.HUDODUHYHDOHGWKHIROORZLQJVDOLHQW¿QGLQJV which will give important feedback that will be useful to all those who are concerned with this type of project: •
•
The wide gap in basic awareness about the computer world / IT world is seen among males and females, married and unmarried LQ GLIIHUHQW UHJLRQV 7R EULGJH WKH µDZDUHQHVV GLYLGH¶ µDUHDZLVH SXEOLF DZDUHQHVV FDPSDLJQV¶FDQEHDUUDQJHG)RUWKLVSXUSRVH the services of educated unemployed males and females of the region can be effectively used. Household surveys constitute the major µVRXUFHRILQIRUPDWLRQ¶IRUIHPDOHVDQGWKRVH married in both developed and developing regions, and for males in developing areas; while the other sourceslike the Internet, newspapers, friends, and so forthalso play DVLJQL¿FDQWUROHIRUPDOHVLQGHYHORSHGUHgions. Therefore, for introduction of any new
43
Bridging the Digital Divide Through Broadband Deployment
•
•
•
•
•
•
44
project/system, to create awareness, directly contacting the household particularly in rural areas will have better impact. A high proportion of females in more developed regions like district headquarters and males in less developed regions like villages are expressing that they are attending the Akshaya training out of compulsion rather than willingness. A separate strategy needs to be adopted to create inspiration among these groups prior to launching the project to enable them to accept the use of the IT services offered. Since females and males together express WKDWWKHH[SHULHQFHLQWKH¿UVW$NVKD\DFODVV is discouraging as the course content was GLI¿FXOW WKH FRQWHQW RI WKH FRXUVH PD\ EH PRGL¿HGWRVXLWDJURXS¶VVSHFL¿FLQWHUHVWV such as games or spiritual concepts, to kindle the interest to understand the simplicity of working with the computers Since different groups of citizenslike married, unmarried, males, and femalesopt for different timings for the training classes, organization of Akshaya classes should be VXLWDEOHWRWKHVSHFL¿FUHTXLUHPHQWVRIWKHVH groups It is also revealed that though there were many members in the family who were willing to attend the Akshaya classes, as the number per family is restricted to one, families decided on one person for the training, out of unity of understanding. Scope for allowing more members from each family could be explored to avoid cutting out an interested member. Varied perceptions are responsible for the WUDLQHHV¶ DWWHQGDQFH LQ WKH $NVKD\D WUDLQing, as revealed by the study. If the goal of launching the project is made clear prior to its launching, there will be a homogeneous group of trainees and the trainers. Since the majority of the females, males, married, and unmarried are using the knowledge acquired through Akshaya training only to inspire others to learn more about computers, and a smaller proportion of these groups are using the knowledge for job creation,
•
•
•
enhancement of knowledge, and so on, there should also be a strategy included in the project at every stage of implementation to make trainees aware of the various avenues available for making use of the knowledge acquired Another interesting feature of the Akshaya Project is that some females and males in both the districts where the project is under implementation are able to generate income ranging from Rs. 1,000 to Rs. 2,000 by using the knowledge acquired. While some are hiring out the systems to their neighbors for rent, some others are purchasing a system, appointing trainers, and converting a room in their house as a computer center. Analysis of aspirations revealed that the majority of trainees want to become the total EHQH¿FLDULHVRIWKH$NVKD\D3URMHFW,WLVOHIW WRWKHSROLF\PDNHUVWRDGRSWDÀH[LEOHVWUDWHJ\ that would unify the goal of the government and the aspirations of the citizens. Analysis of the suggestions made to the government brought to focus that all females and males in both the areas feel that e-vidya training, which is the plan of action for the second phase of implementation, can be consolidated with the basic training of the ¿UVWSKDVHDQGRIIHUWKHVDPHE\LQFUHDVLQJ WKHQXPEHURIFODVVHVRIWKH¿UVWSKDVHRIWKH project
7KHµOHDGHUVKLSE\H[DPSOH¶SROLF\RIWKH6WDWHV of Andhra Pradesh and Kerala is serving as motivation for other areas to embrace the technological transformation and seek ways to utilize technology WRGULYHKRPHWKHHI¿FLHQFLHV7KHDGRSWLRQRID citizen-centric approach by the planners, and the involvement of local bodies and private partners in the entire functioning of the projects, contributed WRWKHVXVWDLQDELOLW\RIWKHSURMHFW¶VLPSOHPHQWDtion. The dependence on locally available educated unemployed youth in the State of Kerala has sown the seeds for generation of income through self-employment, besides promoting the entrepreneurship development in the area.
Bridging the Digital Divide Through Broadband Deployment
For citizens, the accessibility of services through broadband technology has brought an opportunity to become part of the current knowledge revolution, besides bringing about a great technological transformation to the areas where it is implemented; it has thus contributed to bridging the digital divide. Moving from a manual to electronic process with broadband technology as an enabler, the states set an example, which will serve as guide for application of broadband in various geographical settings.
REFERENCES Agarwal, B.B. (2001, March 19). Faculty scarcity at IIT threatens knowledge capital. Retrieved from http://www.ciol.com/content/news/ trends/10103902.asp Arora, A., Arunachalam, V.S., Asundi, J., & Fernandes, R. (2000). The globalization of software: The case of the Indian software industry. Report submitted to the Sloan Foundation, Carnegie Mellon University, USA. Retrieved from http://www. heinz.cmu.edu/project/india/publications.html Davison, E., & Cotten, S.R. (2003). Connection discrepancies: Unmasking further layers of the digital divide. First Monday, 8(3). Retrieved from KWWS ZZZ¿UVWPRQGD\RUJLVVXHVLVVXHKWPO Lenhart, A., Horrigan, J., Allen, K., Boyce, A., 0DGGHQ0 2¶*UDG\( The ever-shifting Internet population: A new look at Internet access and the digital divide. Washington, DC: The Pew Internet and American Life Project. Hambly Odame, H. (2004). Gender and the Information Society. CTA Highlights. Wageningen, CTA. Keniston, K. (2003). The four digital divides. New Delhi: Sage. Modoux, A. (2002). The ‘digital divide’ could lead to a gigantic ‘cyber ghetto’ in developing countries. Geneva.
Olorunda, O., & Olorunda, A. (2006). Bridging the digital dividethe social and cultural impact of VoIP in developing countries: Nigeria as a case study. 3URFHHGLQJVRIWKH3DFL¿F7HOHFRPPXQLFDtion Council 37&¶ +DZDLL OECD. (2005). OECD broadband statistics. Retrieved November 30, 2005, from http://www. oecd.org/document/39/0,2340,en_2649_201185_ 36459431_1_1_1_1,00.html Pierce, J.C. (2003). Technology and social inclusion: Rethinking the digital divide. Comparative Technology Transfer and Society, 2(2), 218-220. Radhakumari, C. (2006, December 15-17). Akshaya a grassroots level IT project in Keralaa unique experiment with broadband. Proceedings of the International Conference on E-Governance, New Delhi. Radhakumari, C. (2006, December 15-17). IT as a source of entrepreneurship development in rural AreasAkshaya Project in the State of Kerala, India. Proceedings of the International Conference on E- Governance, New Delhi. SocialText.com. (2006). Broadband in India. Retrieved November 24, 2006, from http://www. socialtext.net/broadband/index.cgi?india TRAI (Telecom Regulatory Authority of India). (2006a). Telecom Regulatory Authority of India. Retrieved November 30, 2006, from http://www. trai.gov.in TRAI. (2006b). Study paper no.2/2006 on analysis of Internet and broadband tariffs in India. Retrieved November 30, 2006, from http://www. trai.gov.in/StudyPapers_list_year.asp
KEY TERMS Broadband for A.P: High-speed data transmission. For implementation of the A.P Broadband Project, the capacity to deliver a symmetrical bandwidth of at least 512 kbps with a guaranteed latency of less than 100 milliseconds and less
45
Bridging the Digital Divide Through Broadband Deployment
than 1% packet loss, to the end user, with the currently available technologies, is considered µEURDGEDQG¶ ISP Services: The broadband access service ZLOOEHDEOHWRÀH[LEO\SURYLGHXVHUDFFHVVWRDZLGH unrestricted number of ISPs and ASPs providing JHQHUDODQGVSHFL¿FVHUYLFHVHOHFWURQLFDOO\ Minimum Bandwidth: The broadband access service and customer network equipment ID supporting the requirement for delivery of symmetrical 512kbps service.
46
PURA: Providing Urban Facilities for Rural Areas. Voice Services (VoIP): The broadband access service will be capable of the carriage of VoIP WUDI¿FXVLQJDQDSSURSULDWHOHYHORI4R66XLWDEOH customer network equipment will be made available to support the delivery of VoIP; however, the provision of customer premises equipment is not to be an integral part of the scope of the project.
47
Chapter IV
Broadband Policy, Market Competition, and User Adoption in Taiwan Yu-li Liu National Chengchi University, Taiwan
ABSTRACT 7DLZDQKDVEHHQPDNLQJJUHDWHIIRUWVWRSURPRWHWKHIXO¿OOPHQWRILWV1DWLRQDO,QIRUPDWLRQ,QIUDVWUXFture (NII), with broadband network technology and its application having rapidly developed in recent years. In relation to the two major forms of broadband network technology, cable modem and ADSL, two multiple system operators—namely, Eastern Multimedia Company and China Network System—provide broadband network via cable modem, and HiNet and Seednet are the major ADSL service providers. 7KUHHQHZ¿[HGQHWZRUNVKDYHRIIHUHG$'6/VHUYLFHVDVZHOOVLQFHWKH\EHJDQWKHLURSHUDWLRQV7KHUH DUHPLOOLRQKRXVHKROGV ++V RURIWKHKRXVHKROGSRSXODWLRQLQ7DLZDQXVLQJWKH,QWHUQHW DQGPLOOLRQKRXVHKROGVRURIWKH++SRSXODWLRQXVLQJEURDGEDQG 7:1,& 7KLV chapter analyzes the broadband adoption, the competition among providers of broadband, and relevant policies in Taiwan. The research methods adopted include literature reviews, in-depth interviews, and secondary analysis of previous surveys conducted by TWNIC. Since Taiwan’s broadband penetration is WKH¿IWKKLJKHVWLQWKHZRUOGWKHGLVFXVVLRQVRIWKHPDMRUIDFWRUVFRQWULEXWLQJWREURDGEDQGGHSOR\PHQW in this chapter might provide some experiences from which other countries can learn.
INTRODUCTION In May 2002, the ROC government announced its “Challenge 2008” Six-Year National Development Plan to transform itself into a high-tech service island. The six-year plan involves a budget of NT$2.6 trillion (approximately US$75 billion). One of the objectives of this plan is to develop a “Digital Taiwan,” aiming to promote e-life, e-
government, e-commence, and innovative digital LQGXVWULHVLQ7DLZDQ7RWKLVHQGWKHUHDUH¿YH goals that have been set for achieving this, with one of the most important aspects of the plan being to install broadband Internet in every household, with a view to reaching six million households by the end of 2007 (CEPD, 2002). 7KH IRXU ¿[HG QHWZRUNV SODQQHG WR LQYHVW D total of NT$364.4 billion (US$10.5 billion) in
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband Policy, Market Competition, and User Adoption in Taiwan
broadband deployment between 2002 and 2007. The Directorate General of Telecommunications (DGT), the former telecommunications regulator, had been asked to play an active role in opening up the telecommunications market, deregulating the unnecessary rules and controls, lowering the entry barriers, maintaining market order, providing a stable and competitive environment, and removing the obstacles to the broadband infrastructure. However, there exist doubts about whether the goals can be achieved, since it may be impossible to reach 82.36% of the 7.28 million households in Taiwan, in view of the fact that some of the households might not consider it necessary to subscribe to broadband. The purpose of this chapter is to: 1.
Describe the development and current status of broadband in Taiwan 2. Analyze broadband adoption and user demand 3. Analyze broadband strategies among providers $QDO\]HWKHJRYHUQPHQW¶VEURDGEDQGSROLcy 5. Discuss the obstacles to the deployment of broadband 6. Propose solutions to facilitate a more rapid deployment of broadband. While broadband includes ADSL, cable modems, ZLUHOHVV /$1V OHDVHG OLQH RSWLFDO ¿EHU and so forth, this chapter focuses mainly on ADSL and cable modems.
THE BROADBAND DEPLOYMENT According to a Point-Topic report conducted in 0DUFK7DLZDQKDVWKH¿IWKKLJKHVWEURDGEDQG household penetration in the world, only being surpassed by South Korea, Hong Kong, Iceland, and Israel (Cox, 2006). In July 2006, there were 5.25 million households (HHs) using the Internet, representing a 72.11% penetration, and 4.74 million households using broadband, or 65.05% of the HH population (TWNIC, 2006).
48
In Taiwan, the two major forms of broadband network technology are the cable modem and ADSL, with cable modem services having been in operation since the end of 1998. ADSL services started one year later than the cable modem services. Before the government opened up the ¿[HG QHWZRUNV &KXQJKZD 7HOHFRP &+7 WKH incumbent operator, monopolized the telecommunications market and hesitated to promote an ADSL service, fearing that the new broadband service might hurt leased line revenues from the business sector. When the government opened XS WKH ¿[HG QHWZRUNV LQ &+7 UHDOL]HG that it could not postpone the ADSL service any longer. After ADSL came on the market, it grew very fast. By the end of 2000, ADSL penetration surpassed that of cable modem. There were four main reasons for this: 1.
The NII established a penetration goal in July 2000 to reach three million users in three years. 7KH JRYHUQPHQW RSHQHG XS WKUHH ¿[HG networks in March 2000, all of which subsequently provided ADSL service. 3. The residential demand for broadband steadily increased. 4. Many ISPs started to provide ADSL services. Since competition within the broadband market KDVEHFRPHLQFUHDVLQJO\¿HUFHWKHFDEOHPRGHP providers have hesitated to invest more money to upgrade their one-way infrastructure to a two-way infrastructure. The situation resembles that of the chicken and the egg. If the cable modem providers do not upgrade their infrastructure, the subscribers will lose patience and might convert to ADSL services. A one-way cable modem is not only slow but also costs more, because the subscribers still have to pay the circuit fees. GigaMedia (owned by one of the multiple system operators, or MSOs) decided to provide ADSL service in addition to cable modem service in March 2001. In September 2002, there were only 241,495 cable modem subscribers. When the cable operators began to upgrade their systems to two-way,
Broadband Policy, Market Competition, and User Adoption in Taiwan
they had more cable modem subscribers. Eastern Multimedia Co., Ltd. (EMC) and China Network System Co., Ltd. (CNS) were the two MSOs providing broadband network services via cable modem. The two largest MSOs were both acquired by foreign companies in 2006. In September 2006, 3.96 million households subscribed to ADSL, while 415,962 subscribed to cable modem services (NCC, 2006). ADSL occupies 89% of the broadband market, while cable PRGHPLV 7KHLQFXPEHQW&+7¶V+L1HWKDV the largest share of the ADSL market (80.9%). 7KHWKUHHQHZ¿[HGQHWZRUNVDQGRWKHU,63VKDYH GLI¿FXOW\FRPSHWLQJZLWK+L1HW7KHLUWRWDOVKDUH of the ADSL market is only 19.1%. In September 2002, the non-HiNet broadband providers had a 34.07% market share. However, in September 2006, their market share dropped to 28.06%. In addition to the above two kinds RIEURDGEDQGXVHUVWKHUHDUH¿EHUXVHUV 28,541 leased line users, and 52,347 PWLAN users in Taiwan (FIND, 2006; NCC, 2006).
MAJOR FACTORS CONTRIBUTING TO BROADBAND DEPLOYMENT When broadband penetration in South Korea became the highest in the world, a great number of policymakers, industrialists, and scholars ZDQWHGWR¿QGRXWZKDWWKHNH\VXFFHVVIDFWRUV were. They found out that the key success factors
are government policy, industry competition, consumer demand, housing pattern, and cyber EXLOGLQJ FHUWL¿FDWH V\VWHP /HH &KDQ2OPsted, 2004). Broadband penetration in Taiwan ZDVWKH¿IWKKLJKHVWLQWKHZRUOGLQ0DUFK therefore, it is also important to know the key factors contributing to its high penetration. Strictly speaking, Taiwan cannot be considered to have been successful, for it still has some barriers to broadband deployment. Nevertheless, we can at this moment identify the factors that contributed to the rapid deployment of broadband in Taiwan, namely, the broadband policy, market competition, DQGFRQVXPHUV¶GHPDQG
Broadband Policy With regard to broadband policy, six measures are regarded as important to promote the deployment of the broadband network (Chien, 2003; Kao, 2003), as follows. Telecommunications Liberalization The process of telecommunications liberalization in Taiwan can be shown to consist of four stages (Chien, 2003; Hsieh, 2003): a.
b.
First Stage (1987-1995): Opening up customer premises equipment (CPE) and valueadded services Second Stage (1996-1998): Revising the Telecommunications Law and making DGT solely
Table 1. The growth of broadband users in Taiwan (NCC, 2006) ADSL
Cable
Date
Leased Line
PWLAN
Total
Non-HiNet
HiNet
Others
Modem
09/02
1,220,575
376,906
241,495
18,717
2,075
1,859,768
34.07%
09/03
1,806,177
515,682
446,132
21,392
7,750
2,797,133
35.24%
09/04
2,324,460
702,666
504,881
22,218
32,437
3,586,662
35.07%
09/05
2,843,267
787,574
559,910
21,706
41,330
4,253,787
33.07%
09/06
3,202,716
756,839
415,962
28,541
52,347
4,456,405
28.06%
Market Share
49
Broadband Policy, Market Competition, and User Adoption in Taiwan
c. d.
a governing agency and CHT a company that is heading toward privatization, and opening up the mobile and satellite communications markets Third Stage (1999-2000): 2SHQLQJXS¿[HG networks and international cable leasing Fourth Stage (July 2001-2003): Opening up international simple resale (ISR) and the 3G mobile phone market
The ROC government propagated two major Telecommunications Liberalization White Papers to liberalize the telecommunications industry .DR 7KH¿UVWRQHZDVUHOHDVHGLQ and was aimed at providing a fair and competitive environment; promoting technological progress; LPSURYLQJLQGXVWULDOHI¿FLHQF\ SURYLGLQJKLJK quality, diverse, and cheaper telecommunications services; and triggering economic development. The second Telecommunications Liberalization White Paper was propagated in 2002, and was aimed at constructing an international telecommunications environment, promoting universal service, improving the digital divide, promoting competition, providing creative and high-quality telecommunications services, and helping the development of the telecommunications industry (Chien, 2002; Hsieh, 2003). Revising the Telecommunications Law The 7HOHFRPPXQLFDWLRQV/DZZDV¿UVWLQWURGXFHG in 1958, and was later revised in 1977, 1996, 1998, DQG 7KHVLJQL¿FDQFHRIWKH YHUVLRQZDV that it separated the DGT from the CHT, meaning that the DGT would become a pure governing agency and would not manage CHT anymore. CHT has since been ordered to become privatized. In WKH ODZ ZDV UHYLVHG VLJQL¿FDQWO\ WR FRSH with the new telecommunications environment. For example, the foreign ownership restrictions were relaxed in order to encourage foreign investment. For Type I service, the direct foreign ownership ceiling was raised from 20% to 49%, while the indirect foreign ownership ceiling was raised to 60%. In addition, network interconnections constituted an important measure that was stipulated in
50
the 1999 law. The revised Telecommunications Law asked the dominant player CHT to rent its local loop to its competitors so that the consumers ZRXOG KDYH WKH FKDQFH WR FKRRVH WKH QHZ ¿[HG networks. Encouraging Competition The Telecommunications Liberalization White Paper and the Revised Telecommunications Law emphasized competition by opening up the telecommunications market and applying asymmetrical regulations such as network interconnections and unbundling that were imposed on the dominant player CHT. Public Land and Buildings The government also gave public land and buildings priority in order to establish the needed conduit infrastructure. It revised the rules regarding the installation of telecommunications facilities both inside and outside residential buildings in December 2001. It assured consumers that they had the right to choose a local telephone company. If Type I telecommunication services needed to pass through public or private land or buildings, or cable needed to be placed underground, the central and local governments were required to provide assistance. The application process was also shortened. Assistance in Providing Access to the Right of Way 7KHQHZ¿[HGQHWZRUNVVWLOOKDYHSUREOHPVZKHQ it comes to applying for the right of way from local governments. Therefore, the central government has formulated regulations regarding conduit sharing that the local government should adopt. Set Deployment Goal In 2002, the ROC government released the “Challenge 2008” National Development Plan, which consists of 10 projects. One of them involves setting a broadband deployment goal, whereby there are to be six million broadband household subscribers by the end of 2007. The “e-Taiwan” project launched at the end of 2001 has also been incorporated into it. Other goals included:
Broadband Policy, Market Competition, and User Adoption in Taiwan
1. 2. 3. 4. 5. 6. 7.
50% of the population using the Internet 70% of Internet subscribers using broadband A national backbone capacity of 1,250 Gbps An international backbone capacity of 250 Gbps Fiber-to-the-curb (FTTC) penetration reaching 95% Striving for next-generation broadband networks such as IPv6 and wireless LAN Allocating a NT$10.7 billion (US$0.3 billion) budget over six years to develop digital content industries (CEPD, 2002)
Market Competition and Broadband Strategies CHT is the incumbent operator of broadband in WHUPVRI¿[HGQHWZRUNVLQ7DLZDQ$OWKRXJKWKH government has been trying to privatize CHT, it is not yet completely privatized. More than 35% of the shares are still owned by the government. Its subsidiary HiNet provides ADSL service to both residents and the business sector, and dominates WKH '6/ PDUNHW EHFDXVH RI D ODVWPLOH EHQH¿W DQG LWV IRUPHU FKDLUPDQ 0DR &KLK.XR¶V PDQagement agenda. Before Mao Chih-Kuo came to CHT, he was the deputy minister of the Ministry of Transportation and Communications (MOTC). In that capacity, he had the opportunity to become well acquainted with the practices of the telecommunications industry. Realizing that the average revenue per user (ARPU) in the case of ADSL was higher than that in the case of local phone calls, KHDVNHG&+7¶VPDQDJHPHQWWHDPWRIRFXVPRUH RQ+L1HW¶V$'6/GHSOR\PHQWDQGPDUNHWLQJ+H set a goal to reach one million ADSL subscribers in 2001 and two million in 2002. He almost reached his goal (Chen, 2003), for there were over two million ADSL subscribers in January 2003. &+7¶VVWUDWHJ\KDVEHHQWRIRFXVRQEUDQGLQJDQG reliability, and so their ADSL price is higher than that of the ISPs. Although HiNet dominates the local loop, there are many other broadband providers such as the
QHZ¿[HGQHWZRUNV,63VDQGFDEOHPRGHPSURYLGers. The broadband strategies they adopt include promotion, bundling, advertisements, and price competition. They often use price competition and promotion strategies in tandem. Seednet is an ISP whose strategy is to focus on good service and lower prices. $VLD3DFL¿F%URDGEDQG7HOHFRPP $3%7 LV DQHZ¿[HGQHWZRUNWKDWIRUPHUO\KDGDIDPLO\ DI¿OLDWLRQZLWKWKHFDEOH062(0&,WDFTXLUHG DQ,63QDPHG$VLD3DFL¿F2QOLQH $32/ DQGEHcame the second largest provider in the Taiwanese broadband market. It formerly provided both ADSL and cable modem services. Its broadband strategy is to attract subscribers with its broadband content by offering online games and cheaper rates (Jeng, 2003). However, it sold its cable modem subsidiary back to EMC in 2006. GigaMedia not only provides cable modem service but also provides ADSL service. Because it hesitated to invest in upgrading to two-way cable in 2001-2002, it lost some of its one-way cable modem subscribers. As of August 2003, it had only 80,000 cable modem subscribers. Perceiving the drawbacks of one-way cable modem service, it no longer promotes one-way cable. In relation to its cable modem service, it provides a wide variety of cheap packages such as 512 k-64k, 1.5M-128K, and 6M-256K; for ADSL service, it provides 512 k-64k and 1.5M-384K. It provides DSL (512 k-512k) service as well (Chiang, 2003). 2WKHU QHZ ¿[HG QHWZRUNV VXFK DV 7DLZDQ Fixed Network (TFN) and SPARQ (New Century InfoCom Tech Co., Ltd.), because they are new entrants, still have problems regarding access to the last mile. They prefer to promote broadband to small and medium businesses and also Hi-building. Other ISPs such as Seednet and SoNet also promote ADSL actively. For instance, SoNet encouraged HiNet subscribers to switch to its service with gifts such as computers and digital cameras. In sum, the ADSL service area covers the whole country, while cable modem providers have designated service areas and thus have limited coverage. Additionally, people have more choices among ADSL providers. They can choose from IRXU¿[HGQHWZRUNVDVZHOODVRWKHU,63V,ISHRSOH
51
Broadband Policy, Market Competition, and User Adoption in Taiwan
Table 2. Comparison of cable modem and ADSL services Cable Modem
ADSL
Designated service area, limited coverage
Nationwide service area
Less choice
More choice
Not fully two-way
Chunghwa Telecom (HiNet) dominates last mile
One-way cable modem has to pay for the circuit fee
Interconnection takes time
Negative image
Better image
/HVV¿QDQFLDOUHVRXUFHV
0RUH¿QDQFLDOUHVRXUFHV
&DEOH79¶VDJHQGD WULSOHSOD\
+L1HW¶VWRSDJHQGD QHWZRUNHIIHFW
want to subscribe to cable modem, their choices are very limited, because there are only one or two cable TV operators in most cable TV-designated service areas. Since CHT at one time was the only telephone company providing local and long distance phone calls, people are more familiar with CHT than with the cable operators. Although CHT once monopolized the telecommunications market, people had fewer complaints regarding telephone companies than the cable operators. There are only two major cable modem providers, and some of their service areas are not fully two-way deployed. One-way cable modem subscribers have to pay the circuit fee. These subscribers are both time sensitive and price sensitive. When they cannot accept their phone bills, they tend to quickly switch to ADSL service. In recent years, the cable TV systems have been upgrading their services to two-way. It is obYLRXVWKDWWKH¿[HGQHWZRUNVKDYHPRUH¿QDQFLDO resources, but interconnection for the new entrants takes time. In short, the broadband providers have been trying to target different people by providing different packages and bundling strategies (Liu & Chen, 2002).
Consumer Demand: Broadband Adoption The following discussion is based on the secondary analysis of four surveys conducted by TWNIC via telephone between July 2003 and July 2006.
52
The valid samples of the four surveys were 5,523, 3,123, 3,059, and 3,192 (TWNIC, 2006).
7KH%URDGEDQG'HPRJUDSKLF3UR¿OH According to a TWNIC 2006 survey, men (63.82%) use broadband more than women (61.91%), and those people between the ages of 16-20 and 2125 tend to use broadband more than people of other ages (TWNIC, 2006). A total of 85.51% of the economic head of the broadband households KDYHDPDVWHU¶VGHJUHHDQGDERYHIROORZHGE\D university degree (79.96%).
Broadband Users’ Behavior Among the broadband users, people mostly use broadband at home (92.9%); 36.93% are at work, and 23.56% are at school. During the workday, 17.79% of the respondents spent 1-2 hours using broadband daily, while 15.9% spent 2-3 hours. With regard to the factors considered by the users in choosing the broadband suppliers, a survey conducted in July 2003 indicated that the major factors included fast speed (26.73%), cheaper price (20.03%), and stability (18.61%) (TWNIC, 2003). However, in July 2006, low cost was considered as the most important factor, followed by convenience, fast speed, and stability (TWNIC, 2006). In Taiwan, the local phone fee does not consist RIDÀDWUDWHDQGSHRSOHDUHFKDUJHGIRUWKHWLPH
Broadband Policy, Market Competition, and User Adoption in Taiwan
Table 3. Factors considered in choosing broadband suppliers Source: TWNIC (2003-2006)
1
2
3
4
July 2003
July 2004
July 2005
July 2006
Speed
Speed
6XSSOLHU¶V5HSXWDWLRQ
Cost
26.73%
19.84%
17.76%
18.68%
Cost
Cost
Cost
Convenience
20.03%
16.9%
15.03%
17.56%
Stability
6XSSOLHU¶V5HSXWDWLRQ
Stability
Speed
18.61%
12.88%
14.10%
17.23%
High Reliability
Friends or Relative
Speed
Stability
16.28%
Recommendation
12.52%
14.03%
9.84% 5
Friends or Relative
Stability
Friends or Relative
6XSSOLHU¶V5HSXWDWLRQ
Recommendation
9.08%
Recommendation
12.08%
16.19%
10.10%
Table 4. The main purposes for using broadband (multiple choice) (TWNIC, 2003-2006)
1
2
3
4
July 2003
July 2004
July 2005
July 2006
WWW
WWW
WWW
WWW
82.91%
71.71%
63.52%
70.99%
E-mail
E-mail
E-mail
E-mail
63.88%
50.63%
43.82%
47.5%
Online game
Online game
Search information
Search information
24.76%
18.11%
29.59%
34.38%
Download
Chat (MSN)
Online game
Chat (MSN)
15.83%
16.21%
21.16%
25.77%
they use. Therefore, it is very expensive to use dialup or one-way cable modem for a long period of time. One-way cable modem subscribers are time sensitive and price sensitive. When they no longer appreciate the speed of dial-up, they will be easily attracted by the promotion of broadband. What do people do with broadband? The two main purposes for using broadband from 2003 to 2006 were browsing Web pages and using e-mail. It is important to realize that broadband users do not utilize the characteristics of broadband very well, because they do not emphasize the use RI GRZQORDGLQJ DXGLR DQG YLGHR ¿OHV 0RVW RI the broadband users do not want to pay for the
broadband content. They think they can easily download the unlicensed music or movies free of charge (Chen, 2003).
Problems Encountered by Broadband Users Broadband users are still confronted by some problems. Those they encountered most were slow GRZQORDGVSHHGLQVWDELOLW\DQGWUDI¿FFRQJHVWLRQ during peak hours (TWNIC, 2006). Because cable modem users have to share the cable modem capacity in the same hub with other users, the chances RI WKHLU EHLQJ IDFHG ZLWK WUDI¿F FRQJHVWLRQ DQG
53
Broadband Policy, Market Competition, and User Adoption in Taiwan
Table 5. Problems encountered by broadband subscribers (TWNIC, 2003-2006)
1
2
July 2003
July 2004
July 2005
July 2006
Slow download speed
Slow download speed
Instability
Slow download speed
54.46%
26.13%
20.73%
20.65%
Instability
Instability
Slow download speed
Instability (poor connection
38.88%
19.7%
20.29%
quality)
Slow upload speed
Slow upload speed
7UDI¿FFRQJHVWLRQGXULQJ
7UDI¿FFRQJHVWLRQGXULQJ
25.59%
13.4%
peak hours
peak hours
10.89%
13.14%
Too expensive
Worries over virus
Worries over virus
Slow upload speed 12.01%
19.8%
11.03%
9.49%
19.82% 3
4
Table 6. Reasons for not subscribing to broadband (TWNIC, 2003-2006)
1
2
3
July 2003
July 2004
July 2005
July 2006
No demand
No demand
No demand
No demand
39.76%
41.04%
49.25%
60.01%
Lack of time
Lack of hardware
Lack of hardware
Lack of hardware
29.63%
30.87%
22.78%
15.25%
Too expensive
Lack of time
Lack of time
8VHU¶VDJHLVWRRROGRUWRR
21.47%
8.09%
10.01%
young 14.55%
slower speeds are higher than for ADSL users. Theoretically speaking, the ADSL infrastructure is more secure than that of the cable modem. However, it seems that ADSL users are more worried about Web security than cable modem users (Liu & Chen 2002). Among the broadband subscribers, only 9.16% ZHUHQRWVDWLV¿HGZLWKWKHEURDGEDQGVHUYLFHV$V for the reasons why some dial-up users have not adopted broadband, the answers are no demand and lack of hardware (TWNIC, 2006).
BROADBAND CONTENT 'XULQJWKH¿UVWVWDJHRIEURDGEDQGGHSOR\PHQW fast speed was important to the wide deployment
54
of broadband. At that time, the broadband operators provided many promotions regarding speed, and they lowered their prices to attract users. For instance, when CHT started to provide ADSL service in 1999, its online fee for 512 K was NT$1,299 (US$39) per month; it lowered its price to NT$499 (US$15) (Chen, 2003). The circuit fee was NT$595 in 2003; it was NT$410 for 1 M in February 2007. (The subscribers must pay for both online fees and circuit fees each month). Since the penetration of broadband household reached 65% in July 2006, the growth of broadband subscribers will slow down. Therefore the content of broadband will be considered the next important driving force. webs-tv and hiChannel are the two major broadband content aggregators developed in 1999 and 2000. As they grew in the market, the telephone
Broadband Policy, Market Competition, and User Adoption in Taiwan
companies and ISPs started to promote high-speed Internet. The reasons they could grow included: (1) the large bandwidth of the content aggregators, (2) the promotion of broadband connections, and (3) the support of the content providers (Liu, 2006). hiChannel is owned by CHT and Elta. In July 2005, it had 90 partners and 150 channels, including terrestrial channels, cable channels, and other movie channels. Its business model is to split the revenues with the content providers on the basis of a 3:2:5 ratio (CHT, Elta, content providers). CHT is responsible for the video server, storage, bandwidth, and billing; Elta is responsible for aggregating the content and digitizing the content; and the ICP is required to provide the content with the copyright &KHQE 7KH¿[HGQHWZRUNVDQGWKHFDEOH modem operators each have their own Web sites. They all provide content, but are not as attractive as hiChannel. hiChannel is an open platform and KDVWKHDGYDQWDJHRI&+7¶VODVWPLOHDQGEUDQGLQJ Thus, many content providers are willing to come to this platform because of the network effect. webs-tv positioned itself as a Web multiple system operator. It has formed strategic alliances ZLWKVHYHQSDUWQHUVLQFOXGLQJ¿[HGQHWZRUNVDQG ISPs such as anet.tv, sparq.tv, GigaTV, SonetTV, BTV, avplus.tv, and estv. webs-tv is responsible for aggregating the contents and managing the VHUYLFHV7KH¿[HGQHWZRUNVDQG,63VRQO\VXSport the bandwidth and connection. The users can choose any ISP to receive the broadband content service. In order to provide better quality content, webs-tvKDVORFDWHGLWVVHUYHUVLQHDFKSDUWQHU¶V territory (Chen, 2003a).
OBSTACLES TO BROADBAND DEPLOYMENT In order to understand the obstacles to broadband deployment, the author conducted several focus groups between 2001 and 2005. In late November 2006, the author interviewed the vice presidents of WKUHHQHZ¿[HGQHWZRUNVDQGWKHJHQHUDOPDQDJHU of an ISP, SeedNet, to update the status quo of the broadband development. The author found that the problems ADSL providers are confronting now
are the same with the problems they confronted ¿YH\HDUVDJR ODFNRIWKHODVWPLOH ORFDO loop unbundling, (3) wholesale price, and (4) local government attitudes. &+7¶V+L1HWDFFRXQWVIRUPRUHWKDQRI WKH$'6/PDUNHWVKDUHZKLOHWKHRWKHUWKUHH¿[HG networks only have less than 3% of the ADSL market individually (TWNIC, 2006). The main UHDVRQIRUWKLVLVWKDWWKHQHZ¿[HGQHWZRUNVODFN the last mile. It is dominated by CHT. There have been many disputes between CHT and the new ¿[HGQHWZRUNV(YHQWKRXJKWKHUHYLVHG7HOHFRPmunications Law stipulates that the dominant CHT must provide “network interconnection” and lease LWVODVWPLOHWRWKHQHZ¿[HGQHWZRUNV&+7KDV been trying to postpone renting its local loop. In WKHEHJLQQLQJ&+7¶VH[FXVHZDVWKDWLWKDGWREX\ new equipment and upgrade its facilities in order to provide network interconnection . Although CHT is to be privatized, it is still partially state-owned (35%), and its budget has to be approved by the Legislative Yuan. For this reason, it takes time to upgrade its facilities. The former regulator (DGT) WULHGWRDUELWUDWHEHWZHHQ&+7DQGWKHQHZ¿[HG networks, but the contract disputes over rental fees, termination terms, and maintenance responsibility have not yet been resolved. Therefore, the three ¿[HGQHWZRUNVKDYHOHVVWKDQRIWKHORFDOSKRQH VHUYLFH0RVWRIWKHQHZ¿[HGQHWZRUNVVWLOOKDYH to use the HiNet local loop when they provide ADSL service. The second and third obstacles are related to the ¿UVWREVWDFOH:KHQWKHQHZ¿[HGQHWZRUNVDQG RWKHU,63VZDQWHGWRXVH&+7¶VORFDOORRS&+7 did not apply local loop unbundling. Therefore, the applicants were forced to pay more fees than they expected. As for the wholesale price, the new ¿[HGQHWZRUNVDQGRWKHU,63VDOOFRPSODLQHGWKDW &+7¶VZKROHVDOHSULFHIRUWKHORFDOORRSDQGWKH connection fees were too high. The new regulator National Communications Commission (NCC) has been devoted to these issues since its establishment in February 2006. It set up a range for CHT to drop its wholesale price in November 2006 (http://www. QFFWZ :KHWKHUWKHSROLF\ZLOOKHOS&+7¶VFRPpetitors to develop their broadband services, it is still premature to conclude.
55
Broadband Policy, Market Competition, and User Adoption in Taiwan
The fourth obstacle is that local governments RIWHQDVNWKHQHZ¿[HGQHWZRUNVWRKHOSWKHORFDO interests if they want to deploy lines and engage in FRQVWUXFWLRQORFDOO\7KHQHZ¿[HGQHWZRUNVQHHG to apply for licenses or documents from the local government if they wish to dig roads or deploy lines either on the ground or underground. If they do not “contribute” something to the local government, or local interests, they might be confronted ZLWKGLI¿FXOWLHV6HHLQJWKLVREVWDFOHWKHFHQWUDO government has instructed the local governments WRKHOSWKHQHZ¿[HGQHWZRUNVGHSOR\IDVWHU The cable modem providers are often confronted with four major problems: (1) the economies of scale are too small, (2) the service area is restricted to its franchise area, (3) upgrading to two-way cable PRGHPVLVH[SHQVLYHDQG WKHLQGXVWU\¶VLPDJH is more negative. First, the designated service areas for cable television in Taiwan consist of 47 areas. Each cable operator is limited to operating in its franchise area. Since Taiwan covers only 36,000 square kilometers and the population is only 22.5 million, the economies of scale to be achieved in cable television are very small. The same applies to the cable modem operators. Second, the cable modem service is provided by its MSO. Since the cable television service area is restricted to its franchise area, these service providers will usually not serve cable modem to the people who do not subscribe to their cable systems, because they will be confronted with too PDQ\GLI¿FXOWLHVUHODWHGWRPRQRSRO\HI¿FLHQF\ interconnection, and so forth (Chiang, 2003). Third, upgrading to a two-way cable modem is expensive. Since cable operators hesitated to invest money in order to upgrade to a two-way cable modem, some of their one-way cable modem subscribers switched to other services. Therefore, the big MSOs realized the importance of two-way services and started to upgrade their services. Fourth, cable operators formerly had a negative image, because they started to operate before the government granted them franchises. They were regarded as illegal operators before 1993 when the Cable Television Law was passed. In addition, there were many disputes between cable operators and cable channels regarding the program loyalty fees several years
56
ago. The subscribers do not feel very secure in subscribing to cable modem because of the negative past records of cable operators. In 2006, the two largest MSOs were acquired by foreign companies which guaranteed to the regulator NCC that they would continue to invest in upgrading their cable infrastructure.
SOLUTIONS PROPOSED BY THE GOVERNMENT AND THE BROADBAND INDUSTRY The biggest obstacle to broadband deployment is WKDWWKHQHZ¿[HGQHWZRUNVGRQRWKDYHHQRXJK time or budget to build the local loop throughout WKHZKROHFRXQWU\(LJKW\¿YHSHUFHQWRIWKHORFDOORRSRIWKHQHZ¿[HGQHWZRUNVLVUHQWHGIURP HiNet. In order to solve the problems arising from lacking a local loop, the National Information and Communications Initiative Committee (NICI), a cabinet-level committee, coordinated with the Ministry of the Interior (MOI) and the Ministry of Economic Affairs (MOEA) to propose the “M-Taiwan Program” with a budget of NT$37 billion (US$ 1.14 billion) from 2005 to 2009. The M-Taiwan Program is viewed as the key initiative WUDQVIRUPLQJ7DLZDQLQWRDµXELTXLWRXVQHWZRUN VRFLHW\¶ 816 17ELOOLRQZLOOEHDOORFDWHG WR EXLOG WKH VHFRQG EDFNERQH IRU WKH QHZ ¿[HG networks. NT$7 billion will be allocated to build up seamless wireless networks (http://www.nici. nat.gov.tw/). The local government is to maintain WKH IDFLOLWLHV DQG UHQW WKHP WR WKH SULYDWH ¿[HG networks. In this case, the local government can collect rents and have more incentives to help the ¿[HGQHWZRUNV)XUWKHUPRUHWKHQHZ¿[HGQHWworks will not have to wait to negotiate with CHT to get the local loop, for they can rent it from the local government directly. They think that this is a more effective means of encouraging competition. +RZHYHUDOOWKHQHZ¿[HGQHWZRUNVFULWLFL]HWKDW they do not need a second backbone. They only need the last mile to the users. Believing that “content is king,” the government has also propagated its digital content policy. It developed a sequence of digital content training
Broadband Policy, Market Competition, and User Adoption in Taiwan
programs to train the content specialists. In addition, in April 2003 the Executive Yuan decided to HVWDEOLVKDVSHFLDORI¿FHWRSURPRWHDQGGLUHFWWKH GLJLWDOFRQWHQWLQGXVWU\7KHRI¿FHKDVVL[GLYLVLRQV including software, telecommunications application, digital learning, video, animation, and games. The government and the broadband industry hope that the content providers can develop more killer applications of broadband content, which can act as an impetus for additional broadband subscription and deployment.
CONCLUSION Taiwanese experience of rapid broadband deployment indicates that government policy, industry promotion and competition, and the demands of consumers are the major factors contributing to broadband penetration. The consumers are interested in fast speeds and more bandwidth, but are also very concerned about the price. If the prices of broadband services offered by the different providers do not differ too much, the consumers will be more concerned about the branding and stability of the providers. Since the dominant CHT enjoys RIWKHORFDOORRSWKHQHZ¿[HGQHWZRUNVDQG RWKHU,63VFDQQRWFRPSHWHZLWK&+7¶V+L1HWLQWKH broadband market on an equal footing. Therefore, WKHUHJXODWRU1&&VKRXOGEHYHU\¿UPZLWKUHJDUG to executing the interconnection policy stipulated in the Telecommunications Law. To have six million household broadband subscribers by the end of 2007 is a goal set up in the Challenge 2008 National Development Plan. In September 2006, there were 4.45 million broadband households. Since the broadband users in some apartments and communities are not counted, the broadband subscribers should be closer to 4.72 million. It is expected to reach 5.3 million at the end of 2007. It does not seem to be easy to reach 82.36% household broadband penetration by the end of 2007. Now that the second backbone will not help much for the new broadband deployment, the most practical and effective way to solve the interconnection problem is to force CHT to rent its local
loop and provide reasonable wholesale prices to the QHZ¿[HGQHWZRUNVDQGRWKHU,63V0HDQZKLOHLW is also important to develop digital and broadband content. This might serve as a driving force for broadband penetration. However, most surveys indicate that the consumers are not ready to pay for the broadband content (Liu & Chen, 2002; TWNIC, 7KHUHDUHWZRPDLQUHDVRQVIRUWKLVWKH¿UVW being that the bandwidth is not broad enough to support the broadband content, and the second that WKHEURDGEDQGFRQWHQWLVQRWVXI¿FLHQWO\DSSHDOLQJ Therefore, there is much scope for improvement on the part of the broadband providers, namely, through their providing more bandwidth to the broadband subscribers and developing more killer applications of the broadband content.
REFERENCES CEPD (Council for Economic Planning and Development). (2002). Challenge 2008, National Development Plan, Executive Yuan. Retrieved August 25, 2006, from http://www.cepd.gov. tw/2008/challenge2008.pdf Chen, M. (2005a). CEO, webs-tv. Interview by author, January 17, 2005. Chen, S. (2005b). President of Elta. Interview by author, June 9, 2005. Chen, Y.C. (2003). Managing Director, Operations Management Department, Data Communication Business Group, CHT. Interview by author, August 10, 2003; February 13, 2007. Chiang, B. (2003). Technician, GigaMedia. Interview by author, September 1, 2003. Chien, J.T. (2003). How do Taiwanese telecommunications policy and regulations cope with WTO? Teaching Materials, National Chengchi University, Taiwan. Cox, T. (2006). World broadband statistics: Q1 2006. Retrieved August 25, 2006, from www. point-topic.com
57
Broadband Policy, Market Competition, and User Adoption in Taiwan
FIND (Focus on Internet News and Data). (2003). 2002 family connectivity and application survey. Retrieved August 27, 2006, from http://www. find.org.tw/0105/howmany/howmany _disp. asp?id=46 Hewett, K. (1997). Digital subscriber line: The route to broadband. London: Ovum. Hsieh, A. (2003). Telecommunications development and policy in China. In Y.L. Liu (Ed.), Telecommunications. Taiwan: Yeh Yeh Book Gallery. Jeng, J. (2003). President of Eastern Broadband Telecom Co., Ltd. Interview by author, September 1, 2003. Johnson, T., & Butt, H. (2003). Cable competition in broadband markets. Retrieved October 3, 2003, from www.point-topic.com Kao, B. (2003). Telecommunications policy and regulations. In Y.L. Liu (Ed.), Telecommunications. Taiwan: Yeh Yeh Book Gallery. Lee, C., & Chan-Olmsted, S.M. (2004). Comparative advantage of broadband Internet: A comparative study between South Korea and the United States. Telecommunications Policy, 28, 649-677. Liu, Y.L. (2006). Bandwidth, connection and access: Analyzing the business models of Webcasters in Taiwan. In L. Ha & R.J. Ganahl III (Eds.), Webcasting worldwide: Business models of an emerging global medium. Hillsdale, NJ: Lawrence Erlbaum. Liu, Y.L., & Chen, C.H. (2002). Service quality and user satisfaction of the broadband communication services. National Science Council Project, Executive Yuan, Taiwan. Liu, Y.L., Chen, C.H., Cheng, C.W., & Chu,Y.S. (2003). An analysis of narrowband and broadband user behaviors in Taiwan 2001-2002. Information Society, (July), 59-99. TWNIC. (2003-2006). A survey of the Internet users in Taiwan. Taiwan.
58
Yen, C.C. (2003). Former Managing Director, Multimedia Department, Northern Taiwan Business Group, CHT. Interview by author, August 28, 2003.
KEY TERMS ADSL: Asymmetrical digital subscriber line. Cable Modem: Modem used to provide highspeed Internet with a cable distribution system. Challenge 2008: The government of Taiwan in May 2002 launched the Challenge 2008 National Development Plan, which emphasizes such goals as strengthening international competitiveness, XSJUDGLQJWKHSHRSOH¶VTXDOLW\RIOLIHDQGSURPRWing sustainable development. Chunghwa Telecom: 7KH LQFXPEHQW ¿[HG networks in Taiwan. ISP: Internet service provider. LAN: Local area network. A group of devices that can communicate with each other within a limited area. Last Mile: Also known as local loop. It is the telephone line that runs from the local telephone FRPSDQ\WRWKHXVHU¶VSUHPLVH Leased Line: Also known as private or dedicated lines. It can be used to transmit voice, data, or video. Companies that rent leased lines pay a monthly fee for the leased lines that are exclusive to the companies that lease them. M-Taiwan: A key initiative to transform Taiwan LQWRDµXELTXLWRXVQHWZRUNVRFLHW\¶LQZKLFKWKH SXEOLFFDQIHHOWKHEHQH¿WVRIµHHQDEOHPHQW¶ National Communications Commission (NCC): A converged government agency regulating telecommunications and electronic media. It is also an independent government agency established in February 2006.
Broadband Policy, Market Competition, and User Adoption in Taiwan
National Information and Communications Initiative Committee (NICI): A cabinet-level committee that is responsible for accelerating the development of the IT industry.
laborates with international network information organizations such as ICANN and APNIC, as well as other national Internet organizations such as JPNIC, CNNIC, and KRNIC.
Taiwan Network Information Center (TWNIC): Oversees domain name registration and IP address allocation in Taiwan. TWNIC col-
Unbundling: The separation of all the components offered by the telecommunications operators so that the tariff for each component is transparent.
59
60
Chapter V
ICT Competency of Bangladesh to Face Broadband Diffusion Anwarul Islam 3ULPH0LQLVWHU V2I¿FH%DQJODGHVK K.C. Panda Sambalpur University, India
ABSTRACT The bull’s eye of Bangladesh has to achieve the millennium development goal and to adapt the globalization, necessitate pursuing the development. Information is playing a driven force in development. As a developing country, Bangladesh has taken keen initiatives to develop its sustainable information infrastructure. Teledensity and overall IT infrastructure is now in a growing stage. Recently, Bangladesh ZDVFRQQHFWHGZLWK6($0(:(VXEPDULQHFDEOHHVWDEOLVKLQJDQRSWLFDO¿EHUEDFNERQH LWVWHOHdensity is changing in rapid pace. But the broadband diffusion in Bangladesh is not on par with other Asian countries, as it is in an embryonic stage in broadband diffusion. This chapter, therefore, tries to show the initiatives taken and the existing condition of Bangladesh to fetch the countrywide broadband diffusion. Efforts have been made in this chapter to unmask the overall development of ICT infrastructure in Bangladesh, so to judge the environment of broadband diffusion in the country.
INTRODUCTION %DQJODGHVK¶VVWDQGRQWKH,&7GLIIXVLRQLQGH[E\UHgional groupings does not look goodBangladesh remains at the bottom in South Asia with a rank of 164 in 1997, 171 in 2001, and 171 in 2004even though the country actually has greatly progressed in terms of cellular phone penetration (Rahman, 2006). Some of the Asian countries are successful and experienced in ICT, as well as broadband diffusion. Korea, Japan, China, Singapore, and Hong Kong are successful e-society creators. Even
India has taken substantive initiatives in various SXEOLFDQGSULYDWH¿HOGVWRHVWDEOLVKDQH,QGLD It is important to learn from the experiences of successful Asian economies so as to create or foster broadband demand and provide necessary support on regulatory and policy aspects on the supply side. An effective framework of cooperation among Asian economies needs to be developed for deployment of broadband and to bridge the digital divide. Some of the measures that could be considered to create an effective broadband network environment in Asia are: preparation of roadmaps
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
ICT Competency of Bangladesh to Face Broadband Diffusion
for broadband economy at state and regional levels; undertaking collaborative programs for broadband proliferation; and developing school-nets, education and research nets, library nets, community information centers, and IGrid. Bangladesh is a South Asian developing country. Its area is 1, 47, 570 square kilometers (without sea area) with the population of 138.8 million; density of the population is 941 per square kilometer (projected in 2006). Per capita income is US$482, while the per capita GDP (gross domestic product) is US$456. Per the DCI (Direct Calorie Intake) method, the percentage of poverty is 40.9 (Ministry of Finance, 2006, p. xvii). Bangladesh has completed 35 years as an independent country and has been trying hard to get rid of the curse of poverty. Bangladesh comprises six divisions, 64 districts, 507 police stations/upazilas, and about 86,000 villages. The headquarters of the six divisions are governed by City Corporations (a specialized local government). Information and communication technology (ICT) is declared as a thrust sector of Bangladesh (Ministry of Science and Information & Communication Technology, 2002). The government has changed the name of Ministry of Science and Technology to Ministry of Science and Information & Communication Technology, which has revolutionized the globalization and market economy, DV WKH FRXQWU\¶V HFRQRPLF SURJUHVV LV YLUWXDOO\ impossible without the development of this sector. The cabinet has approved the National ICT Policy DQGWKH,&7$FWWRKHOSWKH,&7VHFWRUÀRXULVKLQ the country. In addition, the Copyright Act of 2000 has been amended, incorporating issues relating to software intellectual property right (IPR) (Ministry of Finance, 2005, p. 159). This chapter includes descriptions of literature reviews, the current National ICT Policy of Bangladesh, the enabling ICT infrastructure, national DQG%77%LQLWLDWLYHVWKHFRXQWU\ZLGHRSWLFDO¿EHU network, BTRC in communication development, teledensity, connectivity with submarine cable, ISP, Internet, and broadband.
LITERATURE REVIEWED Consistent policies for competition in telecom include policy measures to induce broadband network building and policy measures to boost demand for broadband. Korea took several measures to nurture the IT industry in 1985-1987. This led to systematic dissemination of broadband vision through various initiatives including the emerging of the National Basic Information System (1987-1996), Korean Information Infrastructure Initiative (1995-2005), National Framework Plan for Informatization Promotion (1996-2000), Cyber Korea 21 (1999-2000), and E-Korea Vision 2006 (2002-2006) (Agrawala & Shera, 2004, p. 16). In December 2005, four countries (Iceland, Korea, The Netherlands, and Denmark) led the OECD (Organization for Economic Cooperation and Development) in broadband penetration, each with more than 25 subscribers per 100 inhabitants. Iceland now leads the OECD with a broadband penetration rate of 26.7 subVFULEHUVSHULQKDELWDQWV.RUHD¶VEURDGEDQG market is advancing to the next stage of development where existing subscribers switch platforms IRU LQFUHDVHG EDQGZLGWK ,Q .RUHD ¿EHUEDVHG broadband communications grew 52.4% during 2005. This switchover effect is evident by the net loss of DSL (-3.3) and cable (-1.7%) subscribers during the year. Iceland takes over from Korea as broadband champ (OECD, 2005). The Global Information Technology Report 2002-2003 monitors the progress in networked readiness that is undoubtedly occurring in various parts of the world, and reveals the obstacles that prevent countries from IXOO\FDSWXULQJWKHEHQH¿WVRI,&7DQGEURDGEDQG The Networked Readiness Index (NRI) is the relationship between networked readiness and key variables such as gross domestic product per capita, ICT expenditure, and Internet usage. The index is composite of three components: the environment for ICT offered by a given country or community; WKHUHDGLQHVVRIDFRPPXQLW\¶VNH\VWDNHKROGHUV (individuals, businesses, and government) to use ,&7 DQG¿QDOO\WKHXVHRI,&7E\WKHVHVWDNHKROGers. The top 10 countries, as per the Networked Readiness Index, are: Finland, the United States, Singapore, Sweden, Iceland, Canada, the United
61
ICT Competency of Bangladesh to Face Broadband Diffusion
Kingdom, Denmark, Taiwan, and Germany. It also showed the Networked Readiness Index of Asian countries. Singapore is the only Asian country; it ranked third in 2003. Korea and Australia are ranked 14 and 15, while Hong Kong and Japan are ranked 18 and 20. Other Asian countries are ranked much lower as far as networked readiness is concerned. But the scenario shows changes in the report as of 2005. The report shows that the U.S. GURSSHGIURP¿UVWWR¿IWKLQWKHUDQNLQJVZKLFK measures the propensity for countries to exploit the opportunities offered by information and communication technology. Iceland, Finland, and Denmark occupy positions two, three, and four out of 104 countries surveyed, with Iceland achieving the most improvement among the top countries, moving up from tenth last year. ,QGLDDQG&KLQDVLJQL¿FDQWO\LPSURYHGWKHLUSRVLtions climbing to numbers 39 and 45, compared to 45 and 51 in 2003, respectively (Forbes, 2005). ICT diffusion is the pre-term of broadband diffusion. ICT diffusion to the poor people of third-world countries is meager, where foremost parts of the population do not have competence in reading and writing. In an interview Onno Purbo, an Indonesian IT teacher, indicated two major factors for this: “For poor people in Indonesia, there are two major barriers (poverty and illiteracy), mainly because of the society. Most of the poor people cannot read and write. So they would be fairly incidental. I see these only as a tool. For [the] listen and talk community, maybe radio and television would be appropriate. For [the] read and write community…Internet and computer would be appropriate.” (Purbo, 2007)
NATIONAL ICT POLICY: A PATHFINDER The National ICT Policy of Bangladesh was passed in October 2002 and was divided into four parts: (a) Preamble, (b) Vision and Objectives, (c) Policy Statements, and (d) Implementation and Monitoring. The preamble mentions that the policy is a guideline for the development of the ICT sector
62
in Bangladesh as a part of the overall national development plan. This policy aims at building an ICT-driven nation comprising a knowledgebased society by the year 2006. In view of this, a countrywide ICT-infrastructure will be developed to ensure access to information by every citizen to facilitate empowerment of people and enhance democratic values and norms for sustainable economic development by using the infrastructure for human resources development, e-governance, e-commerce, e-banking, public utility services, and all sorts of online ICT-enabled services (Ministry of Science and Information & Communication Technology, 2002). The above mentioned major statements of the National ICT Policy may have some drawbacks or may have included more than expectations. HowHYHULWLVRQO\DSUHGH¿QHGJXLGHOLQHIRUDFKLHYLQJ the national expectation to handle ICT for social interest. Especially in a developing country, most RIWKHSROLFLHVDUHµSDSHUGRFXPHQWV¶DVLPSOHmentation of the whole policy remains a myth DQGLVKDUGO\IXO¿OOHG/LNHWKLVDSRUWLRQRIWKH National ICT Policy has been implemented, not as a direction of the policy, but these are considered DVWKHSDWK¿QGHULQDFKLHYLQJWKHJRDOV
BTTB in Information Communication The Bangladesh Telegraph and Telephone Board (BTTB) is playing a vital role in information and communication. For developing the telecommunication, BTTB has been taking various steps and making a steady progress. In the meantime, BTTB has established 71 digital data network (DDN) nodes in 41 district towns. Presently, BTTB is providing Internet service in all the districts (64 districts) and some sub-districts (upazila level) where the digital exchange services are available. For the development and extension of Internet service, BTTB has installed remote access servers (RASs) in 39 districts instead of Nation Wide Dialing (NDW) channel. It has increased the number of international circuits to 8,401; it is providing low-cost telephone service in 25 countries. To develop and modernize the telephone system at the upazila level, small-size digital exchanges at 92
ICT Competency of Bangladesh to Face Broadband Diffusion
Table 1. Major policy statements of the National ICT Policy of Bangladesh Fields Training and Human Resources Development
Focused On Production of skilled manpower in ICT to compete effectively in the global ICT market.
ICT Infrastructure
Appropriate ICT infrastructure should be established both in public and private sectors, so as to establish direct connectivity with international information and communication backbone.
Research and Development in ICT
Contributing in applied research to the improvement of quality DQGHI¿FLHQF\RIDSSOLFDWLRQWRWKH,&7LQGXVWU\
ICT Industry
To encourage the entrepreneur, government is providing
(i) Software Industry
¿QDQFLDODQGORJLVWLFVXSSRUWWRHVWDEOLVKWKHVHLQGXVWULHV
(ii) Hardware Industry (iii) Services Industry E-Commerce
Government shall make preparation to introduce and promote G2G (government-to-government) and G2B (government-tobusiness) transactions.
E-Government/E-Governance
Government shall introduce ICT system within the public DGPLQLVWUDWLRQWRLPSURYHHI¿FLHQF\UHGXFHZDVWHRI resources, enhance planning, and raise the quality of services.
Legal Issues
Software copyright provisions embodied in the Copyright Act of 2000 shall be implemented promptly.
Healthcare
ICT should be used to develop such capabilities, especially in the areas of electronic medical records, telemedicine, medical and health education, and so forth.
Agriculture and Poverty Alleviation
Initiative will be taken to utilize ICT system in agro-based industries and agricultural research.
Social Welfare
ICT system will be implemented for rural development DFWLYLWLHVDJULFXOWXUDOKRUWLFXOWXUH¿VKHULHVOLYHVWRFN extension, career guidance for youth, technology guidance for rural enterprises, micro-level planning, and so forth.
Transportation
Introducing an ICT-based integrated transport management system.
Tourism
Launching tourist attraction activities via the Internet and providing reservation for travel and accommodation booking and shopping.
Judiciary
To enhance the capacity of the judiciary, computer-based MIS with suitable WAN and LAN will be established for the Supreme Court and for the District Courts and Tribunals.
Regional and International Cooperation
The concern ministry (Ministry of Science and ICT) shall explore regional, sub-regional, and international cooperation and ICT with developed and developing countries, as well as with relevant international agencies and development partners.
63
ICT Competency of Bangladesh to Face Broadband Diffusion
upazilas (sub-districts) of the country have already been installed by using the “Resource Mobilization and Services” fund. In the meantime, a total of 216 upazilas and 8 growth centers have already been provided with the facility of digital exchanges by utilizing the fund available through the ongoing project titled, “Installation of Digital Telephone Exchange in Upazilas and Growth Centers.” In addition to normal NWD and ISD facilities, Internet services will also be available to the subscribers of digital exchange (Ministry of Finance, 2006, p. 118) in the days ahead.
6RXWK$VLD .KDQ ,Q WKHRSWLFDO¿EHU network was completed for Bangladesh Railway along with the installation of digital switching (Islam & Rahman, 2006, pp. 135-146). After a few years, Bangladesh Railway leased its 1,800NLORPHWHURSWLFDO¿EHUEDFNERQHWRWKH*UDPHHQ Phone (a private cellular phone operator). In 2001, %77% VWDUWHG LQWHUH[FKDQJH RSWLFDO ¿EHU QHWworking and major city networking respectively. 7KH 1RUWK6RXWK OLQN WKURXJK WKH RSWLFDO ¿EHU backbone has already been completed and/or is being upgraded by BTTB as:
Optical Fiber Network in Bangladesh
• • • •
With the Norwegian untitled grant, worth US$15.99 PLOOLRQ WKLV¿JXUHKDVEHHQFRQYHUWHGIURP%DQgladesh currency taka, where the present rate is US$1 = 66.57 taka) in 1985, Bangladesh Railway EXLOWWKLV¿UVWRILWVNLQGWHOHFRPLQIUDVWUXFWXUHLQ
•
Dhaka-Chittagong Link Bogra-Gobindaganj-Rangpur Link Rangpur-Lalmonirhat Link Syeedpur-Dinajpur-Thakurgaon-Panchagar Link Sirajganj-Pabna Link
)LJXUH%77%RSWLFDO¿EHUEURDGEDQGQHWZRUNLQ%DQJODGHVK
64
ICT Competency of Bangladesh to Face Broadband Diffusion
•
Comilla-Brahmanbaria Link and Feni-Lakshmipur Link
,QDGGLWLRQWRWKHVHOLQNVWKHZRUNRI¿YHPRUH links are in progress and some of them have already been completed. These links, which deserve mention, are: • • • •
Gazipur-Mymensing-Tangail Link Tangail-Sirajganj-Bogra Link &R[¶V%D]DU&KLWWDJRQJ/LQN Brahmnbaria-Moulavibazar-Sylhet Link Kushtia-Satkhira Link
The Bogra-Natore-Rajshahi-Chapai Nawabganj Link and Dhaka-Srinagar-Munshiganj Link are under implementation. The above mentioned links are unable to cover the entire country to establish a smooth information infrastructure. Owing to this, BTTB has successfully established some Synchronous Digital Hierarchy (SDH) microwave networks. Also, some of these areas are costal areas and a lot of tributary rivers expand their nets. As a result, these areas could not be covered easily
E\ RSWLFDO ¿EHU EDFNERQH 6RPH 6'+ 670 microwave links are: • • • • • • • • • • •
Dhaka-Manikganj-Faridpur-Magura Link, Magura-Jenaidha-Kushtia-Pabna-Natore Link Magura-Jessore-Khulna Link Natore-Naogaon-Bogra-Joypurhat-Barishal Link Natore-Rajshahi Link Khulna-Bagehat-Perojpur-Jhalokathi-Barishal Link Khulna-Gopalgamj-Madaripur-FaridpurRajbari Link Madaripur-Sariatpur Link Barishal-Patuakhali Link Barishal-Bhola-Lakshmipur Link Rajbari-Faridpur Link
Submarine Cable: Connecting with Information Superhighway Telecom transmission media has changed radically with the passage of timefrom HF to satellite to
Figure 2. BTTB SDH microwave network
65
ICT Competency of Bangladesh to Face Broadband Diffusion
Figure 3. SEA-ME-WE-4 submarine cable
WRGD\¶VVXEPDULQHFDEOH1DUURZEDQG+)V\VWHPV were being used for international communications. However, since 1975 BTTB has been using Intelsat Satellites for International Telecommunications for voice and data transmission. International voice service is being provided exclusively by BTTB. Data service is being provided mainly in two different ways: through International Private Leased Circuit (IPLC) and through Leased IP Access to BTTB Internet Node. To facilitate a wide range of bandwidth at an affordable price, especially for software exporters and for ICT usage, BTTB has become a member of the South East Asia-Middle East-West Europe-4 (SEA-ME-WE-4) Consortium to connect Bangladesh to the information superhighway. Bangladesh signed the construction and maintenance agreement (C&MA) on March 27, 2004, in Dubai through which the country joined the SEA-ME-WE-4 Submarine Cable Consortium of 16 parties of 14 countries (Hossain, 2005). The work of submarine cable is already complete and was inaugurated in May 2006. It is hoped that, with the advent of SEA-ME-WE-4, use of bandwidth will increase sharply for international telecommunication because of the cheaper bandwidth cost. Submarine cable connectivity in Bangladesh will dramatically
66
improve the volume and quality of data and voice communication, and meet the increasing demand of international bandwidth. The delay factor in satellite communication, which is a barrier for high-speed data transfer, will be eliminated. ICT usage and activities will be promoted. Available capacity for BTTB in this cable is 468,000 MIU Km which may have a maximum of 64 STM-1 (10 Gbps) at the landing station. MIU is Minimum Investment Unit, which is equivalent to one STM-1. 2XWRI%77%¶VWRWDOFDSDFLW\0,8.PKDV already been contributed to the common pool of the consortium for sale. Sales and marketing of this bandwidth will be performed by the consortium in WKH¿UVWWZR\HDUV7KHFDSDFLW\RIWKHFDEOHZLOO be enhanced further in different phases with minimal investment from BTTB. Minimum capacity unit for this cable is STM-1that is, BTTB must utilize the capacity at STM-1 or a multiple of that between any two landing stations.
Submarine Cable Utilization Plan The main task of BTTB is to plan well so as to utilize the huge bandwidth available in the cable in a pragmatic and systematic way. BTTB held meetings and discussions with ISPs, BASIS, PSTN, mobile
ICT Competency of Bangladesh to Face Broadband Diffusion
phone operators, and other users to ascertain their present and future requirement and to share ideas on utilization of the submarine cable. $IWHUIXO¿OOLQJWKHUHTXLUHPHQWRIWKHFRXQWU\ excess capacity in the mainstream may be leased out to overseas operators. For the next two years, capacity can only be sold on a lease basis and not as IRU. To make better and greater utilization of the submarine cable, new services like IP VPN, virtual telephone network abroad, and various
kinds of broadband service may be introduced in the country. Connectivity with different countries may be established through SMW-3 by exchanging equivalent SMW-4 capacity. For utilization, submarine cable BTTB has taken a project to enrich international telecommunication facilities. Demands for international bandwidth have a wide range of variation in terms of capacity, protocols, purpose, use, national policy, and so forth. Taking consideration of these aspects, the
Table 2. Connected with submarine cable (BTTB, 2006b) Name of the project
:
Establishment of international telecommunication through submarine cable
:
US$94.35 million
Consortium
:
16 international telecommunication companies in 14 countries
Member Countries
:
Allocation & Expenditure
Members of the Submarine
Singapore, Malaysia, Thailand, Bangladesh, India, Sri-Lanka, UAE, Pakistan, KSA, Egypt, Italy, Tunisia, Algeria, and France
Total Length of Submarine Cable
:
20,000 km
:
NLORPHWHUV FRQQHFWLRQWR&R[¶V%D]DU
:
10 GB/Sec
Length of Bangladesh Branch
Present Speed in Data Transmission
Table 3. Statistics of different types of licenses issued by BTRC (Ministry of Finance, 2005, p. 224) Types Cellular Mobile Telecom License (Teletalk) Zonal Fixed Telecom (PSTN) License
Number of Companies/ Institutions
Total 1
1
15
37
Mobile Telecom Operator License Revalidation
4
4
Dhaka MEA Fixed Telecom Operator License Revolution (WorldTel)
1
1
Pre-Paid Calling Card Operator License Revalidation
1
1
Rural Fixed Telecom Operator License Revalidation
1
1
Internet Service Provider
181
181
Nationwide Internet Service Provider
30
30
VSAT Users
85
85
VSAT Provider (Sub-Station) Domestic Data Communication Service Provider
4
4
24
24
67
ICT Competency of Bangladesh to Face Broadband Diffusion
customer base needs to be segmented and a number RI SURGXFWV PXVW EH GHYHORSHG WR PHHW VSHFL¿F requirements of the customers. Proper attention should be given to take into account the indirect HFRQRPLF EHQH¿WV DV ZHOO DV ORVW RSSRUWXQLWLHV while pricing these products. IPLC is a dedicated secure digital point-to-point private connection between two locations (usually two LANs) in two different countries that allows transmission of data, large Internet packets, realtime video applications like videoconferencing, and such other information communication services. BTTB has been providing IPLC service through INTELSAT satellites with bandwidth ranging from 64 Kbps to 2 Mbps. Through submarine cable, bandwidth allocation for IPLCs will depend on the requirement and MXVWL¿FDWLRQZKLFKPD\EHIURP.ESVWR0ESV Higher bandwidth up to 45 Mbps will be allocated after scrutiny by BTTB. IPLCs may be provided on a half-circuit as well as on a full-circuit basis. At present, all cities and towns of Bangladesh are covered by digital exchanges. All Upazila exchanges will have digital exchanges by the end of this year. So, all telephone subscribers up to the Upazila level will have voice and Internet connectivity through the submarine cable. Rural growth centers will be brought under digital exchanges by installing a suitable access network to extend EHQH¿WVRIVXEPDULQHFDEOHLQWRWKHUXUDODUHDV
Connectivity to private mobile and land phone operators is being further improved to allow access for all the existing subscribers and ensure access for the future growth. Establishment of two more international gateway exchanges is under process; %77%¶VSUHVHQWGDWDDFFHVVQHWZRUNLVEDVHGRQ the existing DDN, which can support only up to 2 Mbps capacity. An IP access network must be developed where optical interfaces will be deployed in major cities to connect customers through optical local loop at a higher speed. At present, DDN nodes are available in 41 districts only. To ensure a homogeneous growth of ICT in the country, a data network will be extended to all 64 district +4VLQWKHQH[W¿QDQFLDO\HDU Private operators in the country will be allowed to develop an access network for data communication only. Co-location facilities in Dhaka, Chittagong, and other large cities have been included for this purpose. Co-location facilities on BTTB premises in Dhaka and Chittagong are likely to be allowed for customers who require higher capacity, like E3/DS3 or STM-1 or higher. For the purpose of redundancy and restoration of the international segment satellite, the earth station will continue functioning. Initially, STM-1 Internet connectivity is being established through the west and another through the east for redundancy in the main cable. For redundancy in WKHEUDQFKFDEOHDVLJQL¿FDQW,QWHUQHWEDQGZLGWK is being kept as backup through satellite to meet
7DEOH6XEVFULEHUVRI¿[HGDQGPRELOHSKRQH $SULO 0LQLVWU\RI)LQDQFHS Year
Grameen
Aktel
Banglalink
Citycell
Teletalk
Fixed
Total
Phone 2001
471,371
80,368
69,700
41,109
0
564,800
1,227,348
2002
774,881
161,265
112,900
91,348
0
682,000
1,822,394
2003
1,140,531
401,680
186,500
179,058
0
716,721
2,624,490
2004
2,388,158
1,096,620
369,500
296,509
0
831,280
4,982,067
2005
5,540,000
2,072,328
1,026,100
439,389
200,000
1,081,450
10,359,267
2006
6,700,000
2,458,819
1,600,000
600,000
242,617
1,000,483
12,601,919
(April)
68
ICT Competency of Bangladesh to Face Broadband Diffusion
Table 5. Growth of telecommunication of Bangladesh (2000-2006) Series Name
2000
2001
2002
2003
2004
2005
April 2006
Telephone lines
0.58
0.83
1.32
1.56
N/A
4.8
7.03
0.15
0.19
0.34
0.78
N/A
N/A
N/A
0.08
0.14
0.15
0.18
N/A
0.2
0.35
750,775
1,084,880
1,757,000
2,107,048
4,982,067
10,359,267
12,601,919
200,000
250,000
450,000
1,050,000
N/A
N/A
N/A
100,000
186,000
204,000
243,000
N/A
300,000
500,000
and cellular subscribers per 100 population Personal computers per 100 population Internet users per 100 population Telephone lines and cellular subscribers Personal computers Internet users
N/A= Not available
HPHUJHQF\,QWHUQHWWUDI¿FLQFDVHRIDQ\IDXOWRU failure in the branch cable. To ensure domestic redundancy, both Dhaka&KLWWDJRQJDQG&R[¶V%D]DU±&KLWWDJRQJPLFURwave links are planned to be upgraded to a 7+1 STM-1 SDH system. An initiative continues for KLULQJ GDUN ¿EHU IURP 3*&% EHWZHHQ 'KDND &KLWWDJRQJLQWKH¿UVWSKDVH
BTRC in ICT Development The Bangladesh Telephone Regulatory Commission (BTRC) was formed on January 31, 2002,as an independent statutory body under the Telecommunication Act of 2001. All relevant powers, responsibilities, and pertinent matters related to telecommunication regulation have been vested with the commission. The objectives of the commission include inter alia: broadening the efforts for improvement of the sector through better regulation, providing people with easy access to the telecommunication and Internet service at a reasonDEOHFRVWHQVXULQJWKHHI¿FLHQF\RIWKHWHOHFRPmunication system, and enhancing its capability
to compete at both the national and international sphere, preventing and abolishing discrimination in providing telecommunication services. The aim of the commission is to progressively rely on a competitive market-oriented system. In keeping with these objectives, the commission is committed to ensuring effective control of telecommunication, and introducing new services and creating favorable atmosphere for the local and foreign investors who intend to invest in the telecommunication sector in Bangladesh.
Teledensity in Bangladesh The government granted permission to Bangladesh Telecom Private Limited (BTL) to introduce radio trunking, cellular radio telephone, naval radio telephone network, and paging facilities for the ¿UVWWLPHLQ WRVSXUSULYDWHLQYHVWPHQWLQWKLV sector. The segment of mobile telecom operation of WKH%7/¶VOLFHQVHZDVODWHUKDQGHGRYHUWR3DFL¿F Bangladesh Telecom Limited. Subsequently, three private operatorsGrameen Phone Ltd., Telecom Malaysia International (Bangladesh), and Seba
69
ICT Competency of Bangladesh to Face Broadband Diffusion
Table 6. Installation and rental charges for port and modem (Tk=Taka; 1US$=Tk 66.57) (BTTB,
2006a) Yearly Rent for Port and Local
Speed
New Installation Charge
64 Kbps
Tk. 10,000 (US$150)
128 Kbps
Tk. 10,000 (US$150)
Tk 294,000 (US$4,416)
Tk 18,000 (US$270)
256 Kbps
Tk. 10,000 (US$150)
Tk 438,000 (US$6,580)
Tk 27,000 (US$406)
512 Kbps
Tk. 15,000 (US$225)
Tk 684,000 (US$9,734)
Tk 30,000 (US$451)
1 Mbps
Tk. 15,000 (US$225)
Tk 1,176,000 (US$17,666)
Tk 36,000 (US$541)
2 Mbps
Tk. 20,000 (US$300)
Tk 2,050,000 (US$30,795)
Tk 48,000 (US$721)
Loop up to 5 km Tk 181,000 (US$2,719)
Telecom Private Limitedwere awarded license in 1996 to market digital cellular mobile phones. After the formation of the Telecommunication Regulatory Commission, those licenses/contracts were renewed under the Bangladesh Telecommunication Act of 2001. The Commission Granted Mobile Telecom Operator License to BTTB which was subsequently revised to form a separate entity XQGHUWKHQDPHRIµ7HOHWDON¶DQGKDVDOUHDG\EHcome operational. The number of subscribers of WKH¿YHSULYDWHFHOOXODUPRELOHFRPSDQLHVLQWKH country was about 61.06 lakh (6.106 million) as of June 2005. To meet the increasing demand for telephone, 15 private companies have been awarded OLFHQVHVIRUSURYLGLQJ¿[HGWHOHSKRQHVHUYLFHV in four zones of Dhaka (except in the Dhaka multiexchange area) in a competitive environment. As a result, employment opportunities have opened up for thousands of the unemployed. The number of subscribers of BTTB is currently about 8 lakh (0.8 PLOOLRQ 7KHQXPEHURIVXEVFULEHUVRISULYDWH¿[HG telephone companies is expected to exceed 25 lakh (2.5 million) in the next two years. Four cellular mobile companies are operating competitively. Having been permitted, Sheba Telecom and PTBL to transfer shares, and approved private investment in mobile to transfer share, and approved private investment in mobile telecom sector as well as in
70
Yearly Rent for Modem
Tk 18,000 (US$270)
¿[HGWHOHSKRQHVHUYLFH7RJHQHUDWHFRPSHWLWLRQ on the telecom market and to help drive low-cost telecommunication service to people, Teletalk Bangladesh Limited has been issued license to market mobile telecom service. Already the company started offering mobile telecom service. As a consequence, mobile phone call charges are expected to decline, making it affordable to ordinary FLWL]HQV,QFRPSDULVRQWRQHLJKERUFRXQWULHV¶FHOO phone rates, the cell phone rate of Bangladesh is higher: the number of cellular phone subscribers ZDVPLOOLRQLQ$SULO,QFOXGLQJWKH¿[HG phone services, the total number of subscribers is currently 12.6 million. The current teledensity is 7.3, which is more than the target. It is remarkable that the number of the subscribers of cellular phone service is increasing in rapid SDFHLQFRPSDULVRQWR¿[HGSKRQH,QWKH\HDU WKHQXPEHURIWKH¿[HGSKRQHVXEVFULEHULQFUHPHQW rate was 21%, and 30% in 2005. On the other hand, the cellular phone subscriber increment was 72% in 2002 and increased 123% in 2005. The license contracts of WorldTel and the BTTB for the personal Handy Phone System were revalidated in consideration of a huge demand in Dhaka city. The license of Formula One of the introduction of a pre-paid card system has also been revalidated. Recipients of license in telecom-
ICT Competency of Bangladesh to Face Broadband Diffusion
)LJXUH0HWUR1HW¶VRSWLFDO¿EHUEDFNERQHRI0HWUR1HWLQ'KDND Participation of Operators
5%
2% 8%
Grameenphone Aktel
13%
Banglalink Citycell
20%
munication and broadcasting have been allocated appropriate frequencies to enable them to transmit data, picture, and voice.
Internet, ISP, and Broadband Internet came late in Bangladesh with UUCP email beginning in 1993 and IP connectivity in 1996 (Islam & Rahman, 2006, p. 136). In June 1996, the government allowed private entrepreneurs to act as Internet service providers (ISPs) using VSATs (very small aperture terminals). Initially, there were only a few Unix-to-Unix copy protocol (UUCP) accounts in the country and then they were replaced by Internet protocol (IP) accounts. At a later stage, low-bandwidth 64kbps VSAT link became the main Internet backbone of the country. In 1993 Bangladesh had access to e-mail via dialup to bulletin board systems (BBSs) of a few local providers. The combined Internet users of all the e-mail-only service providers were not more than 500. Users were charged by the kilobyte, and mail was transferred from the BBS service providers to the rest of the world by international dialup using UUCP. In June 1996, the government allowed private entrepreneurs to act as ISPs using VSATs. The government allowed VSATs to be operated in the private sector, albeit to be provided solely by the government-owned telephone operator, BTTB. Only a handful of ISPs were connected within WKH¿UVW\HDU+RZHYHUPRUHOLEHUDOJRYHUQPHQW policies followed in the subsequent years which
52%
Teletalk Fixed
led to a rapid expansion of this industry, eventually resulting in over 180 registered ISPs by 2005 (ISPABD, 2007). On May 30, 2006, the submarine cable was opened by the Honorable Prime Minister for services to which several ISPs are already connected. ISPs are currently regulated by the Bangladesh Telecommunication Regulatory Commission through the Bangladesh Telecommunication Act of 2001. BTTB, the state-owned ISP, now allows people to register their domain names and use the .bd extension in their names. The language used by the content developers is mostly English. But efforts to have Bangla (national language) are also growing subsequently.
Charges to Setup an ISP The Bangladesh Telecom Regulatory Commission KDV ¿[HG WKH UDWHV IRU ,63 OLFHQVHV $ FLUFXODU issued by the BTRC commission on December KDV¿[HGWKHVHWXSFKDUJHVIRU,QWHUQHW services. It is alleged that the license fees are also very high. There are a few categories of licenses issued to the applicants to setup an ISP. 7KH¿UVWFDWHJRU\LVWKHOLFHQVHVIRUWKHPDjor cities, namely Dhaka, Chittagong, or Sylhet. License fees for this category are Taka 200,000 (US$3,004) with a renewal fee of Taka 100,000 (US$ 1,502) per year. Apart from the above license fees, there is a bandwidth fee on the basis of the
71
ICT Competency of Bangladesh to Face Broadband Diffusion
Table 7. Broadband Internet service offered by BTTB (for corporate users) City/District
Places
City/District
Places
1. Dhaka
Babubazar, Basabo, Cantonment,
22. Moulavibazar
NM
23. Munshiganj
NM
ChawkBazar, Demra, DEPZ, Fakirapool, Gandaria, Gulshan, Lalbag, Mirpur, Mogbazar, Nilkhet, Ramna, Savar, Sayedabad, Sher-e-BanglaNagar, Uttara 2. Chittagong
Agrabad, Bayezid, CEPZ, Kalurghat, Muradpur, Nandankanon, Sagarika
3. Gazipur
Gazipur, Tongi
24. Mymensingh
NM
4. Narayangonj
Narayangonj, Siddhirgonj
25. Naogaon
NM
5. Khulna
Central, Khalishpur
26. Narshingdi
NM
6. Jessore
Jessore, Sharsa
27. Natore
NM
7. Bagerhat
Bagerhat, Mongla
28. Netrokona
NM
8. B-Baria
NM
29. Nilphamary
NM
9. Barishal
NM
30. Noakhali
NM
10. Bogra
NM
31. Pabna
NM
11.Chapai Nawabgonj
NM
32. Panchagar
NM
12. Comilla
NM
33. Patuakhali
NM
&R[¶V%D]DU
NM
34. Rajshahi
NM
14. Dinajpur
NM
35. Rangamati
NM
15. Faridpur
NM
36. Rangpur
NM
16. Feni
NM
37. Sirajgonj
NM
17. Jhenaida
NM
38. Sunamganj
NM
18. Kushtia
NM
39. Sylhet
NM
19. Lalmonirhat
NM
40. Tangail
NM
20. Magura
NM
41. Thakurgaon
NM
21. Manikganj
NM
NM=Not Mentioned
72
ICT Competency of Bangladesh to Face Broadband Diffusion
bandwidth used by the operator. The annual bandwidth fee for 128kbps is Taka 30,000 (US$451). The fee is Taka 50,000 (US$751) for 512kbps and Taka 100,000 (US$1,502) for 512kbps. In district towns other than the above three cities, the license fee is Taka 100,000 (US$1,502) and the renewal fee is Taka 50,000 (US$751) per year. Bandwidth fees remain the same in every case except the rural areas. A nationwide license is issued for Taka 500,000 (US$7,510) which can be renewed DW7DND 86 ,61WKH¿UVW,63RI Bangladesh, holds a nationwide license and operates in Dhaka and Khulna. Other ISPs are said to have regional licenses. However license fees in the semi-urban and rural areas are absolutely low. There is a license and renewal fee of Taka 1,000 (US$15) only in those areas with no bandwidth fee. In major cities and district towns, cyber cafés need to have licenses. But in the semi-urban and rural areas, no license fees are required to setup a cyber café (Jabbar, 2005). %URDGEDQGPD\EHGH¿QHGDVWKHVHWRIWHFKnologies that provide speed of at least 256 kbit/s, where this speed is the combined upstream and downstream capacity. The number of broadband subscribers refers to the number of active regisWHUHG,QWHUQHWDFFRXQWVLQFOXGLQJDOO¿[HGQHWZRUN broadband access technologies: digital subscribers line service (DSL), cable modem services, satellite EURDGEDQG ,QWHUQHW ¿EHU WR KRPH ,QWHUQHW DFcess, ethernet LANs, wireless local area network services, and so forth (ITU, 2003). Presently, 180 SULYDWH,63VDUHVHUYLQJLQWKLV¿HOGZLWKJRYHUQment-owned ISP (BTTB); all the ISPs are not active. ISPs depended on satellite, with slow speed and high bandwidth charge. But presently, the ISPs are connecting with submarine cable instead of VSAT. Cable TV channel operators are trying to offer the broadband connectivity through their cable netZRUN,WLVDIDFWWKDWWKHHVWDEOLVKHGRSWLFDO¿EHU backbone radio link has lost its popularity. The rent of a broadband line is Taka 1,000 per month for the individual subscriber. BTRC issued license to build, own, and operate (BOO) data communication infrastructure in private sectors. It helps the ISP to provide broadband connectivity to the subscribers easily (Khan, 2005). One of the ISPs,
MetroNet, put more than 125 kilometers of optical ¿EHUFDEOHZLWKLQWKH'KDNDFLW\DQGLWVQHDUYLFLQity to overcome the absence of a high-performance data communication network. MetroNet claims that LWLVWKH¿UVWWRLQYHVWLQEXLOGLQJDQXQRUWKRGR[ high-performance data communication network in the private sector (Khan, 2005). 0HWUR1HW¶VRSWLFDO¿EHUEDFNERQHLVSURYLGLQJ coverage from Tongi to Narayangonj, Sadarghat to Baridhara, Gulshan to Dhanmandi, and Dhaka University area to Mirpur, and in and around Tejgaon, Kawran Bazar, Maghbazar, Eskaton, Mouchak, Rampura, Malibagh, Kakrail to Motijheel, and Dilkusha (Metrobd, 2006). Presently, 180 ISP have licenses, though some of them are not active; but most of the ISPs are providing dial-up, radio link, and broadband connectivity to the customers. For the individual customer with unlimited time, broadband connection costs the least at Taka 1,000 per month, which is still more than the common people can afford. BTTB is offering in 41 districts leased Internet access (broadband Internet) to corporate customers. The rates of BTTB are given in Table 6. The offered broadband connectivity is based on these given terms: • • •
•
•
• •
BTTB is providing the bandwidth, which shall be provided from the nearest DDN node No security deposit 50% discount on total rental charges for accredited universities, colleges, Madrasa, training and research institutes 25% discount on rental charges for government/semi-government/autonomous and QRQSUR¿WUHVHDUFKRUJDQL]DWLRQV 25% discount on rental charges for software incubators and parks under government control and private software exporters earning at least US$100,000 per year Initial payment for six months 50% discount on registration, installation, and testing fee for 5+ connections at a time
Table 7 shows the city, district, and places where BTTB is offering broadband connection services to customers. BTTB is providing the broadband
73
ICT Competency of Bangladesh to Face Broadband Diffusion
FRQQHFWLRQLQWKHDUHDVRQO\ZLWKLQD¿YHNLORPHWHU radius from the nodes; extended facilities are given for the seven cities/districts.
• • •
Cyber Culture in Bangladesh
•
Cyber culture in Bangladesh has its optimum momentum in the capital city of Dhaka, where there are a few hundred cyber cafés within the city apart from individual users (Jabbar, 2005). Most of the cyber cafés are well-furnished, air-conditioned facilities with broadband connectivity. Some of them have private cabins with 50 or more computers. Computers are generally networked. Locations of these cafés are on the main road and in prime locations of the cities/towns. Some cyber cafés serve snacks and other fast-foods to clients. Most of them have scanning, CD writing, printing, and other facilities. Charges for browsing range from Taka 20 to 35 per hour (around US$.50 or less). Students form the largest user community of the cyber café. Cyber culture is gradually increasing at a slow pace. Still the conservative parents are anxious: misuse of the Internet by students is the main problem of the parents.
•
Barriers of Broadband Diffusion of Bangladesh With a poor economy and wondering movement LQWKH¿HOGRUHOHFWURQLFEDVHVRFLHW\%DQJODGHVK is lagging behind in development of broadband diffusion. There is not a clear plan to explore the government to the root level. Lack of poor broadband data transmission infrastructure and the necessity for broadband use are the most important reasons for the lag. The government has taken some initiative, but it is not enough to develop a good broadband network. The government must provide more citizen services through the Internet. If this is possible, then it will be possible to carry out a path for satisfactory movement. Barriers to broadband diffusion in Bangladesh may include: • •
74
Differences in education and literacy levels Gender, age, and income level
• • • •
Level of connectivity network penetration Lack of user training Lack of providing citizen service through the Internet Lack of providing electronic citizen services Lack of initiatives for the side of government and private level Lack of creating deployment of broadband Lack of creating a more equitable information society Lack of cheap service providing Lack of creating competitive marketing
National Initiatives The government has taken some initiatives so to focus on the National ICT Policy. The “ICT Act” based on the United Nations Commission on International Trade Law (UNCITRAL), known as the UNICTRAL Model Law on electronic commerce, has already been passed. The Copyright Act of 2000 has been amended to protect the intellectual property right (IPR) of the computer software. Public and private universities have started graduate- and postgraduate-level ICT courses, Beside this, the Ministry of Science and Information & Communication Technology has started a one-year postgraduate diploma course to produce skilled human resources. Ten public universities have started this course, and it will be H[WHQGHGLQDQDGGLWLRQDO¿YHXQLYHUVLWLHV3UHVently, the government has also started a one-year ICT internship program. Further, the government has taken an initiative to launce Web sites for all the ministries/divisions to implement e-governance service. These sites include all public documents and forms for providing online service easily to citizens. The Web site of the Bangladesh government (http://www. bangladesh.gov.bd) provides links to all governPHQWRI¿FHV7KH³6XSSRUWWRWKH,&77DVNIRUFH´ SURMHFWRIWKHJRYHUQPHQW¶VSODQQLQJGLYLVLRQKDV taken the initiative to launce e-governance among WKH3ULPH0LQLVWHU¶V2I¿FHHLJKWPLQLVWULHVDQG VL[GLYLVLRQDOFLWLHV)RUW\HLJKW¿UPVDUHFXUUHQWO\ working in the government-founded “ICT Incubator Center” at Kawran Bazar in Dhaka.
ICT Competency of Bangladesh to Face Broadband Diffusion
In other efforts, an IT village is planned for set up very close to Dhaka. The government has already made 18 acres of land available to set this up; it plans to spend US$3.76 million on this project (Ministry of Finance, 2006, p. 120). This village would be similar to the software technology parks in India. The entire infrastructure, including highspeed telecommunication facilities (2Mbps link), would be provided (Islam & Rahman, 2006, p. 142), and these would enable small companies to move into buildings with readily available facilities. As a part of its continuing effort to use ICT for development, the government has established an entrepreneur equity fund worth US$45.07 million IRU¿QDQFLQJSULYDWHHQWHUSULVHV7KHJRYHUQPHQW has created a National ICT Taskforce and has initiated some pilot projects, including:
•
•
CONCLUSION
•
•
An interactive Web site for downloadable and submitable forms, online complaints, billing information, and so forth, under BTTB Interactive Web sites for job search and other information for expatriate welfare under the Ministry of Expatriate Welfare and Employment Opportunities Digitization of useful land-related information such as land records, maps, documents, and so forth, under the Ministry of Land, EHLQJ ¿QDQFHG E\ WKH JRYHUQPHQW¶V RZQ resources, under the umbrella project named “Support to the ICT Taskforce,” under the planning division
All these activities by the government are focused on recognizing the power of ICT. Bangladesh is promoting e-government/e-governance and e-commerce as a way of helping government DQGWKHEXVLQHVVFRPPXQLW\LPSURYHHI¿FLHQF\ create job potential, and reduce poverty. Keeping the millennium goals and its target indicators in mind, the government has created an environment to allow the ICT sector to grow by: • •
Removing all taxes and levies on computer hardware and software Localizing VoIP (voice over Internet protocol)
•
Opening up the telecom sector for private investment Enacting appropriate laws to support ICT
It has been realized that the key to moving forward is applying ICT for development through strategies partnerships. With this belief, the Bangladesh government has undertaken several key projects in this sector supported by several international development agencies such as UNDP, World Bank, USAID, JICA, KOICA, CIDA, and SIDA. The major projects undertaken so far, with international assistance, are in the areas of banking sector reform, human resources development, policy formulation infrastructure development, and so forth.
The information infrastructure of Bangladesh in the last few years is the result of public examinations, as have been published on the Internet through 7HOHWDON¶VRQOLQHDSSOLFDWLRQIRUQHZVXEVFULEHUV A large number of Bangladeshi have applied online for the US DV lottery, though the reality about the ,QWHUQHWLQ%DQJODGHVKLVWKDWLWLVFRQ¿QHGPRVWO\ to the urban, sub-urban, and semi-urban areas of the country. Metropolitan cities along with district towns and a few small towns also have Internet connectivity. Though the availability of this technology has reached the small towns and cities, neither the rural poor nor even the rich farmers are able to use this technology for their needs. Dhaka, the capital city, has the most Internet XVHUV,WLVGH¿QLWHO\DQHZZD\RIOLIHIRUDODUJH number of people in this populous city. E-governance is currently in an early stage. Only platform-making work is in progress. E-commerce activities like G2G (government-to-government) and G2B (government-to-business) activities are yet to make satisfactory progress. E-banking facilities are providing a high rate of usage on a small scale in comparison to its neighboring countries. The ICT Act was passed last year, but it will not run smoothly without passing many related acts. Bangladesh is steadily marching onward to achieve
75
ICT Competency of Bangladesh to Face Broadband Diffusion
the optimum goal of providing access to the information superhighway. To reach its destination, it needs speed, a favorable environment, a speedy YHKLFOH JRYHUQPHQW¶V SDWURQL]DWLRQ ZLWK IXHO ¿QDQFH DQGDJRRGGULYHU SULYDWL]DWLRQ VRDV to occupy a special position in the global ICT map DQGWRUHWDLQLWVLGHQWLW\LQWKH¿HOGDWSDUZLWKWKH developed countries. This chapter could not touch the economical and social aspects of the broadband diffusion in %DQJODGHVK 7KH VFHQDULR VLWXDWLRQ ¿JXUHV RI data, perceptions and initiatives, thinking, and so forth are changing day by day. In the future one may study the economical and social aspects of broadband diffusion in Bangladesh.
REFERENCES
Forbes. (2005). IT competitiveness. Retrieved February 15, 2007, from http://www.forbes.com/ technology/2005/03/09/cx_0309wef.html ITU. (2003). ITU Internet report 2003: Birth of broadband. Retrieved November 10, 2006, from http://www.itu.int/osg/spu/publications/sales/ birthofbroadband/exec_summary.html Jabbar, M. (2005). Cyber communities of Bangladesh. Retrieved November 10, 2007, from http:// www.ispabd.org/content.php?article.14 Khan, A.S. (2003). Bangladesh is WSIS: Lies and statistics. The Daily Star, (December 19). Khan, F.A. (2005). 2SWLFDO ¿EHU LQIUDVWUXFWXUH in private sector: Experiences, opportunities & constraints.8QSXEOLVKHGSDSHU3ULPH0LQLVWHU¶V 2I¿FH'KDND%DQJODGHVK
Agrawala, S.P., & Shera, J.S. (2004, January 12-13). Hydrabad, Andhra Pradesh. Discussion paper for $VLD,70LQLVWHU¶VQGVXPPLW(,QIUDVWUXFWXUH and E-Learning Group, Department of Information Technology, Ministry of Communications and Information Technology, Government of India, New Delhi, India.
Metrobd. (2006). Homepage. Retrieved November 2006 from http://www.metrobd.org
BTTB. (2006a). Rates of DDN. Retrieved November 10, 2007, from http://www.bttb.gov.bd/home/main/ rates/retes.htm
Ministry of Science and Information & Communication Technology. (2002). National ICT Policy. Retrieved February 8, 2007, from http://www. bccbd.org/html/itpolicy.htm
BTTB. (2006b). Submarine Cable. Retrieved November 10, 2007, from http://www.bttb.gov. bd/home/main/SubmarineCable.htm Hossain, M. (2005, May 21). 2SWLFDO¿EHUEDFNERQH links in Bangladesh and their interconnection with Submarine Cable for connectivity with world information super highway. Unpublished paper, Prime 0LQLVWHU¶V2I¿FH'KDND%DQJODGHVK Islam, A., & Rahman, A. (2006). Growth and development of information technologies in Bangladesh. The Electronic Library, 24(2), 135-146. ISPABD. (2007). A short history of Bangladesh ISP industry. Retrieved February 13, 2007, from http://www.ispabd.org/news.php?item.4
76
Ministry of Finance. (2005). Bangladesh economic review 2005. Dhaka. Ministry of Finance. (2006). Bangladesh economic review 2006. Dhaka.
OECD. (2005). OECD broadband statistics as of December 2005. Retrieved February 7, 2007, from http://www.lirneasia.net/2006/04/iceland-takesover-from-korea-as-broadband.-camp/ Purbo, O. (n.d.). Barriers to ICT diffusion to poor people: An interview with Onno Purbo, Bunung Institute of Technology. Retrieved February 13, 2007, from http://www.idrc.ca/en/ev-46217-2011-DO_TOPIC.html Rahman, H. (2006). Access and infrastructure issues: Context Bangladesh. Retrieved February 13, 2007, from http://bangladeshictpolicy.bytesforall. net/?q=access_infrastructure
ICT Competency of Bangladesh to Face Broadband Diffusion
KEY TERMS Networked Readiness Index (NRI): The :RUOG (FRQRPLF )RUXP¶V 15, PHDVXUHV WKH propensity for countries to exploit the opportunities offered by information and communication technology. It is published annually. The NRI seeks to better comprehend the impact of ICT on the competitiveness of nations. The NRI is a composite of three components: the environment for ICT offered by a given country or community, WKHUHDGLQHVVRIWKHFRPPXQLW\¶VNH\VWDNHKROGers (individuals, businesses, and governments) WRXVH,&7DQG¿QDOO\WKHXVDJHRI,&7DPRQJ these stakeholders (retrieved February 15, 2007, from http://en.wikipedia.org/wiki/Network_readiness_index). Optical Fiber: A glass thread that acts as a guide for light waves. Fibers used in telecommunications usually have a cladding of glass of a lower refractive index. In a communication system, VHYHUDO¿EHUVDUHPDGHXSLQWRDFDEOH UHWULHYHG February 15, 2007, from http://www.science.org. au/nova/021/021glo.htm). SEA-ME-WE 4: Refers to South-East AsiaMiddle East-Western Europe 4, a submarine telecommunications cable linking those regions (retrieved February 15, 2007, from http://en.wikipedia. org/wiki/SEA-ME-WE_4). Submarine Cable: Submarine cable (under water use) consists of copper conductors insulated with India rubber and varnish. Jute Yarn, which acts as a cushion between the cable-core, the inner sheathing of iron cables, provides mechanical insulation (retrieved February 15, 2007, from http://scard.buffnet.net/gofterms/telegloss.html). Synchronous Digital Hierarchy (SDH): Synchronic transmission and multiplex system for telecommunication networks. It operates at speeds of between 155 Mbits/s and (so far) 2.54 Gbits/s. SDH will be the future transmission system in the
telecommunications network and will gradually replace PDH (retrieved February 15, 2007, from http://www.telenor.no/annual_report/rapporter/ drift/ordliste.html). Synchronous Transport Module (STM-1): The basic rate of transmission of the SDH ITU-T ¿EHURSWLFQHWZRUNWUDQVPLVVLRQVWDQGDUG,WKDVD bit rate of 155.52 Mbit/s and is the SDH equivalent of an OC-3 (SONET) (retrieved February 15, 2007, from http://en.wikipedia.org/wiki/STM-1). Unix-to-Unix Copy Protocol (UUCP): An older protocol for sending e-mail between different Unix machines via regularly scheduled modem and network connections. This is the technology utilized by the Usenet for transmitting news postings. Most Internet mail servers now use the SMTP protocol instead (retrieved February 15, 2007, from http://www.websightsolutions.com/faq_gloss. html#sectU). Voice over Internet Protocol (VoIP): A category of hardware and software that enables people to use the Internet as the transmission medium for telephone calls. Voice data is sent in packets rather than by traditional POTS circuits. One advantage of VoIP is that the telephone calls over the Internet do not incur a surcharge beyond what the user is paying for Internet access, much in the same way that the user does not pay for sending individual emails over the Internet (retrieved February 15, 2007, from http://www.novacon.com/faq_s-z.htm). VSAT: Refers to very small aperture terminal, an earthbound station used in satellite communications of data, voice, and video signals, excluding broadcast television. A VSAT consist of two parts, a transceiver that is placed outdoors in direct line of sight to the satellite and a device that is placed indoors to interface the transceiver with the end XVHU¶V FRPPXQLFDWLRQV GHYLFH VXFK DV D 3& (retrieved February 15, 2007, from http://www. geocities.com/westgladeclcp/a6.html).
77
78
Chapter VI
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society Dal Yong Jin Simon Fraser University, Canada
ABSTRACT This chapter attempts to ascertain the causes of the rapid growth of broadband services in the context of the broader socio-cultural elements. It recognizes technology as a socio-cultural product which has historically been constituted by certain forms of knowledge and social practice, so this chapter explores cultural elements contributing to the diffusion of broadband services in the context of the cultural enYLURQPHQWLQ.RUHD)XUWKHUWKLVFKDSWHUGLVFXVVHVWKHVLJQL¿FDQWUROHRIWKHSHRSOHDVXVHUVLQWKH process of the rapid diffusion and growth of broadband services. In particular, it emphasizes the way in ZKLFKWKH HFRQRPLFFULVLVDVRQHRIWKHPRVWVLJQL¿FDQWVRFLRFXOWXUDOWXUQLQJSRLQWVLQPRGHUQ .RUHDQKLVWRU\KDVLQÀXHQFHGWKHGHSOR\PHQWRIEURDGEDQGVHUYLFHVDVKLJKVSHHG,QWHUQHWFRQQHFWLRQV have developed since 1997.
INTRODUCTION The widespread availability of broadband services KLJKVSHHG,QWHUQHW WKURXJKRXWWKHZRUOGVLJQL¿FDQWO\LQÀXHQFHVSHRSOHLQWKHLUOLIH%URDGEDQG services have made it easier to download digital music and movies, and the wide penetration of broadband services has enabled customers to engage in online stock transactions and online games. Many schools, from elementary to universities, in various countries are also connected to broadband
services and utilize information technology (IT) for education, while Internet broadcasters have rapidly become popular. A wide range of developed countries, such as the United States and the United Kingdom to developing countries in the Third World, have initiated modern telecommunications networks that support broadband access, providing high-speed access and always-on connections to both homes DQG RI¿FHV WR GHYHORS WKH VRFLDO DQG WHFKQLFDO landscape of cyberspace (Han, 2003; Ishii, 2003; Lee & Lee, 2003; Jin, 2007). Korea is among the
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
leading performers in broadband (Lee, Oh, & Shim, 2005). As of January 2006, about 77% of Korean households were connected to broadband services, which is one of the highest throughout the world (MIC, 2006). &RQVHTXHQWO\.RUHDKDVEHFRPHWKHZRUOG¶V best laboratory for broadband services—and a place to look to for answers on how the Internet business may evolve (Taylor, 2006). Much scholarly analysis and discourse (Yun, Lee, & Lim, 2002; /HH2¶.HHIH <XQ 5H\QROGV 6DFNV 2003; Choudrie & Lee, 2004; Lau, Kim, & Atkin, 2005; Lee et al., 2005; Jin, 2005) has focused on the roles of the government and competition among telecommunications companies in developing broadband services in Korea. They emphasize several factors contributing to the rapid growth of broadband services caused by the deregulation and competition policies in the telecommunications sector, such as a variety of promotion policies to boost Internet use and the strategies of broadband Internet providers. Previous studies, however, have not paid much attention to socio-cultural factors, which would be one of the most important contributing factors given that people are the major users of broadband services. Although a few papers examine cultural factors that contribute to broadband services, their discussions are neither comprehensive, nor informative.1 Unlike these preceding studies, this chapter attempts to ascertain the causes of the rapid growth of broadband services in the context of the broader socio-cultural elements. It recognizes technology as a socio-cultural product which has historically been constituted by certain forms of knowledge and social practice, so this chapter explores cultural elements contributing to the diffusion of broadband services in the context of the cultural environment LQ.RUHD)XUWKHULWGLVFXVVHVWKHVLJQL¿FDQWUROH of the people, as users, in the process of the rapid diffusion and growth of broadband services. In particular, it emphasizes the way in which the HFRQRPLFFULVLVDVRQHRIWKHPRVWVLJQL¿FDQW socio-cultural turning points in modern Korean hisWRU\KDVLQÀXHQFHGWKHGHSOR\PHQWRIEURDGEDQG services, as high-speed Internet connections have developed since 1997.
TECHNOLOGY AS CULTURAL FORMS It is generally recognized that technologies are primarily neutral because they operate essentially XQGHUWKHVDPHQRUPRIHI¿FLHQF\LQDOOVLWXDWLRQV Many users of technology argue that technology is essentially amoral and an entity devoid of values (Rescher, 1969; Mesthene, 1970). This instrumental theory, the dominant view of modern governments and the policy sciences on which they depend, argues that “if people use technology for destruction or pollution, as in the case of nuclear weapons and chemical pollution, it should not be blamed on technology, but on its misuse by politicians, the military, big business and others” (Pacey, 1983, p. 2). For many scholars, however, technology is not simply a means to an end, but has become an environment and a way of life; this is its substantive impact (Borgmann, 1984). This substantive theory of technology holds that technology is not neutral, but has a substantive value bias. Substantive theory, best known through the writings of Jacques Ellul, Arnold Pacey, and Martin Heidegger, claims that technology constitutes a new type of cultural system that restructures the entire social world as an object of control. Substantive theory explicates cultural aspects of technology, such as values, ideas, and the creative activity of technology (Feenberg, 1991). This type of cultural system is characterized by an expansive dynamic which ultimately mediates every pre-technological enclave and shapes the whole of social life. Among these, Ellul (1964) pointed out that we have to examine technology through its sociological aspects: “We should consider the effect of technique on social relationships, political structures, and economic phenomena.” Technology is not aloof from the social realm; on the contrary, it is an integral part of the social (Webster, 2002). Pacey (1983, pp. 4-6) also discussed how those who write about the social relations and social control of technology should be made aware of the cultural aspects as well as the organizational and technological aspects.
79
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
In agreement with Pacey, Christians (1989, pp. 124-125) observes: “Technology is the distinct cultural activity in which human beings form and transform natural reality for practical ends with the aid of tools and procedures.” He clearly argues WKDWFXOWXUHVDUHKXPDQV¶GLVWLQFWLYHDQGLPPHGLDWH environment built from the material order by men DQG ZRPHQ¶V FUHDWLYH HIIRUW )XUWKHUPRUH :LOliams (1992, pp. 127-129) stated: “How technology develops is not only a matter of some autonomous process directed by remote engineers, but is a matter of social and cultural processes.” Technologies are developed and used within a particular social, as well as economic and political context (Franklin, 1999). As these theoreticians emphasized, technology does not develop independently, but is part of a particular social-economic and cultural process. This chapter aims to contribute to this ongoing debate of substantive theoretical discourse with a case study of the rapidly growing broadband services in Korea.
Historical Background Information on the Growth in Broadband Services The rapid deployment of broadband services in Korea began in 1995 when the government enacted the Framework Act on Information, which HVWDEOLVKHGWKH¿UVW0DVWHU3ODQZKLFKVHWXSD comprehensive strategy for the Korean Information Infrastructure (KII) (MIC, 2004). The goal of the KII was to construct an advanced nationwide information infrastructure consisting of communications networks, Internet services, application software, computers, and information products and services (Jin, 2007). The KII project aimed at building high-speed networks by providing over 80% of households with more than 20Mbps (megabits per second) broadband access by 2005 through market competition and private sector investment (Lee et al., 2003).2 In the post-1997 economic crisis, efforts bringing about the development of broadband services have accelerated. The economic crisis severely affected the Korean economy due to unprecedented rates of unemployment, corporate bankruptcies, and the demise of the stock market.3 .RUHD¶V
80
economic condition demanded that Korea change its industrial structure from traditional heavy and chemical industries to a more IT-oriented structuretelecommunications and computers. 0RUH LPSRUWDQWO\ .RUHD LGHQWL¿HG LQIRUPDWLRQ technology as its path to economic recovery, folORZLQJWKH ¿QDQFLDOFULVLV .LPD 7KH Korean government has exerted great effort on deploying a high-capacity backbone and pursued a policy of high-speed telecom infrastructure as D IRXQGDWLRQ IRU WKH EHQH¿WV RI ,7 FKDQJHV IRU sustainable economic growth (Lee et al., 2003). Under these circumstances, both government SODQQLQJDQGVWDEOHIXQGLQJKDGSOD\HGVLJQL¿FDQW roles in the rollout of broadband services. The Korean government invested a total of $11 billion into broadband services between 1998 and 2002, while the U.S. government planned to invest only $2 billion in the form of tax breaks as of May 2003 (Belson & Richtel, 2003, p. C1). The Korean government has also initiated a variety of promotion policies expediting the growth of broadband services, such as creating Internet-Friendly Classrooms4 and Ten Million People Internet Education. Clearly, .RUHD¶V UDSLG DQG H[SDQVLYH EURDGEDQG JURZWK can be traced, in large measure, to government initiatives (Lau et al., 2005, p. 351). Competition among telecom companies was also dedicated to the deployment of broadband access throughout Korea. Broadband services in .RUHD ZHUH ¿UVW LQWURGXFHG LQ DQG D IHZ telecom companies, including Korea Telecom (KT), the largest telecom company in Korea, Thrunet, and Hanaro, had severely competed with each other in the market. In May 1999, Thrunet, the introducer of broadband services in Korea, held 63% of the market share, followed by Hanaro (35%) and KT (2%). The market has undergone substantive change in recent years. For example, Hanaro Telecom, controlled by U.S. investors American International Group Inc. and Newbridge, absorbed Thrunet to be in a better economic position to compete with the dominant market player KT in late 2005. LG Powercomm, a subsidy company of the LG group, entered the market in September 2005, and cable system operators have expanded their share in the high-speed Internet market (Kim,
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
2006b). As a result, as of April 2006, KT as the largest player consisted of 49.9% of the market share, followed by Hanaro (28.5%), cable system operators (10.6%), Powercomm (4.4%), and others (Park, 2006, p. 16). The intense competition among telecom companies led to price (monthly fee) reductions and subsequently a rapid increase in demand for broadband Internet services. In other words, competition among telecom companies for increasing their respective market share has resulted in a drop in the price of broadband subscriptions, and it has certainly contributed to the rapid spread of broadband in Korea. Fierce competition resulted in prices decreasing from $40 per month in April 1999 to $30 in February 2003, and again to less than $20 in May 2006 (ITU, 2003a; Taylor, 2006).5 Meanwhile, as a new market trend, Korea has been developing the next generation in high-speed Internet services. The country acknowledges the existence of market saturation, so it has enhanced the capacity of existing broadband services, including its speed and convergence with various types of media. Korea has begun to upgrade its high-speed Internet from less than 5 Mbps to 100 Mbps since 2005 in order to enjoy high-quality Internet telephone service, movies, and online education (Bang, 2005). Further, Korea expects the rapid penetration of 100 Mpbs broadband services will expedite the growth of image-related contents, and thereafter the second wave of high-speed Internet services in 10 years.6 As such, the Korean government along with telecom companies played a pivotal role in the growth of broadband services by providing infrastructure and government initiatives.
Cultural Characteristics Contributing to the Deployment of Broadband Services Although the Korean government and telecom ¿UPVKDYHFROODERUDWLYHO\LQLWLDWHGDQGH[SHGLWHG the deployment of high-speed Internet services, the swift growth of broadband services would have been impossible without people readily accepting new technology more than other nations. The diffusion of services and the widespread use of
high-speed Internet can be attributable to certain distinct characteristics of the Korean people, since the diffusion of broadband services took place with rapid acceptance by most Koreans (Jin, 2007). Regardless of the fact that many countries around the world have initiated and supported the growth of broadband services, the result varies markedly due to the acceptance of new technology by people, the users, as well as government initiatives. 7KHUHIRUHLWLVFUXFLDOWRH[SORUHWKHVLJQL¿FDQFH of assessing the culture in which technology is created and the context in which it is widely accepted. As ITU acknowledges (2003b), Internet user demand indeed contributed most decisively to the rapid explosion of broadband in many places, particularly in Korea. Again Ellul (1964) emphasized that “the development of technology is not an isolated fact in society but is related to every factor in the life of modern humanity.” In particular, cultural forms rooted in Korean history after the 1997 economic crisis should be emphasized DVVLJQL¿FDQWHOHPHQWVZKLFKSURPSWHGWKHUDSLG deployment of broadband, because members of the public are the major customers of broadband VHUYLFHV 7KHUH DUH VHYHUDO VLJQL¿FDQW FXOWXUDO characteristics (i.e., the mentality of the people) rooted in Korean society and its historical context that are important contributing factors to the development of broadband, such as rapid acceptance of new services and technologies, enthusiasm for 'edutainment,' keeping up with their neighbors (“me-too” culture), as well as developing dual personalities of two competing forces, social solidarity and individualism. The primary characteristic among most Koreans is demanding quick change, which eventually expedited the growth in high-speed Internet. Only a few decades ago Koreans were characterized as calm and patient, as one distinguished Confucian characteristic, but in just a few years, the demand for quick change became one of the most distinctive Korean characteristics (Han, 2003, p. 19). Since the 1960s when the country began its high-speed transition from an impoverished state to one of $VLD¶VPDMRUHFRQRPLHV.RUHDQVFDQQRWZDLWIRU long periods of time, nor are they very patient. Koreans have not been known for taking things
81
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
slowly. Instead, Korean society is well known for its impulsivenessthat is, desiring quick communication, quick contact, and quick results (Jin, 2007). Most Koreans would certainly agree that Korea is nothing if not dynamic. In this regard, Jong Suk Lee (2000), director of the Korea Culture and Policy Institute, states: “Nine out of ten Koreans rush into the road as soon as the light changes to green and they cross the street on or even before the green light is about to turn red. In contrast, &KLQHVHZRXOGXVXDOO\ZDLWXQWLODOOWKHWUDI¿FLV cleared and safe even at the clear green light. And if the light is about to turn red, they just stop and wait for the next green light.” Indeed, “balli, balli”—Korean for “hurry, hurry”—is heard incessantly on the crowded streets RI6HRXOUHÀHFWLQJWKHIUHQHWLFSDFHRI.RUHDQOLIH in a global era (Jin, 2007). Another survey conducted in the 1980s is still compelling to explain the same trend. According to the survey (Chinese Cultural Connection, 1987), regarding Confucian dynamism which measures long-term vs. short-term orientation, Japan showed one of the highest scores, compared to the U.S., the UK, and Australia. The high Japanese score LQWKHVXUYH\UHÀHFWHGWKHSDWLHQFHRUORQJWHUP orientation of Japan compared with other developing countries (McFadyen, Hoskins, & Finn, 1998). However, Korea has lost patience and has become addicted to short-term results, and this changing Confucian dynamism in Korea has contributed to the rapid penetration of high-speed Internet. This relentless drive, which has led Korea to FKDON XS D QXPEHU RI VLJQL¿FDQW DFKLHYHPHQWV has driven the rapid spread of broadband Internet connections. Most Koreans have hurried to have broadband Internet connections, and it has been easier to develop these services compared to other countries, including Japan. What most Koreans wanted was the same thing: quick communication, quick games, and quick contact. Everything needs to go “balli, balli,” and this unique characteristic of Koreans has expedited the rapid penetration of high-speed Internet (Shim, 2006). In other words, for most Koreans, everything needs to be done right now. For example, in Korea, instant message-swapping is comparatively popular mainly because not
82
answering a message from a friend with all due speed is considered a faux-pas in Korean society (Taylor, 2006). This quick pace has particularly accelerated with the 1997 economic crisis because Koreans began to realize that quick change and adoption of new technology and information was needed to survive in the globalization era. Korea is full RIµHDUO\DGRSWHUV¶ZKRDUHZLOOLQJWREX\QHZO\ released digital devices for consumer testing. In a survey by Korean ad agencies in 2004, 43% of respondents said that they consider themselves µHDUO\DGRSWHUV¶LQWKHPDUNHW .ZRQ $Fcording to a survey by the U.S. market research ¿UP3DUNV$VVRFLDWLRQLQ1RYHPEHU.RUHD ranked second on the list of 13 countries in the adoption of consumer technologies (Parks Associates, 2005).7 Although balli balli culture has several negative aspects, such as bribery scandals and the collapse of buildings and bridges, in the midst of achieving social and economic successes in the Korean society, rapid adaptability to change is the key to the swift deployment of broadband services. An excessive enthusiasm for edutainment—the combination of education and entertainment—has also greatly contributed to the unique growth in broadband services. Above all, Korea is one of the most developed countries in terms of education. Its overall school enrollment rate (primary, secondary, and tertiary) of 90% is the highest throughout WKHZRUOG.RUHD¶VKLJKUDWHRIOLWHUDF\DQGVFKRRO enrollment are essential as prerequisites for the widespread adoption of ICTs, and these factors have helped contribute to the growing impact of ICT in Korean society (ITU, 2003a). Over-enthusiasm for education in Korea is not new, but after the 1997 economic crisis, Internet skills were considered one of the most important survival tools for many Koreans. They devote their attention to the Internet due in part to the Internet being a necessary technology for their jobs, education, and entertainment. In particular, for most parents broadband is a QHFHVVDU\WRROIRUFKLOGUHQ¶VHGXFDWLRQ -LQ Relatively simple initiatives such as encouraging school teachers to post their homework assignments on their own personal Web site and requiring stu-
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
dents to submit their assignments by e-mail could create a feeling among parents that the Internet is a QHFHVVLW\IRUWKHLUFKLOGUHQ¶VHGXFDWLRQ &KRXGULH & Lee, 2004). The majority of Koreans have become fast adopters of high-tech, information-intensive products, and they are highly receptive to marketing stories that offer education after the 1997 economic crisis. Under the economic crisis, parents started to realize the potential threats and opportunities of globalization. It became a must for their children to acquire both English and Internet skills in order to survive in the era of globalization (Aizu, 2002). Many parents who have children younger than early teens, for example, have focused on cyber-based English education since the late 1990s. They have acknowledged that their children are able to enjoy online reading, music, game, and animation written in English through the Internet due to high speed and their being 24/7 support services, which is a good alternative to learning English in a classroom. In the Korean context, excessive need for education has also developed so-called edutainment, as a form of studying English while enjoying playing in cyberspace (An, 2001). Edutainment is not only for children. Many midcareers have also turned their eyes on the Internet in the midst of a changing working environment. In Korea, the majority of companies have begun to DGRSWD:HVWHUQVW\OH¿YHGD\ZRUNZHHNV\VWHP since the late 1990s and early 21st century. The VSUHDGLQJ¿YHGD\ZRUNZHHNV\VWHPVLQFHWKHQ has boosted information technology, such as DVDs, mobile games, and edutainment. With more spare time available, people began to look for a new type of education containing more entertainment factors, instead of the existing cramming type of learning (Cho, 2002). Therefore, having high-speed access to the Internet at home became an advantage for everyone from children to housewives to career persons, or a disadvantage if they did not have it (Jin, 2007). Broadband is an indispensable component for edutainment, which is one of the major contributing factors in broadband services. Meanwhile, the peculiar Korean “me-too culture,” which means all companies and people end up with the same systems and taste, has sub-
stantially contributed to the spread of broadband access. Consumers in various societies are susceptible to social pressure to keep up with their QHLJKERUV HJ³NHHSLQJZLWKWKH-RQHV¶V´LQWKH 86 KRZHYHU.RUHDQV¶HDJHUQHVVWRNHHSXSZLWK their friends and neighbors is comparatively very high (Jin, 2007). As discussed, Koreans tend to be early adopters of technology; however more they VLJQL¿FDQWO\DUHIDVWIROORZHUVDV6WHSKHQ:DUG consultant manager of the telecommunications group at Deloitte observed: Koreans are always conscious of the need not to be left behind by others, and the young have a great desire to conform with the gadget-carrying norm of their peers (Ward, 2004). The majority of Koreans have not admitted any difference in several sectors. Businessmen should get the same salary if they begin the job in the same company at the same time, and parents cannot accept the fact that their children are left behind others. They must be in the best position in terms of the quality and quantity of their houses, education, and income. Korean people buy new refrigerators when their neighbors buy them, and they buy pianos when their neighbors buy them, although they have refrigerators and pianos good enough to use. Likewise, many Koreans are eager to have high-speed Internet because they want to keep up with their neighbors. They feel that if they have the service, they could be part of the group; if not, they could not be part of the group. Technology DQG IDVKLRQ IDGV WHQG WR VSUHDG OLNH ZLOG¿UH LQ Korea. If one individual adopts an innovation, then others feel a social pressure to follow (Bajaj, 2002; Lau et al., 2005). The me-too culture, which is despicable in some senses,8 works in broadband services. Many old and new apartments are installing broadband Internet connections because the construction companies and apartment complexes do not want to be behind other apartment buildings (Cho Han, 2006). People do not want to be left behind in broadband services because they do not like a feeling of uneasiness as in many other areas. The disdainful me-too culture has ironically contributed to the rapid deployment in broadband services.
83
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
Finally, a growing double personality between social solidarity and individualism, in particular in the younger generation, has played a crucial role in the rapid deployment of high-speed Internet. As well known, Korea was a country which was proud of its tradition of social solidarity (Crotty & Lee, 2002). People do not want to be left out. Due to this, the New York Times reported in 2002, “Korea is a group-oriented society, where socializing in bunches is the preferred form of interaction, and Western-style individualism is frowned upon,” while reporting booming online games (French, 2002, p. 8). To take one documented case, after conducting a case study of online games in Korea, Chee (2006) also emphasized the importance of social solidarity of teens and twenties in the process of growing online game communities. Indeed, it LVYHU\SODXVLEOHEHFDXVHWKHÀRXULVKRIWKHRQOLQH communication matched the culture and emotion of the Korean people and their need to communicate with one another through boards, which are forums to discuss and exchange information, and clubs, which brought together people with similar hobbies (Seong, 2006). However, this conventional wisdom has changed in recent years, and individualismthe degree to which individuals are able to achieve identity and status on their own rather than through membership in groups (McFadyen et al., 1998)is on the rise. In this regards, the New York TimesGLGQRWUHÀHFW the rapidly changing environment in Korea. With the swift change in the economy before and after WKH ¿QDQFLDOFULVLVDVZHOODVUHFHQWPRGHUQL]DWLRQ.RUHD¶VVRFLDODQGFXOWXUDOVLWXDWLRQKDV shifted on a large scale. The younger generation increasingly seeks Westernized culture, regardLQJDQLQGLYLGXDO¶VVXFFHVVSULYDWHSURSHUW\DQG nuclear family as important, instead of the value of the Confucian culture emphasizing patriarchy and large families (Jin, 2007). According to a marketing report by SK Telecom, “Those in their VDQGVWKH¿UVW,QWHUQHWJHQHUDWLRQLQ.RUHD gives priority to their own time and space, rejecting all invasions of privacy. They are also changing the culture of collectivism to one of individualism” (Joongang Ilbo, 2002, p. C3). The growing individualism has resulted in a unique culture for
84
developing broadband services and demanding personal time, space, and privacy. The double personality of many Koreans, on one hand social solidarity and on the other hand growing individualism, has ironically contributed to the rapid growth of broadband services. As Chee (2006) observes, young people enjoyed popular online games sometimes with friends and other times with nobody, but through high-speed Internet. This shows social solidarity on the one hand; however, it also explains growing individualistic mentality, which prefers enjoying games with no virtual IULHQGVLQF\EHUVSDFHWRSOD\LQJZLWKRQH¶VSHHUV on the playground. While seeking social solidarity, many young people enjoy high-speed cyberspace by creating their own world, and this changing social cultural characteristic has become crucial for the rapid growth of high-speed Internet. In sum, the explosion of broadband in Korea has been possible due in large part to various cultural factors, rooted deeply in Korean society and its historical context, in particular the 1997 economic crisis. Korean citizens are the most sigQL¿FDQWFRPSRQHQWLQEURDGEDQGGLIIXVLRQEHFDXVH they are the main users of the service. Without high-speed Internet, nobody can have a normal socio-cultural life, including enjoyable individual cyberspace from online games to edutainment, in contemporary Korea, which is one of the fastest growing information societies. As Aizu (2002) points out, the social and cultural factors, the aggressive mentality of the Korean people, high awareness of the challenges of globalization, and political and historical contexts played the decisive role in its dynamic acceptance of the Internet and acceleration to broadband in Korea.
CONCLUSION Korea presents a unique example with its rapid deployment of broadband penetration. Several VLJQL¿FDQW IDFWRUV KDYH FRQWULEXWHG WR WKH UDSLG development of broadband Internet connections. The government and telecommunications companies, as providers, have played important roles in the rapid development of broadband Internet,
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
in particular to providing infrastructure for its development. Favorable government policies and competition among telecommunications companies became driving forces for the rapid deployment of broadband Internet. The explosion of broadband in Korea, however, was greatly made possible against the backdrop of the 1997 economic crisis and due in large part to various deeply rooted social, historical, and cultural factors. Although political factors can be important driving forces behind broadband penetration, growth also requires an existing receptiveness to using the services and applications that can be provided through broadband (Jin, 2007). In this regards, it was the citizens who actually made LQURDGVLQWRWKHZRUOG¶VPRVWVKUHZGPDUNHWIRU broadband services in Korea. Several socio-cultural factors, which are crucial to the diffusion and use of new technologies, have SOD\HG VLJQL¿FDQW UROHV LQ WKH VZLIW GHSOR\PHQW of broadband services. Cultural characteristics emphasizing quick communication and quick responses, as well as enthusiasm for edutainment, have contributed to the exponential growth of broadband services. A growing complexity of characteristics of the younger generation showing double personality of social solidarity and individuDOLVPZKLFKLVQRWHDV\WR¿QGLQDQ\RWKHUFRXQWU\ has particularly contributed to the rapid growth in broadband services. If the younger generation has only one characteristic of these two, it could not be the strong cultural factor for the rapid growth of broadband services. In the midst of globalization that expedited with the economic crisis in 1997, the younger generation has developed its unique characteristics, which resulted in one of the most VLJQL¿FDQW FXOWXUDO IDFWRUV LQ EURDGEDQG .RUHD The majority of Koreans have been eager to gain new technology, and their unique cultural characteristics have become the crucial initiative of the rapid deployment of broadband services. In conclusion, the introduction and development of new technology cannot be explained solely in terms of political structures and economic policy surrounding technology. As Ellul (1964) stated: “One should be looking at technology in its sociological aspect because technology is not an
isolated fact in society but is related to every factor in the life of modern man.” Technology does not develop independently, but is part of a particular social-economic and cultural setup. The deployment of broadband in Korea has been possible due WRLWVVSHFL¿FVRFLRFXOWXUDOGLPHQVLRQVDVZHOO as political and economic environment. Cultural characteristics rooted in the Korean society greatly LQÀXHQFHWKHJURZWKRIWHFKQRORJ\DQGEURDGEDQG services throughout the country.
REFERENCES Aizu, I. (2002, April 22). A comparative study on broadband in Asia: Development and policy. Proceedings of the RIETI Symposium on Asian Economic Integration—Current Status and Prospects. $Q-+ (QJOLVKHGXFDWLRQVSHFLDO /HW¶V play with English online. HanKook Ilbo, 27(August), 26. Bajai, V. (2002). High-sped connections common in South Korea. The Seattle Times, (October 21). Bang, H.C. (2005). The second round of speed war in high-speed Internet. Weekly Chosun. %HOVRQ. 5LFKWHO0 $PHULFD¶VEURDGband dream is alive in Korea. New York Times, (May 5), C1. Borgmann, A. (1984). Technology and the character of contemporary life. Chicago: University of Chicago Press. Chee, F. (2006). The games we play online and RIÀLQH 0DNLQJ:DQJWWDLQ.RUHDPopular Communication, 4(3), 225-239. Chinese Cultural Connection. (1987) Chinese values and the search for culture-free dimensions of culture. Journal of Cross-Cultural Psychology, 18(2), 143-164. Cho, H.Y. (2002). 5-day workweek to spotlight new software sectors: SERI. The Korea Herald, (November 9).
85
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
Cho Han, H.J. (2006). Internet and culture: About the formation of subject in information society. Retrieved September 2, 2006, from http://about. haja.net/bbs/download.asp?ClubNumber=214 7483569&BoardType=3&file=119%08&code_ generation=Ecce_Homo Choudrie, J., & Lee, H.J. (2004). Broadband development in South Korea: Institutional and cultural factors, European Journal of Information Systems, 13(2), 103-114. Christians, C.G. (1989). A theory of normative technology. In E.F. Byrne & J.C. Pitt (Eds.), Technological transformation: Contextual and conceptual implications (pp. 123-139). Dordrecht, The Netherlands: Kluwer Academic. Crotty, J., & Lee, K. (2002). A political-economic analysis of the failure of neo-liberal restructuring in post-crisis Korea. Cambridge Journal of Economics, 26(5), 667-678. Ellul, J. (1964). The technological society. New York: Vintage. Feenberg, A. (1991). Critical theory of technology. New York: Oxford University Press. Franklin, U. (1999). The real world of technology. Toronto: Anansi. )UHQFK + .RUHD¶V UHDO UDJH IRU YLUWXDO games. New York Times, (October 9), 8. Han, G.J. (2003). Broadband adoption in the United States and Korea: Business driven rational model versus culture sensitive policy model. Trends in Communications, 11(1), 3-25. Ishii, K. (2003). Diffusion, policy, and use of broadband in Japan. Trends in Communication, 11(1), 45-61. ITU (International Telecommunications Union). (2003a). Broadband Korea: Internet cast study. Paris. ITU. (2003b, September 16). Number of global EURDGEDQGVXEVFULEHUVJURZVLQ Press Release, ITU, Paris.
86
Jin, D.Y. (2005). Socio-economic implications of broadband services: Information economy in Korea. Information, Communication & Society, 8(4), 503-523. Jin, D.Y. (2007). Hands on/hands off: The Korean state and the market liberalization of the communication industry. Cresskill, NJ: Hampton Press. Joongang Ilbo (English Edition). (2002). Generation 2030 went from apathy to passion. Joongang Ilbo, (December 21), C3. Kim, D.Y. (2006b). Hanaro, share the Wireline Telecom with KT. Dong-A Ilbo, (January 2), 38. Kim, P.H. (2006a). Is Korea a strong Internet nation? The Information Society, 22, 41-44. Kim, T.G. (2005). South Korea remains king of high-speed Internet penetration. The Korea Times, (May 29). Kwon, K.D. (2006). How to stay ahead as digital powerhouse. The Korea Herald, (February 23). Lau, T., Kim, S.W., & Atkin, D. (2005). An examinaWLRQRIIDFWRUVFRQWULEXWLQJWR6RXWK.RUHD¶VJOREDO leadership in broadband adoption. Telematics and Informatics, 22(4), 349-359. Lee, J.S. (2000). Balanced cultural policy. Retrieved September 3, 2006, from http://www.kcpi. or.kr/database/e_sosik/20000102/news_1.html Lee, Y.K., & Lee, D. (2003). Broadband access in Korea: Experience and future perspective. IEEE Communications Magazine, 30-36. /HH+-2¶.HHIH50 <XQ. 7KH growth of broadband and electronic commerce in South Korea: Contributing factors. The Information Society, 19(1), 81-93. Lee, H.J., Oh, S., & Shim, Y.W. (2005). De we need broadband? Impacts of broadband in Korea. Info: The Journal of Policy, Regulation, and Strategy for Telecommunications, Information and Media, 7(4), 47-56. Lim, H. (2001). Hype of education information policy. Kookmin Ilbo, (April 25), 3.
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
McFadyen, S., Hoskins, C., & Finn, A. (1998). The effect of cultural differences on the international co-production of television programs and feature ¿OPVCanadian Journal of Communication, 23(4), 523-538. Mesthene, E. (1970). Technological change. New York: Signet. MIC (Ministry of Information and Communication). (2004). Broadband IT Korea vision 2007. Seoul. MIC. (2006). 2006 Korea Internet white paper. Seoul. Pacey, A. (1983). The culture of technology. Cambridge: MIT Press. Park, Y.S. (2006). High-speed Internet market is going to June war. Munhwa Ilbo, (May 29), 16. Parks Associates. (2005, November 21). Taiwan and Korea lead in the adoption of consumer technologies. Press Release. Reynolds, T., & Sacks, G. (2003, April 9-11). Promoting broadband: Background paper. Proceedings of the International Telecommunications Union’s Workshop on Promoting Broadband, Geneva. Rescher, N. (1969). What is value change? A framework for research. In K. Baier & N. Rescher (Eds.), Values and the future. New York: The Free Press. Seong, N.Y. (2006). How has Korea become Internet powerhouse? The Korea Times, (September 22). Shim, M.S. (2006). Similar aspects between Korean soccer and Korean information technology. HanKook Ilbo, (June 23).
Webster, F. (2002). Theories of the information society (2nd ed.). London: Routledge. Williams, R. (1992). Television: Technology and cultural form. London: Wesleyan University Press. Yun, K., Lee, H.J., & Lim, S.H. (2002). The growth of broadband Internet connections in South Korea: Contributing factors.3DSHU$VLD3DFL¿F5HVHDUFK Center, Stanford University, USA.
KEY TERMS Confucianism: A philosophy of life developed by Confucius. It stressed the proper relationships in society, such as father/son and ruler/subject. The philosophies of Confucius are emphasizing love for humanity; high value given to learning, and devotion to family, peace, and justice. Convergence: Also known as digital convergence. The concept that all modern information technologies are becoming digital in nature. The technological trend whereby a variety of different digital devices such as TVs, in particular highGH¿QLWLRQ79VDQGPRELOHWHOHSKRQHVDUHPHUJing into a multi-use communications appliance employing common software to communicate through the Internet. Edutainment: The combination of education and entertainment. Many people in the digital age use the Internet as a form of studying several subjects, in particular English while enjoying playing in cyberspace.
Taylor, C. (2006). The future is in South Korea. Retrieved June 14, 2006, from http://money.cnn. com/2006/06/08/technology/business2_futureboy0608/index.htm
Information Technology (IT): Compared to the labor-led technologies and industries, information technology usually includes semiconductors, computers, and telecommunications, although some economists like to use the term “knowledgebased industry” in order to explain IT.
Ward, A. (2004). Where high-speed access is going mainstream the Korean experience: Government policy, high levels of urbanization. Financial Times, (June 9), 4.
1997 Economic Crisis: The Korean economy has gone from being an example of one of the most successful development experiences in modern history to economic stagnation and decline in 1997.
87
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
From the middle of 1997, Korea was beset with DVHULHVRI¿QDQFLDOFULVHV7KHWUHQGRIGHFDGHV of rising incomes reversed, and unemployment and poverty were reaching alarming levels. In particular, factors responsible for the decline in the value of exports then include a dramatic fall in the prices of some electronic and information equipment, in particular semiconductors, which had dire consequences for a number of countries in the East and South Asia region. Online Games: Internet games (also known as online games) are games that are played online via the Internet. They are distinct from video and computer games in that they are normally platform independent, relying solely on client-side technologies. Normally all that is required to play Internet games are a Web browser and the appropriate “plug-in.”
4
Social Solidarity: The degree or type of integration of a society. In simpler societies, solidarity is usually based on kinship ties or familial networks. In more complex societies, solidarity is more organic, referring to the interdependence of the component parts. Thus, social solidarity is maintained in more complex societies through the interdependence of its component parts.
ENDNOTES 1
2
3
88
For example, Yun et al. (2002) and Lee et al (2003) mention that the strong emphasis on education and academic performance has prompted parents to turn to the Internet for educational goods and services; however, they do not extend their discussion further. Since broadband penetration was above 77% as of January 2006, Korea almost achieved one of the major goals; however, the speed of broadband, another major goal, could not EHIXO¿OOHGJLYHQWKDWDQDYHUDJHVSHHGRI broadband was only 2.5Mpbs in May 2005 (see Kim, 2005). In 1997, Korea experienced the worst economic recess in modern history, mainly due WRWKHFULVLVLQ¿QDQFLDODQGFRUSRUDWLRQVHF-
5
tors. Thirteen large conglomerates, including Daewoo and Donga, collapsed or became partially insolvent because of the shortage of currency. The immediate result of these FRPSDQLHV¶ FROODSVH ZDV WKH VXIIHULQJ RI small companies and workers. When they failed, thousands of small-scale companies collapsed. As of February 1998, 3,377 midsize and small companies were bankrupt. When companies went bankrupt, workers lost their jobs. By the end of June 1999, the number of jobless reached 1.78 million. This means the unemployment rate rose sharply from 2.6% in the fourth quarter of 1997 to 8.5% in the second quarter of 1999, and real wages fell by 7.9% (see Jin, 2007). The government placed an emphasis on creating Internet-Friendly Classrooms at every level of schooling by connecting them to high-speed Internet free of charge as the main component of educational information infrastructure. The government spent $1.4 billion for the establishment of school information infrastructure between 1998 and late 2000. Of this amount, the government used $600 million to connect classrooms of 10,064 elementary, middle, and high schools around the country with broadband services (see Lim, 2001, p. 3). The relatively low monthly subscription fees have helped the growth of broadband services; however, it is almost certain that a high subscription fee could not deter the spread of broadband services because the majority of Koreans have not taken the price as a factor. For example, most Koreans over teens have had cell phones regardless of high monthly subscription fees. The subscription fees of mobile phones have greatly risen since 1997; however, the penetration rate has also exponentially increased. In the late 1990s, the .RUHDQJRYHUQPHQWRI¿FLDOO\EDQQHGPRELOH carriers from offering handset subsidies in an attempt to block excessive marketing; however, many local mobile users still wanted to have mobile phones, although the full price of many handsets were priced at more than $500
Socio-Cultural Interpretations to the Diffusion and Use of Broadband Services in a Korean Digital Society
6
7
in 2002. There are major factors other than promotion policies and competition among telecom companies in the Korean context which would be cultural factors. 100Mbps broadband services can download the equivalent of a 32-page newspaper per VHFRQGDQGD¿OP *% ZLWKLQWZRPLQutes, compared to one hour with 3.5 Mbps broadband services. About 80% of customers used broadband services of 1-4Mpbs in Korea as of December 2005. Parks Associates surveyed more than 10,000 households in 13 countries and ranked nations
8
according to their proclivity to adopt and use MP3 players, video-on-demand (VOD), home networks, computers, online services, and similar advanced technologies. The survey found that Taiwan (Digital Living Index: 8.7) and Korea (8.4) lead the world in the adoption of consumer technologies, followed by the U.S. (8.1), Canada (8.0), and Japan (7.9) (see Parks Associates, 2005). It is in some ways despicable because Korean people follow others, not because they need, but because they want to be in the same class and enjoy showing off.
89
90
Chapter VII
Structural Changes and Regulatory Challenges in the Japanese Telecommunications Industry Hidenori Fuke Komazawa University, Japan
ABSTRACT The structure of the telecommunications industry in Japan has been changing revolutionarily. The changes DUHREVHUYHGLQ¿YHSKDVHV GHYHORSPHQWRIFRPSHWLWLRQLQWRWKHORFDOFDOOPDUNHW GLIIXVLRQRI broadband Internet and development of inter-platform competition, (3) rapid growth of cellular services and Internet access via cellular, (4) decline of POTS (plain old telephone service), and (5) structural changes from vertical integration to layered structure and development of media convergence. These changes require total review of the regulatory framework that was formed in the POTS era. In this chapter, I propose to review: (a) essential facilities regulation, (b) a universal service system, and (c) DÀDWUDWHSULFLQJV\VWHPRIWKH,QWHUQHWWRVROYHSUREOHPVWKDWDUHOLNHO\WRGLVWRUWWKHQHZLQGXVWU\ structure and would stress the importance of a regulatory system that is competition, technology, and content neutral.
INTRODUCTION The structure of the telecommunications industry in Japan has been changing revolutionarily. Competition has developed from the long distance call market into the local call market. New competitive carriers started to offer subscriber telephone services by leasing dry copper from 177(DVWDQG177:HVW KHUHDIWHUFDOOHGµ177 ORFDOFRPSDQLHV¶ WKDWZDVPDLQO\XVHGIRU'6/
(digital subscriber line) services. The diffusion of the broadband Internet is impressive. FTTH (Fiber to the Home) has started to grow following spectacular growth of DSL. According to a 2005 International Telecommunication Union (ITU) report, Japanese broadband Internet service is the fastest and cheapest in the world. The changes are not limited to quantity. The nature of competition is changing from intra-platform to inter-platform as the broadband Internet is offered on various
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
platforms. The diffusion of cellular services, especially Internet access via cellular and 3G (thirdgeneration) services, has been remarkable. Cellular carriers are competing to offer various services on terminals including IC card functions, music distribution functions, and digital TV. With the rapid growth of broadband Internet, IP telephony has diffused. The diffusion of broadband Internet accompanying the deployment of IP telephony and cellular services has been drastically decreasing POTS SODLQ ROG WHOHSKRQH VHUYLFH FDUULHUV¶ UHYHQXHV It has become clear that the universal service system based on POTS is no longer sustainable. +RZHYHU,QWHUQHWWUDI¿FKDVEHHQFRQWLQXLQJWR increase due to the diffusion of IP broadcasting services and peer-to-peer exchange of moving pictures between users. It has also become clear WKDWWKHSUHVHQWVLPSOHÀDWUDWHSULFLQJV\VWHPRI the Internet is no longer sustainable due to rapid LQFUHDVHLQ,QWHUQHWWUDI¿F With the diffusion of the Internet, especially that of the broadband Internet, the structure of the infocommunications industry is changing from vertical integration to a layered structure. Telecommunications and broadcasting, which have developed as separate industries, are converging. These changes require urgent reform of regulatory systems that were designed in the POTS era. This chapter focuses on three matters that risk distortion in industry structure in the near future. They are: (a) essential facilities regulations designed for POTS, (b) a universal service V\VWHPIRU3276DQG F DÀDWUDWHSULFLQJV\Vtem for broadband Internet. Although reform of regulationsincluding the copyright system that has discouraged the development of IP broadcasting to respond to the development of media convergenceis also urgently required, reform involves not only economic and legal aspects but also cultural discussions. Therefore, I will focus on the above three topics in this chapter. This chapter is organized as follows. First, a brief history of Japanese telecommunications regulatory reform is given as background information. Second, the Japanese telecommunications market
is analyzed to throw structural changes into relief. Third, viability of the application to the broadband Internet market of the essential facilities regulation to promote competition in the POTS market is discussed. I will argue that the imposition of an REOLJDWLRQWRRIIHUGDUN¿EHURQO\RQLQFXPEHQW NTT local companies will distort competition in the broadband Internet market and retard the development of facilities-based competition. Fourth, the recent reform of the Japanese universal service system is analyzed, and I will propose the reform of the present universal service that ensures only POTS to a new system that ensures access to voice communications services regardless of WKHWHFKQRORJLHVGHSOR\HG)LIWKWKHµWUDJHG\RI WKH FRPPRQV¶ RQ WKH ,QWHUQHW LV DQDO\]HG DQG the introduction of a new charging system based on the quality guarantee and transmission speed is proposed. Finally, as concluding remarks, the importance of designing a new regulatory system that is neutral to competition, technology, and FRQWHQW LV HPSKDVL]HG E\ EULHÀ\ WRXFKLQJ XSRQ the copyright problem.
BACKGROUND The telecommunications industry is going through drastic structural changes with the rapid diffusion of the Internet and mobile services. Regulations of POTS assume that physical, service, and content layers are closely associated and competition is basically within the metallic subscriber line platform. Contrary to this, regulation of the Internet and 3G mobile services should be based on a layered approach, as links between layers are cut off and inter-platform competition becomes VLJQL¿FDQW3DVWUHJXODWLRQWKDWPDLQO\IRFXVHGRQ the control of market power of incumbent carriers in local access markets should be restructured to accommodate the new structure. Most regulations in the telecommunications industry were basically MXVWL¿HGE\WKHH[LVWHQFHRIPDUNHWIDLOXUH7KHUH are three main reasons behind market failure in the telecommunications industry1:
91
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
a. b. c.
Existence of carriers with market power2 Asymmetric information between consumers and carriers Network externality
With the structural changes indicated above, these factors are also changing. The telecommunications industry used to be regarded as a typical example of a natural monopoly. But with technological development, privileges awarded to the former state monopolies were revised and competition was introduced in the 1980s. The introduction of competition itself did not eliminate market failure, as incumbent carriers have continued to maintain a de facto monopoly on local telecommunications facilities, especially metallic subscriber lines. The prices of long distance services and services for business users were set above cost to subsidize local services for residential users whose prices were set below cost. New entry began from long distance and business markets because of this price structure. New entrants offered their services by interconnecting with the local network of incumbent carriers that cannot be practically and reasonably replicated. Incumbent carriers offered monopolistic wholesale products (access to local networks) and competitive retail products (local and long distance call services). Under this market structure, there are two choices for ensuring fair competition. One is structural separation, and the other is conduct regulation. In the case of the United States, structural separation was adopted in 1984. In many other countries, conduct regulation without structural separation was preferred. Japan is one of the few countries that introduced competition without having taken a clear position between these two choices. Although structural separation of incumbent NTT (Nippon Telegraph & Telephone) Public Corporation was discussed, the discussion was deferred and it was SULYDWL]HGDVDXQL¿HGHQWLW\RIIHULQJERWKORFDO and long distance services in 1985. At the same time, one of the most important conduct regulations (i.e., interconnection rules) was not well provided for. The regulatory uncertainty lingered on and informal asymmetric regulation by regulatory
92
authorities to make up for the gap was criticized as non-transparent and arbitrary.3 :KHQWKHUHJXODWRU¿QDOO\VHWWOHGWKHSUREOHP by the introduction of interconnection rules in 1997 and the reorganization of NTT under a holding company structure in 1999, the industry structure itself began to change. The nature of competition changed from intra-platform to inter-platform based on a layered structure. This will change the pattern of market failure. Although the MIC (Ministry of Internal Affairs and Communications)4 introduced a revised TBL (Telecommunications Business Law) from April 2004 to respond to the change, it is still built on the ideas valid in the POTS era and it missed the need to protect consumers by abolishing virtually any regulation on prices. Japan took a step forward toward regulatory reform to accommodate the above changes by WKHSXEOLFDWLRQRIWZRUHSRUWVLQ7KH¿UVW is the report published in June by the Panel on Frameworks of Communications and Broadcasting, which was convened by the Minister for Internal Affairs and Communications. The second is the report published in September by the Study Group on a Framework for Competition Rules to Address the Transition to IP-Based Networks,5 which was organized by the MIC. I appreciate that these reports were moves to reform the regulatory framework based on POTS and on the clear distinction between telecommunications and broadcasting. However, it is true that detailed discussions on the actual implementation of these reports still are not free from the way of thinking of the POTS era. Below I analyze the structural changes in the telecommunications industry in Japan since 1985, and I propose policies to accommodate changes in industry structure.
STRUCTURAL CHANGES IN THE TELECOMMUNICATIONS INDUSTRY IN JAPAN Structural changes in the telecommunications industry in Japan are observed in the following ¿YHSKDVHV
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
Figure 1. Growth of broadband Internet Thousand 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0
3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6
Cable DSL FTTH
2000
a. b. c. d. e.
2001
2002
Development of competition into the local market Diffusion of broadband Internet Diffusion of cellular and Internet access via cellular Decline of POTS Development of media convergence based on a layered structure
Development of Competition into the Local Market Since the introduction of competition in the telecommunications industry in 1985, the share of the incumbent (NTT) started to decrease, especially in the long distance market. With the introduction of interconnection rules in 1997 and the adoption of carrier pre-selection in 2001, competition developed into the local call market. The share of the NTT group was 63.9 % in the local market and its share in the inter-prefecture long distance market was 48.2 % as of March 2006. The only market in ZKLFKWKH177JURXSVWLOOKROGVVLJQL¿FDQWPDUNHW share is the traditional telephone access market, where their share as of June 2006 was 89.3%. In addition, new entrants started to offer teleSKRQHVHUYLFHVE\XWLOL]LQJ177ORFDOFRPSDQLHV¶ unbundled metallic subscriber lines (Dry Copper)
2003
2004
2005
2006
in 2004. Their share in subscriber telephone lines as of June 2006 reached 5.5%.
Diffusion of Broadband Internet $VIRU,QWHUQHWDFFHVVYLD¿[HGOLQHVEURDGEDQG access has grown rapidly (see Figure 1). The number of users of broadband Internet reached 24.2 million in June 2006. Now about 48.6% of households are connected via broadband. Although DSL dominates the market with 14.5 million users, the recent growth of FTTH is remarkable. The number of users of FTTH reached 6.3 million, and WKHQHWLQFUHDVHLQWKHQXPEHURIXVHUVLQWKH¿UVW TXDUWHURIH[FHHGHGWKDWRI'6/IRUWKH¿UVW time (see Figure 2). FTTH has continued to grow, while DSL recorded a net loss in number of users for the second quarter of 2006. About 26.1 % of broadband users are now on FTTH. One of the characteristics of the Japanese FTTH market is that both facilities-based and service-based competition have developed. In the wholesale market, NTT local companies and electric power companies are competing. In the retail market, service providers including Softbank are competing with NTT local companies and electric power companies. DSL services have grown rapidly due to the aggressive marketing of Softbank
93
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
Figure 2. Growth of broadband Internet: Quarterly net increase 1600
Thousand Cable DSL FTTH
1400 1200 1000 800 600 400 0 - 200
3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6
200
2000
2001
2002
BB, which took advantage of cheap, regulated line-sharing charges.6 The MIC also imposed unbundling obligation of FTTH at low charges on NTT local companies. Softbank BB set retail prices of FTTH services very low like DSL.7 To FRPSHWHZLWK6RIWEDQN%%¶VFKHDS'6/HOHFWULF power companies set prices of FTTH services very low and NTT local companies followed. Not only intra-platform competition but also inter-platform competition in the broadband Internet access market has developed, and that has contributed to the growth of cheap and fast Internet service. According to the ITU (2005), broadband service in Japan is the most advanced in the world in terms of price and speed. With the growth of broadband, IP telephony has begun to be widely deployed, and even NTT local companies started IP telephony services on FTTH. The MIC reported that the number of telephone numbers assigned to IP telephony reached 12 million at the end of June 2006 which is over 20% of POTS. To this number we must add IP telephony without telephone numbers.
Diffusion of Cellular and the Internet Access via Cellular The number of subscribers of mobile (cellular and 3+6 WHOHSKRQHVH[FHHGHGWKDWRI¿[HGWHOHSKRQH 94
2003
2004
2005
2006
OLQHVLQ¿VFDODQGUHDFKHG PLOOLRQDWWKH HQGRI0DUFKZKLOH¿[HGOLQH 3276DQG ISDN) subscribers decreased to 58.1 million. The mobile subscribers per 100 inhabitants were 71.6, ZKLOH¿[HGOLQHVHUYLFHVZHUH7KHUHYHQXH IURPPRELOHVHUYLFHVH[FHHGHGWKDWRI¿[HGOLQHV LQZKLFKKDVEHHQGHFUHDVLQJVLQFH¿VFDO 2000. Since NTT DoCoMo started Internet access service via a cellular service named i-mode in February 1999, the number of users of Internet access via cellular phones started to grow rapidly. The number of users reached 81.5 million (about 87.0% of cellular subscribers) by the end of September 2006. Japan is unique in global Internet usage in that access via cellular commands an important UROH177'R&R0RVWDUWHGWKH¿UVW*VHUYLFHLQ the world in October 2001 with W-CDMA, and KDDI followed in April 2002 using cdma2000. The number of users of 3G service is also showing rapid growth. By the end of September 2006, total 3G users exceeded 48.9 million (about 62.6 % of cellular users). KDDI started to offer 2.4Mb/s service in November 2003, and NTT DoCoMo started 3.6Mb/s service in August 2006. Cellular LVEHFRPLQJDVLJQL¿FDQWWRROIRUKLJKVSHHGDFcess to the Internet. The characteristics of the Japanese cellular market are not limited to the above mentioned
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
growth. Cellular carriers are competing to add QHZIXQFWLRQVRQFHOOXODUWHUPLQDOV7KH¿UVWLV GLJLWDOFDPHUDIXQFWLRQDQGLWLVGLI¿FXOWWR¿QG cellular terminals without this function. IC card function has been added to terminals, and now cellular terminals are used for various purposes including electronic money, credit card, and train and air tickets. They also are used for entertainment purposes such as music, TV, and electronic books. As a result, the rate of data transmission revenue LQ¿VFDOKDVLQFUHDVHGWRDERXWRIWRWDO transmission revenue compared to decreasing revenue from voice transmission. The lesson to be learned from this development is the importance of facilities-based competition. As cellular carriers are competing on the facilities base, they have scrambled to add new functions on their network and terminals. Service-based competition based RQDSDUWLFXODUFDUULHU¶VWHFKQRORJ\ZRXOGQRWKDYH realized this kind of innovation.
of hours have continued to decline. The share of ¿[HGWR¿[HG FDOOV LQ WHUPV RI QXPEHU RI FDOOV GURSSHGWRLQ¿VFDO7KLVKDVOHGWRD GHFUHDVHLQ177ORFDOFRPSDQLHV¶UHYHQXHVIURP FDOOV VHH)LJXUH &RPSHWLWLRQLQWKH¿[HGYRLFH call market has developed, and voice services are also offered on various platforms other than metallic subscriber lines. These changes call for reform of the universal service system. Japan Telecom (now called Softbank Telecom) and KDDI, major competitors in the long distance market, started to offer their own subscriber telephone services usLQJ177ORFDOFRPSDQLHV¶GU\FRSSHULQDQG 2005 respectively, and competition has extended from the call market to the local access market. As NTT local companies reduced monthly basic charges in 2005 to respond to new competition, their revenues should decrease further.8
Development of Media Convergence Based on Layered Structure
Decline of POTS :LWKWKHFKDQJHVDQDO\]HGDERYHFDOOVIURP¿[HG telephones in terms of number of calls and number
7KLVLPSDFWLVQRWOLPLWHGWRFDUULHU¿QDQFHV7KH diffusion of broadband Internet and 3G and 3.5G cellular blurs the boundary between telecommu-
Figure 3. Decreasing voice revenue 50,000
Milion US$
NTT- East NTT- West NTT Communications NTT (Total)
45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 2000
2001
2002
2003
2004
2005
95
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
nications and broadcasting. While broadcasting over the Internet is already available, terrestrial digital TV service started in December 2003 and broadcasters started digital services over 3G cellular terminals in April 2006. The info-communications industry used to be vertically integrated with a clear boundary between media. With the diffusion of the Internet, especially broadband Internet, the structure has been changing. Any kind of content can be divided into packets and transmitted on the Internet as long as TCP/IP is observed. On the other hand, packets can be transmitted over any physical network, as 7&3,3LVGH¿QHGRQO\E\VRIWZDUH7KXVWKHVWULFW links between physical, service, and content layers are cut off. Horizontal separation of the media industry according to three layersthat is, physical, network, and content layersis developing. It is becoming an urgent challenge to revise the regulatory framework of the info-communications industry to accommodate these changes. One of WKHOHDGLQJWULDOVLVWKHLQWURGXFWLRQRIWKH(8¶V new regulatory framework.9 With the above-mentioned changes the pattern of competition in the telecommunications industry has been changing from intra-platform competition to inter-platform competition. In the case of POTS, competition is based on incumbent carriHUV¶PHWDOOLFVXEVFULEHUFDEOHV,QWKHFDVHRIWKH Internet, various platforms such as FTTH, cable PRGHP*FHOOXODUZLUHOHVV/$1DQG:L0$; are competing. The regulatory framework for POTS is no longer sustainable.
REGULATORY CHALLENGES Essential Facilities Regulation The changes in industry structure and the pattern of competition call for reforms in telecommunications regulation. I will focus on the following three points: a. b.
96
Essential facilities regulation A universal service system
F $ ÀDWUDWH SULFLQJ V\VWHP IRU EURDGEDQG Internet
DSL and Essential Facilities Regulation As mentioned earlier, the pattern of competition is changing from intra-platform to inter-platform. To accommodate this change, it is necessary to revise essential facilities regulation on incumbent carriers.10 In the case of POTS, physical, service, and content layers are closely associated, and the physical layer in the local market is offered as metallic subscriber lines and local exchanges that accommodate them. At the early stage of the introduction, fair competition policy focused on the LQWHUFRQQHFWLRQEHWZHHQQHZHQWUDQWV¶ORQJGLVWDQFHIDFLOLWLHVDQGWKHLQFXPEHQW¶VORFDOWHOHSKRQH network facilities, which were regarded as essential facilities. Service-based competition based on a single platform (i.e., metallic subscriber cables) has developed. Here access regulation has rationale regardless of the trade off between the promotion of competition and investment incentive. Diffusion of the Internet has started to change the pattern of competition. Access to the Internet is obtained through various physical networks, including metallic subscriber lines, FTTH, cable PRGHPFHOOXODUZLUHOHVV/$1DQG:L0$;$V a result, the pattern of competition is moving from intra-platform to inter-platform. This is especially true in the case of Japan. DSL service started to grow through the introduction of line-sharing obligation on NTT local companies and the extensive marketing of new entrants such as Softbank BB. Total DSL users as of March 2006 reached 14.5 million. Softbank acquired market share of 34.8%, compared to the combined share of NTT local companies of 39.2%. However, DSL service is dependent on a single platform, that is, 177ORFDOFRPSDQLHV¶PHWDOOLFVXEVFULEHUOLQHV
Is FTTH Essential? In the case of FTTH, the situation is rather different. Electric power companies entered the market RQDIDFLOLWLHVEDVLV177ORFDOFRPSDQLHV¶VKDUH LQWKHZKROHVDOHPDUNHWDWWKHHQGRI¿VFDO
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
was 78.6%, compared to 94.8 % share in the metallic market. Their share of the retail market for the same time was 74.3 % for independent houses DQGEXVLQHVVPDUNHWDQGIRUÀDWVPDUNHW Their combined shares were less than 50% in some prefectures where electric companies are active. Regardless of the change in the pattern of competition, the MIC has imposed the same kind of regulation for wholesale FTTH as metallic subscriber lines. The regulation includes the REOLJDWLRQWRRIIHUGDUN¿EHUDWUHJXODWHGFKDUJHV and margin squeeze regulation. Retail prices of NTT local companies are not regulated. As the retail market is very competitive, they set retail prices very low to compete with electric power companies and other service providers. Margin squeeze regulation requires that the margin between retail and wholesale prices covers the cost of retail (Cave, 2004). The strategic price setting in the retail market obliges NTT local companies to set wholesale charges below cost.
It is not reasonable to apply this kind of access regulation to the market where inter-platform competition has developed. This will not only disturb the development of facilities-based competition but also prevent the diffusion of new technologies VXFKDVZLUHOHVV/$1DQG:L0$;7HFKQRORJLFDO neutrality is one of the key factors in newly developing markets such as broadband Internet.
A UNIVERSAL SERVICE SYSTEM Universal Service System in Japan The second challenge is reform of the universal service system designed for POTS. With the introduction of competition, the universal service system in the telecommunications industry has been changing from an internal cross-subsidy system of monopolistic incumbent carriers to an industry-level subsidy mechanism. In Japan, to
Table 1. Outline of 2006 reform
97
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
respond to developments since the introduction of competition in 1985, a universal service fund (USF) system was introduced in 2001 by the revision of the Telecommunications Business Law (TBL), which regulates telecommunications businesses. The system did not function at all because it stipulated that subsidies would be given to NTT local companies when the total cost for all services GH¿QHG DV XQLYHUVDO VHUYLFH H[FHHGHG WKH WRWDO revenue from the services. The system was revised LQDVWKHSUR¿WDELOLW\RI177ORFDOFRPSDQLHV was expected to further deteriorate. The outline of the reform is given in Table 1. First, as no change was added to eligible telecommunications carriers, the problem of the 2001 system still remains. Second, local calls were excluded from universal services, and only acFHVV WR ¿[HG WHOHSKRQH QHWZRUNV LV NHSW DPRQJ universal services. Third, under the new system the amount of subsidies is calculated as the difference between cost for high cost areas estimated by TERLIC and national average cost for universal services. High cost areas are those where the upper 4.9% of subscribers belong in terms of cost. More VLJQL¿FDQWO\WKH0,&H[FOXGHGIURPVXEVLG\DUHDV those areas where carriers other than NTT local companies have entered, for the reason that it is not competitively neutral to give subsidies only to NTT local companies in competitive areas (MIC, 2005, p. 46). Fourth, a big change was made in the contribution obligation. Under the new system, carriers that are assigned telephone numbers should contribute to USF according to the number of telephone numbers assigned.
Actual USF Operation The TCA (Telecommunications Carriers Association), which supervises USF, published in September 2006 that the amount of contribution from January 2007 would be 7 Yen (5.9 cents) per number assigned (TCA, 2006). This amount would be revised every six months thereafter. The total amount of subsidy is $128.55 million. The management cost of the USF (i.e., $1.05 million) is added to this amount for a total of $129.6 million. As the
98
number of total telephone numbers assigned is 165.9 million, the amount of contribution to the USF is calculated as 7 Yen (5.9 cents) per month.
Evaluation of 2006 Reform It is an important step forward that the USF in Japan is likely to function by revision of the subsidy system based on cost and revenue to a kind of benchmark method. NTT local companies applied for the designation as eligible telecommunications carriers in January 2006. The MIC estimated that the size of the subsidy would be between $166 and $234 for 2006 and would increase thereafter. However, there are still some problems to be VROYHG 7KH ¿UVW RQH LV WKH VFRSH RI XQLYHUVDO service. As analyzed in the above section, the contribution to USF based on the number of telephone numbers assigned is not competitively neutral. This is because the MIC put simplicity before competiWLYHQHXWUDOLW\7KHEDVLFIDFWRUEHKLQGWKHFRQÀLFW between simplicity and competitive neutrality is the scope of universal service. The MIC limited WKHVFRSHRIXQLYHUVDOVHUYLFHWR¿[HGWHOHSKRQH service. But as shown in earlier sections, voice communication services are offered on various platforms including cellular and the Internet. The concept of universal services is to ensure certain services regardless of the physical network on which they are offered. If voice communication services are to be ensured, the physical network should not be restricted to PSTN. The diffusion of cellular and the Internet, especially broadband ,QWHUQHWKDYHVWDUWHGIURPSUR¿WDEOHXVHUVLQXUEDQ DUHDV $V RQO\ XQSUR¿WDEOH XVHUV LQ UXUDO DUHDV would be left on the metallic telephone network LQWKHQHDUIXWXUHLWZLOOEHGLI¿FXOWWRPDLQWDLQ universal service within the framework of POTS. As voice services are offered on various platforms, universal service should accommodate this change. It is inevitable to change universal service to access to networks on which voice services are available. ,WLVFRQVLGHUDEO\GLI¿FXOWWRUHDOL]HWKLVZLWKWKH system of universal service that gives subsidies to carriers. As various facilities-based carriers and service-based providers are competing in the market, new universal service subsidies should be
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
targeted toward end users. If end users are allowed choice among carriers or providers, it is possible to ensure universal service under an environment characterized by intra-platform competition. The second problem is related to the policy goal of universal service given by the MIC. The MIC explained that the policy goal of universal service is to prevent a geographical gap in telecommunications. However, both economic and geographical affordability are to be taken into account. There are huge differences in cost between exchange areas due to subscriber density. If we set prices according to this cost trend, the price for high cost exchange areas should be prohibitive. However, if we integrate exchange areas into message areas11 and set prices according to average cost for each message area, the difference should not be so big. $WWKHOHDVW177ORFDOFRPSDQLHV¶PRQWKO\EDVLF charges that are set against cost trends should be UHEDODQFHG WR UHÀHFW WKHVH WUHQGV12 Because it LV QRW HFRQRPLFDOO\ HI¿FLHQW WR JLYH VXSSRUW WR everyone in high cost areas, subsidies should be targeted toward low-income families that cannot afford to get access to voice services at competitive prices. However, the above discussions were within the framework of POTS. While discussions on universal service have been dragging on, the structure of the telecommunications industry has shown spectacular changes. NTT announced that it would offer FTTH services to more than 30 million users by 2010 in its medium-term management strategy (NTT, 2004). If this plan is realized, about half of the present POTS subscribers would move to FTTH. This move will make it almost impossible to maintain universal service for users that remain on PSTN. It has become evident that a universal service system that takes only POTS into account is no longer sustainable. Reform of the universal service system from a system that secures POTS to one that secures access to IP networks is urgently required.
NEW UNIVERSAL SYSTEM DESIGN I propose to design a new universal service system WRVROYHWKHDERYHSUREOHPVEDVHGRQ1RDP¶V¿YH criteriacompetitive neutrality, technological neutrality, structural neutrality, applications and content neutrality, and geographical neutrality (Noam, 1997, pp. 115-116). The basic framework of my proposal here is to give subsidies to lowincome consumers in high cost areas who cannot afford services included in universal services at competitive prices. This will eliminate the presHQWHFRQRPLFLQHI¿FLHQF\E\WDUJHWLQJVXEVLGLHV toward users in high cost areas. It is necessary to designate carriers of last resort (COLRs), to specify services, low-income users, high cost areas, amount of subsidy, and contribution method. First, it is necessary to designate a COLR or COLRs that accept a commitment to supply services as universal service to all consumers in the area, to ensure that there is at least one carrier that offers universal service. Second, services subsidized as universal service VKRXOG EH UHGH¿QHG DV DFFHVV VHUYLFHV WR YRLFH transmission services, regardless of the physical network on which they are offered and technologies deployed. The reason why voice call services are excluded is that voice call services have already been offered very cheaply or virtually at no cost in the case of IP telephony, and economic and geographical affordability has already been realized. Therefore, the affordability problem left is with access to voice services. As this proposal is not affected by the kind of technologies employed, it is technologically neutral. Further, as the proposal does not question whether content and applications DUHRIIHUHGRYHU¿[HGWHOHSKRQHFHOOXODURUWKH Internet, content neutrality is ensured. 7KLUGWKHGHFLVLRQUHJDUGLQJLGHQWL¿FDWLRQRI low-income consumers is primarily a political and social matter. Although I will not elaborate on it here, my tentative proposal is to adopt criteria for livelihood subsidy. Fourth, high cost areas are selected based on the actual cost of the least costly COLR for each area. There is trade off between the size of areas and the possibility of cream skimming, when ac-
99
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
tual cost varies. If we adopt individual exchange areas, the difference is too large. Therefore, I propose to adopt message areas that integrate several exchange areas as a basis for the calculaWLRQRIFRVW7KHQKLJKFRVWDUHDVDUHGH¿QHGDV areas where the cost of the most cost-effective COLR in the area exceeds the national average of the most cost-effective COLR in each area. This will ensure geographical neutrality. However, the precondition for the introduction of this system is WKHUHEDODQFLQJRI177ORFDOFRPSDQLHV¶PRQWKO\ basic charges. Present monthly charges that are inversely set against cost will promote cream skimming entry into urban areas, and NTT local FRPSDQLHV ZLOO ORVH SUR¿WDEOH XVHUV 7KH SULFH of dry copper is set at 1,334 Yen ($11.4)/month in the case of NTT-East and 1,405 Yen ($12.0) in the case of NTT-West. Comparing these prices with monthly basic charges, it is clear that it will QRWEHGLI¿FXOWIRUQHZHQWUDQWVWRHDUQSUR¿WVE\ targeting business users in urban areas. NTT loFDOFRPSDQLHV¶EXVLQHVVHVZLOOIDFHGLI¿FXOWLHVLI they do not rebalance basic monthly charges. To UHPRYHSROLWLFDOUHEDODQFLQJGLI¿FXOWLHVWKLVQHZ universal service system is required. Fifth, the amount of subsidy given to each low-income user is the difference between the price of the cheapest access service of the COLR or COLRs in the area and the national average of the cheapest access service of COLR or COLRs in each area, excluding high cost areas. This will ensure competitive neutrality. Sixth, relevant low-income consumers will be given vouchers for the above difference and will be able to use it for any access service offered in the area by any carrier. This will ensure competitive and technological neutrality. Seventh, the subsidy will be covered by a fund in which every telecommunications carrier will contribute according to their net revenue, calculated by reducing charges paid to other telecommunications carriers from their total revenue.13 It is competitively neutral because all telecommunications carriers contribute to the fund, and it is also structurally neutral because the amounts to be contributed are the same between vertically integrated carriers and non-integrated carriers that purchase input from
100
other carriers. It is desirable to exclude small carriers to avoid excess burden.14
Flat-Rate Pricing System for Broadband Internet Internet and the Tragedy of the Commons With the development of the broadband Internet DQG,3EURDGFDVWLQJGLVFXVVLRQVRQµIUHHULGHRQ WKHQHWZRUN¶DQGµQHWZRUNQHXWUDOLW\¶DUHDWWUDFWLQJ many people in Japan. However, the discussions originated in the United States, and I doubt the appropriateness of the application of the American discussions to Japan, where the structure of broadband business is different. Even so, discussions in Japan are not free from the criticism that they lack the proper understanding of the layered structure of the Internet and the actual status of the interconnections among ISPs. I would stress that the problem should be discussed from the viewpoint of the pricing system inherent in the ,QWHUQHW WKDW LV OLNHO\ WR OHDG WR WKH µWUDJHG\ RI WKHFRPPRQV¶ One of the contributing factors behind the rapid GLIIXVLRQRIWKH,QWHUQHWLVWKHÀDWUDWHFKDUJLQJ system. This system can be broken down into two elements: the interconnection structure between ISPs and the connection between end users and ISPs. There are two patterns in the interconnection between ISPs. One is peering and the other is transit.153HHULQJLVDNLQGRIµELOO NHHS¶V\VWHP and no interconnection charge is paid between ISPs concerned. Under transit charging, ISPs requesting interconnection to another backbone ISP must pay interconnection charges. The charges are negotiated between ISPs concerned and hidden behind non-disclosure clauses, so it is hard to know actual conditions. My interviews with some Japanese ISPs support that the charges are generally at a ¿[HGUDWHEDVHGRQWKHFDSDFLW\RILQWHUFRQQHFtion lines. In both cases, the incremental cost of sending extra packets into interconnecting ISPs LV ]HUR 7KHUH LV OLWWOH LQFHQWLYH WR OLPLW WUDI¿F between ISPs. At the same time, there exists little
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
incentive to expand network capacity on the part RIXSSHUWLHU,63VDVLQFUHDVLQJWUDI¿FGRHVQRW generate extra revenue. Internet access in Japan that started by dial-up PRYHGWRÀDWUDWHZLWKWKHGLIIXVLRQRIWKH,QWHUQHW DSL, which has grown rapidly, has been offered E\ÀDWUDWHDVLWLVRIIHUHGRQPHWDOOLFVXEVFULEHU lines without passing through telephone exchanges. )77+ZKLFKKDVVWDUWHGWRJURZVLJQL¿FDQWO\LV DOVRRIIHUHGRQDÀDWUDWHEDVLV:LWKWKHGLIIXVLRQ of broadband Internet, various applications such as IP telephony and IP broadcasting services have been offered over the Internet. They are offered IUHHRUE\ÀDWUDWHDVIDUDVFRQWHQWWUDQVPLVVLRQLV FRQFHUQHG8QGHUWKHÀDWUDWHV\VWHPLQFUHPHQWDO cost for users of receiving and sending extra packets is zero, and there is little incentive for them to limit WUDI¿F7KHUHUDWKHUH[LVWVLQFHQWLYHWRLQFUHDVHWUDI¿FDVEHQH¿WVIRUXVHUVLQFUHDVHZLWKWKHTXDQWLW\ of packets they send and receive. ISPs design and build their network capacity EDVHGRQWKHHVWLPDWHGDYHUDJHWUDI¿FRIDYHUDJH users and offer services by the best effort principle. If some of their users send and receive an extremely large volume of packets within a limited time, most of their network capacity will be consumed by these XVHUVDQGFDXVHGHOD\LQRWKHUXVHUV¶FRPPXQLFDtions or degrade communication quality. ISPs, which have been competing intensely since the start of broadband services, claim that they cannot afford to neither introduce a measured system nor expand their network capacity at pace ZLWKWKHUDSLGLQFUHDVHLQ,QWHUQHWWUDI¿F7KHODFN of a centralized governing Internet body helps them with this strategy. The total quality of transmission on the Internet is not managed, and matters are left to the best effort of each ISP. In principle, if one of their networks is overrun by concentrated big WUDI¿FSDFNHWVDUHURXWHGWRRWKHU,63V¶QHWZRUNV This is another reason why ISPs lack incentive to expand their own networks. From the above analysis it is clear that the LQFHQWLYH WR FRQWURO WUDI¿F GRHV QRW ZRUN ERWK RQWKHSDUWRI,63VDQGXVHUV7KXVWKHµWUDJHG\ RI FRPPRQV¶ WKDW ZDV UDLVHG E\ +DUGLQ becomes a real threat to the Internet. Goods and services that are rivalrous but are not excludable
DUHGH¿QHGDVFRPPRQUHVRXUFHV7KH,QWHUQHWFDQ be regarded as one of these common resources, as it is hard to exclude the use of network capacity while consumption of network capacity by users is rivalrous. The Internet has intrinsic cause to induce the same kind of tragedy. The Internet has developed as an open and autonomous network by the interconnection of many networks. In the case of POTS, telecommunications operators centrally manage the quality of the network. The quality of the Internet is maintained by the best effort of each provider. There is no central body to control the quality of the network as a whole. The capacity of each ISP is limited. If an individual user pours too many packets into the network, the capacity for other users decreases and that causes negative externality in the form of decrease in transmission speed or deterioration in transmission quality. If an individual ISP gains as its client a content provider offering moving picture services, too many packets will be sent out to upper-tier ISPs and the same kind of negative externality will appear. There is little economic incentive for individual users and ,63VWRFRQWUROWUDI¿FDVDQDO\]HGDERYH However, there is one big difference between the commons that Hardin observed and the Internet. Land resources (i.e., commons) cannot be expanded. In the case of the Internet, network capacity can be expanded. If we succeed in giving ISPs enough incentive to expand their network, we may be able WRDYRLGWKHµWUDJHG\RIWKHFRPPRQV¶
Diffusion of the Broadband Internet and Network Neutrality $OWKRXJK WKH µWUDJHG\ RI WKH FRPPRQV¶ RQ WKH Internet has long been pointed out, it has not been regarded as a real threat when access to the Internet is via narrowband. When the contents of the data transmitted over the Internet were text based, the sizes of packets were small compared to music or moving pictures. Even if many texts are sent, the load for the network is small. The rapid diffusion of the broadband Internet has changed the situation, as music and moving pictures are starting to be exchanged over the
101
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
Internet. The form of the circulation of music and moving pictures are not limited to downloading from content providers. Users have started to exchange music and moving picture data in the IRUPRISHHUWRSHHUGDWDH[FKDQJHE\XVLQJ¿OH H[FKDQJHVRIWZDUH,Q-DSDQWUDI¿FRYHUWKH,QWHUnet is growing rapidly and ISPs are facing demand to expand network capacity. Expansion of capacity means cost increases on the part of ISPs, and if they can recover, the cost from users there will be no problem. However, this is not an easy task in a competitive market ZKHUHÀDWUDWHFKDUJLQJLVHVWDEOLVKHG)DFLQJWKLV trouble, American carriers started to accuse content SURYLGHUVIRUWKHµIUHHULGH¶RQWKHLUQHWZRUNV)RU example, Edward Whitacre, the CEO of SBC, raised the problem in an interview with Business Week (2005). He claimed that content providers should bear the fair network cost, as telecommunications carriers and cable operators have made a huge investment to construct networks. Lessig, Wendel, and Carlsmith (2006) criticize this attitude, fearing that this will discriminate against particular services offered by content and service providers by imposing extra charges (access-tiering). 7KH HFRQRPLF VROXWLRQ WR WKH DERYH FRQÀLFW between telecommunications carriers and content and application providers is rather simple. I propose a system to give investment incentives to telecommunications carriers and give users incentives to FRQWUROWKHLUWUDI¿F
Problems for Each Layer 7KH$PHULFDQWHOHFRPPXQLFDWLRQVFDUULHUV¶DVsertion that content and application providers should bear the fair burden of network costs has been copied by some Japanese telecommunications carriers. Are these discussions rightly based in the Japanese broadband environment? The majority of broadband access in the United States is offered by cable modems. As no obligation to open their networks is imposed on cable operators, basic transmission functions and ISP functions are bundled. In the case of DSL, the share of incumbent local telephone companies amounts to 96.3% (FCC, 2006). Verizon, which merged with
102
MCI, and the new AT&T, born by the merger of SBC and the old AT&T, are also offering Internet access services together with local telephone services. Physical layer and network layer services are bundled, but discussions on the free ride on networks have been made without clear distinction of layers. However, Japanese and American market structures are different. The Japanese broadband market is characterized by the layered structure. In the case of the physical layer, various services such as DSL, FTTH, cable modem, and wireless LAN are competing. As for the network layer, ISPs other than physical network providers are competing, and various content and application services are offered on the network layer service. :KHQZHGLVFXVVWKHµWUDJHG\RIWKHFRPPRQV¶ on the Internet, it is necessary to take this layered structure into account. Leased lines used for the backbone transmission of ISPs are offered by telecommunications carriers such as NTT local companies, NTT Communications, and KDDI. These carriers recover the cost of leased lines E\ÀDWUDWHDQGWKHFRVWLV¿[HGUHJDUGOHVVRIWKH mode and degree of use. Therefore, there arises no incremental cost on the part of carriers even LI,63WUDI¿FLQFUHDVHVE\WKHH[FKDQJHRIPXVLF and moving picture data on leased lines. Blame for a free ride is not accurate in the case of the physical layer. It is on the network service layer where the µWUDJHG\RIWKHFRPPRQV¶FDQEHDWKUHDW:LWK WKHJURZWKRI,QWHUQHWWUDI¿FFRQJHVWLRQRFFXUV unless ISPs increase their network capacity. If a IUDFWLRQ RI XVHUV SRXU WRR PXFK WUDI¿F LQWR WKH Internet where transmission capacity is limited, services for other users will deteriorate because of the delay or the loss of data. This is a typical H[DPSOHRIWKHµWUDJHG\RIWKHFRPPRQV¶
Realistic Solution to the Tragedy of the Commons It seems that it is more than easy to propose an economic solution to this problem. The solution is to give ISPs incentives to expand network capacity DQG JLYH XVHUV LQFHQWLYHV WR FRQWURO WUDI¿F 7KH
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
pricing system by which the cost of facilities used exclusively by a particular user is recovered by ¿[HGUDWHDQGWKHFRVWRIIDFLOLWLHVWKDWDUHVKDUHG by more than one user is recovered by usage based UDWH LV HFRQRPLFDOO\ HI¿FLHQW 8QGHU WKH XVDJH based charge, each user consumes up to the point ZKHUHWKHEHQH¿WDQGFRVWIRUKLPRUKHUPDWFKHV DQGWKLVZRUNVDVDQLQFHQWLYHWRFRQWUROWUDI¿F ISPs are given incentives to invest in expansion of network capacity, as their revenues increase with WKHDGGHGWUDI¿F,QWKHZRUOGRI3276DWZR SDUWWDULIIV\VWHPRI¿[HGEDVLFPRQWKO\FKDUJHV and usage-based call charges has been adopted in many countries.16 However, it is not easy to adopt a two-part pricing system for the Internet. One of the contributing factors in the rapid diffusion of the Internet is the network design principle to eliminate as many functions as possible, including the usage-based charging function. If we add new charging functions to the Internet, it might damage operational DQGHFRQRPLFHI¿FLHQF\7KH,QWHUQHWLVIRUPHG by the autonomous and decentralized interconnection of innumerable LANs differently from
PSTN, which is centrally designed and controlled by telecommunications carriers. If one route is FRQJHVWHGWUDI¿FZLOOEHGLYHUWHGWRRWKHUURXWHV,I we wish to introduce a usage-based pricing system DQGUHÀHFWWKHXVDJHRQWKHVHWWOHPHQWEHWZHHQ ISPs, it is necessary to add charging functions to all routers and to exchange charging information between many ISPs. This is not realistic. Although Guputa, Dale, and Whinston (2005) propose a kind of priority pricing model,17 the actual implementation of the system is not easy, as it entails large FRVWIRUUHFRQ¿JXUDWLRQRIWKHQHWZRUN My proposal is to set monthly charges based on the combinations of service guarantee and transmission speed for end users, including application and content providers. My idea is to divide Internet access services into two classes. One is the besteffort class and the other is the guaranteed class. Each class is subdivided into some classes based on access speed. Under the system, those users that constantly send and receive moving pictures will subscribe to the guaranteed high-speed class. Those users that do not send and receive moving pictures so frequently will subscribe to the non-
Figure 4. Interconnection between GyaO and ISPs
103
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
guaranteed high-speed class. Those users that send and receive mainly e-mails will subscribe to the non-guaranteed low-speed class. ISPs will get extra revenues to expand their network capacities by the introduction of this scheme. As a result, those users who want to send and receive large amounts of data at a guaranteed speed can use the network without interfering with RWKHUXVHUV¶WUDI¿FDQGZLWKRXWEHLQJLQWHUIHUHG ZLWKE\RWKHUXVHUV¶WUDI¿F,QWKLVZD\WKHUHLV DQRSWLRQIRUOHVVWUDI¿FLQWHQVLYHXVHUVWRHQMR\ economical services. Of course, it is up to ISPs whether to adopt or not to adopt this system, as they are doing business in the unregulated world. Therefore, it is necessary to build incentives to adopt this kind of system on the Internet. The incentives have already been built in the interconnection system between ISPs. The problem is that ISPs have not made use of it. I will elaborate based on the Japanese Internet TV service named GyaO18 offered by USEN Corporation (see Figure 4). GyaO is not a simple content provider. As it has an AS (autonomous system) number, it is regarded as an autonomous network like ISPs. GyaO buys transit from other ISPs in some cases or concludes peering arrangements with other ISPs in other cases. In the case of transit arrangement, the solution is simple. If users that belong to national tier 1 ISPs from whom GyaO is buying transit subscribe to GyaO, they may charge GyaO based on the capacity and transmission speed of the interconnecting circuits (a in Figure 4). In some cases GyaO users belong to secondary ISPs that buy transit from national tier 1 (b in Figure 4). It is necessary to change the transit arrangement between national tier 1 and secondary ISPs as above. When GyaO buys transit from secondary ISPs (c in Figure 4), the ISPs concerned may change the contract to a usage-based one. In this case it is possible that RYHUÀRZHGWUDI¿FLVURXWHGWRQDWLRQDOWLHUYLD ,; ,QWHUQHWH[FKDQJH GLQ)LJXUH 7KHVROXWLRQLVWRUHÀHFWWKHWUDI¿FRQWKHWUDQVLWFRQWUDFW between GyaO and national tier 1, and to change the contract between GyaO and secondary ISPs at the same time. To introduce this kind of system, it is necessary for national tier 1 to measure the
104
WUDI¿FEHWZHHQ*\D2DQGLQWHUFRQQHFWLQJ,63V As this is limited to the interconnection with networks having AS numbers, it will not become a large burden for national tier 1. In the case of national tier 1 ISPs that are peering with GyaO (e in Figure 4), it is necessary to change the contract to transit. If the change is not possible because of competition with other ISPs, they must accept the situation and rearrange the relationship with secondary ISPs as above. National tier 1 and secondary ISPs are facing the trade off between expanding their network capacity without extra revenues and expanding network capacity by the extra revenues generated by bearing the cost of introducing a new charging system. It is up to their business decision. It is clear from the above discussion that there is a means to give economic incentives to avoid the tragedy of the commons resulting from the increase LQ,QWHUQHWWUDI¿F,WLVXSWRWKHEXVLQHVVGHFLVLRQ of the ISPs in the competitive market. If they think that the cost of expanding network capacity will OHDGWRGHWHULRUDWLRQLQWKHLU¿QDQFLDOSHUIRUPDQFH their rational business decision should be the introduction of a new charging system.
CONCLUSION I have pointed out that the regulations designed for POTS are not applicable to broadband Internet based on the analysis of changes in the Japanese info-communications industry structure. I also propose the revision of essential facilities regulation, the universal service system, and the introduction of a new charging system for the Internet based on quality guarantee and transmission speed. However, the area in which regulatory reform is required with the development of media convergence with the rapid diffusion of the broadband Internet is not limited to the above matters. It is necessary to redesign the regulatory framework that has been developed with a clear distinction between telecommunications and broadcasting. A copyright system is among them. Because a copyright system applied to IP broadcasting is categorized as telecommunications, it is neces-
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
sary to get consent from all the people concerned, including actors or actresses that appeared only a few seconds in a movie. In the case of TV broadcasting, a package approval system is adopted and it is easy to rebroadcast movies over TV. Thus IP broadcasting is disadvantaged compared to TV broadcasting. Although many people, including myself, are discussing the need to revise, many people on the broadcasting side are reluctant to accept the revision by emphasizing the cultural value of broadcasting. This is against the principle of competitive, technological, and content neutrality. To conclude, I will mention how the Japanese experience could help in the diffusion of broadband Internet. First, it is important to impose line-sharing obligation on an incumbent carrier at considerably low prices where the broadband market is characterized by intra-platform competition. At the same time, it is advisable to promote facilities-based competition once the market starts to follow a growth curve, as service-based competition basically lacks innovative function. Second, the diffusion of broadband will stimulate the development of various applications on the Internet and this will require the revision of regulations designed for POTS.
REFERENCES Baldwin, R., & Cave, M.E. (1999). Understanding regulation. Oxford: Oxford University Press. %XVLQHVV:HHN $W6%&LW¶VDOODERXW³VFDOH and scope.” CEO Edward Whitacre talks about the $7 7ZLUHOHVVDFTXLVLWLRQDQGKRZKH¶VPRYLQJ to keep abreast of cable companies. BusinessWeek Online Extra, (November 7). Retrieved September 1, 2006, from http://www.businessweek.com/magazine/content/05_45/b3958092.htm?chan=search Cave, M.E. (2004). Remedies for broadband services. Journal of Network Industries, 5(1), 23-49. Crandall, R.W. (2005). Competition and chaos: U.S. telecommunications since the 1996 Telecom Act. Washington, DC: Brookings Institution Press.
Crandall, R.W., & Waverman, L. (2000). Who pays for universal service? When telephone subsidies become transparent. Washington, DC: Brookings Institutions Press. European Union. (2002). Directive 2002/22/EC of the European Parliament and of the Council of March 2002 on universal service and users’ rights relating to electronic communications networks and services (Universal Service Directive). FCC. (2000). The digital handshake: Connecting Internet backbones. OPP Working Paper No. 32, Federal Communications Commission, Washington, DC. FCC. (2005). Policy statement (adopted: August 5, 2005; released: September 23, 2005). Washington, DC. FCC. (2006, July). High-speed services for Internet access: Status as of December 31, 2005. Retrieved September 1, 2006, from http://hraunfoss.fcc.gov/ edocs_public/attachmatch/DOC-266596A1.pdf Fuke, H. (2003a, August). Evaluation of new regulatory framework of European Unionfrom Japanese perspective. Proceedings of the International Telecommunications Society European Regional Conference, Helsinki, Finland. Fuke, H. (2003b). Spectacular growth of DSL in Japan and its implications. Communications & Strategies, (52), 175-191. Fuke, H. (2005). Structural changes in telecommunications industry in Japan and the regulation. Communications & Strategies, (59), 113-139. Geradin, D., & Kerf, M. (2003). Controlling market power in telecommunications: Antitrust YV VHFWRUVSHFL¿F UHJXODWLRQ Oxford: Oxford University Press. Guputa, A., Dale, O., & Whinston, A.B. (2005). 3ULFLQJWUDI¿FRQLQWHUFRQQHFWHGQHWZRUNV ,VVXHV approaches, and solutions. In S.K. Majumdar, I. Vogelsang, & M.E. Cave (Eds.), Handbook of telecommunications economics—voume2: Technology evolution and the Internet. Amsterdam: North-Holland.
105
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
Hardin, G. (1968). The tragedy of the commons. Science, 162, 1243.
KEY TERMS
Hausman, J.A. (1998). Taxation by telecommunications regulation: The economics of the e-rate. Washington, DC: AEI Press.
Essential Facilities:,QFXPEHQWFDUULHUV¶IDFLOLties that are essential for the new entrant. When it is neither practical nor reasonable to copy the LQFXPEHQW FDUULHUV¶ IDFLOLWLHV WKH\ PXVW DOORZ new entrants to access the facilities on rational FRQGLWLRQV ,QFXPEHQW FDUULHUV¶ ORFDO QHWZRUN facilities are regarded as essential facilities, and such obligations as unbundling and charging based on TERLIC are imposed.
ITU. (2005]). ITU Internet reports: The Internet of things. Geneva. Kovacic, W. E. (1992). The antitrust law and essential facilities in public utility regulation. In M.A. Crew (Ed.), Economics innovations in public utility regulation. Kluwer Academic. Lessig, L., Wendell, C., & Carlsmith, E.M. (2006, February 7). Testimony of Lawrence Lessig, C. Wendell and Edith M. Carlsmith, Professor of Law, Stanford Law School, Senate Committees on Commerce, Science and Transportation, hearing on “network neutrality.” Retrieved September 1, 2006, from http:// commerce.senate.gov/pdf/lessig-020706.pdf Majumdar, S.K., Vogelsang, I., & Cave, M.E. (Eds.). (2002). Handbook of telecommunications economics— volume 1: Structure, regulation and competition. Amsterdam: North-Holland. MIC. (2005, October 25). The review of the Universal Service Fund system (Report of the Telecommunications Council) (in Japanese). Mitchell, B.M., & Vogelsang, I. (1991). Telecommunications pricing theory and practice. Cambridge: Cambridge University Press. NTT. (2004, November 10). NTT group’s mediumterm management strategy. Press Release, NTT. Noam, E.M. (1997). Reforming universal service: A prerequisite to competition and interconnection. In E.M. Noam & A.J. Wolfson (Eds.), Globalism and localism in telecommunications. Amsterdam: North-Holland. Noam, E.M. (2001). Interconnecting the network of networks. Cambridge, MA: MIT Press. Telecommunications Carriers Association. (2006, September 15). The application for permission to MIC on the Universal Service Fund (in Japanese). Press Release.
106
Facilities-Based Competition: When new entrants build their own network to compete with incumbents, it is called facilities-based competition. Inter-Platform Competition: When carriers compete with each other using various platform WHFKQRORJLHVVXFKDVPHWDOOLFFDEOHRSWLFDO¿EHU wireless LAN, or cable modem, it is called interplatform competition. Intra-Platform Competition: When new entrants compete with incumbents by leasing facilities from incumbents, the competition develops on a single platform like metallic cable. This type of competition is called intra-platform competition. Layered Structure: When the industry structure is divided according to the division of layers, it is called layered structure. In the case of the Internet, the industry is divided into physical, network, and content layers. POTS (Plain Old Telephone Service): Traditional public telephone service is called POTS when compared to the Internet. Service-Based Competition: When new entrants compete with incumbents by leasing such facilities as local access networks from incumbents, it is called service-based competition. Tragedy of the Commons: As the use of the commons is rivalrous but it is hard to exclude use by members of the community, it is likely to be RYHUFRQVXPHGDQG¿QDOO\EHUXLQHG7KLVNLQG of negative externality is called the tragedy of
Structural Changes and Regulatory Challenges in the JapaneseTelecommunications Industry
the commons. The Internet is a kind of commons facing the same kind of risk.
5
6
Vertical Integration::KHQD¿UPLVRSHUDWing in both upstream and downstream markets, it is regarded as vertically integrated. In the case of the traditional telephone industry, physical, network, and content layers are closely associated, and it is regarded as a typical pattern of vertical integration.
7
8
9 10
11
ENDNOTES 12 1
2
3
4
As for telecommunications and regulation, see Baldwin and Cave (1999), Geradin and Kerf (2003), and Majumdar, Vogelsang, and Cave (2002). The existence of market power by itself does QRWMXVWLI\H[DQWHUHJXODWLRQE\VHFWRUVSHFL¿F law. For background discussions, see Fuke (2005). The MIC is the successor ministry of the MPT (Ministry of Posts and Telecommunications) created at the time of restructuring of Japanese government organizations in 2001. It was formerly called MPHPT (Ministry of Public Management, Home Affairs, Posts and Telecommunications) in English.
13
14
15
16
17
18
Available at http://www.soumu.go.jp/joho_ tsusin/eng/pdf/060926_1.pdf See Fuke (2003b) for details. Some doubt the long-term success of this strategy (Crandall, 2005, p. 144). NTT local companies estimate that total decrease in revenue will be 140 billion Yen ($1.192 million). See Fuke (2003a) for details. As for essential facilities, see Kovacic (1992). Japan is divided into 567 message areas (i.e., local call areas). The MIC itself admits the need for rebalance (MIC, 2005, p. 43). This system is similar to the NetTrans Account System proposed by Noam (2001, pp. 237-238). The MIC excluded those carriers whose revenue is under 1 billion Yen ($8.5 million). As for peering and transit, refer to FCC (2000). As for two-part tariffs, see Mitchell and Vogelsang (2001, pp. 73-117). Guputa, Ostal, and Whinston (2005, pp. 422431) proposed to introduce a service class that offers priority transmission service in the case of congestion and to impose extra charges for the service. http://www.gyao.jp/
107
Division III
Australia and New Zealand
109
Chapter VIII
$QDO\VLVRIWKH,QیXHQFHRI Australia’s Government Policy on Broadband Internet Access Qiuyan Fan University of Western Sydney, Australia
ABSTRACT Like many other governments in the world, the Australian government has taken a multi-faceted approach to promoting broadband Internet access. This chapter provides an in-depth analysis of the impact of policy issues on broadband Internet access in Australia. The primary goal of this chapter is to develop a holistic understanding of Australia’s national approaches pertaining to broadband Internet access. This research has clearly indicated that the state of broadband Internet access is closely related to the government’s policy and regulatory framework. The government has been basing its actions on market forces as a principal driver for broadband Internet connectivity. However, market forces only play their roles in improving broadband Internet access in the major cities and have little effect in regional and rural Australia. The research has indicated that the regulatory competition regime, by and large, has failed to address concerns of market dominance and market power in the telecommunications sector as is evidenced by a relatively lower price-performance ratio of broadband services in Australia.
INTRODUCTION Australia is a wealthy country with a highly educated population and high dial-up Internet penetration. It might be expected that Australia would be among the leaders in broadband Internet access. However, in the past few years, it has been overwhelmingly reported that Australia is lagging behind in terms of broadband access penetration. The International Telecommunication Union (ITU) data indicates that as of January 2005 Aus-
tralia was ranked outside of the top 20 economies worldwide in broadband uptake per 100 inhabitants (ITU, 2005). The Organization for Economic Cooperation and Development (OECD) statistics show that Australia has slipped from 13 in 2000 to 21 in RQWKHOLVWRI2(&'FRXQWULHV¶XVHRIEURDGband (Working Party on Telecommunications and Information Services Policy, 2001; OECD, 2005). In December 2001, Australia was ranked 16th in the OECD, with broadband penetration of 0.8 per 100 inhabitants, compared to Korea at 17.2, Canada at
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
8.8, and the OECD average at 2.9 (OECD, 2005). By December 2004, the OECD ranked Australia 21st out of 30 countries in broadband uptake per 100 inhabitants. Australian broadband subscribers per 100 inhabitants were 7.7, compared to Korea at 24.9, Canada at 17.8, and the OECD average of 10.3 (OECD, 2005). Australian government data indicate that around 28% of home Internet users were connected to broadband in May 2004; this compares with the leading economies (e.g., Hong Kong 84% and U.S. 48%) (DCITA, 2004b). $OWKRXJKWKHUHDVRQVIRU$XVWUDOLD¶VVORZEURDGband growth and uptake are complex, broadband Internet access is determined by a combination of the widely available telecommunication infrastructure and the accessibility and affordability of Internet services, which are closely related to government policies. As the OECD suggests: If countries wish to take advantage of the enabling capabilities of networks such as the Internet, to implement information infrastructure initiatives, they need to urgently address the extent to which WKHXQGHUO\LQJFRPPXQLFDWLRQSROLFLHVLQÀXHQFH current growth rates.” (OECD, 1996) Other literature suggests that the underlying reason for Internet access can be traced back to the regulatory regime in each country, because government policies governing the telecommunications service market and promoting information LQIUDVWUXFWXUHKDYHDVLJQL¿FDQWLPSDFWRQWKHDIfordability and availability of broadband access. Governments around the world have moved from developing policy frameworks that encourage the growth of e-commerce to involvement in broadband deployment. The consensus reached by 175 countries taking part in the World Summit on WKH,QIRUPDWLRQ6RFLHW\ LQ*HQHYDFRQ¿UPV that broadband connectivity is a central enabling agent in building the information society, and information and communication infrastructure is an essential foundation for an inclusive information society. More recently, the 2004 Global Symposium for Regulators (GSR) encouraged political support at the highest government levels for the promotion
110
of low-cost broadband connectivity, and stressed that regulatory policy must be developed to promote universal access (GSR, 2004). In common with other national governments, the Australian government recognizes that timely access to affordable high-speed networks is critical to long-term economic development. Many national approaches to promoting Internet development have been taken since the early-1990s. How have those strategic and regulatory developments impacted on broadband access? To answer this question, this chapter examines: • • •
National strategies aimed at supporting broadband deployment The Internet market and the regulatory context within which the market operates Some main initiatives and funding programs that have been put in place by government to promote broadband Internet access
This chapter also analyzes the impact of various policy issues on Internet broadband access. As the OECD (Working Party on Telecommunications and Information Services Policy, 2000) suggested, the main policy areas examined are: • • • • •
Policy support for infrastructure competition Policy support for unbundling local loop (ULL) Policy support for Internet interconnection, Policy support for the development of broadband/high-speed access options Policy support for equitable and affordable access
The goal of this chapter is to develop a holistic XQGHUVWDQGLQJRI$XVWUDOLD¶VQDWLRQDODSSURDFKHV pertaining to broadband Internet access. The objectives deriving from the goal are to: • • •
Explore national approaches to promoting broadband Internet access Examine the key policy issues surrounding broadband Internet access Understand the impact of policy issues on broadband Internet access
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
This chapter investigates the relationships between policy and broadband access by providing a detailed and comprehensive examination of policy development and the impact of national approaches RQEURDGEDQG,QWHUQHWDFFHVV7KHUHVHDUFK¿QGings in this study suggest the relationships between the elements of policy and broadband access, and indicate how the various policy aspects interplay to affect broadband access. 0RUHRYHUWKHUHVHDUFK¿QGLQJVRIIHULQVLJKWV into the regulatory outcome and effectiveness of national approaches, which will be helpful to inform and improve current practice in Australia. “National governments are not isolated actors and thus their actions and policies increasingly must take into account what is happening in other countries” (OECD, 1997). The researcher views broadband Internet access in Australia as a revelatory case to the extent that it provides an opportunity to identify and describe problems that may be common to other countries.
NATIONAL STRATEGIES AIMED AT SUPPORTING INTERNET INFRASTRUCTURE DEVELOPMENT The Australian government has developed some ICT policies for the information economy since the HDUO\ V+RZHYHUQRQHRIWKHPVSHFL¿FDOO\ addresses the challenge of promoting broadband infrastructure development. Despite the fact that the Strategic Framework for the Information Economy (1998) and the 2004-06 new Strategic Framework proposed by the current Australian government set the priority areas for advanced telecommunication infrastructure, the key priorities for action in the telecommunications sector rely heavily on private sector and market development (DCITA, 2004a). However, given that very large investment is required in broadband rollout, it is not possible to leave it up to market forces. The Senate report pointed out: There is clearly a need for long-term government involvement and leadership on telecommu-
nications infrastructure, particularly in relation to rural and regional Australia, as market forces will not be able to deliver broadband networks that reach all citizens. The future development of Australia’s telecommunications network is too important to be left solely to the decisions of SUR¿WGULYHQSULYDWHVHFWRUV´ (Australian Senate, 2004b) The Australian Competition and Consumer Commission (ACCC, 2005b), in its report on telecommunications infrastructure, notes that during 2003-04, carrier investment in xDSL services only accounted for 6% of the total investment in telecommunications infrastructure. Telstra, the incumbent carrier, and other major operators planned to undertake modest levels of investment in their networks for 2004-05 (Australian Senate, 2005). In his address to the Australian Telecommunications Users Group on March 10, 2005, the ACCC Chairman, Samuel (2005) pointed out that “Telstra admits it has no current plans for any VLJQL¿FDQW LQYHVWPHQW LQ ¿EHU WR WKH KRPH DQG claims the existing copper network has another 15-20 years of useful life in front of it.” However, WKH H[LVWLQJ FRSSHU ¿[HGOLQH QHWZRUN LV YHU\ limited and will not be able to provide the level of services which the public and businesses will need in the not too distant future (Australian Senate, 2004b). Even Telstra described the longevity of its copper network, through which ADSL services DUHGHOLYHUHGDVEHLQJµ¿YHPLQXWHVWRPLGQLJKW¶ (Lundy, 2004). Decisions about the rollout of broadband in Australia have largely been made on the basis of commercial considerations. Government agencies VXFKDVWKH$&&&DQGWKH$%6GH¿QHEURDGEDQG around speeds of 2Mbps, which is about the FDSDELOLWLHVRI7HOVWUD¶VADSL service and much less than on the capacity of telecommunications networks operating at international best practice levels (ACCC, 2005a; ABS, 2005). This has resulted in a slow broadband transmission rate. For example, the Australian market widely accepts a minimum download transfer rate of 256 kbps and upload rate of 64 kbps as a broadband speed.
111
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
Compared to the best overseas networks (e.g., Korea, 6Mbps/640kbps; Sweden, 4Mbps/400kbps; Canada, 3Mbps/640kbps; HongKong, 6Mbps/ 256kbpsVHH7DEOH 7HOVWUD¶V['6/EURDGEDQG Internet capability lags well behind. Frieden (2005) states: “The acquisition of comparative advantages in ICT development appears impossible without some degree of government involvement.” As Cutler (2001) commented: “The world faces the major task of reinvesting in next generation infrastructure to deliver high capacity bandwidth. Broadband infrastructure is the future. Only governments and public administration can re-allocate public resources to establish next generation infrastructure in the public interest.” Many countries have developed the public policy necessary both to invest in and deploy broadband networks or upgrade their existing networks. One of the most important factors accounting for the successful rollout of broadband in Korea was government strategic commitments and investment. This research has clearly indicated that the current national strategies are not achieving the rollout of a world-class broadband infrastructure and the range of more affordable broadband services that exist in other OECD countries such as Korea, Canada, and Japan.
REGULATORY AND COMPETITION ISSUES The laws and regulations applied in competition policy in the telecommunications market broadly take two forms: telecommunications regulation and competition law (ITU, 2002). The government introduced open competition into the telecommunications industry in 1997. The telecommunications regulatory framework that was introduced in 1997 consists of the Telecommunications Act of 1997 DQGWKHWHOHFRPPXQLFDWLRQVVSHFL¿FFRPSHWLWLRQ provision in the Trade Practices Act (TPA) of 1974. The telecommunications competition provisions in WKH7UDGH3UDFWLFHV$FWDUHFRQWDLQHGLQ3DUW;,% DQWLFRPSHWLWLYHFRQGXFW DQG3DUW;,& DFFHVV regime). This section will explore some regulations DQG FRPSHWLWLRQ SROLFLHV LQÀXHQFLQJ EURDGEDQG Internet access, and examine the linkages between regulatory regimes, market environments, and broadband Internet access along the levels of telecommunications supply.
Liberalization of the Australian Telecommunications Sector The Telecommunications Act of 1997 and its related package of legislation came into force in July
Table 1. Comparison of maximum ADSL speeds for residential customers in the selected economies (source: company Web sites, accessed November 2005) Maximum Connection Speeds Country
Company Upload
Download
Korea
Korea Telecom
640 kbps
8 Mbps
Sweden
Telia
500 kbps
8 Mbps
Hong Kong
PCCW
256 kbps
6 Mbps
Canada
Bell Canada
800 kbps
4 Mbps
Sympatico
112
United States
SBC
384 kbps
4Mbps
Australia
Telstra
256 kbps
1.5 bps
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
1997. In the legislation, past regulatory barriers to market entry were removed and open competition was introduced. According to this new regulatory framework, there are no restrictions on entry to any telecommunications market. There is no limitation on the number of carrier licenses that may be issued by the Australian Communication and Media Authority (ACMA). ISPs are not subject to licensing requirements. As a result, a number of carriers and service providers have entered the Australian telecommunications market since the opening up of the Australian telecommunications market to full competition in 1997. However, most of them have focused their infrastructure investment on the CBD areas of the major capital cities. There are many players, but they usually offer only one type of service or offer services only in VSHFL¿FJHRJUDSKLFDUHDV,QRWKHUZRUGVWKHUHPD\ not be inter-modal competition in some areas. Apart from inter-modal competition, another key element in a competitive broadband market is the presence of strong competitive carriers (ITU, 2003b). In Australia, Telstra remains the major wholesale and retail supplier of telecommunications services, including (Willett, 2004): • • • • •
Dial-up and broadband Internet Local, national, long-distance, international, and mobile telephony Data Printed and online directories Pay-TV (through its 50% ownership interest in Foxtel)
Based on its research on the EU and the U.S. telecommunications market, the ITU (2002) concluded that “a market share of 40-50% is highly LQGLFDWLYHRIGRPLQDQFH´7KHH[WHQWRI7HOVWUD¶V dominance of the sector was demonstrated by the fact it had about 68% of the broadband market, which is almost three times the market share that its closest competitor, SingTel Optus, had (Australian Senate, 2004a). It is reported that Telstra receives almost 60% of total industry revenue, which is four times as much as the revenue that Optus receives (Australian Senate, 2004b).
The ITU (2003a) concluded that “these countries that have prospered often have a deep-pocketed second carrier that provides a real competitive challenge to the incumbent, such as Hanaro Telecom in Korea or Yahoo BB! in Japan.” The ITU research evidence indicates that without strong competitive carriers, networks will not be developed and used to their full potential. New services (such as highspeed Internet) will not be introduced as early as they otherwise would. Services will not be provided HI¿FLHQWO\DQGDWOHDVWFRVWIRUFRQVXPHUV
Regulatory Issues about Industry Structure The Network Economics Consulting Group (NECG, 2004) stated that “an important issue in national competition policy and related reforms has been the structure of industries that were formerly dominated by government business enterprises.” :LQVORZ KHDGRIWKH2(&'¶V&RPSHWLWLRQ Policy Outreach Program (1996-2002), stressed that “the OECD has encouraged its member state to give serious consideration to full structural separation in the course of regulatory decisions, especially in the course of privatisation and liberalisation.” However, in Australia, telecommunications reform did not follow this advice. The ACCC and the NCC (National Competition Committee), the two authorities responsible for implementing competition policy reforms in Australia, have both raised concerns related to “an apparent failure of the Commonwealth Government to conduct a review of the merits of breaking up Telstra prior to partial privatization” (NECG, 2004). As discussed in the previous sections, Telstra is one of the most integrated telecommunications companies in the world and has dominated all the major telecommunication service markets, which has anti-competitive implications. For instance, Telstra did unfairly favor its own retail business units over its wholesale customers (the details can be found in the following sections). Despite this, the ACCC noted that: The Australian government explicitly instructed the Productivity Commission not to consider
113
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
Telstra’s structural separation in the review of WHOHFRPPXQLFDWLRQVVSHFL¿F FRPSHWLWLRQ UHJXlation that commenced in June 2000, and abandoned a parliamentary inquiry into the question of structural separation in 2003. Consequently, issues surrounding the possibility of structural separation of Telstra have not been subject to a comprehensive assessment. (ACCC, 2003) In addition to the concerns about the integration issues, the cross-ownership issue is also of major concern in the telecommunication industry. Ownership of a network is a key determinant of market conduct. Telstra owns two of the three major local access networks outside the CBDs of major cities. They are the copper (PSTN) network that connects virtually every household and business in Australia and the largest cable (HFC) network, ZKLFKSDVVHVPLOOLRQKRPHV)HZRI7HOVWUD¶V competitors have any real alternative to using 7HOVWUD¶V FXVWRPHU DFFHVV QHWZRUN IRU QDWLRQDO coverage, including the second largest carrier in Australia, Optus, which owns the other HFC network, which passes approximately 2.2 million homes (Australian Senate, 2004a). Most competitors are dependent on Telstra and Optus to provide Internet services and they are competing at the UHWDLOOHYHOE\UHVHOOLQJ7HOVWUD¶VVHUYLFHV 7HOVWUD¶VFURVVRZQHUVKLSRIWKHODUJHVWFDEOH network and the ubiquitous customer access network means that Telstra has little incentive to roll out both DSL and cable modem technologies, since the two would compete against each other. This has resulted in the lack of potential price and VHUYLFH EHQH¿WV WKDW VXFK FRPSHWLWLRQ EHWZHHQ 7HOVWUD¶VFRSSHUDQG+)&QHWZRUNVFRXOGGHOLYHU to consumers. For instance, Telstra lowered the speed of its most basic cable modem service to 256 kbps from 512 kbps when it launched DSL services, since the then most basic cable offer at 512 kbps was not providing better value than the most basic DSL offer to be launched at 256 kbps. This example highlights the disadvantages of one company owning both infrastructures (Working Party on Telecommunications and Information Services Policy, 2001). Incumbent telecommunication carriers own cable networks in only nine countries out of 30 114
OECD countries, and Australia is one of the only two countries out of the nine countries where the incumbent owns the largest cable network and is also the largest broadband service provider (Working Party on Telecommunications and Information Services Policy, 2002). Countries are increasingly encouraging competition between the local loop and cable TV infrastructure as a viable means of promoting facilities-based competition (ITU, 2002). As early as 1999, the EU adopted a direcWLYHWKDWUHTXLUHG¿UVWDFFRXQWLQJVHSDUDWLRQDQG then the legal separation of cable TV networks from incumbent telecommunications operators. The United States Telecommunications Act of 1996 restricts cross-shareholdings between cable and telephony operators. The ITU analysis suggests that separation of cable TV and the copper networks may help infrastructure competition (ITU, 2002). In Australia, however, there is “no hope for LQIUDVWUXFWXUH FRPSHWLWLRQ ZLWK WKH 0LQLVWHU¶V refusal to accept the ACCC recommendation that Telstra divest itself of its share in Foxtel” (Budde, 2003). OECD experience suggests that competition between DSL and cable modem services has been the main driver of overall broadband penetration, particularly in the residential market (Working Party on Telecommunications and Information 6HUYLFHV 3ROLF\ )RU LQVWDQFH &DQDGD¶V rapid development of broadband services largely resulted from competition between independently owned competing networks or platforms. The ITU study on promoting broadband concludes that “countries where no cable TV network is available, or where the incumbent cross-owns both the telephone and cable TV networks, generally lag behind their counterparts in developing broadband” (ITU, 2003b).
The Competition Regulatory Regime 7KH¿UVWDQGODVW stage of the provision of Internet services is the connection between end users and ISP networks, which can be established through use of dial-up or other technologies such as ADSL, cable modem, and satellite, and interconnection
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
between different ISP networks. Few of the players in the market have any real alternative but to VHHNDFFHVVWRRULQWHUFRQQHFWZLWK7HOVWUD¶VORFDO access network for the supply of Internet access service nationwide. The key areas for dealing with WKHSUREOHPVSRVHGE\VLJQL¿FDQWPDUNHWSRZHU in broadband competition are access to the ubiquitous local loop owned by Telstra, and Internet interconnection.
Unbundling the Local Loop (ULL) Customer access networks, or local loops, are a ¿QDOOLQNEHWZHHQWKHFXVWRPHU¶VWHUPLQDOHTXLSment and ISP networks. Competition at this layer is less well developed in Australia. Telstra almost totally dominates local access services with 94% PDUNHWVKDUH)RUVWHU LGHQWL¿HGWKHODFNRI competition at access level as the primary inhibitor to broadband access in Australia. Budde (2002a) argued that “the country is lagging behind in the rollout of broadband as there is simply no one competing with Telstra.” Broadband access can be provided by infrastructure such as copper (using xDSL technology), ¿EHURSWLFDQGVDWHOOLWHRUWHUUHVWULDOPLFURZDYH Given that the PSTN network is already in place, many countries regard competition in the local loop to be a prerequisite for effective competition in the provision of broadband services. This is particularly true in Australia where the incumbent, Telstra, owns a ubiquitous local access QHWZRUN)HZRI7HOVWUD¶VFRPSHWLWRUVKDYHDQ\UHDO DOWHUQDWLYHWRXVLQJ7HOVWUD¶VQHWZRUNVHUYLFHVDVDQ input for providing their own services, and many are simply competing at the retail level by reselling 7HOVWUD¶V VHUYLFHV ,Q WKH FRQWH[W RI FRPSHWLWLRQ policy, access to the unbundled local loop of the incumbent is critical to increase competition in the local loop. The ACCC decided to regulate the supply of ULL in July 1999, which would enable competitors WRXVH7HOVWUD¶VFRSSHUOLQHVWRGHOLYHUEURDGEDQG service via xDSL technologies. The declaration of ULL in 1999 by ACCC was regarded as an important regulatory development in Australia, DV$&&&¶VGHFLVLRQHQDEOHVFRPSHWLWRUVWRDWWDFK
WKHLU RZQ IDFLOLWLHV WR 7HOVWUD¶V FXVWRPHU DFFHVV networks. It is particularly useful for new entrants to deliver high-speed data service such as ADSL (Besley, 2000; ITU, 2001). It also enables access seekers to build residential ADSL networks in FRPSHWLWLRQ ZLWK 7HOVWUD¶V RZQ $'6/ QHWZRUN However, local loop unbundling appears to have KDGQRVLJQL¿FDQWHIIHFWLQDFFHOHUDWLQJWKHUROORXW of xDSL in Australia, as is evidenced by the lower level of broadband uptake. The ULL does not seem to be doing its job. Consequently, xDSL service quality is lower and pricing is higher since the maMRULW\RIFRPSHWLWRUVDUHVLPSO\UHVHOOLQJ7HOVWUD¶V broadband products and are not able to compete on the price and quality of broadband services because they do not have access to the underlying costs of the network (Willett, 2005).
Internet Interconnection While the PSTN and ULL relate to the provision of Internet services to end users, Internet interconnection refers to the access arrangements for technical interoperability and pricing among ISPs. Given that the Internet consists of local and regional networks interconnected with one another, HI¿FLHQWInternet interconnection is crucial to the operation of the Internet. This section examines how the current Internet interconnection arrangements have impacted on the provision of Internet access, and how the government has addressed Internet interconnection issues. Australia has used the regulatory mechanism to address anti-competitive activities since the LQWURGXFWLRQ RI WKH WHOHFRPPXQLFDWLRQVVSHFL¿F SURYLVLRQLQ3DUW;,%DQG3DUW;,&RIWKH7UDGH 3UDFWLFHV$FWRI $OWKRXJK3DUW;,&LVGHsigned to deal with interconnection, the ACCC mainly uses the competition notice regime set out in 3DUW;,%WRGHDOZLWKLVVXHVDULVLQJIURP,QWHUQHW interconnection. For instance, Telstra was taking advantage of market power to refuse or impose higher interconnection charges on interconnection until the ACCC issued two competition notices XQGHU3DUW;,%RIWKH7HOHFRPPXQLFDWLRQV$FWRI 1997 in May and June 1998 (ACCC, 1998). Under the pressure of the ACCC notices, Telstra signed
115
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
interconnection peering agreements with OzEmail, Optus, and Connect.com. However, none of the smaller ISPs have established a peering relationship with any of these Big Four (Telstra, Optus, 2](PDLODQG&RQQHFWFRP VLQFHWKH$&&&¿UVW tackled the issue in 1998 (ZDNet, 2003). The ACCC has received complaints from ISPs who were excluded from the domestic peering agreements in recent years (ACCC, 2004a). Some smaller ISPs indicated that “they were being charged excessive WUDQVLWUDWHVIRUSDVVLQJWUDI¿FRYHUWKH%LJ)RXU ,63V¶QHWZRUNVDQGWKHUHZDVOLWWOHRUQR¿QDQFLDO recognition provided by the Big Four in relation to the investments they had made on their own smaller networks” (ACCC, 2004b). Others claimed that “they were equivalent in some respect to one or more of the Big Four networks and should enter into peering arrangement to avoid the transaction costs involved with transit arrangements” (ACCC, 2004b). With transit agreements all these Big Four ISPs charge downstream ISPs for Australian data, but the downstream ISPs are not able to charge the larger ISPs, which puts many smaller downstream ISPs at a competitive disadvantage. Many of the smaller ISPs have called on the ACCC to intervene in the matter (ACCC, 2004a, 2004b). However, WKH$&&&FRQ¿UPHGLWVGHFLVLRQnot to declare DQLQWHUFRQQHFWLRQVHUYLFHXQGHU3DUW;,;RIWKH Trade Practices ActLQ LWV ¿QDO UHSRUW HQWLWOHG Internet Interconnection Service (ACCC, 2004b) in January 2005.
Internet Access Pricing and Competition Internet access pricing is an important factor affecting access to the Internet. If people do not view Internet access pricing as affordable, they will not use it. Australia has high levels of penetration of telephone lines and personal computers and dial-up Internet services but, as has been pointed out, low levels of broadband take-up. One factor that contributes to the disparity of broadband access among countries is the cost of accessing Internet services. Generally, an end user pays ISP subscription fees to access the Internet, and
116
an ISP builds, buys, or leases domestic and international transmission links required to provide a retail Internet access service, or it relies on the transmission networks of other ISPs via Internet interconnection relationships. Therefore, the main cost components of an Internet access service are domestic and international transmission capacity, access to an Internet interconnection point via a transmission tail connecting to other networks, and provision of Internet access to end users via a customer access network or wireless technology, many of which will be purchased from Telstra or other telecommunications providers such as Optus and AAPT. It is apparent that the cost depends to a large extent on the conditions of competition in an infrastructure market (Umino, 2002). Higher prices from the carriers directly translate into higher prices for consumers, which in turn retard Internet expansion and use. In short, Internet access cost is directly related to the regulatory framework and level of competition permitted in the local telecommunications sector.
Impact of the Level of Market Concentration on Internet Access Pricing &RPSHWLWLRQKDVDVLJQL¿FDQWLPSDFWRQ,QWHUQHW access pricing. The level of market concentration is one indicator of the degree of competition within a market. In 2004 the Big Four had 72% of customers with permanent connections, 89% of domestic IP network capacity, and 87% of international IP transmission capacity (ACCC, 2004a). The market share of the Big Four ISPs accounted for 51% of the retail subscriber customer base and 80% of the wholesale market (ACCC, 2004a), in contrast to more than 500 ISPs offering retail Internet access service and approximately 50 ISPs offering some form of Internet interconnection in Australia. Four RXWRI¿YH,63VLQWHUFRQQHFWZLWKRQHRUPRUHRI the Big Four. Since the Big Four compete in the retail markets along with their competitors, smaller ISPs, they have naturally been inclined to delay or refuse to interconnect with competitors on a peering EDVLV*LYHQWKHVLJQL¿FDQWPDUNHWVKDUHWKH%LJ
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
Four ISPs in general and Telstra in particular have DVXEVWDQWLDOLQÀXHQFHRQWKHWHUPVRIZKROHVDOH Internet interconnection. The ACCC notes:
competition rule was only $6.5 million, instead of the sum of $10 million plus $1 million for each day until the Competition Notice lifted.
Fifteen to twenty per cent of retail price is attributable to the cost of interconnection borne by ,63VZKRDUHWUDQVLWFXVWRPHUVDQGWKLV¿JXUHLV EHWZHHQKLJKHUWKDQIRUWKH%LJ)RXUDQG other transit providers. The Big Four have lower interconnection costs due to their peering arrangePHQWVIRUQDWLRQDOWUDI¿FDWOHDVWDQGSRVVLEO\IRU VRPHLQWHUQDWLRQDOWUDI¿F (ACCC, 2004a)
Backhaul Pricing in Non-Metropolitan Areas
ISPs such as Comindico, PowerTel, MCI, and Request believed in 2004 that “Internet interconnection is having a negative impact on downstream markets as it is an essential input in the provision of these same services” (Comindico, 2004; PowerTel, 2004; MCI, 2004; Request, 2004). Primus (2004) commented that current Internet interconnection pricing is offered on both an arbitrary and discriminatory basis by the Tier 1 ISPs.
Telstra’s Wholesale and Retail Broadband Pricing and Competition Notice Access regimes are rarely concerned with refusal to grant access per se but rather about the terms and conditions of accessprice being the most important (Productivity Commission, 2001). Telstra has been using predatory pricing, which means it offers broadband access to its own retail customers more cheaply than it sells to its rivals that have access to its network. For instance, on )HEUXDU\7HOVWUDDQQRXQFHGVLJQL¿FDQW price reductions for its retail broadband services without any reductions in its wholesale prices. On March 19, 2004, the ACCC served Telstra with a competition notice, as it claimed that Telstra had been engaging in anti-competitive behavior by offering a broadband retail price that was below the fees it charged its rivals for access to its copper wire network. However, the competition notice was revoked in February 2005 after almost one year, and the penalty payable for a contravention of the
Gaining access to the local loop is only part of the cost of providing an end-to-end broadband service to consumers. In rural and regional areas, WKHUHDUHVLJQL¿FDQWFRVWVLQYROYHGLQSURYLGLQJD FRQQHFWLRQ IURP WKH UHJLRQDO VHUYLFH SURYLGHUV¶ infrastructure back to its metropolitan point of presence, as connection to backbone networks is required regardless of broadband access technology solutions. Although the ULL declaration enables smaller ISPs to connect their own infrastructure to the Telstra local loop, the backhaul to high-speed backbone networks problem remains. Current backhaul charges are prohibitive to those supplying broadband services to many regional areas. The Australian Telecommunications User Group (ATUG) pointed out: 7KHEDFNKDXOFRVWLVWKHVLJQL¿FDQWEDUULHUWR bush broadband. The current cost per megabit per month from Telstra for interstate backhaul is approximately 20 times too high to make those circuits viable for small regional operators to provide broadband services as there is little infrastructure competition down there. (ATUG, 2005)
GOVERNMENT INITIATIVES AND FUNDING PROGRAMS Regulatory Safeguards The government has attempted to ensure that customers in rural and remote Australia have equitable access to Internet services through a number of safeguards, such as Digital Data Service Obligation (DDSO), Special Digital Data Service Obligation (SDDSO), Extended Zones Agreement, and Internet Assistance Program, all of which are enshrined in legislation (Coonan, 2004). These regulatory initiatives guarantee Australians ac-
117
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
cess to basic Internet services. For example, all Australians can access the Internet either via a dial-up or satellite service, and most dial-up Internet users are guaranteed at least a 19.2kbps connection speed. However, higher speed Internet access is no longer a luxury but an essential service for many Australians in all parts of Australia. The Senate report argues that reliance on the current regulatory initiative is not considered to be adequate to provide reasonable access to all Australians (Australian Senate, 2004b). These safeguards need to be upgraded to take account of improvements in technology and thus the expectations of the majority of users.
Funding Programs for Broadband Services in Regional Australia 7KH JRYHUQPHQW¶V VWUDWHJLHV UHO\LQJ RQ PDUNHW forces as the primary driver have resulted in inequitable access between the metropolitan areas and rural and regional Australia. As the Broadband Advisory Group (BAG) reported, broadband services were not available in all areas at an affordable price, largely due to geographic location (BAG, 2003). While Australia has published numerous policy statements and national strategies aimed at supporting Internet development, there are too few examples of initiatives truly addressing high-speed Internet access nationwide. The recommendations made by the Telecommunications Service Inquiry (TSI) Report, the Broadband Advisory Group Report, and the Regional Telecommunications Inquiry (RTI) Report led to a series of government funding programs over the last six years. However, the funding programs such as Networking the Nation (NTN) and the Internet Assistance Program (IAP) have often tried to address basic Internet service needs rather than broadband service and also have tried to meet the short-term demands of rural people. Indeed, there was no program that was aimed at addressing the need for affordable access to high-speed Internet services until the government implemented the Higher Bandwidth Incentive Scheme (HiBIS) in April 2004 and announced in
118
August 2005 the Broadband Connect program for the next four years. The HiBIS ($107.8 million over four years), a key component of the National Broadband Strategy action plan, was established out of Recommendation 6.3 of the RTI (Estens, 2002) as an incentive scheme for the provision of higher speed Internet services to regional, rural, and remote areas. However, HiBIS seems not to be a comprehensive solution to the problems many consumers experience as a result of their inability to obtain access to affordable broadband. For example, the HiBIS subsidized monthly price is $69.95 for 500MB with Telstra Broadband two-way 256/64 kbps satellite plan, compared to $59.95 for unlimited usage with Telstra Broadband ADSL plan (Telstra, 2005). 7KHSULFHRIWKHµEHQFKPDUN¶VHUYLFHLQUXUDODQG remote areas is not really comparable to that of such a service in metropolitan areas, given that the Telstra Broadband Satellite plan is the only choice in most rural and remote areas. Moreover, most HiBIS providers are only offering threshold services with a monthly usage allowance of 500 MB at connection speeds of 256/64 kbps in the non-metropolitan area. 2QH FRPPRQ FULWLFLVP RI WKH JRYHUQPHQW¶V funding programs is that much of the funding has ÀRZHG LQGLUHFWO\ WR 7HOVWUD ZKHQ LW FRXOG KDYH been used to assist alternative providers and foster greater competition (Australian Senate, 2004b). This conclusion is supported by Optus in its submission to the Environment, Communications, Information Technology and the Arts References Committee in April 2005: 2YHURIWKHDOORFDWLRQRI PLOOLRQRI Government funding for rural and regional telecommunications since 1996 have been used directly by Telstra. In addition, it is fair to assume that a large bulk of the $347 million, not attributable to VSHFL¿FFDUULHUVLQGLUHFWO\VXSSRUWHG7HOVWUDLQIUDstructure and services. On the other hand funding directed toward non-Telstra carrier constitutes RQO\RIWKHIXQG´ (Optus, 2005) Despite the fact that the government aims to increase competition in the provision of broadband
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
LQUXUDODUHDV7HOVWUD¶VVLJQL¿FDQWLQYROYHPHQWLQ the government initiatives and funding programs enhances its position as the dominant carrier in the Australian telecommunications industry. Lundy (2002) argues that “the longer-term impact is that other potential providers have fewer opportunities to establish a business case to compete effectively with Telstra in rural and remote Australia.” It is apparent that the government is failing to support and promote the growth of viable competitors through a variety of policies and programs, DVQRWKLQJLVGLVSODFLQJ7HOVWUD¶VGRPLQDQFHZLWK the growth of competing companies. In fact, most of these government funding programs have been operating in conjunction with Telstra.
CONCLUSION AND IMPLICATIONS This research has demonstrated that the government has taken a multi-faceted approach to promoting Internet access, including reliance on market forces, regulations, regulatory safeguards, and grant programs to fund Internet access in regional Australia. The analysis in this chapter has clearly indicated that the Australian government has the strategic framework to drive towards an information-based economy, and recognizes the economic and social EHQH¿WVRIEURDGEDQGFRQQHFWLYLW\DQGVHUYLFHV but it has failed to develop an effective action plan to address national concerns about the speed of broadband infrastructure and take-up of broadband services, preferring to leave the rollout of high-speed Internet service to market forces. This research has indicated that Australia has lagged behind other comparable nations in terms of broadEDQG DGRSWLRQ DQG WKH JRYHUQPHQW¶V KDQGVRII approach and belief in market channel as a primary driver has been seen as one of the barriers to the growth of broadband. In general Australia has sound regulatory safeguards, which ensures that all Australians have basic Internet access services. However, these regulatory initiatives may need to be adjusted over WLPHWRUHÀHFWXSGDWHG,QWHUQHWDFFHVVWHFKQROR-
gies (e.g., broadband) when the market does not deliver these. Despite the fact that the National Broadband Strategy (NBS) and the associated action plan were released in 2004, a closer analysis in this research shows that there is no new funding commitment or VSHFL¿FWDUJHWVLQWKH1%6WRLPSURYHEURDGEDQG UROORXWDQGDGRSWLRQ7KHJRYHUQPHQW¶V5HVSRQVHV to the Recommendations of the Regional Telecommunications Inquiry (DCITA, 2003) were reused as a major part of the strategy. Since 1997, the government has introduced an open and full competition regulatory frameworkthe Telecommunications Act of 1997. It contains various measures such as no limits on the number of licenses, access regime, anti-competitive conduct regime, and removal of barriers to entry to facilitate competition in the telecommunications and the Internet sector. The analysis in this chapter has clearly indicated that while there is some infrastructure competition in the telecommunications industry in the capital cities, it has in general not spread far beyond metropolitan areas. The lack of IDFLOLW\EDVHGFRPSHWLWLRQKDVDVLJQL¿FDQWLPSDFW on the provision of broadband services in terms of SULFHÀH[LELOLW\DQGVHUYLFHTXDOLW\ Market concentration and vertically integrated provision of the key inputs of the provision of an ,QWHUQHWDFFHVVVHUYLFHDUHDUHÀHFWLRQRIDQXQFRPpetitive Internet market. The degree of inter-modal competition and cross-ownership of infrastructure KDV VLJQL¿FDQW LPSOLFDWLRQV ZLWK UHVSHFW WR WKH need for access regulation. The current regulatory competition regime does not appear to address effectively concerns of market dominance and market power in the telecommunications sector. )RUH[DPSOH3DUW;,%DQG3DUW;,&RIWKH73$ have not produced timely outcomes in the context of the take-up of ULL and the recent experience of the broadband competition notice, despite the IDFWWKDWUHJXODWRU\GHFLVLRQVKDYHVRPHLQÀXHQFH over retail and wholesale broadband price. The very large number of Internet service providers suggests that competition is strong in the retail Internet access service market. Competition among ISPs is driving their prices toward
119
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
cost, much of which is made up by payments to the backbone providers. However, the wholesale broadband market is less competitive. The Internet interconnection market structure is characterized by the four large backbone providers supplying a VLJQL¿FDQWSDUWRIWKHPDUNHWDQGKDYLQJVSHFL¿F peering arrangement that they do not provide to other parties (ATUG, 2004). The analysis in this research has clearly shown that these large backbone providers such as Telstra and Optus exercise strong market power when negotiating terms and conditions with other smaller ISPs. When ISPs pay higher charges to backbone providers, they must pass these costs on to business and residential users. This partly explains why the cost of broadband access is relatively high, in particular in rural and regional Australia. Structural issues impose limitations on the effectiveness of the current regulatory competition UHJLPH*LYHQWKDW7HOVWUD¶VXELTXLWRXVORFDOORRS is highly likely to remain the dominant customer access network in Australia for the foreseeable future, it is necessary to strengthen the powers of the ACCC to protect the interest of consumers and allow other service providers to compete on a level SOD\LQJ¿HOG0RUHHIIRUWVVKRXOGEHVKLIWHGWRZDUGV structural separation to achieve a more competitive environment where all Australians have high-speed Internet access at an affordable price. This research has clearly indicated that the state of broadband Internet access is closely related to WKHJRYHUQPHQW¶VSROLF\DQGUHJXODWRU\IUDPHZRUN It suggests that strong government involvement in broadband rollout, strong inter-modal competition both among providers of the same broadband technologies and between providers of different broadband technologies, as well as various initiatives and projects that are integrated to develop infrastructure-based competition and to provide more equitable and affordable broadband options to rural and regional areas are essential to the take-up of broadband services in Australia.
Implications for Other Countries The researcher initiated this study partly in a belief that the experience of Australia and its policies
120
in introducing and managing competition and in developing national information infrastructure, as well as dealing with rural and regional access issues, could be expected to provide some guidance and implications for other governments, especially relatively inexperienced ones in developing countries. While no two countries will ever have identical policies, lessons can be learned from the issues examined in the case of Australia. This section highlights three key areas for policymakers in other countries to consider their ongoing attempts to make broadband Internet access available for all at an affordable price. It aims to provide a more complete picture of Internet services to policymakers in order to allow them to make better informed user choices.
Rural and Regional Access Issues Like many other governments in the world, the Australian government has used a number of policy tools such as funding programs and Special Digital Data Obligations to address the remote, rural, and regional Internet access issue. However, while increasing concern is being focused on the digital GLYLGHLVVXHVVXFKDVLQVXI¿FLHQWLQIUDVWUXFWXUH KLJKFRVWRIDFFHVVDQGLQHI¿FLHQFLHVLQWKHSURYLsion of telecommunications networks and Internet services in rural and regional areas, the digital delay factor has been overlooked. Digital delay refers to the implications of the time effect inherent in the diffusion of technologies among different geographic areas and social groups. The national rollout of broadband services always starts from the metropolitan cities and is then progressively rolled out across the country. Some broadband LQIUDVWUXFWXUHVOLNHKLJKVSHHGRSWLF¿UHDFFHVV networks, may never really reach outside the CBD of the major cities. For example, the national rollout of ADSL service took years to reach 75% of the Australian population and this service was not available to most parts of regional Australia XQWLOWKRXJKLWZDV¿UVWODXQFKHGLQ The question remains of how long the rest of the population will have to wait. It appears that the government remains unaware of digital delay. Cutler pointed this out as long ago as 2001:
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
This time effect now matters more than before. In a global information economy, delay and being late into a market really matters. Delayed access to new network devices and applications imposes a huge new economic penalty on regional and non metropolitan areas. (Cutler, 2001) Minimizing digital delay is not something that should happen after addressing digital divide challenges. It should be a key component of national broadband strategy, as digital delay has some important implications for regional economic development. A fundamental requirement for reducing digital delay is for policymakers to give priority to the development of their telecommunications infrastructure and to provide R&D incentive to the development of next-generation, low-density broadband technologies for rural and regional areas. As broadband is moving from being a luxury towards being a necessity, faster access to richer information and online services becomes more important to all citizens.
Structural Issues The lesson learned from the case of Australia is that the current regulatory problems have more to do with the structure of the industry than with individual services (Budde, 2002c). The current telecommunications framework has not sought to address directly structural issues in the industry. The government has failed to review the advantages of the structural reform in the telecommunications industry prior to introducing competition into the affected markets or privatizing Telstra (NECG, 2004). The ITU research indicates that “the structural separation of dominant telecommunications operators has proved effective in facilitating competition in highly concentrated markets” (ITU, 2002). The right industry environment is therefore crucial in assuring positive outcomes of market forces. Evidence from the experience of successful developed and developing countries suggests that the government should make greater efforts to ensure a competitive industry structure, as it is a prerequisite in developing a well-functioning market, one which has the greatest effect on
prices, and in turn on broadband adoption (ITU, 2003b).
Government Leadership 7KH$XVWUDOLDQJRYHUQPHQW¶VVWURQJUHOLDQFHRQ the market has been seen as a key impediment to broadband growth. One of the main obstacles to EURDGEDQG GHYHORSPHQW LQ $XVWUDOLD LV 7HOVWUD¶V ability to provide innovative broadband service. $VWKH&RPSHWLWLYH&DUULHUV¶&RDOLWLRQFRPPHQWHG in 2005: The implications of commercial decisions by individual suppliers and the retarding effect that this has on broadband deployment and adoption and on innovation in Australian telecommunications markets must be seen in the broader context of Australia’s international competitiveness. &RPSHWLWLYH&DUULHUV¶&RDOLWLRQ Broadband infrastructure is the future. National broadband networks should be considered a public infrastructure asset. The leading countries with high broadband penetration worldwide are characterized by strong government commitment to broadband rollout. There has been a common call for the government to take an action-oriented vision to broadband development in Australia. If no action is taken at this point in time, the potential costs of inaction are greater than ever before, DVEURDGEDQGLVDVLJQL¿FDQWGULYHURIHFRQRPLF growth and innovation. Government leadership in the future broadband rollout will not only accelerate broadband deployment, but also reduce a potential geographical digital divide and the deleterious effects of digital delay.
FUTURE RESEARCH This study provides a framework for understanding the complexity of policy development and examining policy issues related to Internet access. In the face of the complicated landscape of current ICT policies, the framework provides a way in which to critically analyze how government policies af-
121
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
fect Internet access. It also provides the necessary feedback on how polices are actually working in practice. However, comparative studies of other FRXQWULHV¶HIIRUWVFRXOGDVVLVWLQEHWWHUXQGHUVWDQGing the impact of government policies on Internet access, as national governments are not isolated actors. Their actions and polices increasingly must take into account what is happening in other FRXQWULHV$V1RUWK¿HOG S SRLQWHGRXW “Comparisons of national experiences identify some common lessons for nations to review their policies and provide a mirror for countries to examine results of existing approaches.” The comparative study of public policy has a long and honorable tradition within the social sciences, but the comparative study of information and communications policy has been rather limited (Kahin & Wilson, 1997). Much of the ZRUNLQWKH¿HOGKDVIRFXVHGPDLQO\RQLQWHUSUHWing, examining, and understanding policy issues in selected countries in isolation, rather than on drawing conclusions or providing implications for other countries. Regrettably, the literature to date has little discussion of the methodological issues and lacks even a conceptual framework for policy analysis. As Rowlands (1997) pointed out, political and social scientists have developed a number of models, frameworks, and theories for analyzing public policy, but few of these have been explicitly applied in the context of ICT policy analysis. Future research can seek to address some of these GH¿FLHQFLHV
REFERENCES ABS. (2005). 8153.0 Internet activity, Australia. Retrieved August 29, 2005, from http://www.abs. gov.au/Ausstats/
[email protected]/0/EF49373E136F82FD CA256A15007E5EDC?Open ACCC. (1998). Telstra competition notice. Retrieved November 11, 2002, from http://www.accc. gov.au/content/index.phtml/itemId/87302 ACCC. (2003). Report to Senator Alston, Minister for Communications, Information Technology and the Arts, on emerging market structures in
122
the communications sector. Retrieved June 11, 2004, from http://www.accc.gov.au/content/index. phtml/itemId/324108 ACCC. (2004a). Internet interconnection service: Draft report on whether or not an Internet interconnection service should be declared under Part XIC of the Trade Practices Act 1974. Retrieved August 10, 2005, from http://www.accc.gov.au/ FRQWHQWLWHPSKWPO"LWHP,G QRGH,G ¿OH 416dcf10bd522&fn=Internet%20interconnection %20declaration%20Draft%20report.pdf ACCC. (2004b). Internet interconnection service: Final report on whether or not an internet interconnection service should be declared under Part XIC of the Trade Practices Act 1974. Retrieved August 10, 2005, from http://www.accc.gov.au/content/ LWHPSKWPO"LWHP,G QRGH,G ¿OHHGFF GGI IQ ,QWHUQHWLQWHUFRQQHFWLRQ¿Q al%20report%20(January%202005).pdf ACCC. (2005a). Snapshot of broadband development as at 31 March 2005. Retrieved October 10, 2005, from http://www.accc.gov.au/content/index. phtml/itemId/693170/fromItemId/356751 ACCC. (2005b). Telecommunications infrastructure in Australia 2004. Retrieved August 10, 2005, from http://www.accc.gov.au/content/index.phtml/ itemId/690299/fromItemId/356751 ATUG. (2004). Regulatory and market developments report, June-August 2004. Retrieved August 10, 2005, from http://www.atug.org.au/RegDevReports/ATUGRegulatoryandMarketDevelopmentReportJulAug04.pdf ATUG. (2005). Opinion—regional broadband, March 2005. Retrieved August 10, 2005, from http://www.atug.com.au/article.cfm?newsid=334 &newstype=1 Australian Senate (Australian Senate Environment, Communications, Information Technology and the Arts References Committee). (2004a). Competition in broadband services. Retrieved August 15, 2005, from http://www.aph.gov.au/senate/committee/ ecita_ctte/completed_inquiries/2002-04/broadband_competition/report/report.pdf
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
Australian Senate (Australian Senate Environment, Communications, Information Technology and the Arts Reference Committee). (2004b). The Australian telecommunications network. Retrieved August 15, 2005, from http://www.aph.gov.au/ Senate/committee/ecita_ctte/completed_inquiries/2002-04/tele_network/report/telecommunications.pdf Australian Senate (Australian Senate Environment, Communications, Information Technology and the Arts References Committee). (2005). The performance of the Australian telecommunications regulatory regime. Retrieved November 1, 2005, from http://www.aph.gov.au/Senate/committee/ ecita_ctte/trr/report/index.htm Besley, M.A. (2000). Connecting Australia: Report of the Telecommunications Service Inquiry. Retrieved August 10, 2005, from http://www. WHOLQTXLU\JRYDX¿OHV¿QDOBUHSRUWSGf BAG (Broadband Advisory Group). (2003). Australia’s broadband connectivity: The Broadband Advisory Group’s report to government. Retrieved August 10, 2005, from http://www.dcita.gov. DXBBGDWDDVVHWVSGIB¿OH%$*BUHSRUWSGI Budde, P. (2002a). Australia—regional government policies. Retrieved February 14, 2002, from http://www.budde.com.au Budde, P. (2002b). Australia—regional statistics overview. Retrieved February 14, 2002, from http://www.budde.com.au Coonan, H. (2004). Australian telecommunications summit 2004—keynote address—Sydney Monday 26 July 2004. Retrieved August 2004 from http:// www.minister.dcita.gov.au/media/speech01.html Comindico. (2004). Comindico submission to the ACCC’s Internet Interconnection Discussion Paper. Retrieved January 2005 from http://www.accc. gov.au/content/index.phtml/itemId/477041 &RPSHWLWLYH&DUULHUV¶&RDOLWLRQ Revitalising Competition in Australian Telecommunications: Proposals for Policy Reform: Discussion Draft. Retrieved August 10 2005 from http://www. dcita.gov.au/_data/assets/word_doc//26603/CCC_ paper_1.doc
Cutler, T. (2001). Connecting the continent: Telecommunications (speech to the Institute of Public Administration, Adelaide 22 March 2001). Retrieved August 10, 2005, from http://www. connectingthecontinent.com/ctcwebsite/pdf/terryCutlerSpeech.pdf DCITA. (2003). The government’s response to the recommendations of the Regional Telecommunications Inquiry. Retrieved August 10, 2005, from http://www.dcita.gov.au/tel/regional,_rural_and_remote_communications/telecommunication_inquiry_-_government_response DCITA. (2004a). Australia’s strategic framework for the information economy 2004-2006. Retrieved August 10, 2005, from http://www.dcita.gov. au/__data/assets/pdf_file/20457/ New_SFIE_ -XO\BB¿QDOSGI DCITA. (2004b). The current state of play 2004. Retrieved August 10, 2005, from http://www.dcita. JRYDXBBGDWDDVVHWVSGIB¿OH&63B pdf Estens, R. (2002). Connecting regional Australia: The report of the Regional Telecommunications Inquiry. Retrieved August 10, 2005, from http://www.telinquiry.gov.au/rti-report/ rti%20report%20text%20f-a%2018.pdf Forster, J. (2001). Broadband regulation. In S. Barns (Ed.), Broadband for what? Driving demand. Sydney: Network Insight Group. Frieden, R. (2005). Lessons from broadband development in Canada, Japan, Korea and the United States. Telecommunications Policy, 29(8), 595-613. GSR (Global Symposium for Regulators). (2004). Best practice guidelines for the promotion of low cost broadband and Internet connectivity. Retrieved August 10, 2005, from http://www.itu. int/ ITU-D/treg/Events/Seminars/2004/GSR04/ consultation.html ITU. (2001). Case study: Broadband: The case of Australia. Retrieved August 10, 2005, from http://www.itu.int/osg/spu/ni/broadband/workVKRS$XVWUDOLD¿QDOSGI
123
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
ITU. (2002). Competition policy in telecommunications: Background paper. Retrieved August 10, 2005, from http://www.itu.int/osg/spu/ni/competition/background/Final%20background%20paper. pdf ITU. (2003a). ITU Internet reports: Birth of broadband. Retrieved August 10, 2005, from http://www.itu.int/osg/spu/publications/sales/ birthofbroadband/ ITU. (2003b). Promoting broadband: Background paper. Retrieved August 10, 2005, from http://www. itu.int/ osg/spu/ni/promotebroadband/PB03-PromotingBroadband.pdf ITU. (2005). ITU’s new broadband statistics for 1 January 2005. Retrieved September 16, 2005, from http://www.itu.int/osg/spu/newslog/ITUs+New+B roadband+Statistics+For+1+January+2005.aspx Kahin, B. & Wilson, E.J. (1997). The What, Why, Where and How of National Information Initiatives. In National Information Infrastructure Initiatives: Vision and Policy Design, MIT Press, Cambridge, MA. Kennard, W. (1999). Connecting the Globe: a Regulator’s Guide to Building a Global Information Community, Federal Communications Commission, Washington, DC. Lundy, K. (2002). The broadband divide. Proceedings of the Rural and Regional Telecommunications Summit 2002, Melbourne. Retrieved from July 2003 from http://www.katelundy.com. au/broadband.htm
1RUWK¿HOG' The Information Policy Maze. CIRCIT: Melbourne. OECD. (1996). Information infrastructure convergence and pricing: The Internet (OCDE/GD(96)73). Retrieved from August 15, 2005, from http://www. oecd.org/dataoecd/37/34/2758574.pdf OECD. (1997). The OECD report on regulatory reform: Synthesis. Retrieved April 22, 2000, from http://www.oecd.org/dataoecd/17/25/2391768.pdf OECD. (2005). Broadband statistics, December 2004. Retrieved September 16, 2005, from http:// www.oecd.org/ document/60/0,2340,en_2825_ 495656_2496764_1_1_1_1,00.html Optus. (2005). Submission to Australian Senate: Environment, Communications Information Technology and the Arts References Committee on the inquiry into the performance of the Australian telecommunications regulatory regime. Retrieved August 15, 2005, from http://www.aph. gov.au/Senate/committee/ecita_ctte/trr/submissions/sub12.pdf PowerTel. (2004). PowerTel submission to the ACCC’s Internet Interconnection Discussion Paper. Retrieved January 26, 2005, from http://www. accc.gov.au/content/index.phtml/itemId/477041 Primus. (2004). Request submission to the ACCC’s Internet Interconnection Discussion Paper. Retrieved January 2005 from http://www.accc.gov. au/content/index.phtml/itemId/477041
Lundy, K. (2004). Broadband competition. Proceedings of the ATUG 2004 Conference. Retrieved August 10, 2005, from http://www.katelundy.com. au/Broadbandcomp.htm
Productivity Commission. (2001). International benchmarking of remote, rural and urban telecommunications services. Melbourne: Productivity Commission. Retrieved August 2005, from http:// ZZZSFJRYDXUHVHDUFKEHQFKPUNUDUWV¿QDOUHport/rarts.pdf
MCI. (2004). MCI submission to the ACCC’s Internet Interconnection Discussion Paper. Retrieved January 11, 2005, from http://www.accc. gov.au/content/index.phtml/itemId/477041
Request. (2004). Request submission to the ACCC’s Internet Interconnection Discussion Paper. Retrieved January 2005 from http://www.accc.gov. au/content/index.phtml/itemId/477041
NECG. (2004). Submission to the Productivity Commission Inquiry into national competition policy. Retrieved November 1, 2005, from http:// www.pc.gov.au/inquiry/ncp/subs/sub134.rtf
Rowlands, I. (ed). (1997). Understanding Information Policy. Bowker-Saur: London.
124
Samuel, G. (2005, March 10). Address of ACCC chairman to Australian Telecommunications
,QÀXHQFHRI$XVWUDOLD¶V*RYHUQPHQW3ROLF\RQ%URDGEDQG$FFHVV
Users Group. Retrieved August 15, 2005, from http://www.accc.gov.au/content/item.phtml?item ,G QRGH,G ¿OHI H IQ 50310%20ATUG.pdf Telstra. (2005). Telstra broadband plans. Retrieved May 6, 2005, from http://www.bigpond.com/internet-plans/broadband/pricing/default.asp Umino, A. (2002). Broadband infrastructure deployment: The role of government assistance (DSTI/DOC(2002)15). Retrieved August 15, 2005, from http://www.olis.oecd.org/olis/2002doc. nsf /43bb6130e5e86e5fc12569fa005d004c/ 42158ef983225772c1256bc100560c01/$FILE/ JT00126526.PDF Willett, E. (2004, June 25). Challenges in telecommunications competition and regulation (speech to SPAN Conference). Retrieved August 15, 2004, from http://www.accc.gov.au/content/item. SKWPO"LWHP,G QRGH,G ¿OHDGGD 2&fn=20040625%20SPAN.pdf Willett, E. (2005, May 30). National Telecommunications Summit 2005, Melbourne: The state of competition in the telecommunications industry. Retrieved August 15, 2005, from http://www.accc. gov.au/content/item.phtml?itemId=609202&node ,G ¿OH IHFG IQ 1DWLRQDO %20Telecommunications%20Summit.pdf Winslow, T. (2002). Latest developments in international competition law/policy. Retrieved December 11, 2004, from http://203.127.220.112/ content/apec/apec_groups/committees/committee_on_trade/competition_ policy.downloadlinks.0002.LinkURL.Download.ver5.1.9 Working Party on Telecommunications and Information Services Policy. (2000). Local access pricing and e-commerce. Retrieved August 15, 2005, from http://www.oecd.org/dataoecd/9/63/27129228.pdf Working Party on Telecommunications and Information Services Policy. (2001). The development of broadband access in OECD countries. Retrieved August 15, 2005, from http://www.oecd.org/dataoecd/48/33/2475737.pdf
Working Party on Telecommunications and Information Services Policy. (2002). Broadband access for business. Retrieved August 10, 2005, from http://www.olis.oecd.org/olis/2002doc. nsf /87fae4004d4fa67ac125685d005300b3/ a963ab2ca9617affc1256c85005d7190/$FILE/ JT00136306.PDF World Summit on the Information Society. (2003). Declaration of principles: Building the information society: A global challenge in the new millennium. Retrieved August 15, 2005, from http://www.itu. LQWZVLVGRFVJHQHYDRI¿FLDOGRSKWPO ZDNet Australia. (2003, February 24). ACCC tackles Internet interconnection. Retrieved August 15, 2005, from http://www.zdnet.com.au/news/ business/soa/ACCC_tackles_Internet_interconnection/0,39023166,20272319,00.htm
KEY TERMS Broadband Internet Access: Often shortened to “broadband Internet” or just “broadband,” refers to a high data-transmission rate Internet connection (retrieved from http://en.wikipedia. org/wiki/Broadband_Internet_access). Internet Backbone Provider: A company that usually has substantial network infrastructure and provides ISPs with aggregated access to many or all networks accessible via the Internet protocol. Internet Interconnection: The arrangement through which Internet backbone providers and Internet services providers physically connect to each RWKHU¶VQHWZRUNVDQGWUDQVIHU,QWHUQHWWUDI¿F Internet Service Provider (ISP): $¿UPWKDW provides Internet access services to end users. Policy: A set of government activities, which can be broadly characterized as regulations, national strategies/initiatives, and funding programs. Unbundling Local Loop (ULL): Term used to describe the access provided by local exchange carriers so that other service providers may buy or lease portions of its network elements, such as interconnection loops to serve subscribers (Kennard, 1999). 125
126
Chapter IX
Broadband for the Mass Market Roger Saunders Saunders Associates, Australia
ABSTRACT This chapter suggests there is no new application to stimulate adoption of broadband by the mass market. Many new applications have been introduced but have not created the desired growth. One application that could have mass market attraction is Voice over Internet Protocol (VoIP), and it is the most likely killer application. But failure by major communications carriers to develop VoIP is slowing broadband penetration to this larger market segment. This resistance results from the risk to current carrier call revenue from VoIP, and that infrastructure to support high-speed broadband may not generate adequate UHYHQXHLQWKHVKRUWWHUPWRVDWLVI\VKDUHKROGHUVRU¿QDQFLDOPDUNHWV1RLQWHUQDWLRQDOVWDQGDUGVKDYH yet been set, and VoIP between the Internet platforms is not fully integrated. Also the multitude of broadband packages offered by the various competing carriers creates confusion in the mass market which, as a result, defers purchase decisions.
INTRODUCTION While Internet and broadband access has been promoted as a fast universal data transfer and information providing media, the mass of information available and the relative complexity of accessing the right information for the neophyte Internet user is a deterrent to adoption. It is suggested that despite the many arguments that access to broadband alone or various segmentVSHFL¿F DSSOLFDWLRQV DUH WKH NH\ WR EURDGEDQG adoption, there is no new magic bullet or killer application to stimulate rapid adoption of broad-
band. The only immediate and easily recognizable application of broadband for the mass market is voice communication, that is, voice over Internet protocol (VoIP). This places incumbent telecommunication providers in an invidious position where their primary UHYHQXHVRXUFHLVIURPYRLFHWUDI¿F To accelerate penetration of broadband and reach a level of return from the required capital expenditure, these organizations would be sacri¿FLQJ WKHLU WUDGLWLRQDO UHYHQXH EDVH IRU D ORZHU revenue stream and opening up their customer base to competition. Therefore they are reluctant
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband for the Mass Market
to adopt applications that are likely to have a negative effect on their traditional markets and their prime revenue sources. While the reluctance by these companies to accept the fact that new technology has rendered much of their existing infrastructure and key products obsolescent, build appropriate technical capability, agree to cooperate in the development of international standards, and promote applications such as VoIP, the diffusion of broadband to the wider population is likely to be slower than would otherwise be the case.
BACKGROUND Currently broadband is being offered as a universal means of providing data, information, education, entertainment, and a myriad of other applications, each of which appeals to particular and different segments of the market. However, for the broadband provider, maximization of penetration to the total population is needed in order to recover costs and meet the expectations of their shareholders. It is also evident that many governments around the world are pressuring incumbent communications carriers to provide broadband to cover the majority of the population, even when this means expending large amounts of capital in locations where a return on that investment is unlikely. This seems to be based on a belief that access is the primary stimulant necessary to encourage adoption of broadband and satisfy the general population. However, the assumption that having access to EURDGEDQGLVVXI¿FLHQWWRVWLPXODWHDGRSWLRQE\ the mass market as suggested by Middleton (2003) appears to be fallacious when comparing actual demand relative to broadband availability. A counter to the argument that access alone is the killer application is that expressed by Smith and Leung (2002), who believe that high-speed ,QWHUQHWDFFHVVLVLQVXI¿FLHQWWRGULYHEURDGEDQG connectivity and that there is still a need to overcome a number of technological, economic, and social issues before any key killer application or applications necessary to promote demand can be realized.
The providers who invest capital in broadband infrastructure need these key "killer" applications if they are to attract a broad mix of customers and DWWKHVDPHWLPHHQFRXUDJHVXI¿FLHQWXVHRIWKH service to justify the investment needed. To date, growth in broadband use has been at a steady rate. However, stock market investors and analysts have been conditioned by what they see as the rapid market penetration of mobile communication to expect rapid adoption of this new technology. But broadband is not the same as the mobile service. Broadband is essentially a simplex means of communication where information is either downloaded from the Net or uploaded from the customers. While it does have text capability, it is not a direct two-way communication medium. There is some duplex voice carriage capability with VoIP, but it is not the same as mobile voice and it currently does not have the same functionality or quality as the existing wired telephone service or the mobility element of mobile network, so direct comparison of growth expectations is somewhat spurious. The apparent high rate of market penetration by mobile communications is somewhat illusory since mobile communication was available in the 1960salbeit limited in availability and quality with high access and usage pricesand it should not be used as a benchmark for new technology take-up. Take off in mobile communications in the mass market did not occur until cellular techQRORJ\ZDVLQWURGXFHGZLWKVLJQL¿FDQWO\JUHDWHU availability, performance, and at acceptably lower call prices. It is likely that broadband will follow the same path and require more technology to stimulate adoption by the general public. Technology was the “killer” factor for mobile communications, and a key question for broadband SURYLGHUVLVWR¿QGWKHIDFWRUWKDWZLOOVWLPXODWH adoption of broadband by the market so that expenditure on the needed technology can be justi¿HG+RZHYHUZLWKEURDGEDQGPDQ\RIWKHPDMRU incumbent communications carriers, having not \HWLGHQWL¿HGRUDFFHSWHGWKHH[LVWHQFHRIDNLOOHU application, are reluctant to invest the high level of capital necessary to support the required technol-
127
Broadband for the Mass Market
ogy or attempt to achieve a common worldwide standard for Internet platform access. The outcome of this reticence is that some potential applications have yet to reach their full potential and may well be a factor in the slower-than-expected diffusion of broadband to the wider population.
DISCUSSION Broadband adoption by both business and domestic customers is directly affected by a number of key IDFWRUV,QJHQHUDOWKHVHFDQEHLGHQWL¿HGDV • • • •
System availability, quality, and performance Customer application, use, and need Customer value perception of the service provided Price limits set by customers
While availability, quality, value, and price can in principle be managed by a broadband infrastructure provider, application is determined ¿UVWO\E\WKHFXVWRPHUVDQGVHFRQGO\DWWKLVVWDJH of development, by content providers who are OLWWOH FRQFHUQHG DERXW WKH ¿QDQFLDO DQG WHFKQLcal constraints associated with the provision of broadband. The following sections consider each of these broadband factors in its current state of development.
System Availability and Capability Access to the Internet can be achieved by dial-up and modem capability. However, speeds across the standard phone line are somewhat limited. To provide broadband proper, there are a number of technologies that can deliver better access and performance to the household or business, each requiring greater capital investment and commitment to continued use by the customers. While the lowest level of access to the network (dial-up) is available, it cannot be considered as broadband proper although it provides Internet access capability. A digital subscriber line (DSL)
128
using the existing copper network is in fact the ¿UVWPDUJLQDOOHYHORIEURDGEDQGDQGUHTXLUHVWKH least investment by network builders, provided the core infrastructure already exists and the customHUVDUHQRWORFDWHGVLJQL¿FDQWGLVWDQFHVIURPWKH exchange. Coaxial cable is a higher speed capable access, and provision in many countries is packaged with cable television in order to defray the cost and provide a better chance of reaching adequate revenue return to justify the investment. Wireless access and satellite is available where customers are located away from suitable infrastructure, but it has some operational limitations and capacity LVVXHV2SWLFDO¿EHUSURYLGLQJ³Fiber to the Home” (FTTH) or “Fiber to the Node” (FTTN) is currently regarded as the best option for high-speed and fully capable broadband delivery. It is however the most costly for the infrastructure builder since it requires replacement of existing copper customer access networks (CANs) and accounts for the reluctance of some communications companies who are most likely to be the infrastructure providers for implementation. Each of these technologies requires a progressive increase in capital expenditure by the network providers, as each increase in speed and functionality increases the cost of the service delivery. In order for broadband providers to even consider provision of these higher levels of service, they need some expectation of return on their investment, and in the current environment, speed of capital return is critical to satisfy the expectations of shareholders. As stated above, incumbent carriers will need WRPDNHVLJQL¿FDQWFKDQJHVWRWKH&$1WRWKHLU product strategies and their operating systems. In a paper in the Telecommunications Journal of Australia (TJA), the journal of the Telecommunication Society of Australia (TSA), Darling (2005) of Pondarosa Communications stated: “If the CAN of the future is to provide a full range of services, it will have to support truly broadband access. For many customers already using broadband this is not the case, particularly so in non-metro areas both across Australia and in other countries.”
Broadband for the Mass Market
For broadband builders, high market penetration, high usage, or a combination of both is required to achieve a desired level of return, and this requires a degree of faith and commitment by them, especially if, at this stage of development of adoption by the mass market, they are not meeting ¿QDQFLDOO\ GHVLUDEOH UDWHV 3HQHWUDWLRQ FDQ RQO\ be achieved if appropriate levels of service are available to the majority of the population, but this requires capital investment where the majority of the population lives. This is where the issue becomes both problematic and political. Political imperatives seldom coincide with WKH¿QDQFLDOFULWHULDRILQGXVWU\SDUWLFXODUO\DQ industry that is in essence a provider of utility services. Electricity, gas, water, and other services including communications are considered essential in a developed society. Therefore, provision of broadband communications to the total population is in most countries a political imperative, none more so than in a country such as Australia where the population is concentrated around the coastal areas and is sparse and widely dispersed in the rest of the country. A comparison of broadband adoption between metropolitan and regional populations in Australia was presented to the TJA by the University of Queensland (Cameron, 2004) summarizing a number of studies that have been undertaken by various bodies in various localities. They all LGHQWL¿HGWKDWWKHODFNRIDYDLODEOHLQIUDVWUXFWXUH to support broadband is a key factor in the nonproliferation of broadband. A further point not indicated in the research is that even where some broadband capability was available, the level of technology used was such that the service was slow, intermittent, and in some cases, expensive. The studies also indicated that socio-economic factors had a bearing on whether some members of the population would even consider adoption of EURDGEDQGHYHQLILWZDVDYDLODEOH7KLVLVUHÀHFWHG both in the metro and regional areas, and potential customers who represent the poorer members of WKHFRPPXQLW\KDYH\HWWRLGHQWLI\MXVWL¿FDWLRQ for adoption of broadband. To provide broadband with any level of capability to these remote locations would be costly and
because of the small population the likelihood of return on the investment is unlikely. Consequently, broadband providers are reluctant to provide such services. Cross-subsidization from the high-density metropolitan centers to the remote and rural communities, similar to that which has been applied to the telephone services before, is one possible approach. That however requires higher prices for connection and usage in the cities and in an environment where government holds the view that competition and free market forces are desirable; this is not politically supported. The Australian Competition and Communications Commission (ACCC) has been considering a proposal to introduce differential pricing for both the existing communications network and for future broadband services. That is, the customer prices for services in the metro and non-metro areas should be priced according to the cost of provision and delivery. To date no outcome of this review has been forthcoming. Under this and the competitive market policy, competition in the high-density metro areas would preclude any price balancing availability to a broadband provider considering subsidizing services in the low-density areas. Without regulation and political acceptance of something less than a free market condition, the potential for higher prices DQGVLJQL¿FDQWGLIIHUHQFHVLQFKDUJHVIRUQRQPHWUR compared to metro locations further discourages take-up of broadband outside the metro areas. Therefore government policy and communications regulation play critical roles in whether broadband diffusion takes place to the total population in a country or not. In support of this view, Alcatel and KPMG (Fitzpatrick & Perin, 2006) recently produced a MRLQWUHSRUWLGHQWLI\LQJWKH¿YHUHJXODWRU\OHYHUV used by governments around the world to promote broadband. These are: • • • • •
Government subsidy and/or incentives Safe harbor Access price certainty Deployment based on return on investment Vertical disaggregation
129
Broadband for the Mass Market
Government subsidy and support has been the approach taken by the Korean Government and this LVUHÀHFWHGLQWKHOHYHORISHQHWUDWLRQUHSRUWHGE\ the OECD (2005), but the broadband capability provided is at the lowest level and at a low price to customers. The Safe Harbor policy has been used in the United States to provide a honeymoon period for an infrastructure builder. This approach restricts or stops competitors to the provider from either directly accessing the new broadband network or LQVWLWXWLQJ SULFLQJ WKDW GRHV QRW UHÀHFW WKH WUXH cost of capital for a set period of time, thus limiting the ability of the builder to compete in the marketplace and receive an appropriate return on its investment. No country explicitly using the access price¿[LQJDSSURDFKKDVEHHQLGHQWL¿HGDOWKRXJKWKH safe harbor approach may be doing so by implication. Where limited direct access or competition LVDOORZHGWKHDFFHVVSULFHLVLQWKLVFDVH¿[HGE\ regulation preventing unfair competition. Basing the establishment of a broadband network purely on a return on investment approach may be an acceptable approach in dense and heavily populated countries, such as Japan and some European countries. However, under this approach there would be pockets of population in such countries as Australia, the United States, and India that would not gain access to broadband for a considerable time if ever. For many large countries with scattered and low-density population centers such as these and much of Southeast Asia, this is not generally an acceptable political outcome. Vertical disaggregation involves the separation of core infrastructure from that of the retail operations of a communications operation. In effect, a broadband provider would become a wholesaler of broadband to all customer interface organizations. This is an approach that has been proposed since the worldwide move to sell government ownership of incumbent communications operations and the introduction of competition. The issue has been that all network infrastructure requires considerable capital to build and maintain. Consequently the existing copper pair networks as well as new developments required to support high-speed
130
broadband need capital investment. In a competitive market where market prices are constantly under GRZQZDUG SUHVVXUH VKDUHKROGHUV DQG ¿QDQFLDO institutions still expect expedient returns on their investments. Therefore the boards of directors and the businesses they manage are severely constrained on what the can spend on new and as yet low penetrating technology. The option to place the high-cost infrastructure operation into a separate FRPSDQ\KDVWZRSRLQWVLQLWVIDYRU7KH¿UVWLV that there are precedents in the operations of utilities, where capacity is on-sold to retail operations with premium prices to the high-usage areas. The second is that it avoids duplication of assets which would occur in a country if all competing forces decided to construct alternative networks. There DUHVLJQL¿FDQWGLVDGYDQWDJHVWRWKLVKRZHYHULQ such a diverse operation as communications, such as separation from the using customer base and the dissociation from non-network communications developments. The outcome would be that the wholesale operation would not be in a position to identify use and demand trends in the market and would be reluctant to increase investment where they could not identify any requirement or any increase in revenue return to the operation. Each of these factors, however, focuses on the technology and not on the market, and while they might stimulate availability of service to the population, they do not necessarily guarantee that adoption of broadband will automatically follow. A study conducted by Wale (2002) at the University of Melbourne, as a part of his honors degree thesis, indicates that technology is an element in broadband use stimulation by existing residential customers in that they seek better broadband attributes rather than more applications. The attriEXWHVLQFOXGHµDOZD\VRQ¶QHWZRUNLQJRIKRXVHKROG computers and to a lesser extent speed. The study LGHQWL¿HG WKDW DW SUHVHQW WKHUH DSSHDUV WR EH QR real need for broadband apart from streaming video and that the broadband advantage at this stage is that it enables e-mail, chat, and browsing WREHXVHGPRUHHI¿FLHQWO\7KHUHIRUHLWVHHPVWR LQIHUWKDWZKLOHWKHUHLVVRPHEHQH¿WWRH[LVWLQJ XVHUVZKRKDYHLGHQWL¿HGDSSOLFDWLRQVRIXVHWR WKHPWKHUHLVLQVXI¿FLHQWVWLPXOXVLQWKHFXUUHQW
Broadband for the Mass Market
available applications to attract mass market customers who have no experience on which to base their requirements. To some degree, the proposal by Luber (2001) that e-mail is the already available killer application but is currently constrained by the relatively low bandwidth available on some broadband links supports this view. However, for broadband providers HPDLOLVXQOLNHO\WRVWLPXODWHVXI¿FLHQWUHYHQXH WR MXVWLI\ WKH SURYLVLRQ RI LQ¿QLWH EDQGZLGWK WR the mass market. In summary therefore, three factors have a direct impact on availability and the growth and adoption by the general public of broadband availability. These are the technology and quality level provided, the level of regulation required to support government policy, and the willingness of the broadband infrastructure owners, shareholders, DQG¿QDQFLDOLQVWLWXWLRQVWRDFFHSWWKHULVNDQGWKH possibility of deferred reimbursement on the capital that is required to build, support, and provide maintenance by the infrastructure builders.
Customer Application, Use, and Need Broadband can provide advanced applications to customers, but for universal application to meet the needs of the mass market, a “killer” application is still sought. Kwok (1995) proposed that to achieve the “grand vision” of the residential broadband market, platforms capable of delivering thousands of applications would be necessary. This supported the idea promoted by network providers and communications companies that broadband is, to quote Adams (1982), the answer to “life, universe and everything.” Unfortunately that is not the case. While content providers since 1995 have delivered a mass of information and applications to customers, the mass of data itself is one of the issues that appear to be constraining new users from acquiring broadband access or using the service even when they have access to some level of broadband. The content providers have particular market segments to which they target their product. One product may be sport, which is one that has to date achieved the greatest appeal in Australia, or online
auctions, music downloads, or a myriad of other potential services. One of the biggest money spinners for these content providers is advertising, and the increase in unsolicited promotion is a further addition to the mass of data a broadband customer must negotiate. By virtue of the diversity of information, some appeal to particular groups of users but not to all. In fact, the very mass of available information or entertainment services available PDNHVLWLQFUHDVLQJO\GLI¿FXOWWR¿QGWKHVSHFL¿F subject an individual may want. A typical reaction by new users of the Internet either via broadband or dial-up services, when unable to access particular information using one of the search engines, is that they tend to abandon the search and because of the negative outcome choose not to adopt broadband or limit its future use. There have been a number of psychological research projects that have linked stress to the masses of information that is now available in the public domain, and the question of whether humans can perform multiple new tasks and process this data. Swartz (2003) in his book The Paradox of Choice suggests that while more choice means better options and greater satisfaction, that choice overload can cause decision-making paralysis. This applies to both the information that is being sought and the initial broadband connection decision. This latter point is discussed in the sub-section on value perception. Information and content providers are in the market to sell their products or services, and are likely to be little concerned about the cost of broadband delivery or the cost to the customer of broadband accessing or attempting to access their Web sites. Also the cost recovery required by the network provider is not an issue for them. The broadband provider has little control over the content that is needed if adoption and usage is to DFFHOHUDWHWRWKHOHYHOWKH¿QDQFLHUVDQGVKDUHKROGers expect, and if the content claims are considered risible it is more likely to deter possible broadband subscribers than to encourage growth. So the situation is that in principle, the network capability may be available but the broadband proYLGHUKDV\HWWR¿QGRUKDYHWKHDELOLW\WRHQFRXUDJH greater participation or the incentive to provide
131
Broadband for the Mass Market
DQ\ PRUH WKDQ EDVLF IXQFWLRQDOLW\ VXI¿FLHQW WR deliver a return on its investment. Apart from the ZHOOLGHQWL¿HGHDUO\DGRSWHUVDQGLQGLYLGXDOVZKR KDYHLGHQWL¿HGVSHFL¿FQHHGVQRHVVHQWLDOVHUYLFH available from the content providers appears to have been offered to the mass market so far. Video and video streaming has frequently been cited as a key application for broadband, but to date issues relating to revenue sharing and copyright between content providers, video producers and distributors, and service aggregators has tended to restrict the potential application as indicated by Lim (2002). Also, broadband bandwidth in many cases KDVPDGHGRZQORDGLQJRI¿OPVORZWKXVUHGXFing the possible attractiveness of this application. However, Price Waterhouse Coopers (2004) has proposed that by adding video capability to existing applications such as conferencing and e-mail, there is a potential short-term gain in usage. This may be one application enhancement that could EHRIEHQH¿WWREXVLQHVVEXWWKHDWWUDFWLRQWRWKH residential mass market seems marginal. A relatively recent application has been developed, not by content providers or the network providers, but by broadband users themselves. Weardon (2001) predicted that users would develop peer-to-peer applications that would stimulate use by existing customers. This has to some extent EHHQ VWXOWL¿HG E\ WKH ODFN RI Internet protocol (IP) standards, but users have adapted the network to provide their own content in the form of self-produced broadcast videos, music exchange groups, and Weblogs (blogs). While this has the potential to attract some new customers, it is still not the killer application the networks really need to increase adoption by new customers from the mass market. Therefore to provide a reason for an increase in infrastructure investment, these providers still need a key application, and the infrastructure providers have little expertise or control over the current type of content carried on their broadband DQGOLWWOHLQÀXHQFHLQLGHQWLI\LQJDFRQWHQWSURYLGHU to deliver such a product. Yet there is one application that does have universal application to the mass market, and that is the plain old mundane voice communication with
132
voice over Internet protocol, VoIP, which is currently being provided to some degree by organizations in most countries other than the builders of the broadband network. Provision of user-friendly voice communication using broadband would provide the motive to deliver an essential service alternative to the current standard telephone network, and infrastructure providers would at least have some experience with this application. A paper by Bachelet (2005) of OVUM based on a report, “Altnet Broadband Strategies,” indicated that they believed that VoIP was the killer application, not TV over DSL (TVoDSL) or Video on Demand (VoD), and that they expected only a small minority of their customers to adopt these higher level services. In fact, they believed that they had an opportunity to capture more of an LQFXPEHQW¶VPDUNHWE\GHYHORSLQJDSDFNDJHRI broadband access plus VoIP. However, for incumbent and current communications companies to combat this potential attack on their market share by promoting VoIP is a two-edged sword. Most broadband infrastructure providers are telephone service providers, and to date most have resisted applying a service that both erodes their current revenue source and provides competitors access to their existing customer base. Yet this offers the greatest chance for market penetration and dispersion of broadband adoption, particularly since the carriers themselves will at least have some degree of control, albeit needing to RIIHUYRLFHDWVLJQL¿FDQWO\ORZHUSULFHVWRPLJUDWH their existing customer base to broadband. But, since these incumbent communications companies build the broadband network and in the majority of cases own and operate the CAN, they have the technological capability to deliver enhanced services using the more sophisticated broadband applications and provide faster upload and download capability such that video/voice combinations and online interactive video entertainment could become a reality. In practice these higher level services could be offered at higher usage prices to offset the inevitable decline in their current telephone revenue stream that would be a given if broadband had this capability. Until these incumbent and competing communications
Broadband for the Mass Market
carriers are prepared to relaunch their communications infrastructure, there is little chance that there will be any move to establish appropriate world standards that are necessary for such applications as VoIP to communicate between the Internet provider platforms to enable peer-to-peer communication and other capabilities to be viable. Therefore this and other potential applications are unlikely to happen quickly. As it stands, the promotion of broadband as all things to all people is highly dependant on content, and if broadband adoption is to grow and usage to increase, carriers will ultimately need to take steps to review their technology, infrastructure, products, systems, and pricing structures. In summary, the current content available to customers on broadband is extremely diverse and complex. The need for better, faster, and more convenient service by individuals is dependent on the content they need and value they place on it. While businesses can identify practical operational reasons for broadband, the domestic market and the growth on which infrastructure providers depend is an application that is required all the time as an essential everyday service. The needs for business are different to domestic needs, but as yet the delivery of broadband is undifferentiated and much of the business activity involves inter and intra-company data transfer ZKLFK EURDGEDQG VDWLV¿HV +RZHYHU IRU WKH GRmestic customer, the applications may be accessing GDWDRUVSHFL¿FLQIRUPDWLRQEXWWKHDELOLW\IRUSHHU to-peer access is limited. As previously indicated, WKHEURDGEDQGSURYLGHUVKDYHOLWWOHLQÀXHQFHRQ the content currently carried or how the network is used. Consequently they have little power at this stage to foster adoption and use by the mass market, unless they make a strategic move to accept the changes that are inevitable as broadband adoption increases and commence the shift from dependence on their current communications network product revenue to broadband.
The Value Perception Value to the user of broadband relates directly to the content available and the ability of the service
provided to deliver that content in an appropriate manner. In other words, does broadband meet a VSHFL¿FFXVWRPHUQHHGGRHVLWSHUIRUPLQDPDQner that enables the customer to satisfy that need in a way the customer considers acceptable, and is the service seen as good value? Broadband providers around the world deliver different levels of speed and functionality, and prices are set for each packaged level. Korea has a reasonably high level of adoption by delivering a relatively low-speed download broadband and minimal if any upload capability to a large number of areas. This was in line with the Korean *RYHUQPHQW¶VDLPRIHQDEOLQJDFFHVVWREURDGEDQG by the majority of the population for which they ZHUH SUHSDUHG WR ¿QDQFLDOO\ VXSSRUW WKH EURDGband provision across Korea and to minimize the resistance to price associated with the adoption of this new technology. The Japanese telecommunications industry on the other hand has implemented high-speed, highTXDOLW\)77+EURDGEDQGVHUYLFHDWDVLJQL¿FDQWO\ higher price to customers than in Korea, and to date, penetration in Japan has been somewhat slower than that in Korea. Therefore the link between customer need, service quality, adoption, and willingness to pay is blurred. Some Japanese customers appear willing to pay high connection and/or usage prices for higher value capability, whereas others either do not have EURDGEDQGRUDUHVDWLV¿HGRUKDYHXVHZKLFKFDQ EHPHWZLWKORZHUFRVW'6/1RVSHFL¿FPDUNHW UHVHDUFKKDVEHHQLGHQWL¿HGWKDWGH¿QHVWKHGLIIHUence between adopters and non-adopters. However it seems reasonable that the socio-economic factors LGHQWL¿HGLQWKH$XVWUDOLDQFDVHZLOOH[LVW In Korea, a lower price and a lower value broadband service has attracted a relatively high penetration, but does not necessarily satisfy the total mass market since the applications must be constrained by the lower level of broadband capability. Once again, QRVSHFL¿FVWXG\RQEURDGEDQGXVHUVDWLVIDFWLRQ LQ.RUHDKDVEHHQLGHQWL¿HG However, in an increasingly global communications world, what is being provided in one country is easily comparable to that in another. Therefore, a Korean customer may question why their service
133
Broadband for the Mass Market
is not as good as that in Japan, while a Japanese customer may be concerned that prices in Korea DUH VLJQL¿FDQWO\ ORZHU WKDQ WKH\ DUH UHTXLUHG to pay, with the outcome being that both have perceptions of value that is based on unrealistic expectations. For a business operation the differences between the two may be irrelevant or understandable, but the mass market tends to be conditioned by information gained from casual and possible illinformed sources, coloring their value perception and expectations. To add to the confusion of the relatively inexperienced potential Internet users represented in the mass market, broadband packages being offered by the various competitors appear inconsistent and prices being charged seemingly unrelated to the offerings. If international comparisons are made, this becomes even more apparent. Equally, as LGHQWL¿HGE\6ZDUW] WKHPDVVRIDOWHUQDtive packages merely adds to the confusion and the tendency to defer choice. This decision paralysis is supported by a recent Australian study by AdDPV LQZKLFKRQHSRLQWLGHQWL¿HGZDV³WKH complexity of the purchase decision acting as a barrier to adoption.” The 2(&'¶VSDSHURQEHQFKPDUNLQJ 2(&' 2004) reports the diversity of broadband packages and offers in the OECD, and shows clearly that for a new relatively unsophisticated customer to select an appropriate broadband to meet their as-yetXQLGHQWL¿HGQHHGVZRXOGEHGLI¿FXOWFRQIXVLQJ and likely to result in customers deferring making a purchase decision. This is further discussed in the next section on price. The outcome would be that if the wrong package is chosen, the value to the customer would be compromised. In summary, value appears dependant on the value customers place on the needs they identify but can be conditioned by information input from a wide variety of sources. Where their needs can or PLJKWEHVDWLV¿DEOHDWORZHUSULFHVHOVHZKHUHRULW appears that good value is not being offered compared to alternative options, adoption of broadband would be for them unnecessary or too expensive for what they get. At this stage, a universal application for the mass market requiring base-level broadband
134
LVQRW\HWLGHQWL¿HG(TXDOO\WKHUHLVDWWKLVVWDJH of development little differentiation of broadband capability based on the value and functionality of the content or applications currently on offer. The sole differentiation to date appears to be based on the technology of volume and speed of the usage required by the individual customers and little else, all of which for a new mass market customer are unknown requirements.
Price Limits The price setting regime around the world is varied. In some cases there is no charge for connection to the broadband network, in others excess usage prices apply while others throttle use when a prepaid limit is exceeded. In still other cases, download speed offers are priced depending on availability of the technology to support increased speed. The Telecommunications Journal of Australia (Saunders, 2005) review of the OECD Broadband Benchmarking Paper suggested that few if any RI WKH FXUUHQW SULFLQJ DUUDQJHPHQWV UHÀHFWHG D customer perspective of pricing for the value of the service supplied. The variables in any broadband service, such as whether upload capability was offered, the speed of download, number of Mbytes included, and so on makes comparison of UHODWLYHYDOXHVRIWKHYDULRXVSDFNDJHVGLI¿FXOWLI QRWLPSRVVLEOH,WDOVRLGHQWL¿HGWKDWZKLOH.RUHD has a good adoption history, it has yet to make a SUR¿W ZKLFK PD\ EH D IDFWRU UHVXOWLQJ IURP DQ inability to charge higher prices for relatively low levels of service value or delivery. Japan has built a high value service but a lower history of broadband take-up, and the question of return on their investment has yet to be determined. The question then arises, is the high price charged by -DSDQDGHWHUUHQWWRDGRSWLRQZKLOH.RUHD¶VSULFH regime and customer use are too low to enable an acceptable return on investment? In either case the balance to achieve satisfactory outcomes for both customers and infrastructure providers, if wider diffusion of broadband is to take place, is critical and needs to be developed if providers and broadband are to achieve their potential.
Broadband for the Mass Market
7KHGLI¿FXOW\IRUDWUDGLWLRQDOFRPPXQLFDWLRQV company reliant on phone call revenue is to make the transition from the current hybrid copper network WRRSWLFDO¿EHUDQGRWKHULQIUDVWUXFWXUHQHFHVVDU\ to adequately support a high-quality, high-speed broadband network, which is the key requirement for them to move any of their current services to broadband. Consequently their broadband pricing structures and the capability they provide is colored by the fear of migration from their existing network and the need to either retain or develop a new revenue source. Customers appear to be only willing to pay a certain price for what they consider is the use and value of the service they require from broadband as opposed to the standard network services already available; and if the capability of fast and quality application delivery is not possible because of the technology, the value perception must be limited. If they want access to sporting Web sites, that is what they want to pay for, but they want realtime and quality service. That means availability where they want it, available when they want it, and at a price better or at least comparable to any existing communication or information provision available. As far as the average mass market customer is concerned, the means to access that Web site and the broadband network is a totally transparent entity. Therefore to price access to broadband based purely on technology is not visible to the customer and not relevant to his or her perception of pricing for valueit is what they can do with that access and whether what they do is at a lower economic cost to them than any other alternative. In summary, it is not what the broadband provider sees as price for value because of the technolRJ\LWLVWKHFXVWRPHU¶VGHFLVLRQRQZKHWKHUZKDW they are being offered is good value at the price offered, and that is in effect based on content or capability available to them from this new media. Therefore for the mass market there appears to be a need for a commodity product, such as standard voice calls at a lower price than currently available on the existing phone network, to attract them to broadband adoption. It is then possible to up-sell greater capability as the customer develops a
need for enhanced services, such as voice/video or real-time and interactive entertainment, but the technology needs to be in place to handle the higher levels of service capability beforehand and before the demand is evident in practice.
FUTURE TRENDS At this stage of development of the market and technology, the future of broadband is clouded. That it will develop is unquestioned; it is the way it develops that is uncertain. There is a need for the establishment of a set of international interface standards to enable connection between IP platforms, such that peer-to-peer communication in some utility form can be achieved without incompatibility reminiscent of the VHS/Beta situation that occurred with the introduction of videotape recording. An Australian Industry Forum (Evans &KLUJZLQ GLVFXVVLRQSDSHULGHQWL¿HVWKH complexity and variety of the technology solutions possible with the provision of peer-to-peer communication and the alternative approaches that could be adopted if there were some level of agreement between the players in the market. The discussion paper clearly states that some standards need to be set by the European Telecommunications Standards Institute (ETSI) and the International Telecommunications Union (ITU) if platformto-platform connection is to take place, opening up the full capability of broadband as a complete communications and information capability. This is where the question of the rate of broadband expansion into the broader market is indeterminate, since platform-to-platform interface UHTXLUHV¿UVWWUXVWEHWZHHQWKHFRPSHWLQJIRUFHV and second complex revenue transfer and billing, similar to the international calling agreements between competitors and countries. In a highly FRPSHWLWLYHHQYLURQPHQWWKLVPD\EHGLI¿FXOWWR do since all the competing players are now required to satisfy not only their domestic markets and governments, but also the international stock markets, WKHVKDUHKROGHUVDQGWKH¿QDQFLDOLQVWLWXWLRQV To some degree, this situation is unprecedented in the communications market, in that in follow-
135
Broadband for the Mass Market
ing the new technology developments, falling prices, and the increasing commodity view of the FRPPXQLFDWLRQ FRPSDQ\¶V FDVK FRZ SURGXFWV WKHSUHVVXUHLVRQWRLQFUHDVHUHYHQXH¿QDQFLDO performance, and dividends. In this environment, it may be that free market competition is in itself a constraint on both the willingness of communications companies to take risks by investing in new infrastructure where the return on capital may be over a long period or where it provides an alternative service that erodes the current revenue stream.
CONCLUSION For existing communications carriers the transition from one network technology to another is an expensive and risky venture. Equally it is a major shift in traditional pricing and marketing VWUDWHJLHV:LWKEURDGEDQGSULFHLVIDUEHWWHU¿Wted to access pricing as the major revenue stream that has always been secondary to call revenue in traditional communications services. A further issue is that increase in adoption of broadband by the mass market increases the risk to these traditional carriers of both losing their long-held revenue stream with no guarantee of attaining the same level of return from the new technology and the opening up of their customer base to competitors. Their biggest problem from their perspective is that they no longer control customer use or the reason for that use unless they at least move their existing products, such as local, long distance, and international calling, plus all the support products and existing infrastructure to the new technology. This will require careful and well-planned pricing strategies to minimize loss of revenue, and more important, loss of customer base if they are to survive. It is clear that adoption of broadband by the mass market will occur with or without the active involvement of the traditional carriers. However, the transition and adoption of broadband will not be as fast or as dramatic as the adoption of mobile communication is reputed to have been. But with the active acceptance by the traditional carriers,
136
the adoption by the market will occur as customers adapt to the newer technology. It would seem that ultimately these traditional communications companies will have to move GRZQWKLVSDWKUDWKHUWKDQ¿JKWDUHDUJXDUGDFWLRQ that can at best only slow the migration of voice WUDI¿FZKLOHORVLQJWKHPEURDGEDQGDFFHVVUHYHQXH as other providers move in. Without doubt, the challenge is whether the traditional and incumbent carriers can persuade WKHLQYHVWLQJFRPPXQLW\WKDWWDNLQJD¿QDQFLDO risk is necessary and the returns may be over the longer term. The setting of prices and balancing the revenue transition will be a more delicate operation than ever before for these telecommunication companies to both keep their customers and to maintain shareholder satisfaction.
REFERENCES $GDPV 3 ,VRODWLQJ µZK\¶ $XVWUDOLDQ households adopt broadband. Telecommunications Journal of Australia, 56(3/4), 27-36. Adams, D. (1982). Life, universe and everything. London: Pan Books. Bachelet, C. (2005). Broadband’s so-called killer application, Wireline Strategy @OVUM service. Retrieved December 28, 2006, from http://www. ovum.com/go Cameron, A. (2004). Factors affecting broadband adoption and demand. Telecommunications Journal of Australia, 54(2), 53-60. Darling, P. (2005). The 21st century CAN. Telecommunications Journal of Australia, 55(2), 3-14. Evans, S., & Chirgwin, R. (2006). Industry discussion paper QOS-based VOIP service interconnectivity. Market clarity. Sydney: Australian Communications Industry Forum. Fitzpatrick, D., & Perin, A. (2006). Alcatel and KPMG identify regulatory levers used by governments around the world to promote broadband infrastructure development. Retrieved May 5, 2006, from http://www.kpmg.com.au
Broadband for the Mass Market
Lim, J. (2002). Killer applications for broadband ,QWHUQHWDUHSURPLVLQJLQ$VLD3DFL¿F (pp. 1-6). Hong Kong: ITU Telecom Asia. Luber, A. (2001). The ultimate (surprise) broadband killer application. Capitalism Magazine. Retrieved December 23, 2006, from http://www.capmag. com/articlePrint.asp Middleton, C.A. (2003). What if there is no killer application? Journal of Information Technology, 18(4), 231-245. OECD. (2004). Benchmarking prices in the OECD. OECD Committee for Information, Computer and Communications Policy, DSTI/ICCP/TISP/ FINAL. OECD. (2005, June). Broadband subscribers per 100 inhabitants, by technology. Retrieved January 6, 2007, from http://www.oecd.org/16/0 PriceWaterhouseCoopers. (2004). White paper details key applications for broadband Internet, short-term opportunities are addition of video to existing applications, not distribution of video entertainment. Retrieved December 23, 2006, from http://www.pwc.com Saunders, R. (2005). Benchmarking broadband. Telecommunications Journal of Australia, 55(2), 20-24. Schwartz, B. (2003).The paradox of choice. Canada: HarperCollins. Smith, M.W., & Leung, H.-T. (2002). Finding the killer application: The role of the Broadband Applications Laboratory. BT Technology Journal, 20(1), 11-21. Wale, C. (2002). In search of the elusive ‘killer application’ that will drive broadband uptake in Australian households. Department of Information Systems, University of Melbourne, Australia. Wearden, G. (2001). Broadband’s killer apps will be created at home. Retrieved December 23, 2006, from http://www.zdnet.com.au/news/comunications
KEY TERMS Australian Competition and Communications Commission (ACCC): The government body established to regulate the industry and ensure fair competition in the Australian market. Similar to the U.S. Federal Communications ComPLVVLRQ )&& DQGWKH2I¿FHRI&RPPXQLFDWLRQ (OFCOM) in the UK. Blogs: Web logs, personal diaries, comments, or other material posted on the network by individual broadband users for comment or information. Customer Access Network (CAN): The wiring or cable link dedicated to communication customers from the local exchange to the customer premises. Digital Subscriber Line (DSL): The CAN PRGL¿HG WR FDUU\ GLJLWDO DV RSSRVHG WR DQDORJ signals. European Telecommunications Standards Institute (ETSI): The regulatory body established to ensure common international communications standards to enable country-to-country communication. Fiber To The Home (FTTH): The use of optiFDO¿EHUDVDUHSODFHPHQWIRUWKHVWDQGDUGFRSSHU CAN wiring. Fiber To The Node (FTTN): The use of existLQJRSWLFDO¿EHUWROLQNWKHLQGLYLGXDOVZLWFKLQJ nodes within the communications network. International Telecommunications Union (ITU): A union of the majority of the international communications providers dedicated to the development of communications and inter-exchange of technology and standard setting. Internet Platform: A server used by an Internet provider to support Internet access by their customers. Internet Service Provider (ISP): Usually a commercial organization selling access to the Internet.
137
Broadband for the Mass Market
Organization for Economic Cooperation and Development (OECD): The European body established to support free trade and cooperation across all the member countries. Telecommunications Journal of Australia (TJA): The quarterly journal supported and published by the communications industry in Australia. Telecommunications Society of Australia (TSA): The society established and supported by
138
all of the competing and developing communications operations in Australia. Television over DSL (TVoDSL): Television broadcasting or data cast television delivered over broadband DSL lines. Video on Demand (VoD): Customer-selected FRPPHUFLDOYLGHRDQG¿OPGHOLYHU\ Voice over Internet Protocol (VoIP): The transmission of voice communication across the Internet using data packaging technology.
139
Chapter X
Competition, Regulation, and Broadband Diffusion: The Case of New Zealand Bronwyn Howell Victoria University of Wellington, and New Zealand Institute for the Study of Competition and Regulation, New Zealand
ABSTRACT New Zealand offers a through-provoking case study of the effects of different competition and regulatory policies on broadband diffusion rates. Despite having one of the highest rates of Internet connection and usage in the OECD, widely available broadband infrastructure, and low broadband prices, broadband uptake per capita languishes in the bottom third of the OECD. While low uptake has typically been attributed to competition and regulatory factors associated with New Zealand’s ‘light-handed’ regulatory regime, this chapter proposes that a more credible explanation lies in a combination of New Zealand’s legacy of demand-side regulations, in particular the retail tariff options for voice telephony, and the limited value being derived by New Zealand residential consumers from the small range and narrow adoption of applications used currently that necessitate broadband connections. The New Zealand case illustrates the effect that legacy regulations can have on both the diffusion of new technologies per se and the choices made by consumers between different generational variants within that technology. The case indicates a need for more research on the effect of telecommunications industry regulations on demand-side uptake factors.
INTRODUCTION Despite being one of the leading countries in the adoption1 and use of Internet access, and having (1) a sophisticated e-commerce infrastructure and internationally very high numbers of e-commerce transactions per capita; (2) a population regarded as avid early adopters and users of applications such as electronic funds transfer and electronic
commerce (Boles de Boer, Evans, & Howell, 2000); (3) widespread and early deployment of a variety of high-speed broadband technologies; and (4) internationally low prices for broadband products, 1HZ=HDODQG¶VEURDGEDQGXSWDNHSHUFDSLWDKDV been consistently in the lower third of the OECD2 (see Table 1). The simplistic explanation typically offered for low broadband uptake has been New =HDODQG¶V WHOHFRPPXQLFDWLRQV FRPSHWLWLRQ DQG
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Competition, Regulation, and Broadband Diffusion
Table 1. Relative OECD Internet connectivity and uptake rankings, 2003 (Howell, 2003, p. 39)
regulatory framework (Ministry of Economic Development, 2006). ,QDQLQGXVWU\ZKHUHVHFWRUVSHFL¿FUHJXODWLRQ and local loop unbundling (LLU) in particular, have been strongly advocated and widely adopted internationally3ZLWKWKHVSHFL¿FREMHFWLYHRIDFFHOerating broadband uptake rates,4 New Zealand has stood apart from most of its OECD counterparts. Principal differences are its historic reliance upon competition law and minimal regulation to shape SDUWLFLSDQWV¶DFWLRQV5 and its reluctance to impose competitive access obligations on its solitary priYDWHO\RZQHGQDWLRQDO¿[HGOLQHQHWZRUNRSHUDWRU (Boles de Boer & Evans, 1996). In the absence of GHWDLOHGPDUNHWDQDO\VHVµFRPSHWLWLRQSUREOHPV¶ attributed to the different regulatory approaches have become convenient scapegoats upon which to lay the blame (e.g., Network Strategies, 2006) for WKH1HZ=HDODQGEURDGEDQGXSWDNHµSUREOHP¶6 7KHFUHGLELOLW\RIWKHµFRPSHWLWLRQSUREOHPUHJXODWRU\GLIIHUHQFHV¶H[SODQDWLRQIRUORZEURDGEDQG uptake begins to fail, however, when juxtaposed DJDLQVW1HZ=HDODQG¶VOECD leadership in virtually every other indicator associated with uptake and utilization of the Internet. The same competitive and regulatory framework applying during
140
the rollout of broadband infrastructures appears not to have impeded investment in, or uptake and utilization of, all other non-broadband Internetrelated infrastructures and applications.7 Neither does it appear to have impeded investment in or internationally competitive pricing of broadband services.8 The primary characteristics that competitive markets are presumed to deliver, and which are the primary objectives of regulatory intervention when the competitive process is impededlow prices, timely introduction of new services and higher product qualities, and universal availability and prices9 have already been achieved. The adGLWLRQDOEHQH¿WVWKDWDUHW\SLFDOO\H[SHFWHGIURP increased competitive intensitygreater provider variety and consequent marginal improvements in service qualityGRQRWDSSHDUWREHVXI¿FLHQWWR account for the substantial differences in broadband diffusion per capita between New Zealand and countries with smaller Internet-using populations, higher absolute prices for broadband access, and less widely available infrastructures.10 The New Zealand case study begs the question RIZKDWUROHVSHFL¿FUHJXODWRU\LQWHUYHQWLRQVSOD\ in determining broadband diffusion rates. Firstly, how has New Zealand been able to achieve sup-
Competition, Regulation, and Broadband Diffusion
ply-side conditions in the provision of broadband access that have yet to be achieved by the majority of countries employing extensive regulatory LQWHUYHQWLRQLQ¿[HGOLQHDFFHVVPDUNHWV WKDWLV various forms of local loop unbundling), despite its reliance upon a privately owned network provider with substantial market power, and only competition law and a handful of very light-handed regulations (and in particular, an absence of LLU) with ZKLFKWRVKDSHSDUWLFLSDQWV¶LQWHUDFWLRQV" Secondly, if increases in the rate of broadband diffusion are the primary objectives of regulatory intervention, then does the New Zealand case data, by its apparent absence of supply-side impediments to the timeliness, availability, and price of broadband services, yet extremely low levels of broadband uptake, support the contention that in the diffusion of this technology, supply-side regulatory intervention, even if it is successful in inducing wider availability, lower prices, and higher qualities of broadband products supplied, has minimal effect on the diffusion rate? Rather, is there some other factor on the demand side of the market that is acting as a brake on broadband GLIIXVLRQ",IWKHDQVZHUWRWKLVTXHVWLRQLVµ\HV¶ then is substantial policy reliance upon supply-side regulatory measures such as LLU and local and national government subsidies to infrastructure providers in order to accelerate national broadband diffusion rates misplaced? Thirdly, but also quite importantly, what does WKHIDFWWKDW1HZ=HDODQG¶VORZDQGVORZGLIIXsion pattern is occurring in a population of highly sophisticated, widely experienced Internet users with an internationally very high rate of Internet diffusion say about not just the diffusion of broadband per se, but about the rate of substitution by sophisticated, experienced users of legacy Internet access technologies (i.e., dial-up) with frontier technologies (i.e., broadband)? This chapter addresses these questions. Firstly, the nature of market interactions in determining technology diffusion rates is discussed. The naWXUHDQGHI¿FDF\RIWKHUHJXODWRU\LQWHUYHQWLRQV
that have been utilized in telecommunications markets are examined in light of their focus of interventionthe supply side of the market or the demand sideand their effects on the diffusion of both Internet and broadband access. The New Zealand case data indicate that, as there is no eviGHQFHRIDVLJQL¿FDQWVXSSO\VLGHLPSHGLPHQWWR broadband diffusion relative to the rate observed in other countries, the low diffusion levels may be attributable to factors on the demand side of the market. As demographic factors do not appear to account for the differences, are factors associated with the high rates of Internet use via dial-up access implicated? This leads to a development of a demand-side model that, holding all supply-side and demand-side factors other than the method of Internet access constant, assesses when, or even if, a particular Internet access technology will be adopted. 7KHPRGHOLVWKHQDGDSWHGWRUHÀHFWWKHWDULII characteristics prevailing in the New Zealand Internet access marketVSHFL¿FDOO\ÀDWUDWHXQmetered pricing of dial-up Internet use. The effect of this tariff instrument, the result of regulatory requirements predating the commercialization of the Internet, has been implicated in the rapid diffusion of dial-up Internet access in New Zealand. The model illustrates that the tariff instrument may also be implicated in a slow rate of substitution to broadband by existing dial-up Internet users if the range of applications used is small or the value derived from them is small relative to the increase in cost of broadband access. Furthermore, the tariff instrument and other regulations are shown to also KDYHKDGDVLJQL¿FDQWLQÀXHQFHXSRQWKHLQFXPEHQW WHOHFRPPXQLFDWLRQVSURYLGHU¶VVWUDWHJLFDSSURDFK to the supply of ADSL services. Thus, New Zealand regulatory intervention is implicated positively in the diffusion of Internet access via its effect on dial-up access, but negatively in the diffusion of broadband Internet access, GHVSLWHKDYLQJSRVLWLYHO\LQÀXHQFHGWKHFUHDWLRQ of a desirable set of supply-side conditions in the market.
141
Competition, Regulation, and Broadband Diffusion
COMPETITION, REGULATION, AND DIFFUSION At the crux of the questions addressed by this chapter is the extent to which supply-side or demand-side factors are the principal drivers of, or impediments to, the New Zealand broadband diffusion rate. In a perfectly competitive market, neither side will dominate. The diffusion rate observed will be the consequence of the systemic interaction of suppliers and consumers. Changes in consumer demand will be matched by changes in producer supply. New providers enter the market to satisfy increasing demand, and producers continually inQRYDWHWR¿QGQHZFKHDSHUDQGPRUHQRYHOZD\VRI varying the technology in order to capture a greater share of the returns from the demand increase. While supply-side factors make the purchase of a technology possible (i.e., no uptake is possible if the technology is not supplied), uptake (diffusion) rates measure the demand-side response to its availability. Consumers will buy a technology at DJLYHQSULFHRQO\LIWKHEHQH¿WVGHULYHGIURPWKH technology exceed the costs of its purchase. No matter how widely available the technology LVLIWKHEHQH¿WVGHULYHGE\WKHPDMRULW\RIFRQsumers do not exceed the costs, then widespread purchase will not occur, and the diffusion rate will be low. Low diffusion rates when the technology LVSULFHGYHU\ORZLQGLFDWHWKDWWKHEHQH¿WVDUHVR small that they do not exceed the very low price at which the technology is offered (i.e., its cost of production). Conversely, high diffusion rates, especially when the technology is high priced, LQGLFDWHWKDWWKHFRQVXPHUEHQH¿WVDUHFRQVLGHUDEOHDVFRQVXPHUVDUHSUHSDUHGWRVDFUL¿FHDODUJHU amount of their limited resources to acquire it. However, markets may deviate from perfect competition. The most common deviation in telecommunications markets arises in the balance of market power favoring suppliers. If incumbent suppliers face little or ineffective competition (e.g., a supplier is the only provider or there are barriers to other providers entering), they may be able to restrict the supply of the technology, raising prices above the competitive level and thereby
142
DUWL¿FLDOO\GHSUHVVLQJWKHXSWDNH GLIIXVLRQ UDWH relative to the counterfactual where new providers can freely enter. The consequence of suppliers choosing to H[HUFLVH PDUNHW SRZHU LV SHQWXS XQVDWLV¿HG demand. Many consumers would gain from the SXUFKDVHDVWKHLUEHQH¿WVH[FHHGWKHFRVWRISURduction, but only a few are able to do so as supply LVFRQVWUDLQHG8QVDWLV¿HGWKXVFRQVXPHUVIRUIHLW SRWHQWLDOLQFUHDVHVLQEHQH¿W)DFWRUVVXFKDVVXQN costs and natural monopoly (where the economy RIVFDOHHIIHFWIURPYHU\KLJK¿[HGFRVWVDQGYHU\ low variable costs relating to the infrastructures required for service provision is so great that it is HFRQRPLFDOO\HI¿FLHQWWRKDYHRQO\RQHSURYLGHU serving the entire market) mean that these issues DUHSDUWLFXODUO\UHOHYDQWLQWKHSURYLVLRQRI¿[HG line telephony services (e.g., ADSL), although the risk is reducing as technological innovation reduces WKHH[WHQWRIWKH¿[HGDQGVXQNFRVWVRIVHUYLFH provision infrastructure (Melody, 2005). The typical response is regulatory intervention aimed at preventing the suppliers with market power from exercising it. If the regulatory interventions are successful, then supplier market SRZHULVQRWH[HUWHGSHQWXSGHPDQGLVVDWLV¿HG by increases in supply, the price in the market falls, and the diffusion rate increases to the competitive OHYHOUDWKHUWKDQWKHGHSUHVVHGOHYHO7KHEHQH¿WV are accrued principally by those additional customers who now buy the product because supply and demand are now in equilibrium.
Supply and Demand in New Zealand The apparent absence of supply-side impediments in the New Zealand market suggests that an explanation for slower rates of broadband diffusion may lie in demand-side factors. Broadband is widely available, low priced, and of high quality. Moreover, there is already a large population of Internet users in New Zealand, with evidence of high usage. Demand-side factors implicated in the diffusion of broadband include wealth (e.g., GDP per capita or household income), population density, the degree of urbanization, and user characteris-
Competition, Regulation, and Broadband Diffusion
tics such as gender, age, and education (Wallsten, 2006; Rappoport, Kridel, & Taylor, 2002; Network Strategies, 2006; among others). Yet none of these DSSHDUVXI¿FLHQWWRDFFRXQWIRU1HZ=HDODQG¶VORZ rate relative to that of other countries with similar demand-side characteristics (e.g., Network Strategies, 2006). Even when multivariate models are used to account for multiple differences between GHPDQGSUR¿OHVRIFRXQWULHV1HZ=HDODQGVWLOO appears to have an inexplicably low diffusion rate. For example, New Zealand data applied to :DOOVWHQ¶V PRGHO VXJJHVW DQ DFWXDO UDWH about 10% lower than the model would predict from income, population density, and urbanization characteristics, as well as observed prices and available supply of broadband technologies. 2QHVLJQL¿FDQWGHPDQGVLGHGLIIHUHQFHWKDWKDV not been examined systematically is the effect that 1HZ=HDODQG¶VKLJKLQLWLDOUDWHRIGLDOXS,QWHUQHW users might have had on broadband diffusion rates. The apparent reluctance of experienced Internet users in New Zealand to buy broadband connections even in the apparent absence of any supplyside impediments suggests that factors associated with the demand side of the pre-existing dial-up Internet market may be at least in part contributing to the low broadband diffusion rate. The focus upon regulatory intervention on the supply side of the broadband market may be distracting from the ability to detect these effects. A risk exists that regulators focused on improving the supply-side factors in the broadband market are ill-equipped to identify demand-side effects. This information asymmetry leads to risks of over-attribution of the effect of regulatory intervention on uptake statistics for a given variant of a technology. Underpinning this presumption is that the new technology has emerged in isolation from existing technologies. In practice, this is rarely the case. The new technology may offer a FRPSOHWHO\QHZVHWRIEHQH¿WVHLWKHULQWKHLURZQ ULJKWRUDVFRPSOHPHQWVWREHQH¿WVIURPH[LVWLQJ technologies, or may enable consumers to substitute EHWZHHQPHWKRGVRIDFTXLULQJWKHVDPHEHQH¿WV The extent to which a complement or substitute technology will be adopted depends therefore not RQWKHDEVROXWHEHQH¿WVRIWKHQHZWHFKQRORJ\EXW
RQWKHUHODWLYHEHQH¿WVFRPSDUHGWRWKHH[LVWLQJ technology. By holding all supply-side and demand-side factors apart from those directly associated with the two different technologies constant, a demand-side model of broadband purchase both in its own right for a new Internet user, and as a substitute for existing dial-up users, can be developed. The model, based upon the utility derived by users from a de¿QHGVHWRI,QWHUQHWDSSOLFDWLRQVDOORZVWKHHIIHFW of different regulatory instrumentsprincipally price-setting and mandatory tariff structuresto be examined. Moreover, it leads to questions about WKHHI¿FDF\RIWKHXVHRIGHPDQGVLGHVWDWLVWLFVWR measure the success or otherwise of supply-side regulatory intervention.
Supply-Side Regulatory Intervention The argument typically offered to justify the regulation of telecommunications markets, and incumbent telecommunications operators in particular, LVWRFRQVWUDLQRSHUDWRUV¶XVHRIPDUNHWSRZHUWR PD[LPL]HWKHLUSUR¿WVE\UHVWULFWLQJWKHVXSSO\RI services to consumers and charging prices above the point where welfare is maximized. Regulatory intervention principally seeks to reduce these supply-side constraints (including the use of price controls), resulting in a focus upon the use of metrics indicating changes in uptake per capita in order to assess the performance of regulatory intervention in constraining supplier market power. Increases in uptake per capita are thus the principal demand-side indicators to regulators of increases in total welfare resulting from reductions in the exertion of supplier market power, inasmuch as it has been previously established that the supply VLGHZDVLQIDFWEHLQJFRQVWUDLQHGE\VXSSOLHUV¶ actions. The inference drawn is that regulatory intervention has been the driver of the ensuing welfare increase represented by increased uptake. This is distinct from the welfare inferences that can be drawn from increases in uptake (i.e., the natural diffusion pattern) of a completely new technology. $QLQFUHDVHLQXSWDNHLQWKLVFDVHVLJQL¿HVWKDWWKH EHQH¿WVRISXUFKDVLQJWKHWHFKQRORJ\H[FHHGWKH costs for an increasing number of customers who
143
Competition, Regulation, and Broadband Diffusion
are gradually choosing to purchase. This may occur as the price reduces (e.g., as more suppliers enter the market or as a growing market leads to economies of scale and improved production processes, resulting in lower costs passed on to consumers as lower prices); as new applications utilizing the technology are developed, thereby increasing the size of the potential market; or as more potential FXVWRPHUV EHFRPH LQIRUPHG DERXW WKH EHQH¿WV that the technology offers them, relative to other calls on their budgets.
Risks from Demand-Side Information Asymmetries Melody (2005) notes that the current body of regulatory knowledge has been accumulated principally from regulating suppliers whose exertion of market power is well established, and where it is has been established that there is a pent-up demand for products and services that is not being met VSHFL¿FDOO\EHFDXVHRIVXSSOLHUDFWLRQV+HQFHLQ the conduct of regulatory activities focusing upon constraining these supply-side powers, “attention to demand is not a priority” (p. 28). The consequence is an approach to market development “fostered by the fascination of many old and new operators with the technical capabilities of new technologies” and a prevailing view that “if we build it, they (consumers) will come,” at the expense of understanding the role of the demand-side factors in the diffusion of technologies supplied by the regulated providers. Thus, when new technologies are diffusing in industries that have been regulated because of the presence of historic supply-side market power, a risk exists that uptake increases arising from the natural patterns of a new technology offered by the regulated providers diffusing may be inappropriately attributed to supply-side regulatory LQWHUYHQWLRQVVLPSO\EHFDXVHWKHUHLVLQVXI¿FLHQW understanding of the relevant demand-side factors that are simultaneously contributing to the observed diffusion pattern. Likewise, it may be inappropriately assumed that diffusion rates can be accelerated with supply-side interventions at YHU\HDUO\VWDJHVRIWKHQHZWHFKQRORJ\¶VGLIIX-
144
sion, simply because the same interventions have been instrumental in raising uptake, and therefore welfare, in more mature product markets. Such approaches risk distorting the normal competitive interactions that occur in the early stages of any new technology, where there are substantial uncertainties for both suppliers and consumers about the short- and long-term merits and uses of the technology. The diffusion of Internet access in the historically highly regulated telecommunications industry provides an example where the risks of distortions arising from regulatory intervention DUHVLJQL¿FDQW([LVWLQJUHJXODWRU\PHDVXUHVZLOO affect diffusion patterns, with differences in initial conditions and regulations necessarily contributing to differences in patterns between countries being observed. Melody (2005) further notes that “typically, both regulatory rules and tariff restrictions prevent experimentation and add to the risks of innovation, particularly by intermediaries and endusers on the demand side of market development” (p. 33). In order to determine whether regulatory, supply-, or demand-side factors are responsible for REVHUYHGGLIIHUHQFHVWRDVVHVVWKHOLNHO\HI¿FDF\ of any regulatory intervention and to improve regulatory quality, an understanding of existing demand-side factors and their contribution to diffusion patterns is essential.
A Demand-Side Model of Broadband Uptake In the case of Internet access technologies, demand LVGHULYHGIURPWKHEHQH¿WVWKDWFRQVXPHUVDFFUXH from the applications utilizing Internet access (Crawford, 1997). As with all telecommunications technologies, Internet demand generates a requirement for both access and usage (data transmission) (Laffont & Tirole, 2002). Dial-up and broadband diffusion statistics measured as connections per capita are capturing only the component that relates to Internet access. When making a decision about whether to purchase a connection, and which connection type to purchase, a user must consider the effect of both access and usage charges required to
Competition, Regulation, and Broadband Diffusion
GHULYHEHQH¿WVIURPDGH¿QHGVHWRIDSSOLFDWLRQV While all Internet users must pay an access charge, users of applications generating large amounts of GDWDWUDI¿FZLOOLQFXUJUHDWHUXVDJHFKDUJHVWKDQ users of applications generating lower amounts of data transfer. The user will purchase an Internet FRQQHFWLRQRQO\LIWKHEHQH¿WVIURPWKHDSSOLFDWLRQV used exceed the sum of access and usage charges (Wenders, 1990), plus other incidental costs. The issue of diffusion of broadband is complicated by the fact that it is simply one variant in a series of technologies enabling access to the Internet. If the Internet is considered a general purpose technology (GPT) as per Helpman and Trajtenberg (1996), then dial-up access and broadband access are two generations of Internet access technolRJ\7KHYLQWDJHWHFKQRORJ\ LHWKH¿UVWWREH developed) is dial-up access. The frontier (i.e., the most recent technology developed in respect of the GPT) is broadband. The frontier is substantially more capable than the vintage, but is introduced in a market where the vintage has already been diffusing. Both versions offer capability to access, but the frontier will offer advantages that may offer JUHDWHUEHQH¿WVWRVRPH EXWQRWQHFHVVDULO\DOO of the vintage users. Assuming that the set of applications used is held constant, vintage users will substitute to the IURQWLHUZKHQWKHDGGLWLRQDOEHQH¿WVRIXVLQJWKH frontier technology exceed the additional costs IRU WKH GH¿QHG VHW RI DSSOLFDWLRQV 7KH IURQWLHU technology is said to be dominant when all new users of the GPT purchase it, but only conditionally dominant when some new users choose to purchase the vintage technology, despite the frontier being available (Helpman & Trajtenberg, 1996). 7KH VSHFL¿F EHQH¿WV WKDW EURDGEDQG DFFHVV confers are savings in time (due to faster data transfer) and the ability to use applications that are physically impossible to use on dial-up (Howell & Obren, 2003). Thus, the user will substitute if the EHQH¿WVRIEURDGEDQGDUHJUHDWHUWKDQWKHDGGLWLRQDO costs incurred to support use of the existing application base. Moreover, if broadband offers the RSWLRQ RI DFFHVV WR DGGLWLRQDO EHQH¿WV WKDW ZLOO accrue from the use of additional applications that can only be utilized with broadband access, and
WKHVHSURMHFWHGEHQH¿WVH[FHHGDQ\DGGLWLRQDOFRVWV that their usage incurs, substitution will likewise occur (as per Boyan & Jovanovic, 2000).11
Purchase and Substitution Holding all other supply-side and demand-side factors constant, the broadband purchase and substitution decision can be modeled thus.12 Assume an Internet customer i uses a set of j = {1, 2, 3, ..., n} applications each requiring Xj megabytes of EDQGZLGWKWRJHQHUDWHDEHQH¿W QHWRIDOOFRVWVRWKHU than those associated with information transfer) of 6 nj 1 ij 7KHEHQH¿WFDQEHJHQHUDWHGXVLQJHLWKHUD broadband or a dial-up Internet connection. Fixed costs are FD for dial-up access and FB for broadband (e.g., equipment, telephone line, DSL line, cable subscription, ISP subscription). Assuming that dial-up access is metered per minute connected and broadband per megabyte transferred, variable costs are VD per minute for dial-up access (including both telephone and ISP time-based charges) and VB per megabyte transferred for broadband. 7KHXVHU¶VWLPHWDNHQWRWUDQVIHULQIRUPDWLRQIRU application j is TDj minutes using dial-up and TBj minutes using broadband (TDj > TBj > 0). The user has a value of time Ji per minute. The user can choose to purchase no connection, a dial-up connection, or a broadband connection, depending upon which option renders the highest utility 3 Ki PHDVXUHGDVEHQH¿WVOHVVFRVWV$FRQQHFWLRQ will be purchased only if 3 Ki ! 0, K D, B . If all other user costs are independent of the information WUDQVIHUPHWKRGWKHQWKHXVHU¶VQHWXWLOLW\IURP Internet use will be: 3 Di
n
6 j 1 (Eij ( J i VD )TDj ) FD using dial-up
(1)
and 3 Bi
6 nj 1 (
ij
i TBj VB j ) FB
using broadband.
(2)
Users will continue to add applications to their portfolios, and increase the usage of exist-
145
Competition, Regulation, and Broadband Diffusion
ing applications (incurring additional costs in minutes of time and data transfer consumption), DVORQJDVWKHDGGLWLRQDOEHQH¿WVH[FHHGWKHDGditional costs. Dial-up access will be purchased where 3 Di ! 3 Bi ! 0 and broadband where 3 Bi ! 3 Di ! 0 . Substitution of dial-up with broadband will occur when 3 Bi > 3 Di . If one application j=n cannot physically operate using dial-up, then broadband will be purchased only if 3 Bi exceeds 3 Di 6 nj 11 ( ij ( i VD )TDj ) FD . Broadband will be preferred to dial-up when: ( FD FB ) i 6 nj 1 (TDj TBj ) 6 nj 1 (VD TDj VB j ) ! 0 (3)
Equation (3) shows that broadband substitution for dial-up will be more likely to occur if the ¿[HGFRVWVIRUGLDOXSDUHODUJHUHODWLYHWRWKH¿[HG cost of broadband, the value of time for the user is KLJK DQGRUEURDGEDQGLVVLJQL¿FDQWO\IDVWHUWKDQ dial-upconnection speed is implicitly captured in the user value of time), or the volume of information transfers is large. It is less likely to occur if the per-megabyte charge is large relative to the per-minute charge, the value of user time is low, or the relative speed difference is small. If broadband was truly a dominant technology, the applications predominantly used were feasible only on broadband, or if time is valued highly and the time savings from broadband use are large because data transfer volumes are large, then broadband would quickly come to dominate dial-up access. However, empirical data in both the New Zealand and other contexts does not appear to support this contention. New dial-up connections continue to be sold to new Internet users in the United States (Horrigan, 2006), New Zealand (Howell, 2006a), and Europe (EU, 2006). Different user time valuations, however, may lead to conditional dominance of broadband, at least for some user classes. Different time valuations plausibly account for observations of a greater likelihood of higher income users (presumed to have higher valuations of time) purchasing broadband connections than lower income users (Horrigan, 2006; Rappoport et al., 2002), even though in some surveys they appear
146
to use the connection less frequentlyHorrigan (2006) notes that higher income users spend less time per month on Internet activities than lower income users. Low valuations of time and/or low data volumes are also consistent with experimental (Varian, 2002) and survey (Horrigan, 2006; EU, 2006; Point Topic, 2006) data, suggesting that the majority of consumers are not prepared to pay large premiums for faster connections, given the current application base and usage patterns.13 Together, these data suggest that the applications currently available and routinely used to generate EHQH¿WV DSDUWIURPWKRVHVXSSRUWLQJvoice over Internet protocol telephony, and audio and video streaming and gaming, for which a broadband connection is mandatory) may not, as yet, necessitate the purchase of broadband connections for the majority of Internet users. Rather, they suggest that broadband technology choice is largely determined by relative prices, especially at low usage volumes. Usage data from the United States, Australia, and New Zealand suggests that the average consumption of megabytes transferred via broadband is small, and in some countries may be decreasing across time as more broadband connections are sold, suggesting that the consumption of new broadband users is actually less than that of the early adopters, as occurs typically in telephony markets as the technology becomes more widely diffused (Howell, 2006a). This would support the contention of a limited number of applications being used and a relatively static number of transfers14 occurring.
Flat-Rate Tariffs and the Demand-Side Substitution Model 7HFKQRORJ\ FKRLFH PD\ DOVR EH VWURQJO\ LQÀXenced by tariff structures. If historic regulatory intervention in the telecommunications market KDVUHVXOWHGLQSULFLQJSDWWHUQVWKDWLQÀXHQFHWKH relative connection and utilization charges faced by customers, then in the absence of any other obvious supply-side distortions in the market, it cannot be discounted that these interventions are affecting the rate of diffusion of the broadband frontier access technology. Of particular interest
Competition, Regulation, and Broadband Diffusion
in respect to Internet access is the use of two-part DQGÀDWUDWHWDULIIV )ODWUDWHSULFLQJWDULIIVLQÀXHQFHXVHUV¶FRQsumption levels by equalizing the prices paid by both light and heavy users. Whereas in order to minimize total costs, a heavy user will typically IDYRU D KLJK ¿[HG ORZ XVDJH WDULII DQG D ORZ XVHUDORZ¿[HGKLJKXVDJHWDULIIXQGHUÀDWUDWH tariffs both pay the same fee, irrespective of the level of usage (Carlton & Perloff, 2000). Once a FRQQHFWLRQLVSXUFKDVHGWKHXVHUIDFHVQR¿QDQcial consequences of varying usage levels. Light users consuming less than the average quantity subsidize heavy users consuming more than the average. Flat-rate tariffs enable heavy consumers, in particular, to use more of the product than they would had they faced a charge equal to the marginal cost of their usage. Welfare is reduced relative to marginal cost pricing, as the additional costs incurred must be recovered either from all XVHUVLQWKHIRUPRIDKLJKHU¿[HGIHH15 or from another source (e.g., cross-subsidy from other parts of the business, such as voice telephony line rentals, content distribution, or advertising). Flat-rate pricing regimes for Internet access are prevalent across the OECD. Combined with an unmetered telephony component, they have been widely advocated as means of encouraging increases in both Internet access and use (OECD, 2000). On the one hand, it is claimed they offer EXGJHWµLQVXUDQFH¶E\HOLPLQDWLQJWKHULVNRIXVHUV receiving unexpectedly large bills when unaware of the quantity of data transfer their use is generating, especially when using new Internet technologies where megabytes transferred per application may be opaque (Anania & Solomon, 1997; Brownlie, 1997).16 2QWKHRWKHUKDQGDVWKH\IDFLOLWDWHµFRVWOHVV¶ WR WKH FRQVXPHU LQFUHDVHV LQ XVDJH ÀDWUDWH tariffs are attributed with engendering the higher levels of both dial-up Internet uptake and usage REVHUYHG¿UVWLQFRXQWULHVZKHUHXQPHWHUHGORcal telephony calling was the norm (Australia, Canada, New Zealand, and the United States) (OECD, 2000; Howell, 2003). Miravete (2003) suggests that, in respect of voice telephony plans at least, while consumers facing two-part tariffs
PD\ODFNVXI¿FLHQWLQIRUPDWLRQWRVHOHFWWKHRSWLPDO plan initially, learning is rapid and they substitute quickly to a more cost-effective plan if the initial choice is non-optimal. (TXDWLRQV FDQEHPRGL¿HGWRDFFRXQW IRUWZRW\SHVRIÀDWUDWHWDULIIV:KHQORFDOWHOHphony calls are unmetered (i.e., VD = 0), equation (1) becomes: 3 Di
6 nj 1 (
ij
i TDj ) FD .
(4)
When broadband usage is unmetered (i.e., VB = 0), equation (2) becomes: 3 Bi
6 nj 1 (
ij
iTBj ) FB
(5)
With no additional infrastructure usage charge incurred with increasing transaction volumes XQGHUÀDWUDWHWDULIIVXSWDNHRIQHZDSSOLFDWLRQV and additional use of existing applications incur only the cost of user time. If new applications and WUDQVDFWLRQVDFFUXHDQHWEHQH¿WRYHUWKHDGGLWLRQDO user time costs incurred, they will be adopted and used increasingly, irrespective of the infrastructure access method chosen. The substitution choice EHFRPHV D VLPSOH WUDGHRII EHWZHHQ ¿[HG FRVWV the value of user time, the number of applications used, and the relative speed of the connections: ( FD FB ) i 6 nj 1 (TDj TBj ) ! 0
(6)
8QGHUÀDWUDWHSULFLQJRIERWKDFFHVVPHWKRGV EURDGEDQGZLOOEHSUHIHUUHGLIWKH¿[HGFRVWVRI dial-up are large relative to those of broadband, if EURDGEDQGLVVLJQL¿FDQWO\IDVWHUWKDQGLDOXSDQG the value of time for the user is high, and if the number of information exchanges is large. Once WKH ¿[HG SULFHV DQG LQIUDVWUXFWXUH VSHHGV KDYH been set by infrastructure providers, assuming users initially purchase the optimal technology for their volumes of usage, growth in the number of broadband connections is dependent solely upon LQFUHDVHV LQ XVHUV¶ YDOXDWLRQ RI WLPH DQGRU WKH volume of information transfers. +RZHYHU WUXO\ ÀDWUDWH GLDOXS WHOHSKRQ\ charges as per equation (6) are uncommon.17 Even where unmetered calling has prevailed, connec-
147
Competition, Regulation, and Broadband Diffusion
WLRQVWRWKH,QWHUQHWPD\LQFXUD¿[HGFKDUJHSHU call (e.g., in Australia). New Zealand has been one of the few countries where the regulatorymandated local residential telephony calling tariff has resulted in users incurring no charge for either dial-up Internet connection or utilization, although such tariffs are commonly offered voluntarily by providers in the United States (Miravete, 2003). From equations (1), (4), and (5), it now becomes clear why the substitution of broadband for dial-up connections has occurred at a substantially slower rate in New Zealand and the United States than in countries with metered dial-up telephony. When dial-up access is metered, but broadband is unmetered, broadband is preferred when: i
them), substitution from the vintage to the frontier will occur more rapidly when there is a metered usage price for the vintage technology. Conversely, where there is no usage charge for the vintage, current and future expectations of application usage contribute to the technology purchase decision RQO\LQDVPXFKDVWKH\DUHLQÀXHQFHGE\WKHXVHU¶V YDOXDWLRQRIWLPH%HQH¿WVGHULYHGIURPLQFUHDVHV in usage must be greater in this case to justify the decision to substitute, simply because there is no infrastructure usage cost incurred as a result of increasing transaction activity. Thus, both initial purchases of and substitution to the frontier will be less likely to occur than if a vintage infrastructure usage charge is present.
6 nj 1 (TDj TBj ) VD 6 nj 1TDj ! ( FB FD ) . (7)
IMPLICATIONS 7KHVPDOOHUWKHGLIIHUHQFHEHWZHHQWKH¿[HG FRPSRQHQWVRIHDFKWHFKQRORJ\¶VDFFHVVSULFHDQG the higher the usage price per metered dial-up minute, the greater the likelihood that broadband will be purchased for a given application base and transaction volume. Unlike in equation (6), an increase in the volume of usage contributes directly to the technology trade-off decision via an infrastructure usage charge for the vintage technology (dial-up) which is not incurred for the frontier (broadband). Thus, all other factors being held constant, where transaction volumes are increasing (that is, where new applications and uses are continuing to be developed and users are continuing to adopt
The demand-side model illustrates the weakness of relying upon uptake statistics alone in isolation from utilization statistics when assessing the effects of regulatory intervention on markets. 5HJXODWLRQV LQÀXHQFLQJ XSWDNH DQG XVH RI WKH vintage technology (in this case, dial-up) must be considered when interpreting uptake statistics for the frontier technology. While slow uptake of the frontier might indicate market power abuse, it PLJKWHTXDOO\EHUHÀHFWLQJDUHJXODWRU\GLVWRUWLRQ favoring continued use of the vintage technology. Conversely, rapid uptake of the frontier might just DVHDVLO\UHÀHFWGHPDQGVLGHUHVSRQVHVWRSULFLQJ
Table 2. Comparison: Broadband and dial-up costs, 2003 (Howell, 2003, p. 42)
148
Competition, Regulation, and Broadband Diffusion
structures of the vintage technology as the consequence of regulatory interventions in the frontier WHFKQRORJ\¶VVXSSO\VLGHSULFLQJ
Application to New Zealand $VVXSSO\VLGHFRQVWUDLQWVDSSHDULQVXI¿FLHQWWR H[SODLQ1HZ=HDODQG¶VORZEURDGEDQGXSWDNHWKH demand-side model offers a credible alternative hypothesis. While multivariate regressions on countries suggest that low GDP per capita, low population density, and uneven urbanization patterns may contribute (Wallsten, 2006), the New Zealand case study implicates regulation of vintage technologies. Together, the demand-side model and the New =HDODQGGDWDVXJJHVWWKDWWKHEHQH¿WVRIEURDGEDQG DFFHVVIRUWKHYDVWPDMRULW\RI1HZ=HDODQG¶VH[perienced Internet users do not, at present, justify the additional costs of substituting away from very low-priced dial-up access to very low-priced broadband access (see Table 2, which shows the relative price of New Zealand broadband access to dial-up access, compared to a sample of other countries). This may be because New Zealanders place very low valuations on their time, or do not, on the whole, place a high value on the range of applications currently available to them that necessitate broadband connections (that is, audio and video streaming and gaming) relative to other calls on the residential budget (e.g., substitute applications).18 This is consistent with data from other countries such as Australia and the United States, suggesting that where bandwidth used per consumer has been measured, the average bandwidth consumed per customer is low,19 with a small proportion of consumers utilizing a disproportionate share of available bandwidth. $V1HZ=HDODQG¶V*'3SHUFDSLWDLVUHODWLYHO\ ORZ WZHQW\¿UVW LQ WKH 2(&' WKH LQGLYLGXDO valuation of time may play a role. However, Howell ¿QGV WKDW EXVLQHVV XSWDNH RI EURDGEDQG connections relative to other OECD countries is high, and consistent with all of the other Internet access and usage statistic rankings in the upper quartile. This suggests that valuations of business time, at least, are not delaying broadband uptake
noticeably compared to other countries. As the number of households exceeds the number of businesses by a factor of around 8.6:1, national uptake statistics are dominated by residential user patterns. The evidence suggests that residential uptake is primarily responsible for the differences IURPRWKHUFRXQWULHV$VÀDWUDWHWHOHSKRQ\WDULIIV have applied to residential telephony users, but not EXVLQHVVXVHUVWKHGHPDQGVLGHPRGHORIÀDWUDWH pricing of both infrastructures (equation (6)) appears to offer the most plausible explanation for the low levels of broadband uptake despite high levels of Internet uptake. The differences between equation (6) and equation (7) appear to account for the lower levels of broadband uptake in New Zealand, as well as in Australia and the United States,20 relative to other countries where metered telephony usage has prevailed.
Strategic Interaction It is apposite to consider at this point the extent WRZKLFKWKHÀDWUDWHUHVLGHQWLDOWHOHSKRQ\WDULII might have contributed towards the development of the supply-side characteristics exhibited in New =HDODQG¶VEURDGEDQGPDUNHW When the incumbent telecommunications provider was corporatized in the late 1980s and privatized in 1990, all legislative barriers to entry in the New Zealand telecommunications market were removed. While the incumbent faced no requirements to allow access by other operators to its infrastructures, it had no protections from competitive entry. Under competition law, it faced the risk of legal action by competitors from exerting its market power. As a former state-owned enterprise in a light-handed regulatory environment, it also faced a very real political risk that LQGXVWU\VSHFL¿FUHJXODWLRQFRXOGEHUHLQWURGXFHG at any time. Thus, competitors and customers had access to two powerful mechanisms in addition to normal competitive and regulatory forces via which to constrain Telecom from acting in the manner typically associated with unregulated monopolies (Evans & Quigley, 2000). The threat of such actions has undoubtedly shaped the strategic choices made by the company.
149
Competition, Regulation, and Broadband Diffusion
In the absence of LLU, new technology providHUVKDGQRRSWLRQRIXVLQJ7HOHFRP¶VLQIUDVWUXFWXUH to offer broadband services. Stand-alone infraVWUXFWXUHVZHUHUHTXLUHG,QWKHDEVHQFHRIVSHFL¿F entry barriers, these could be installed whenever FRPPHUFLDOO\LQGLFDWHG1HZ=HDODQG¶V¿UVWEURDGband provider, CityLink (ethernet LAN), entered in 1995, followed by iHug (satellite) and Saturn &RPPXQLFDWLRQV ¿EHURSWLFFDEOH LQ :KHQ Telecom offered ADSL in January 1999, it was the fourth entrant into the broadband market. While each of the competitors (apart from iHug) ZDVFRQ¿QHGWRUHJLRQDORSHUDWLRQFRPSHWLWLRQ IURP DOWHUQDWLYH WHFKQRORJLHV SURYLGHG VLJQL¿cant competitive pressure on Telecom. iHug was initially the price leader (Howell & Obren, 2003). As satellite broadband could potentially be offered nationwide, Telecom was required, right from the start, to offer its ADSL services throughout the country at one (universal) price. To have adopted any other pricing strategy would have risked political action in response to accusations of strategic pricing.21 The risk of political and/or regulatory LQWHUYHQWLRQZDVUHDODQGVLJQL¿FDQWJLYHQ1HZ =HDODQG¶VORQJKLVWRU\RIJHRJUDSKLFDOO\DYHUDJHG telephony charging, consumer expectations, and WKHµXQLYHUVDOVHUYLFH¶REOLJDWLRQVRQ7HOHFRPDV DFRQVHTXHQFHRIWKHµ.LZL6KDUH¶ VHHIRRWQRWH 7). Thus, early investment in, and low universal pricing of, broadband by Telecom are in part explained by competitive entry and threat of legal and regulatory action. However, competitive entry and legal and regulatory threats alone were not the only forces acting on Telecom. These forces have interacted with another key factormandatory unmetered local residential telephony tariffs. Flat-rate residential tariffs have prevailed from the earliest days of state-owned telephony provision in New Zealand, and were included in the Kiwi Share as a point of political pragmatism in order to minimize consumer (and voter) dissent when Telecom was privatized in 1990. While designed to address voice telephony pricing, unmetered local calling SURYLGHG UHDO EHQH¿WV WR UHVLGHQWLDO FRQVXPHUV with the advent of Internet access.
150
As illustrated in equation (4), unmetered tariffs place few restrictions on increases in consumer usage. Howell and Obren (2003) show that between DQGWUDI¿FPLQXWHVRQWKH¿[HGQHWZRUN trebled as a consequence of new dial-up data traf¿F E\ FRQWUDVW GLDOXS ,QWHUQHW WUDI¿F UHVXOWHG LQOHVVWKDQGRXEOHWKHWUDI¿FSDVVLQJRYHU%ULWLVK 7HOHFRP¶VYRLFHQHWZRUNVRYHUWKHVDPHSHULRG :KLOHWKHµ.LZL6KDUH¶UHJXODWRU\DUUDQJHPHQWV allowed for Telecom to increase residential line rentals annually in line with the CPI, partly because of the real risks of political intervention arising from consumer dissatisfaction, line rentals were adjusted only once between 1990 and 2003. Consequently, Telecom was forced to bear practically all of the costs of additional use of the network for residential dial-up Internet access. The costs to Telecom of this were substantial, given that New Zealand dial-up Internet users were on average among the heaviest in the OECD (35 hours per month on average for 850,000 residential users) (Howell & Obren, 2003). On the one hand, the tariff structure was directly instrumental in accelerating the initial adoption of Internet access via the vintage dial-up techQRORJ\7KHVXEVWDQWLDOFRVWWR7HOHFRPRIµIUHH¶ (to the user) residential dial-up usage most likely SURYLGHGWKHVWURQJHVWVWUDWHJLFMXVWL¿FDWLRQIRUWKH ¿UPWRLQYHVWHDUO\UDSLGO\DQGZLGHO\LQEURDGband infrastructure. The sooner high-consuming residential Internet users could be migrated from dial-up use, which could not be charged on a usage basis, to broadband, the use of which would generate additional income (in either or both of access and usage fees), the sooner the substantial losses arising from the huge increases in the volume of dial-up Internet usage could be stemmed. A high-quality ADSL product was introduced (2 Mbps, when the entry-level residential offering in most other OECD countries was 256 kbps) early (January 1999, third in the OECD) in order to create clear quality differences between the two technologies, rollout around the country was rapid (80% of telephony customers were connected to ADSL-capable exchanges by 2002, 94% by 2005), and prices were set low to encourage early substitution. A range of two-part tariffs were offered,
Competition, Regulation, and Broadband Diffusion
providing incentives for light users to substitute broadband without having to subsidize heavy users. Heavy users, however, paid a lower per-megabyte usage charge. All users were required to pay usage FKDUJHV DOEHLWYLDµEXQGOHV¶RIGRZQORDGLQJFDSDFLW\ LQSDUWUHÀHFWLQJWKHIDFWWKDWRYHU RIGDWD consumed came from offshore, via the monopoly Southern Cross cable. All providers, including those with their own infrastructures, were required to charge in this manner, as all are required to utilize the monopoly Southern Cross cable.22 On the other hand, however, the tariff structure that induced early adoption of the vintage is now acting as a brake on the diffusion of the frontier, at least inasmuch as informed customers knowing the value derived from current usage patterns face no incentive to substitute to the frontier unless the levels of usage or their valuation of time indicate that this is the method via which they will increase WKHLUEHQH¿W([WHQGLQJ0LUDYHWH¶V ¿QGLQJV in telephony markets to Internet access markets, when faced with a menu of tariff options, experienced consumers are remarkably adept at choosing the right option for their usage pattern, and very quick to change once they learn that an alternative WDULIIZLOORIIHUJUHDWHUEHQH¿WV That the low New Zealand broadband diffusion rate has been exhibited in a market with no apparent supply-side impediments and one of the ZRUOG¶VPRVWNQRZOHGJHDEOHDQGH[SHULHQFHG,QWHUnet-using populations lends further credence to the contention that it is a combination of demand-side factors (either or both of a shortage of value-adding applications and a low user valuation of time) and legacy regulation of the voice telephony market that has spilled over into the diffusion of patterns of dialup Internet access that more plausibly accounts for WKHGLIIHUHQFHVEHWZHHQ1HZ=HDODQG¶VEURDGEDQG diffusion rate and that of other countries.
Answers and Implications ,QDQVZHUWRWKH¿UVWTXHVWLRQSRVHGLQWKHLQWURduction, regulatory factors are implicated in New =HDODQG¶VORZEURDGEDQGGLIIXVLRQUDWH,URQLFDOO\ however, it is the combination and interaction of a set of factors, including the regulatory legacy
of retail tariff obligations and universal service obligations in the dial-up telephony market, that have led to New Zealand achieving supply-side conditions in the provision of broadband access that have yet to be achieved by the majority of countries HPSOR\LQJH[WHQVLYH¿[HGOLQHUHJXODWRU\LQWHUYHQtions to induce wide availability, low prices, and high uptake of broadband. Importantly, the case study illustrates that competition is a dynamic, interactive process involving suppliers, consumers, regulators, and legislators, and that there are many ways in which desired supply-side outcomes can be DFKLHYHGZLWKRXWQHFHVVDULO\KDYLQJPDQ\¿UPV in a market, or without recourse to LLU. In answer to the second question, while some of the circumstances in the case study may be unique WR1HZ=HDODQGDNH\¿QGLQJLVWKHVLJQL¿FDQW effect that regulations relating to a legacy vintage technology can have on the supply conditions of a frontier technology. It is impossible to assess the HI¿FDF\ RI DQ\ UHJXODWRU\ LQWHUYHQWLRQ ZLWKRXW taking into account both supply and demand factors as they relate to both technologies. Moreover, it is negligent to ignore the effect of regulations on the demand side of the market (as price caps, universal service obligations, and retail tariff requirements are) upon the measures of regulatory performance, simply because they were not FRQVLGHUHG VLJQL¿FDQW LQ WKH SDVW 7KH ULVN WKLV invokes is that new regulations may be imposed on the suppliers of the frontier technology that DUHERWKXQMXVWL¿HGDQGPRVWOLNHO\LPSRWHQWLQ addressing the factors associated with the vintage technology that are the real causes underlying the observed outcomes. In answer to the third question, the demand-side model and the New Zealand case study illustrate that, in the absence of supply-side constraints, it is not at all clear that experienced Internet users will adopt broadband simply because it exists. There is no reason to presume that the decision to purchase broadband is any different from that governing any other product or service with derived demand (e.g., physical transport). Consumer purchase is determined by increases in utility, relative to other RSWLRQV,IWKHEHQH¿WVRIWKHIURQWLHUDUHQRWODUJH then relative prices are pivotal.
151
Competition, Regulation, and Broadband Diffusion
The New Zealand data suggest that, given the current range of applications and usage patterns, the additional utility of broadband is possibly much lower relative to dial-up than may have been presumed by regulators. It may be that the historical emphasis on technological capabilities and the focus on supplier factors have created an over-reliance on assessments of user requirements by vested supplier interests, at the expense of genuinely consumer-determined assessments. With regulatory effort focused on infrastructure providers in markets where wholesale and access FXVWRPHUVKDYHEHFRPHWKHSULPDU\µXVHUV¶RID UHJXODWHGWHOHFRPPXQLFDWLRQVFRPSDQ\¶VVHUYLFHV the usage of residential consumers, in particular, may be more easily overlooked by regulators. The ULVNLVWKDWLQWKHVSLULWRIµEXLOGLWDQGXVHUVZLOO FRPH¶ VXSSO\ PD\ RXWSDFH GHPDQG HVSHFLDOO\ in the case of new technologies where it may take time for consumers to develop applications using its capacity, with uninformed regulators incorrectly interpreting the lack of consumer uptake DVHYLGHQFHRIVXSSO\VLGHµSUREOHPV¶MXVWLI\LQJ further regulatory intervention. A further danger emanating from the lack of demand-side understanding is that uninformed DQG XQMXVWL¿HG UHJXODWRU\ LQWHUYHQWLRQ PD\ GHWHUPLQHWKHWHFKQRORJ\WKDWµZLQV¶WKHPDMRULW\RI the market, thereby leading to increased pressure on legislators and regulators from incumbent and new entrant suppliers. As the New Zealand case VWXG\ LOOXVWUDWHV ÀDWUDWH WHOHSKRQ\ SULFLQJ KDV UHVXOWHGLQGLDOXS,QWHUQHWDFFHVVµZLQQLQJ¶WKH market share battle, at least given current range and usage of applications. Without doubt, the rate of substitution from dialup to broadband will be accelerated if this tariff option is replaced with metered dial-up tariffs. But is such a regulatory change warranted? As with all other regulatory interventions, the ¿QDODUELWHUPXVWEHWRWDOZHOIDUH,ILWLVFKHDSHU IRUDQHWZRUNSURYLGHUWRFDUU\WKHVDPHWUDI¿FRQ broadband than dial-up networks, then welfare is higher if substitution occurs. Without tariff restrictions, in these circumstances network providers would voluntarily change the tariff structure in order to accelerate substitution. Mandatory tariff
152
regulations preventing suppliers from making WKHVHDGMXVWPHQWVPD\DOORZLQHI¿FLHQWRYHUFRQsumption on the vintage technology to continue for longer than would have been the case absent the regulations. It is possible that this is the welfare outcome in New Zealand, if broadband data transfer really is cheaper per megabyte transferred than dial-up. The further lesson from this chapter is that PDQGDWRU\ÀDWUDWHWDULIIVIRUFXUUHQWEURDGEDQG technologies may in the future impede the rate of substitution to the next generation of Internet technologies when broadband is supplanted by another frontier and it too becomes a vintage technology. 6LPLODU¿QGLQJVZLOODSSO\WRRWKHUPDUNHWVWRR)RU H[DPSOHÀDWUDWHFDOOLQJWDULIIV RFFXUULQJZLWKLQ XVDJHµEXQGOHV¶ PD\LPSHGHWKHVXEVWLWXWLRQ* mobile handsets for 2G handsets among the user groups who place a low value on the additional functionality of 3G applications.
CONCLUSION The New Zealand case study illustrates that regulatory interventions have interacted with other IDFWRUVWRKDYHVLJQL¿FDQWHIIHFWVRQWKHVXSSO\ side market conditions and broadband diffusion rate. The New Zealand data call into question the extent to which supply-side factors and regulatory intervention in other countries can explain differences in broadband uptake rates. Demand-side factors, and in particular, regulations affecting GHPDQGVLGH EHKDYLRU VXFK DV ÀDWUDWH WDULIIV price caps, and universal service obligations, may also be instrumental. Further analysis such as that undertaken in this chapter will likely reveal other demand-side factors affecting broadband diffusion. More research enabling regulators to understand the demand side of telecommunications markets, which has long been opaque, is indicated.
REFERENCES ABS. (n.d.). Retrieved from http://www.abs.gov. au/ausstats/
[email protected]/e8ae5488b598839cca25682
Competition, Regulation, and Broadband Diffusion
0001316126445f12663006b83ca256a150079564d !OpenDocument
Working Paper 31, Brookings Institution, Washington, DC.
Anania, L., & Solomon, R. (1997). Flatthe minimalist price. In L. McKnight & J. Bailey (Eds.), Internet economics (pp. 91-120). Cambridge, MA: MIT Press.
Hausman, J. (2002). Internet-related services: The results of asymmetric regulation. In R. Crandall & J. Alleman (Eds.), Broadband; should we regulate high-speed Internet access? (pp. 129-156). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies.
Arnold, T., Boles de Boer, D., & Evans, L. (2003). The structure of New Zealand industry: Its implications for competition law. In M. Berry & L. Evans (Eds.), Competition law at the turn of the century: A New Zealand perspective. Wellington, NZ: Victoria University Press. Boles de Boer, D., & Evans, L. (1996). The economic HI¿FLHQF\RIWHOHFRPPXQLFDWLRQVLQDGHUHJXODWHG marketthe case of New Zealand. Economic Record, 72(216), 24-39. Boles de Boer, D., Evans, L., & Howell, B. (2000). The state of e-New Zealand. Wellington, NZ: New Zealand Institute for the Study of Competition and Regulation (http://www.iscr.org.nz). Brownlee, N. (1997). Internet pricing in practice. In L. McKnight & J. Bailey (Eds.), Internet economics (pp. 77-90). Cambridge, MA: MIT Press. Carlton, D., & Perloff, J. (1999). Modern industrial organization. New York: Addison-Wesley. Cave, M. (2006). Encouraging infrastructure competition via the ladder of investment. Telecommunications Policy, 30, 223-237. Crawford, D. (1997). Internet services: A market for bandwidth or communication? In L. McKnight & J. Bailey (Eds.), Internet economics (pp. 379400). Cambridge, MA: MIT Press. EU (European Union). (2006). E-communications household survey. Retrieved November 6, 2006, from http://ec.europa.eu/public_opinion/archives/ ebs/ebs_249_en.pdf
Hausman, J., & Sidak, H. (2004, December). Did mandatory unbundling achieve its purpose? EmSLULFDOHYLGHQFHIURP¿YHFRXQWULHV Working Paper 04-40, Department of Economics, Massachusetts Institute of Technology, USA. Hausman, J., Sidak, G., & Singer, H. (2001). Cable modems and DSL: Broadband Internet access for residential consumers. American Economic Review, 91, 302-307. Helpman, E., & Trajtenberg, M. (1996). Diffusion of general purpose technologies. Working Paper 5773, National Bureau of Economic Research. Horrigan, J. (2006). Home broadband adoption 2006. Pew Internet and American Life Project. Retrieved November 6, 2006, from http://www.pewinternet.org/pdfs/PIP_Broadband_trends2006.pdf Howell, B. (2003). Building best practice broadband in New Zealand: Bringing infrastructure supply and demand together. Wellington, NZ: New Zealand Institute for the Study of Competition and Regulation (http://www.iscr.org.nz). Howell, B. (2006a, August 11). Submission: Telecommunications Amendment Bill. Presented to the Finance and Expenditure Select Committee, New Zealand House of Representatives. Available from http://www.iscr.org.nz
Evans, L., & Quigley, N. (2000). Contracting, incentives for breach, and competition law. World Competition, 23(2).
Howell, B. (2006b, October 16). Broadband penetration: Does it matter and does it require local loop unbundling. Public Lecture, New Zealand Institute for the Study of Competition and Regulation, Victoria University of Wellington. Available from http://www.iscr.org.nz
Ferguson, C. (2002). The United States broadband problem: Analysis and policy recommendations.
Howell, B., & Marriott, L. (2001). The state of eNew Zealand: 12 months on. Wellington, NZ: New
153
Competition, Regulation, and Broadband Diffusion
Zealand Institute for the Study of Competition and Regulation (http://www.iscr.org.nz). Howell, B., Mishra, V., & Ryan, L. (2004). The state of e-New Zealand: 2004. Wellington, NZ: New Zealand Institute for the Study of Competition and Regulation (http://www.iscr.org.nz). Howell, B., & Obren, M. (2002). Broadband diffusion: Lags from vintage capital, learning by doing, information barriers and network effects. Wellington, NZ: New Zealand Institute for the Study of Competition and Regulation (http://www. iscr.org.nz). Howell, B., & Obren, M. (2003). Telecommunications usage in New Zealand: 1993-2003. Wellington, NZ: New Zealand Institute for the Study of Competition and Regulation (http://www.iscr. org.nz). ITU (International Telecommunications Union). (2006). World communication/ICT development report: Measuring ICT for social and economic development. Geneva. Jovanovic, B., & Stolyarov, D. (2000). Optimal adoption of complementary technologies. American Economic Review, 90(1), 15-29. Laffont, J.-J., & Tirole, J. (2002). Competition in telecommunications. Cambridge, MA: MIT Press. Melody, W. (2005). Regulation and network investment: a framework for analysis. In A. Mahan & W. Melody (Eds.), Stimulating investment in network development: Roles for regulators (pp. 19-37). Lyngby, Denmark: World Dialogue on Regulation for Network Economies. Ministry of Economic Development. (2006). Telecommunications stocktake cabinet paper. Wellington, NZ: New Zealand Government. Retrieved November 6, 2006, from http://www.beehive.govt. nz/Documents/Files/Cabinet%20paper%20and% 20minute.pdf Miravete, E. (2003). Choosing the wrong calling plan? Ignorance and learning. American Economic Review, 93(1), 297-310.
154
Network Strategies. (2006). The broadband divide: Achieving a competitive international ranking. Report prepared for the New Zealand Ministry of Economic Development. Retrieved November 6 2006, from http://www.med.govt.nz/templates/ ContentTopicSummary____20541.aspx OECD. (2000). Local access pricing and e-commerce. Report DSTI/ICCP/TISP(2000)1/FINAL, OECD, France. OECD. (2001). The development of broadband access in OECD countries. Report DTSI/ICCP/ TISP(2001)2/Final, OECD, France. OECD. (2003). Developments in local loop unbundling. Report DSTI/ICCP/TISP(2002)5/FINAL, OECD, France. OECD. (2005). Communications outlook 2005. Paris. Point Topic. (2006). KDSS\ZLWKWKHTXDOLW\ of home Internet services. Retrieved October 30, 2006, from http://www.point-topic.com/content/ dslanalysis/061017consats.htm Rappoport, P., Kridel, D., & Taylor, L. (2002). The demand for broadband: Access, content and the value of time. In R. Crandall & J. Alleman (Eds.), Broadband; should we regulate high-speed internet access? (pp. 57-82). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Sacher, S., & Wallsten, S. (2006). What U.S. broadband problem? (Policy Matters 06-18). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Shelanski, H. (2002). Competition and regulation in broadband communications. In R. Crandall & J. Alleman (Eds.), Broadband; should we regulate high-speed Internet access? (pp. 157-196). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Sutherland, E. (2006). An exit from telecommunications regulation. Available from http://3wan.net/ Varian, H. (2002). The demand for bandwidth: (YLGHQFH IURP WKH ,1'(; SURMHFW ,Q 5 &UDQdall & J. Alleman (Eds.), Broadband; should we
Competition, Regulation, and Broadband Diffusion
regulate high-speed Internet access? (pp. 39-56). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Wallsten, S. (2006). Broadband and unbundling regulations in OECD countries. Working Paper 06-16, AEI–Brookings Joint Center for Regulatory Studies, Washington, DC. Wear, S., & Duncan, I. (2005). Telecommunications pricing in New Zealand: A comparison with OECD countries. Wellington, NZ: New Zealand Institute for Economic Research. Retrieved November 6, 2006, from http://www.nzier.org.nz/SITE_Default/ 6,7(B3XEOLFDWLRQV[¿OHV SGf Wenders, J. (1990). Two views of applied welfare analysis: The case of local telephone pricing. Southern Economic Journal, 57(2) 340-348.
Market Power: 7KH DELOLW\ RI D ¿UP WR VHW SULFHV SUR¿WDEO\ DERYH FRPSHWLWLYH OHYHOV PDUginal cost). Monopoly: A single seller in a market. Two-Part Tariff: $¿UPFKDUJHVDFRQVXPHU DIHH WKH¿UVWWDULII IRUWKHULJKWWREX\DVPDQ\ units of the product as the consumer wants at a VSHFL¿HGSULFH WKHVHFRQGWDULII 6HH&DUOWRQDQG Perloff (2000, pp. 297-331). Utility Function: A numerical representation RIDQLQGLYLGXDO¶VSUHIHUHQFHVRYHUGLIIHUHQWSRVsible choices or situations. Variable Cost: A cost that increases in direct proportion to the number of units produced (suppliers) or consumed (consumers).
ENDNOTES KEY TERMS 1
Fixed Cost: A cost that does not vary with the number of units produced (for suppliers) or consumed (for consumers).
2
Flat-Rate Tariff: A variation of a two-part tariff where the second tariff is set at zero. Local Loop Unbundling (LLU): A form of regulatory access requirement whereby an incumbent network provider is required to lease elements of the proprietary network (principally the twisted FRSSHUSDLUFRQQHFWLQJHQGXVHUV¶SUHPLVHVWRWKH local switching infrastructure) to competitors, to enable the competitors to provide a service in competition to the incumbent. Competitors may WKHQDGGWKHLURZQHTXLSPHQWWRWKHLQFXPEHQW¶V equipment, and/or differentiate other aspects of the service provided, such as sales, product, and service bundles. For a full description, see Cave (2006). Information Asymmetries: Differences among individuals in their information, especially when this information is relevant to determinLQJ DQ HI¿FLHQW SODQ RU WR HYDOXDWLQJ LQGLYLGXDO performance.
3
At September 2005, 76.3% of the New Zealand population had access to the Internet (ITU, 2006). 'HWDLOV RI 1HZ =HDODQG¶V UHODWLYH SHUIRUmance in Internet and e-commerce metrics across time are contained in Boles de Boer et al. (2000), Howell and Marriott (2001), and Howell, Mishra, and Ryan (2004). Broadband investment history, diffusion, and utilization is documented in Howell and Obren (2002, 2003) and Howell (2003, 2006a). Comparative international pricing analyses are in Howell (2003), Wear and Duncan (2005), and Network Strategies (2006). Table 1 provides a summary. The OECD has advocated since 2001 that, in the absence of competition between different technological platforms, local loop XQEXQGOLQJ PD\ SOD\ D VLJQL¿FDQW UROH LQ facilitating the development of competition for former monopoly incumbent telecommunications providers (OECD, 2001). As of October 2006, Mexico remained the only OECD nation not to have mandated some form of local loop unbundling.
155
Competition, Regulation, and Broadband Diffusion
4
5
156
Local loop unbundling was mandated by the European Union in Regulation 2887/2000/EC in order to boost competition in the provision RIEURDGEDQG 2I¿FLDO-RXUQDO/SS 30.12.2000 as cited by Sutherland, 2006). New Zealand mandated bit stream unbundling in 2004, and legislation for full LLU is pending. Full LLU is proposed to address low levels of broadband uptake (Ministry of Economic Development, 2006). Between 1990 and 2001, the solitary indusWU\VSHFL¿F µUHJXODWLRQV¶ FRYHULQJ WKH 1= telecommunications sector related to the µ.LZL6KDUH¶UHWDLQHGE\WKH1HZ=HDODQG government following the privatization of the formerly state-owned network operator, 7HOHFRP1HZ=HDODQG/WG7KHµ.LZL6KDUH¶ capped residential line rentals at the price charged at the date of privatization (taking into account consumer price index (CPI) changes), a universal service obligation equalizing line rental charges for rural and urban users, and an REOLJDWLRQWRPDNHDYDLODEOHDµÀDWUDWH¶WDULII option to all residential consumers. Following UHLQWURGXFWLRQ RI VHFWRUVSHFL¿F UHJXODWLRQ in 2001, the costs of the universal service obligation became shared among all network operators. A Telecommunications Service Obligation (TSO) payment is levied annually (retrospectively) by the Telecommunications Commissioner. Despite the ability to raise residential line rental charges in accordance with the CPI, Telecom has very rarely altered prices, resulting in decreases in the real price of residential access across the 1990s that have been greater than the OECD average (Howell & Obren, 2003). The popularity of the ÀDWUDWHUHVLGHQWLDOWDULIIZKLFKZDVWKHRQO\ tariff option offered under state ownership, is evidenced by the fact that fewer than 10% of residential consumers chose the two-part µ(FRQRP\¶WDULIIRIIHUHGE\7HOHFRPDWDQ\ time when it was available between 1997 and 2004. This tariff option has now been withdrawn. Wear and Duncan (2005) show that, as a consequence of the popularity of the ÀDWUDWHWDULIIFDOOLQJPLQXWHVSHUUHVLGHQWLDO
6
7
8
account in New Zealand are substantially higher than in comparator OECD countries where local calls are metered per minute. $WWULEXWLRQRIµEURDGEDQGXSWDNHSUREOHPV¶ to differences in competitive and regulatory regimes has also occurred in the United States (e.g., Hausman, Sidak, & Singer, 2001; Ferguson, 2002; Hausman, 2002; Shelanski, 2002; Wallsten, 2006; Sacher & Wallsten, 2006). New Zealand has historically exhibited a very high proportion of Internet users per capitagenerally in the top quartile in the OECD (Howell, 2003)despite ranking in the lower quartile in GDP per capita, a measure strongly correlated with higher levels of diffusion of all other telecommunications technologies (ITU, 2006). In 2006, it was estimated that 79 per 100 population were accessing the Internet (Howell, 2006a). New Zealand has also been at the forefront of the OECD diffusion of infrastructures supporting e-commercefor example, it has never been lower than fourth in the number of secure servers per capita since the OECD began measuring this statistic. Similarly high ¿JXUHV DUH HYLGHQFHG LQ RWKHU HOHPHQWV RI e-commerce utilization, such as the number per capita of routed autonomous systems, URXWHG,3YDGGUHVVHVDQGV\QWKHWLFSUH¿[HV (Howell, 2006a). The very early achievement of very high levels of both Internet connection and utilization in New Zealand compared to other OECD countries were remarkable as early as 2000 (OECD, 2000). 1HZ=HDODQGZDVDPRQJWKH¿UVW2(&'QDtions to complete digitization of the telephone network (1994). New Zealand was third in the OECD (after the United States and Canada) to offer a commercial ADSL product (January, %\\HDUHQGRI¿[HGOLQH customers could access ADSL connections. By year-end 2003, it was available to 85% RI¿[HGOLQHFXVWRPHUVFRPSDUHGWRLQ Norway; 75% in Australia, Canada, and the United States; 81% in Finland; 85% in the United Kingdom and The Netherlands; and
Competition, Regulation, and Broadband Diffusion
9
10
90% in Iceland (OECD, 2003). In 2006, ADSL ZDVDYDLODEOHWR RI¿[HGOLQHWHOHSKRQ\ customers. The initial speed offered in 1999 (2 Mbps) was one of the highest base-level product qualities offered in the OECD. In 1=¶V $'6/ SURGXFWV ZHUH WKH WKLUG lowest-priced (megabyte per second downloaded) in the OECD (Howell, 2003). It has been suggested that data caps are depressing uptake as they constrain potential usage (e.g., Network Strategies, 2006). Data caps arise as a consequence of the high volumes of data transmitted from offshore, which must pass over the monopoly Southern Cross trans3DFL¿F RFHDQLF FDEOH +RZHOO D
11
12
13
14
15
16
having a population of more than 500,000, and it is geographically isolated. New Zealand markets typically lack advantages from economies of scale, exhibit substantially higher concentration levels than comparator countries, and have relatively higher levels of capital intensity than comparator countries (Arnold, Boles de Boer, & Evans, 2003). Network effects that accrue from having more individuals connected to the Internet network (irrespective of connection type) are LQFOXGHGLQWKLVEHQH¿W The form of the model derives from Wenders (1990). EU data in January 2006 indicate that only 19% of surveyed broadband users would be prepared to pay 10% more for faster connections and only 10% of users 20% more. The EU (2006) indicates 40% of EU dial-up users, and Horrigan (2006) indicates 50% of U.S. dial-up users are happy with existing dial-up speeds and have no plans to purchase broadband connections in the foreseeable future. Point Topic (2006) indicates 82% of UK Internet households (both broadband and GLDOXS DUHVDWLV¿HGZLWKGRZQORDGWLPHV It is noted that these observations were made before the widespread use of Skype and other non-proprietary VoIP applications, and before the widespread use of IPTV for broadcast-type as well as video-on-demand video streaming. Fewer light users purchase a connection, and all users pay a higher fee, thereby lowering welfare relative to the counterfactual of the heaviest users consuming beyond the point ZKHUHWKHEHQH¿WVRIWKHLUXVDJHIDOOEHORZ marginal cost. /LNHZLVHÀDWUDWHSODQVREIXVFDWHWKHIDFW that light users are consuming very small quantities of data transfer, thereby reducing the likelihood of light consumers transferring to a more cost-effective two-part tariff plan. 'HVSLWHWKHHDUO\SRSXODULW\RIÀDWUDWHEURDGband packages, introduction of competition in broadband supply is resulting in more twoSDUWWDULIIVEHLQJRIIHUHG µSURGXFWYDULHW\¶
157
Competition, Regulation, and Broadband Diffusion
17
18
158
as predicted by Wenders (1990). As suppliers gain more information about differences in customer usage patterns, they differentiate their offerings in order to maximize their share of surplus (while simultaneously inFUHDVLQJWRWDOVXUSOXVE\UHGXFLQJLQHI¿FLHQW FURVVVXEVLGLHVIURPÀDWUDWHWDULIIV With the exception of Australia, New Zealand, the United States, and Canada, per minute charging has prevailed, although it is noted that some operators are now offering unmetered data calls on voice lines, even though voice calls are still metered (OECD, 2005). 6LJQL¿FDQWO\1HZ=HDODQGFXUUHQWO\KDVQR IPTV services. The sole digital pay television product is provided via satellite. Fiber-based cable television is available only in two localities and has a very small market share. µ7ULSOHSOD\¶ FDEOH EXQGOHV RIIHULQJ YLGHR content simply resell wholesaled content from the satellite provider. The incumbent telecommunications provider offers a pseudo µWULSOHSOD\¶EXQGOHZKHUHE\WHOHYLVLRQFRQtent is provided by the satellite company over
19
20
21
22
its infrastructure, but with billing aggregated with voice and data services by the telco (Howell, 2006a). Australian data shows residential broadband usage was 1.3 Gb/month in the September quarter of 2003, compared with 1.5 Gb/month in New Zealand in January 2003 (Howell, 2003) to 2.2 Gb/month in the March quarter of 2005 (ABS, n.d.). Canada has a different set of local telephony characteristics, given the fact that access agreements were associated with geographic µGHDYHUDJLQJ¶RISULFHVPHDQLQJWKHVLPSOH model cannot be applied directly to the Canadian market (OECD, 2003; Howell, 2006a; Cave, 2006). When Saturn offered lower-priced telephony VHUYLFHV RYHU LWV ¿EHU FRQQHFWLRQV DQG Telecom moved to match them in the region covered by Saturn, this is precisely what occurred. Most opted to charge different rates for naWLRQDODQGLQWHUQDWLRQDOWUDI¿FUHÀHFWLQJWKH cost differences.
Division IV
Europe
160
Chapter XI
Digital Divide and Broadband Access:
The Case of an Italian Region Enrico Ferro Istituto Superiore Mario Boella, Italy J. Ramon Gil-Garcia &HQWURGH,QYHVWLJDFLȩQ\'RFHQFLD(FRQȩPLFDV0H[LFR Natalie Helbig State University of New York-Albany, USA
ABSTRACT Reducing the digital divide in order to build an information society for all is one of the top priorities for European policymakers. A better understanding of the determinants of broadband access at the individual level represents a key starting point for any e-inclusion policy. Based on a review of the literature on digital divide and broadband access, we document different approaches to understanding the digital divide and argue that these perspectives can also help to understand broadband access. Combining the digital divide and broadband literature provides a systematic and theory-based approach to the selection and inclusion of variables in different models. This chapter presents the results of a survey conducted LQDQ,WDOLDQUHJLRQ:HSURYLGHVRPHLPSOLFDWLRQVRIRXU¿QGLQJVDQGDUJXHWKDWSROLF\PDNHUVVKRXOG explore the relationship between IT skills acquisition, broadband access, and Internet use in order to develop more effective policies and programs.
INTRODUCTION Over the last decade academic scholars, public of¿FLDOVDQGSULYDWHVHFWRUOHDGHUVKDYHDUJXHGWKH positive social and economic impacts of broadband technologies as a key ingredient for the economic and social development of nations. Evidence of
this considerable expectation is demonstrated by the prominent and consistent role the issues of broadband diffusion have played in the last several years in the various strategic development plans devised at the European level (eEurope 2002, 2005, i2010). In his recent best seller, The World Is Flat, Thomas Friedman, New York Times columnist,
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Digital Divide and Broadband Access
argued that in the year 2000 the world entered a third and new era of globalization. According to )ULHGPDQWKH¿UVWWZRSKDVHVZHUHVSHDUKHDGHG by countries signing international trade agreements and organizations turning into multinational corporations, whereas the latest phase is, and will be, built around individuals “globalizing,” a process that requires fast and convenient access to people and knowledge regardless of their location. )ULHGPDQ¶VZRUOGYLHZVWUHVVHVWKHLPSRUWDQFHRI the key role individuals play as dynamic agents in an information-based economy. His perspective adds an interesting frame to the issues surrounding uneven distribution, access, and usage of broadband technology. The public policy problems of the digital divide, or e-inclusion, when viewed from this perspective are shifted from a focus on pure social inequality to an issue of the strategic use of information and communication technologies in the economic and social development of countries competing in a global market. At present, the different globalization patterns individuals may pursue are still vague and surely require further investigation. Nevertheless, the new direction hypothesized by Friedman and others offers an interesting point of departure for exploring new ways of conceptualizing the policy issues surrounding the diffusion of broadband. While little doubt remains about the role broadband plays in sustainable economic development (OECD, 2004; Annis, 2005), it is still unclear how to obtain a widespread and economically accessible diffusion of such networks (Cantamessa et al., 2005). In this respect, the liberalization trend of telecommunication (TLC) markets that occurred in the nineties throughout most of the Western world represented an important turning point that led to a shift in logic. The logic moved away from strategies based on an acceptance of a natural monopoly toward a more market-driven approach (Hulsink, 1999). The privatization of telecommunication markets was one of the ingredients that led to the Internet boom during which infrastructure expansion was mainly driven by an unconditional faith in the future SUR¿WDELOLW\RI,QWHUQHWUHODWHGVHUYLFHV6LQFHWKH burst of the speculative bubble, a strong emphasis
RQWKHHI¿FLHQF\RILQYHVWPHQWDOORFDWLRQSROLFLHV for building network infrastructures has emerged. However, recent studies continue to demonstrate that the expansion of broadband networks was, and to a large extent still is, driven by technological push rather then demand pull dynamics (Ferro, 2006). In this respect, the considerable marketing efforts currently made by the main European TLC operators to raise the adoption rates among residential users somehow supports this thesis. The lack of a demand pull driving the diffusion of broadband infrastructure poses a number of issues. First, it raises the minimum threshold in terms of total inhabitants a municipality has to meet in order to attract investments in infrastructure (Howell, 2002), thus leaving unreached a higher portion of smaller municipalities. Second, LW H[WHQGV FRPSDQLHV¶ SD\EDFN WLPHV IRU VXQN FRVWLQYHVWPHQWVDQG¿QDOO\LWKLQGHUVWKHELUWK of complementary services (Pentland, Fletcher, & Hasson, 2003). In simple terms, it fosters the formation of uneven access and usage opportunities (i.e., in rural and urban areas) and lowers the PDUNHWSUR¿WDELOLW\WKXVKLQGHULQJWKHYLDELOLW\RI broadband technologies as an economic and social development tool. Therefore, it is evident why it is extremely important for both the private and public sectors to understand how to enlarge the broadband customer base. In this respect, there are two PDLQREMHFWLYHVWREHFRQVLGHUHG7KH¿UVWLVDQ increase in the total number of Internet users, and the second is the maturation of usage patterns of current Internet users (i.e., make them broadband dependent). Efforts aimed at understanding the gap between access and usage patterns present among different social groups and geographical areas ZLOO EH EHQH¿FLDO 7KH REMHFWLYH RI WKLV FKDSWHU is to explore how the digital divide literature may help identify some important areas for investigation and could prove useful in understanding the needs of broadband-enabled populations and how to increase broadband use. The rest of the chapter, starting with the next section, highlights the current situation of broadEDQGGLIIXVLRQLQ(XURSHDQGEULHÀ\UHYLHZVWKH broadband literature. We then review the digital
161
Digital Divide and Broadband Access
divide literature and present the methodology of the study. We provide an analysis of different models for expanding the discussion about broadband and conclude with some remarks on policy issues.
there is some variance between Northern and Southern countries, Figure 1 shows an overall picture in which demand is struggling to take off. This suggests the necessity for further investigation of broadband demand in order to understand how it may be stimulated. For this reason, we use an Italian region named Piedmont as a case to explore some of the possible paths leading to the creation of broadband demand. Although the region in question may not be considered representative of each individual European country, its current situation is very close to that of Europe as a whole. As of 2005, the xDSL coverage and penetration in Piedmont were respectively 81% and 21% (Regional Observatory on ICT, 2005).
BACKGROUND: BROADBAND AND THE DIGITAL DIVIDE The European Broadband Outlook The emphasis on the importance of broadband infrastructure posed by European policies in the last decade culminated in 2005 with an average coverage rate, in terms of population reached by xDSL services, of 85% (European Commission, 2006). Despite the presence of a portion of individuals still excluded from easy broadband access, the current situation may be regarded as a positive result. In fact, the reduction in infrastructurerelated costs due to the fast mutating scenario in ZLUHOHVVWHFKQRORJLHVVXFKDV:L¿+,3(5/$1 DQG:L0$;DOORZVXVWREHRSWLPLVWLFDERXWWKH near future coverage of underserved areas. )LJXUHUHSRUWVVSHFL¿FDOO\RQWKHEURDGEDQG penetration rates present in the different European countries. The average adoption rate for the European Union (EU25) in 2005 was 23%. Although
Understanding Broadband: A Brief Literature Review The diffusion of broadband connections represents a complex phenomenon. The relationship of broadband development to economic activities is also important (Ferro et al., 2007). However, three main streams of research are represented in the broadband literature and document the growing complexity of the phenomenon. 7KH¿UVWVWUHDPRIUHVHDUFKWDUJHWVWKHUROHRI socioeconomic variables (such as income, educa-
Figure 1. Broadband penetration Share of households having a broadband connection 2005 (Eurostat) 70% 63% 60%
54% 51%
50% 41%
41%
40%
40%
36% 33%
32%
30% 30% 23%
25%
23%
23%
21%
20% 19%
20%
14%
13% 10%
16% 12%
11% 7%
7%
5%
4% 1%
0% eu25
162
eu15
be
cz
dk
de
ee
gr
es
ie
it
cy
lv
lt
lu
hu
nl
at
pl
pt
si
sk
se
uk
is
no
Digital Divide and Broadband Access
tion, age, and gender, etc.) on adoption and usage of broadband connections. The end goal is to understand and control for the formation of potentially limiting information asymmetries among various population groups that could contribute to the rise in well-documented inequalities in information access and usage (Flamm, 2004; Flamm & Chaudhuri, 2005; Grubesic, 2004; Varian, 2001). Recent progress in this stream of research extends the conceptualization of individuals as just users to active consumers. In particular a few attempts have been made to develop various conceptual PRGHOVDLPHGDWLQYHVWLJDWLQJFRQVXPHUV¶DGRStion, usage, and impact of technology (Venkatesh & Brown, 2001; Oh, Ahn, & Kim, 2003; Dwivedi & Choudrie, 2004; Choudrie & Dwivedi, 2004a, 2004b, 2004c). The second stream of research focuses on the role government and policy play in the diffusion process. In particular, since the liberalization of the telecommunications industry, research investigated the role that local governments have played as catalysts for attracting infrastructure investments by telecommunication couriers. Factors leading local communities to undertake broadband initiatives and the consequences of these initiatives on the local economy are particularly emphasized in this stream of research (Gillett, Lehr, & Osorio, 2003; Lehr, Osorio, Gillett, & Sirbu, 2005). Finally, the third stream of research applies systemic approaches with the intent to identify macro factors contributing to broadband diffusion at the country or regional level. Some of these studies (Ferro, 2006) analyze broadband demand and supply as complementary diffusion processes. Some others tackle the problem from the perspective of multiple stakeholders (Dwivedi, Choudrie, & Gopal, 2003; Choudrie et al., 2003) or try to identify superior applications that could boost the diffusion of broadband-related services (Middleton, 2002; /HH2¶.HHIH <XQ +RZHYHULWLVVWLOO debated whether broadband diffusion and demand DUHVROHO\GXHWRNLOOHUDSSOLFDWLRQVRUGXHWRXVHUV¶ mature use of the Internet. $W¿UVWUHVHDUFKHIIRUWVWDFNOHGLVVXHVUHJDUGing infrastructure diffusion, and subsequently expanded in scope to consider demand aspects
such as adoption, usage, and purchase. Heretofore, most of the demand-related research analyzed socioeconomic differences between adopters and non-adopters and the impact that access to broadband connections had on Internet usage. In doing so, little attention was paid to the paths leading to the acquisition of necessary IT skills to become a broadband-enabled user. This chapter intends to contribute to the existing body of knowledge on broadband demand by using the literature on the digital divide to shed some light on the path leading to broadband adoption, including IT skills.
A Framework to Understand the Digital Divide The last 15 years have yielded three main approaches to understanding the digital divide: an access divide, multi-dimensional digital divide, and a multi-perspective digital divide (see Helbig, Gil-Garcia, & Ferro, 2005, for an extended review; Gil-Garcia, Helbig, & Ferro, 2006). Broadband is one type of access to the Internet, and the broadband literature evolved investigating diffusion rates, demand, and usage issues, as well as calling into question the role of public policy. The purpose of presenting the broadband and digital divide literature together is an attempt to identify some similarities and assess the usefulness of a digital divide framework (Helbig et al., 2005) in understanding broadband access and Internet use. The following sections outline the viewpoints and assumptions taken by different scholars about the digital divide.
Characterizing the Digital Divide The digital divide is a complex social phenomenon. The term generally evokes a broad set of implications. Robinson, DiMaggio, and Hargittai (2003) ZULWH ³7KHGLJLWDOGLYLGHLPSOLHVWKDWVLJQL¿FDQW minorities of the population are effectively denied access to a technology that, like other public facilities like libraries and super highways, is thought to be open to anyone” (p. 2). Therefore, the impact of the digital divide on certain segments of the
163
Digital Divide and Broadband Access
population implies differing life chances and opportunities for those who are not technologically savvy (DiMaggio & Hargittai, 2001; Nelson, 2002; Servon, 2002, p. 14). Debates continue as to the magnitude of these consequences. Some scholars argue there is no digital divide, while others argue WKDWWKHFRQVWUXFWµGLJLWDOGLYLGH¶LVQRWSURSHUO\ framed.
Access, Multi-Dimensional, and Multi-Perspective Digital Divides 2QHRIWKH¿UVWZD\VLQZKLFKWKHGLJLWDOGLYLGHZDV characterized was as an access dividebetween those who have access to technology and those who do not. This perspective suggests that the digital divide is a one-dimensional issue. The PDLQLVVXHLVDQµDFFHVVWRWHFKQRORJ\SUREOHP¶ and the debate is framed as an inherent delay in the diffusion of technology among different groups of people (Adriani & Becchetti, 2003; Compaine, 2001). Another assumption is that “once online, there is no gap” (Walsh et al., 2003, p. 281) and that everyone can use the Internet for the same purposes. Therefore, this perspective evaluates access to the Internet as the only important determinant of the digital divide and sees no difference between having access and also being able to use the technologies (DiMaggio & Hargittai, 2001). Proponents emphasize market-driven forces will close the gap over time and government intervention is not needed (Compaine, 2001). A second stream of research focuses on multiple dimensions. This multi-dimensional viewpoint argues that the basic building block of Internet use is access (Servon, 2002; Norris, 2001) but recognizes the importance of other dimensions which impact access and use such as race/ethnicity, income, skills, geography, cultural content, education, and training (Bimber, 2000; Norris, 2001; Servon, 2002). DiMaggio and Hargittai (2001) state: “As the technology penetrates into every crevice of VRFLHW\ WKH SUHVVLQJ TXHVWLRQ ZLOO EH QRW µ:KR FDQ¿QGDQHWZRUNFRQQHFWLRQDWKRPHZRUNRU in a library or community center from which to ORJRQ"¶EXWLQVWHDGµ:KDWDUHSHRSOHGRLQJDQG ZKDWDUHWKH\DEOHWRGRZKHQWKH\JRRQOLQH"¶´
164
as important factors in understanding the digital divide. More recently, Ferro et al. (2005) added another dimension to this picture by highlighting the presence and the interrelation of demand- and supply-related divides. Proponents emphasize government intervention in access- and use-related issues and do not believe market forces will close the gap over time (Chin & Fairlie, 2004; Cole et al., 2004; Mossberger, Tolbert, & Stansbury, 2003). This view has similarities to broadband research that focuses on socioeconomic variables on the adoption and usage of broadband. A third stream of research looks at many perVSHFWLYHVDQGDVNVµ,VWKHGLJLWDOGLYLGHSUREOHP IUDPHGSURSHUO\"¶6FKRODUVDUJXHWKHUHLVDQHHG to understand how marginalized groups interact with technology and how those differing relationships to social, political, and environmental factors (Mossberger, Tolbert, & Gilbert, 2006; Castells, 2001; Kennedy, Wellman, & Klement, 2003; Warf, 2001) impact their access and use of technology. In this view, the experiences with technology by marginalized populations should be the focal point for understanding the problem, and consequently, for providing solutions (Crensahw, 1999; Servon, 2002; Kennedy et al., 2003). Scholars criticize solutions that mirror the needs of privileged group members in society (in any dimensionage, gender, race, income, location, world) (Winner, 1988). Scholars reject that any one group of individuals inherently uses technologies differently than the majority. Hines, Nelson, and Tu (2001) suggest: “Individuals and communities HPSOR\WHFKQRORJLHVIRUYHU\VSHFL¿FJRDOVOLQNHG often to their histories and social locations” and that “barriers to access [and use] operate on many levels and therefore solutions must take multiple approaches” (p. 5). Proponents of this perspective emphasize government intervention targeted to VSHFL¿F LVVXHV DQG SUREOHPV DQG WR UHH[DPLQH the current discourse (Chin & Fairlie, 2004; Cole et al., 2004; Mossberger et al., 2003).
Digital Divide and Broadband Access
RESEARCH DESIGN AND METHOD An important methodological question in the DQDO\VLVRIEURDGEDQGGLIIXVLRQLVWKHGH¿QLWLRQ of the object of study. As technology advances, WKH GH¿QLWLRQ RI EURDGEDQG LV FRQVLGHUHG DV D PRYLQJ WDUJHW 7KXV ¿QGLQJ D ZLGHO\ DFFHSWHG DQGGXUDEOHGH¿QLWLRQPD\UHVXOWLQDQDUGXRXV task. In the literature, three main interpretations DUHIRXQGZLWKUHVSHFWWRWKHGH¿QLWLRQRIEURDGEDQG7KH¿UVWGH¿QHVEURDGEDQGWKURXJKQRPLQDO bandwidth and is expressed in terms of kilobits or megabits per second (FCC, 2002). The second GH¿QHVEURDGEDQGE\SURYLGLQJDGHWDLOHGOLVWRI supported applications (Industry Canada, 2000). )LQDOO\WKHWKLUGDSSURDFKGH¿QHVEURDGEDQGDV “the technological environment that enables the use of digital technologies at their highest level of interactivity” (Italian Broadband Task Force, 2002). 1RQH RI WKH GH¿QLWLRQV OLVWHG DERYH PD\ EH considered, in principle, as optimal. Each carries ERWKDGYDQWDJHVDQGGLVDGYDQWDJHV7KH¿UVWGH¿QLWLRQLVDQXQDPELJXRXVGH¿QLWLRQEXWEHFRPHV REVROHWHYHU\IDVW7KHODVWGH¿QLWLRQLQFRQWUDVW does not require continual update, but leaves room for interpretation and thus renders comparative analyses very inaccurate, if not impossible. For WKHSXUSRVHVRIWKLVVWXG\WKH¿UVWDSSURDFKZDV preferred because it allowed a higher level of methodological homogeneity in the treatment of the data collected from different sources. In this chapter, broadband will be referred to as Internet connections with a downstream nominal speed equal to, or greater than, 2 Mbps. The empirical analysis presented in this chapter is based on a survey of 2,206 Italians who live in the region of Piedmont. The sample used for the purpose of this chapter was created from a database provided by the Italian National Statistical Institute (ISTAT), whose data refer to the last periodical census carried out in 2001. The entire data set was collected via computer-aided telephone interview &$7,V E\ WKH ,&7 2EVHUYDWRU\ RI 3LHGPRQW¶V Regional Government. Thus, people without a ¿[HGOLQHDUHQRWUHSUHVHQWHGLQWKHVDPSOH7KH VWUDWL¿HGVDPSOHZDVFUHDWHGXVLQJDGLIIHUHQWLDWHG
probability approach in order to over-represent segments with a higher variance in terms of technology adoption and usage (i.e., young vs. older SHRSOH 7KHYDULDEOHVDGRSWHGIRUWKHVWUDWL¿FDWLRQ of the sample were age, gender, and size of town of residence. Following the guidelines provided by the European Statistical Institute, people less than 16 years old were excluded from the sample. Respondents were asked questions about computer ownership, Internet access, and Internet use. Relevant individual demographics and household characteristics were also collected. The main analytical tools used for analysis and interpretation of data are multiple linear regression models, hierarchical cluster analysis, and cross-tabulations.
BROADBAND ACCESS AND INTERNET USE This section has two main purposes. First, it constructs models based on two of the three approaches presented in the digital divide literature review section and applies them to the phenomenon of broadband access: (1) access divide model and (2) multi-dimensional divide model (see Gil-Garcia HW DO 7KLV ¿UVW HIIRUW SURYLGHV HYLGHQFH of the importance of some individual- and household-level factors as determinants of broadband access, as well as evidence of the importance of broadband access as a determinant of the extent of Internet use. Second, this section proposes that ,QWHUQHWXVHUVFDQEHFODVVL¿HGLQWRDGYDQFHGLQtermediate, and sporadic or non-users. Using this FODVVL¿FDWLRQZHDUJXHWKDWWKHGLYLGH LQJHQHUDO is widening and policymakers should pay attention to the problem of lack of demand for broadband use, particularly to the issue of IT skills acquisition. Together these two subsections highlight the importance of broadband access and suggest some areas for future exploration.
Determinants of Broadband Access Table 1 shows the results of the access divide and multi-dimensional divide logistic regression mod-
165
Digital Divide and Broadband Access
Table 1. Determinants of broadband access: 2 Mbps and above (dummy variable) Independent Variables
Access Divide
Multi-
Model
Dimensional
-21.987
-39.895
(<0.001)
(<0.001)
<0.001***
<0.001
Divide Model Constant
Income
Age
Education
Attitude towards Computers
Nationality (Italian = 1)
Location (Town = 1)
Location (Village = 1)
Gender (Female = 1)
(12.656)
(1.736)
-0.032***
-0.010
(23.769)
(0.939)
0.303**
0.077
(4.717)
(0.243)
0.275***
0.128
(11.584)
(2.135)
18.476
17.949
(<0.001)
(<0.001)
0.042
-0.270
(0.013)
(0.451)
-1.093***
-1.381***
(7.784)
(10.746)
-0.402**
-0.336
(4.142)
(2.593)
Other Language (English)
0.444* (3.266)
PC at Home
18.153
PC Use
1.167***
(<0.001)
(6.989) IT Skills
-0.021 (0.010)
Household Size
0.231** (5.670)
Occupation (Employee = 1)
-0.070 (0.043)
Occupation (Self-Employed = 1)
-0.122 (0.082)
Occupation (Unemployed = 1)
-0.155 (0.069)
Occupation (Other = 1)
0.020 (0.002)
-2 Log Likelihood
740.095
651.262
Cox & Snell R-Square
0.112
0.168
Nagelkerke R-Square
0.231
0.346
Chi-Square
161.144***
249.541***
1RWH :DOGVWDWLVWLFVDUHLQSDUHQWKHVHVXQGHUFRHI¿FLHQWYDOXHV7KRVHFRHI¿FLHQWVIROORZHGE\ DUH VLJQL¿FDQWDWWKHOHYHOWKRVHIROORZHGE\
DUHVLJQL¿FDQWDWWKHOHYHODQGWKRVHIROORZHGE\
DUHVLJQL¿FDQWDWWKHOHYHO
166
Digital Divide and Broadband Access
els, where broadband access was measured as a dichotomous variable (whether a person reported a connection of 2 Mbps or more). Income is positively associated with broadband access, but it is RQO\VLJQL¿FDQWLQWKHDFFHVVGLYLGHPRGHO$JHLV negatively associated with broadband access in the DFFHVVGLYLGHPRGHODQGQRWVWDWLVWLFDOO\VLJQL¿FDQW in the multi-dimensional divide model. Similarly, education is positively associated with broadband DFFHVVLQWKHDFFHVVGLYLGHPRGHOEXWQRWVLJQL¿FDQW in the multi-dimensional divide model. Attitude towards computers was positively associated with EURDGEDQGDFFHVVEXWLWLVVLJQL¿FDQWRQO\LQWKH access divide model. Location (city vs. village) was WKHRQO\YDULDEOHWKDWZDVDVLJQL¿FDQWGHWHUPLQDQW RIEURDGEDQGDFFHVVLQERWKVSHFL¿FDWLRQVRIWKH PRGHO %HLQJ ORFDWHG LQ D YLOODJH VLJQL¿FDQWO\ decreases the probabilities of having broadband access to the Internet. Some variables related to the multi-dimensional YLHZZHUHDOVRVLJQL¿FDQW)RULQVWDQFHWKHDELOity to speak English was positively associated ZLWKEURDGEDQGDFFHVV3&XVHZDVDVLJQL¿FDQW determinant of broadband access and the sign was positive. Finally, household size is positively associated to broadband access and the relation is VWDWLVWLFDOO\VLJQL¿FDQW7KH&R[DQG6QHOO5VTXDUH improved from 0.112 to 0.168, and the Nagelkerke R-square went from 0.231 to 0.346 suggesting that the additional variables in the multi-dimensional model have an impact on the percentage of variance explained. However, given the low numbers in both R-squares, it seems that other important variables are missing. Ferro et al. (2005) suggest that individual-level variables have limited explanatory power in relation to broadband, and more household-level variables should be included in future studies.
Broadband and Internet Use Following the three approaches proposed in the literature review, this section presents the results IURPVHYHUDOVSHFL¿FDWLRQVRI,QWHUQHWXVHPRGHOV and assesses the impact of broadband as one of the determinants. Overall, the multi-dimensional divide model has greater explanatory power and
untangles the complex relationships in a more VSHFL¿F PDQQHU 7DEOH SUHVHQWV WKH UHVXOWV RI three models using the extent of Internet use as the dependent variable. The extent of use is represented as the number of activities an individual SHUIRUPVXVLQJWKH,QWHUQHW7KH¿UVWUHJUHVVLRQ model is based purely on the access divide view and therefore considers Internet access (including broadband access) as the only relevant factor affecting Internet use directly. The second model includes the factors mentioned in the access divide view, but tests direct relationships from all of them to Internet use, instead of assuming that they are determinants of access only. Finally, the third model incorporates additional variables related to the multi-dimensional divide view. Overall, there is an important improvement in adjusted R-square, which went from 0.395 in the access divide model to 0.537 in the extended access divide model, and then to 0.699 in the multi-dimensional divide model. Internet access is positively DVVRFLDWHGZLWK,QWHUQHWXVHLQDOOVSHFL¿FDWLRQV %URDGEDQGDFFHVVLVDOVRVLJQL¿FDQWDQGSRVLWLYHO\ associated with extent of Internet use, even when controlling for access. Income is positively associated with Internet use in the extended access divide PRGHOEXWEHFRPHVQRWVWDWLVWLFDOO\VLJQL¿FDQWRQFH controlling for other variables. Age is negatively associated with Internet use. Education and attitude towards computers are positively associated with Internet use. Being female is negatively associated with Internet use. There were several variables related to the PXOWLGLPHQVLRQDOGLYLGHWKDWZHUHVLJQL¿FDQWO\ associated to Internet use. For instance, speaking English was positively associated with Internet use. Having a PC at home was negatively associated with Internet use, but individual use of a PC was positively associated with Internet use. Similarly, individual Internet experience was positively associated with the extent of Internet use, but household Internet experience was negatively associated with the extent of individual Internet use. The models presented in this section represent two of the three approaches to understanding the digital divide and how they can be applied to the study of broadband access. The next sections go
167
Digital Divide and Broadband Access
Table 2. Determinants of Internet use (extent of use) Independent Variables
Access Divide
Access
Multi-Dimensional
Model
Divide Model
Divide Model
(Extended) Constant
Internet Access
0.376***
-0.826*
-0.120
(6.674)
(-1.669)
(-0.274)
2.660***
1.702***
1.417***
(31.250)
(14.876)
(8.914)
Broadband Access (2 Mbps and
1.315***
0.832***
0.411***
above)
(9.343)
(5.185)
(3.125)
Income
Age
Education
Attitude towards Computers
Nationality (Italian = 1)
Location (Town = 1)
Location (Village = 1)
Gender (Female = 1)
Other Language (English)
<0.001***
<0.001
(2.656)
(0.830)
-0.022***
-0.007*
(-7.451)
(-1.942)
0.543***
0.138**
(7.780)
(2.236)
0.249***
0.099***
(7.851)
(3.712)
0.230
-0.057
(0.700)
(-0.212)
0.060
0.059
(0.299)
(0.361)
0.073
0.066
(0.366)
(0.406)
-0.533***
-0.279***
(-5.805)
(-3.668) 0.189** (1.999)
PC at Home
-0.470***
PC Use
1.158***
(-3.373)
(9.291) IT Skills
-0.099 (-1.026)
Household Size
-0.030 (-0.817)
Occupation (Employee = 1)
-0.060 (-0.357)
Occupation (Self-Employed = 1)
-0.211 (-1.042)
Occupation (Unemployed = 1)
0.152 (0.654)
168
Digital Divide and Broadband Access
Table 2. (cont.) Occupation (Other = 1)
-0.171 (-0.834)
Individual Internet Experience
0.362*** (18.094)
Household Internet Experience
-0.088*** (-3.496)
R-Square
0.395
0.541
0.703
Adjusted R-Square
0.395
0.537
0.699
F-Statistic
712.875***
156.939***
146.727***
1RWH 7VWDWLVWLFVDUHLQSDUHQWKHVHVXQGHUFRHI¿FLHQWYDOXHV7KRVHFRHI¿FLHQWVIROORZHGE\ DUHVLJQL¿FDQWDWWKHOHYHOWKRVHIROORZHGE\
DUHVLJQL¿FDQWDWWKHOHYHODQGWKRVHIROORZHGE\
DUHVLJQL¿FDQWDWWKHOHYHO
further, using different analytic techniques to look at the complexities of acquiring IT skills. In doing so, the analysis is an attempt to explore a multi-perspective model (derived from the spirit of WKHGLJLWDOGLYLGHOLWHUDWXUH ZKHUHDQLQGLYLGXDO¶V preferences, skills, and aptitudes are highlighted. The combination of the three models together is an effort to add to the empirical literature, for both broadband access and the digital divide.
Internet Use and Basic IT Skill Acquisition The adoption of broadband services has been linked to more intense, advanced, and varied uses of Internet applications. While the presence of this relationship has been documented in a multitude of studies (Carriere, Rose, Sirois, Turcotte, & Christian, 2000; Horrigan et al., 2001; Anderson, Gale, Jones, & McWilliam, 2002; Dwivedi & Choudrie, 2003a; Lebo, 2001, 2003), the causal associations still remain unclear. This is similar to the multidimensional digital divide view that says access and use are both equally important. From a policymaker point of view, being able to understand to what extent the adoption of broadband leads to advanced Internet behavior or whether broadband adoption occurs only as a consequence of mature Internet
usage is key for the creation of effective broadband access and use policies. Although the answer probably lies somewhere in the middlesince it would be reasonable to assume the presence of a recursive two-way relationshipit may be worthwhile to investigate the issue further in order to have a deeper understanding of its dynamics. The presence of basic IT skillsacquired either through formal or informal processesmay be considered an essential prerequisite for both advanced usage of the Internet and the adoption of broadband connections. In addition, the motiYDWLRQIRUXVH EDVHGRQDSHUVRQ¶VSHUVSHFWLYH may also be an important factor. For this reason, this section will start by exploring the process of basic IT skills acquisition and motivation for use among different user types. The analysis below reveals a user typology UHSUHVHQWLQJ WKUHH PDLQ W\SHV 7KH ¿UVW W\SH LV FKDUDFWHUL]HGE\µDODFNRI¶RUµYHU\OLPLWHGXVH¶ of the Internet and is labeled as non/sporadic users (51% of the population). The second group represents more regular use mainly based on information search and e-mail exchange, and is labeled as basic users (25% of the population). Finally, the last cluster of users demonstrates a much wider range of use including Internet applications that require fast connections (i.e., videoconferencing, VoIP, e-shopping, video-blogging, and auctions) and is
169
Digital Divide and Broadband Access
Figure 2. Basis: All respondents Self-Learning by Usage Level 80% 70% 60% 50% 40% 30% 20% 10% 0% Non/Sporadic Users
Basic Users
GH¿QHGDVDGYDQFHGXVHUV RIWKHSRSXODWLRQ The last cluster reveals the concurrence of mature Internet behavior and the use of broadband-enabled applications. If we go further and investigate how the different types of users acquired their IT skills, some interesting details are exposed. For example, as shown in Figure 2, a good portion of IT skills acquisition appears to occur through an informal process of learning by doing. This is evidenced by the important role self-learning plays among both basic and advanced users. Moreover, this result does not seem to stem from regional considerations. (XURVWDW¶VUHFHQWO\SXEOLVKHGGDWDRQHVNLOOVLQ fact, shows that the percentage of individuals who obtained IT skills through formalized training in educational institutions in 2005 was as low as 20% (Eurostat, 2006). But Figure 2 does not simply show the importance of informal and self-learning. It also suggests the presence of different approaches to technolRJ\ DGRSWLRQ DQG XVDJH RU XVHUV¶ PRWLYDWLRQV Thus it suggests that a correct interpretation of the relationship between broadband adoption and mature Internet usage may require us to consider a number of different individual perspectives. This adds support for looking at the presence of a multi-perspective divide as documented in the literature and discussed above. In other words, it seems plausible to assert that the presence of a sig-
170
Advanced Users
QL¿FDQWOHYHORIVHOIOHDUQLQJLQEDVLFDQGDGYDQFHG users renders the availability of broadband as an enabler or represents the strategy of “engaging the engaged.” For none or sporadic users, instead, the acquisition of IT skills through formal IT training seems to play a more important role in moving users toward more mature usage. 7KHHYROXWLRQRIWKHXVHUV¶GLVWULEXWLRQRYHUWLPH to a certain extent supports the thesis unfolding above by showing the presence of two different maturation speeds between clusters. Figure 3 shows the distribution of user types for the years 2004 and 2005. The presence of a U-shaped distribution is noticeable. The fourfold difference between the migration rates from basic users to advanced users and the one from sporadic users to basic users represents a clear signal. While part of the population is gradually progressing along the LQIRUPDWLRQVRFLHW\SDWKDQRWKHUVLJQL¿FDQWSRUtion is struggling (or is not motivated) to join this development process. The increase in the digital valley between different users represents a serious hurdle in the process of e-inclusion envisioned by recent strategic plans issued at the European level. A possible explanation of this phenomenon is the presence of positive and negative feedback LQZKDWKDVEHHQGH¿QHGDVD³networked society” (Shapiro & Varian, 1998). These concepts, mainly SUHVHQWHGLQWKHOLWHUDWXUHLQUHODWLRQWRFRPSDQLHV¶
Digital Divide and Broadband Access
Figure 3. Basis: All respondents
SHUIRUPDQFHPD\EHDSSOLFDEOHWR,QWHUQHWXVHUV¶ behaviors as well. From a policy standpoint, a counterbalance to the polarization created by positive and negative feedback may be necessary. This could happen WKURXJKSROLFLHVWKDWSURGXFHÀDWWHQLQJHIIHFWVRQ the user-type distributions. Such policies would require actions aimed at fostering the enlargement of the total Internet population, thus contributing to increase the migration rate from non-users to basic users. In order to do this, policymakers need to gain a better understanding of the different motivations and socioeconomic and demographic characteristics of the different user types in order WRGHYLVHDGKRFDFWLRQVWDLORUHGWRWKHVSHFL¿FXVHU types where they want change to occur.
FUTURE TRENDS The analysis conducted above highlights three different types or levels of inconsistencies: (1) inconsistency in coverage, (2) inconsistency in adoption, and (3) inconsistency in usage. This categorization is not exhaustive and provides only a partial view of the whole phenomenon. Other authors have proposed to describe the phenomenon from different (and not necessarily opposing) points
of views (e.g., Van Dijk, 2005). Nevertheless, it is a useful starting point to discuss the evolution of the issues that national and local policymakers will have to deal with as time continues (and most likely, in the near future). The coverage inconsistency has to do with the QHFHVVLW\WR¿QGQHZDQGPRUHFRVWHIIHFWLYHZD\V to deliver broadband services to rural areas or, more generally, to less densely populated areas. Although currently this type of divide clearly represents a pending issue for policymakers, it is bound to decrease in importance as penetration increases. As a matter of fact, technological progress, which is led by companies racing for a FRPSHWLWLYH DGYDQWDJH LV VLJQL¿FDQWO\ GULYLQJ down infrastructure-related costs, but this trend LVQRWXQLIRUPIRUDOOORFDWLRQV6SHFL¿FSROLFLHV and programs may be needed in order to solve this important element. Coverage is necessary, but is not the only condition for the successful adoption of broadband access and meaningful uses of the Internet. The second type of inconsistency has to do with the low demand levels in served areas. This may be attributable to a multiplicity of factors; however, not all factors can be addressed by local or national policymakers. Nevertheless, government action should be aimed at breaking the feedback loop
171
Digital Divide and Broadband Access
between the lack of demand and the existence of broadband-related services that currently hinders the creation of more broadband-related activities. For instance, the fact that the ability to speak EngOLVKKDVDVLJQL¿FDQWSRVLWLYHLPSDFWRQEURDGEDQG access suggests that the content for meaningful uses is limited in other languages. Some policies and programs could promote the creation of meaningful FRQWHQWLQVSHFL¿FODQJXDJHV In the medium/long term, inconsistency that should concern policymakers the most is usage. As a matter of fact, the analysis presented above shows a polarization trend, leading to a U-shaped distribution with a clear separation between none/ sporadic users and advanced users. The situation FDOOVIRUPHDVXUHVWKDWZLOOSURGXFHDÀDWWHQLQJ effect on the distribution. A good starting point may be accounting for the importance self-learning plays in the acquisition of necessary IT skills, in addition to recognizing the presence of complementary skills. Digital divide policies should take into consideration the differences and similarities among social groups and have target programs for different types of users. Some people need training, others need motivation (meaningful content and transactions) and opportunities to experience the potential of using the Internet for a broader range of activities. The models discussed in this chapter are cuPXODWLYH)LUVWWKH¿QGLQJVDQGOLWHUDWXUHVXJJHVW that access to technology (i.e., computers) is important, as is access to the Internet (broadband or QRW 7KHQWKHVHFRQGPRGHODQGOLWHUDWXUH¿QGV that broadband access is an important element, but another important factor is use. This builds on the idea that it is not just about access, but also about how individuals use the Internet. Access to broadband is a category of access to the Internet, and it is suggested that a better connection allows for more extensive and sophisticated use. The third model, a multi-perspective model, says, once you have access and use, meaningful use depends on the context of the user and his or her level of skill, overall motivation for using the medium, DQG WKH H[SHFWHG EHQH¿WV SHUFHLYHG 7KHUHIRUH any meaningful development of comprehensive
172
policies needs to address concerns highlighted by all three models.
CONCLUSION In general terms, the research presented in this chapter recognizes that the digital divide is a complex phenomenon transcending simple information access problems. Broadband access is an important determinant of the extent of Internet use, and therefore, its diffusion could potentially have a VLJQL¿FDQWLPSDFWRQVRFLDODQGHFRQRPLFGHYHORSment. However, broadband access is only one of the important factors to take into consideration. Age, education, gender, attitudes toward computers, the availability of the necessary equipment, and previous experience are also important aspects contributing to the complexity of this phenomenon. If the main goal is to increase Internet use and the value for individuals and jurisdictions, policymakers should consider integrative strategies that include not only broadband diffusion mechanisms, but also some policies targeted at some of the other dimensions of the problem. Even when thinking about broadband access as the goal itself, there are still many factors that need to be taken into account. In fact, this study provides evidence of the importance of multiple determinants of broadband access such as location, ability to speak English, household size, and familiarity with PCs. Consistent with previous research, this study found that location is important and individuals living in villages are less likely to have broadband access. In addition, individuals who can speak English, which is the language of a great number of Web sites, are more likely to have broadband access. In addition, the language difference may indicate, to a certain extent, the value that the Internet has for individuals; the more Web sites an individual is able to use in meaningful ways, the more this individual is willing to invest in better ways to use the Internet, including broadband access and IT training. In addition, different approaches to basic IT skills acquisition emerged and uncovered factors WKDWLQÀXHQFHGLYHUVHXVDJHOHYHOV,QIDFWDERXW
Digital Divide and Broadband Access
one-fourth of our respondents are advanced users (many of them also broadband users), about 20% are characterized as basic users, while the remaining 50% make sporadic use of the Internet or do not use it at all. Moreover, the analysis carried out over a two-year period depicted the presence of a widening gap in terms of Internet use between non/sporadic users and advanced users. Taking into consideration that Internet use is fundamental for individual and jurisdictional development, national and local policymakers should direct part of their efforts to offset this usage polarization, which in fact may be related to broadband access among other factors. In order to do so, understanding how people approach and use technology and the different paths leading to the acquisition of the necessary IT skills represents one fundamental aspect. A careful and close management of the evolution of digital gaps by policymakers seems to be desirable and necessary. At the same time, attention should be put toward avoiding technological deterministic approaches aimed at fostering technology adoption and use per se. Rather, the use of technology should be positioned as an important enabling tool supporting individuals in their main everyday activities (production, social, political, consumption, savings activities) (Selwyn, 2003). Hence, this should translate into opportunities to re-frame public policy debates and direct attention to the problems from a multitude of perspectives. %URDGEDQGDFFHVVPD\VLJQL¿FDQWO\LPSDFWVRPH aspects of the problem described throughout this chapter, but it is only one element in a much more complex web of social, economic, and political issues.
REFERENCES Anderson, B., Gale, C., Jones, M.L.R., & McWilliam, A. (2002). Domesticating broadbandwhat FRQVXPHUVUHDOO\GRZLWKÀDWUDWHDOZD\VRQDQG fast Internet access. BT Technology Journal, 20(1), 103-114. Adriani, F., & Becchetti, L. (2003). Does the digital divide matter? The role of ICT in cross-country level and growth estimates. CEIS Tor Vergata.
Annis, R. (2005). Industry Canada broadband economic impact study. Rural Development Institute, Brandon University, Canada. BAG. (2003). Australia’s broadband connectivity. Retrieved November 9, 2006, from http://www. noie.gov.au/publications/NOIE/BAG/report/index. htm Bouvard, P., & Kurtzman, W. (2001). The broadband revolution: How superfast Internet access changes media habits in American households. Retrieved November 9, 2006, from http://www. arbitron.com/downloads/broadband.pdf Cantamessa, M., De Leonardis, D., Ferro, E., & Paolucci, E. (2005). Wireless technologies and broadband diffusion: The Piedmont case. Wireless Magazine, (December), 44-49. Cantamessa, M., De Leonardis, D., Ferro, E., & Paolucci, E. (2006). Wi-Fi: A hurdle race. Wireless Magazine, (May), 48-52. Cantamessa, M., Ferro, E., Molino, S., & Paolucci, E. (2004). Analysis of the broadband market in Piedmont, Italy. Turin: Regional Observatory on ICT. Cantamessa, M., Ferro, E., & Paolucci, E. (2005). Urban vs. regional divide: Comparing and classifying digital divide. Proceedings of TCGOV 2005 (pp. 81-90). Berlin: Springer-Verlag (LNAI 3416). Carriere, R., Rose, J., Sirois, L., Turcotte, N., & Christian, Z. (2000). Broadband changes everything. Retrieved November 8, 2006, from http:// www.mckinsey.de/_downloads/knowmatters/telecommunications/broadband_changes.pdf Castells, M. (2001). 7KH,QWHUQHWJDOD[\5HÀHFtions on the Internet, business, and society. New York: Oxford University Press. Chin, M.D., & Fairlie, R.W. (2004). The determinants of the global digital divide: A cross-country analysis of computer and Internet penetration. Economic Growth Center, Yale University, USA. Choudrie, J., & Dwivedi, Y.K. (2004a). Analyzing the factor of broadband adoption in the household.
173
Digital Divide and Broadband Access
Proceedings of the 12th European Conference of Information Systems (ECIS), Turku, Finland. Choudrie, J., & Dwivedi, Y.K. (2004b). Investigating the socio-economic characteristics of residential consumers of broadband in the UK. Proceedings of the American Conference on Information Systems (AMCIS), New York. Choudrie, J., & Dwivedi, Y.K. (2004c). Towards a conceptual model of broadband diffusion. Journal of Computing and Information Technology, 12(4), 323-338. Choudrie, J., & Lee, H.J. (2004). Broadband development in South Korea: Institutional and cultural factors. EJIS, 13(2), 103-114. Cole, J.I. et al. (2004). The digital future report. Center for the Digital Future, USC, USA. Compaine, B.M. (Ed.). (2001). The digital divide: Facing a crisis or creating a myth? Cambridge, MA: MIT Press. Dadayan, L., & Ferro, E. (2007) Broadband and e-government diffusion. Proceedings of the 40th Hawaii International Conference on Computing Conference. IEEE Computer Society Press. Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. DiMaggio, P., & Hargittai, E. (2001). From the ‘digital divide’ to ‘digital inequality’: Studying Internet use as penetration increases. Support from the Center for the Arts and Cultural Policy Studies, Woodrow Wilson School, Princeton University, USA. Dwivedi, Y.K., & Choudrie, J. (2004). Developing a model of broadband adoption in the household. Proceedings of the ETHICOMP Conference, Syros, Greece. Dwivedi, Y.K., & Choudrie, J. (2003a). The impact of broadband on the consumer online habit and usage of Internet activities. In M. Levy et al. (Eds.), Proceedings of the 8th UKAIS Annual Conference on Coordination and Co-Opetition: The IS Role, Warwick, UK.
174
Dwivedi, Y.K., & Choudrie, J. (2003b). Considering the impact of broadband upon the growth and development of B-2-C electronic commerce. In R. Cooper et al. (Eds.), Proceedings of the ITS Asia- Australasian Regional Conference, Perth, Australia. Dwivedi, Y.K., Choudrie, J., & Gopal, U. (2003). Broadband stakeholders analysis: ISPs perspective. Proceedings of the International Telecommunication Society’s Asia-Australasian Regional Conference, Perth, Australia. European Commission. (2006). Bridging the communication gap. Retrieved November 9, 2006, from http://europa.eu.int/information_society/ eeurope/i2010/docs/digital_divide/communication/com_bridging_bb_gap_2006_en.pdf Eurostat. (2006). Eurostat. Retrieved November 9, 2006, from http://epp.eurostat.ec.europa.eu/portal/ page?_pageid=1996,45323734&_dad=portal&_ schema=PORTAL&screen=welcomeref&open=/ & p r o d u c t = E U_ M A ST E R _ i n fo r m a t i o n _ society&depth=2 FCC. (2002). United States Federal Communications Commission report FC-02-33A. Retrieved November 9, 2006, from http://hraunfoss.fcc. gov/edocs_public/attachmatch/FCC-02-33A1.pdf Ferro, E. (2006). Broadband diffusion dynamicsa systemic approach. International Journal of Electronic Business, (Special Issue), 1470-6067. Flamm, K. (2004). The role of economics, demographics, and state policy in broadband competition: An exploratory study. Proceedings of the 32nd Telecommunications Policy and Research Conference, Arlington, VA. Flamm, K., & Chaudhuri, A. (2005). An analysis of the determinants of broadband access. Proceedings of the 33rd Telecommunications Policy and Research Conference, Arlington, VA. Gil-Garcia, J.R., Helbig, N., & Ferro, E. (2006, September 4-8). Is it only about Internet access? An empirical test of a multi-dimensional digital divide. Proceedings of the eGov06 Conference, Krakow, Poland. Berlin: Springer-Verlag (LNCS 4084).
Digital Divide and Broadband Access
Gillett, S., Lehr, W., & Osorio, C. (2003). Local government broadband initiatives. Proceedings of the 31st Research Conference on Communication, Information and Internet Policy (pp. 537-558), Arlington, VA. Grubesic, T.H. (2004). “The geodemographics correlates of broadband access and availability in the United States. Telematics and Information, 21(4), 335-358. Helbig, N., Gil-Garcia, J.R., & Ferro, E. (2005). Understanding the complexity in electronic government: Implications from the digital divide literature. Proceedings of the 11th Americas Conference on Information Systems, Omaha, NE. Hines, A.H., Nelson, A., & Tu, T.L.N. (2001). Hidden circuits. In A. Nelson, T.L.N. Tu, & A.H. Hines (Eds.), Technicolor. New York: New York University Press. Howell, B. (2002). Broadband uptake and infrastructure regulation. Working Paper, ISCR. Horrigan, J.B., & Rainie, L. (2001). Online communities: Networks that nurture long-distance relationships and local ties. Pew Internet and American Life Project. Hulsink, W. (1999). Privatization and liberalization in European telecommunication: Comparing Britain, The Netherlands and France. London: Routledge, Taylor and Francis Group.
Lebo, H. (2003). Surveying the digital future, year two, the UCLA Internet report. Retrieved November 9, 2006, from http://www.digitalcenter.org/ Lehr, W.H., Osorio, C.A., Gillett, S.E., & Sirbu, 0$ 0HDVXULQJ EURDGEDQG¶V HFRQRPLF impact. Proceedings of the 33rd Research Conference on Communication, Information, and Internet Policy, Arlington, VA. /HH+2¶.HHIH% <XQ. 7KHJURZWK of broadband and electronic commerce in South Korea: Contributing factors. The Information Society, 19, 81-93. Middleton, C.A. (2002). Exploring consumer demand for networked services: The importance of content, connectivity and killer apps, in the diffusion of broadband and mobile services. Proceedings of the 23rd International Conference on Information Systems. Mossberger, K., Tolbert, C.J., & Stansbury, M. (2003). Virtual inequality: Beyond the digital divide. Washington, DC: Georgetown University Press. Nelson, L. (2002). Protecting the common good: Technology, objectivity, and privacy. Public Administration Review, 62(Special Issue), 69-73. Norris, P. (2001). Digital divide: Civic engagement, information poverty, and the Internet worldwide. New York: Cambridge University Press.
Industry Canada. (2006). Broadband RI¿FH Retrieved November 9, 2006, from http://largebande. gc.ca/pub/technologies/bbdictionary.html
OECD. (2004). Information technology outlook. Information and Communication Technologies publication series. Paris.
Italian Broadband Task Force. (2002). Italian Ministry of Innovation and Technologies. Retrieved November 9, 2006, from http://www.innovazione. gov.it/ita/intervento/banda_larga/ task_force/ GH¿QL]LRQHKWP
Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitude. Journal of Information Technology, 18, 267-280.
Kennedy, T., Wellman, B., & Klement, K. (2003). Gendering the digital divide. IT & Society, 1(5), 72-96. Lebo, H. (2001). Surveying the digital future, year two, the UCLA Internet report. Retrieved November 9, 2006, from http://www.digitalcenter.org/
Pentland, A., Fletcher, R., & Hasson, A. (2003). A road to universal broadband connectivity. Cambridge, MA: MIT Media Lab. Regional Observatory on ICT. (2005). Piedmont report. Retrieved November 9, 2006, from http:// www.sistemapiemonte.it/osservatoriICT
175
Digital Divide and Broadband Access
Robinson, J.P., DiMaggio, P., & Hargittai, E. (2003). New social survey perspectives on the digital divide. IT & Society, 1(5), 1-22. Servon, L.J. (2002). Bridging the digital divide: Technology, community, and public policy. Oxford: Blackwell. Skogseid, I. (2005). Market driven development of broadband infrastructure in rural areas. Western Norway Research Institute. Shapiro, C., & Varian, H.R. (1998). Information rules: A strategic guide to the network economy. Boston: Harvard Business School Press. Varian, H.R. (2001). The demand for bandwidth: Evidence from the INDEX project (pp. 1-19). Berkeley: University of California–Berkeley. Venkatesh, V., & Brown, S. (2001). A longitudinal investigation of personal computers in homes: Adoption determinants and emerging challenges. MIS Quarterly, 25(1), 71-102. Walsh et al. (2001). The truth about the digital divide. In B.M. Compaine (Ed.), Reexamining the digital divide: Facing a crisis or creating a myth? Cambridge, MA: MIT Press. Warf, B. (2001). Segueways into cyberspace: Multiple geographies of the digital divide. Environment and Planning B: Planning and Design, 28, 3-19.
KEY TERMS Access Divide: A common viewpoint which assumes that the digital divide is a symptom of an µDFFHVVWRWHFKQRORJ\SUREOHP¶DQGDVDQLQKHUHQW delay in the diffusion of technology among different geographic areas and social groups.
176
Digital Divide: Often characterized as a set of relationships between information and communication technologies (ICTs) and groups of individuals, who are situated within a complex arrangement of social, environmental, political, and economic issues. Popularly, the term itself, digital divide, carries the connotation of those who have access to technology and those who do not have access to technology. Hiperlan (HIgh PErformance Radio LAN): A wireless LAN standard; a European alternative IRU$PHULFD¶V,(((VWDQGDUGV7KHVWDQGDUGVDUHGH¿QHGE\WKH(XURSHDQ7HOHFRPPXQLcations Standards Institute (ETSI) by the BRAN (Broadband Radio Access Networks) project. IT Skills: A varied concept, ranging from skills describing information-retrieval and searching activities to skills regarding the synthesis of information and productive use of information in daily activities. Multi-Dimensional Divide: A competing viewpoint which assumes that the digital divide is a mirror of other types of socioeconomic inequalities, and suggests access and use are important determinants of the digital divide. In addition, other dimensions such as income, age, education, location, and IT skills impact access and use. Multi-Perspective Divide: A competing viewpoint which rejects that any one group of individuals inherently uses technologies differently than the majority, and suggests any questions about access and use should be evaluated based on the needs and problems of the most disadvantaged groups, in addition to how the intersections of race, gender, class, worldview, and so forth come together. Systemic Approach: Accounting for the interactions and interdependencies present among a number of items forming a unifying whole (broadband diffusion).
177
Chapter XII
Improving Broadband Access in Rural Areas Ingjerd Skogseid Western Norway Research Institute, Norway
ABSTRACT The chapter explores the characteristics of rural broadband infrastructure development. Taking the existing installed base into consideration, small rural communities can initiate bottom-up cultivation of broadband infrastructure. Such initiatives are important contributions to overcoming the disparity in broadband access. In effect they aggregate demand by creating a larger total market for suppliers than the individual needs of the actors. The proposal is to use descriptive clusters as a way to reveal the LQVWDOOHGEDVH$VHWRITXHVWLRQVDQGDQVZHUVZLOOEHGH¿QHGWRUHYHDOWKHLQVWDOOHGEDVH7KLVFDQEH used to acquire an overview of the types of resources available and the choices that need to be made. The use of descriptive clusters places emphasis on the local context and culture. With a bottom-up strategy, questions must be answered in relation to the local context. The responses and lessons learned may vary from one location to the next, making blueprint implementations impossible.
INTRODUCTION To the question “Is it possible to improve broadband access in rural areas?” we must reply, “Yes.” However, the question of “How?” must be answered by taking into consideration a combination of policies, strategies, collaboration, and costs involved, not only on a national level but also on the local level. This chapter explores the development of broadband access in a deregulated market and in relation to areas that are least favored. In countries with a deregulated telecommunications market, the process of developing the broadband infrastructure differs from earlier
infrastructure processes. In a competitive market, telecommunications operators are likely to develop broadband infrastructure in urban areas, while rural areas with lower demand are developed at a much slower rate because of higher costs and LQVXI¿FLHQW GHPDQG $FFHVV DQG DGRSWLRQGHPDQG RI EURDGEDQG WHFKQRORJ\ LV D µFKLFNHQ RU HJJ¶SUREOHP%HIRUHDEURDGEDQGLQIUDVWUXFWXUH may be adopted, it must be made available, while WKHDFFHVVSURYLGHUVUHTXLUHDVXI¿FLHQWOHYHORI demand before extending their services. However, an OECD (2004) status report on broadband provision in rural and remote areas suggests that the competitive market works even in these areas.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Improving Broadband Access in Rural Areas
Why is this true, and what are the characteristics of rural broadband development? This chapter explores the role of the installed base in the development of two rural broadband initiatives in Sogn og Fjordane County, Norway, and discusses how different installed bases resulted in different processes and different infrastructures (see Table 3 for more details). The chapter is divided into six sections: background, research methodology, key theoretical perspectives, a description of the cases, a discussion, and a conclusion, including practical implications. It explores the development of access to broadband in a rural region1 in the context of a deregulated PDUNHW,WDOVRH[SORUHVIDFWRUVWKDWLQÀXHQFHDERWtom-up development of broadband infrastructure in rural areas as opposed to the traditional rollout of infrastructure.
GLI¿FXOW:KHQLQIRUPDWLRQDQGVHUYLFHSURYLGHUV deliver services requiring broadband capacity, the inhabitants and businesses must have access to a high-speed connection in order to take advantage of such services. As a result of this situation, a myriad of broadband providers have come into existence. A VWXG\FRPSOHWHGLQLGHQWL¿HGEURDGEDQG providers in Norway (Norsk_Telecom, 2004), and LQDVLPLODUVWXG\LGHQWL¿HGSURYLGHUV (Post_og_teletilsynet, 2006b). In 2004, about 10 of the 130 were categorized as national providers, delivering broadband services with national coverDJH$ERXWZHUHGH¿QHGDV³UHJLRQDODFWRUV´ and the remaining 80 were characterized as local providers serving local communities. The many small local providers were the result of local initiatives. The municipalities participated as owners IRU DERXW RI WKH SURYLGHUV LGHQWL¿HG LQ (Norsk_Telecom, 2004).
BACKGROUND Broadband Access At the time the telecommunications market was deregulated, the Norwegian government chose a demand-driven strategy for the development and delivery of broadband Internet access. The competition among telecommunications providHUVZDVWKRXJKWWREHVXI¿FLHQWWRHQVXUHJHQHUDO access to broadband services. In the context of PDUNHWGULYHQGHYHORSPHQWLWLVQRWSUR¿WDEOHWR roll out broadband to all parts of the country, and the national providers did not plan to develop a universal service. In March 2004, roughly 77% of Norwegian households had access to broadband infrastructure (Teleplan, 2003a), while by WKHHQGRIWKLV¿JXUHKDGLQFUHDVHGWR (Post_og_teletilsynet, 2006a). The government initiated a program called Høykom2 to increase the demand for broadband services, which also applies to rural areas; this program supports local and regional government by partially funding the acquisition of broadband access and the development of broadband services in the public sector. Although this support is welcomed, local and regional governments are still required to provide a substantial part of the investment required, which can sometimes prove
178
Under the regulated market, new segments of the telecommunications infrastructure were often developed as a top-down rollout and were designed for VSHFL¿FSXUSRVHV7KHVHLQIUDVWUXFWXUHVHJPHQWV were more or less uniform, and existed in markets which might be characterized as monopolies or duopolies (OECD, 2004). In Europe, this changed with the deregulation of the telecommunications market in 1998. As a result, the telecommunications infrastructure is now being developed in the context of a competitive market. One drawback of this development is that areas with smaller competitive markets, such as rural and remote areas, may not gain access to broadband infrastructure (Grubesic & Murray, 2002; OECD, 2004; Stanton, 2004). The constantly increasing use of information and communication technology (ICT) and the economic transformation currently underway (Grubesic, 2003) make access to broadband techQRORJ\HVVHQWLDOIRUSHRSOH¶VOLYHV In a competitive market, the basic assumption is that telecommunications operators are less likely to develop broadband infrastructures in rural areas EHFDXVHRIKLJKHUFRVWVDQGLQVXI¿FLHQWGHPDQG
Improving Broadband Access in Rural Areas
(OECD, 2004), even when they are based on an existing infrastructure, such as telephone lines. This is supported by a study carried out in the United 6WDWHV ZKLFK H[DPLQHV WKH IDFWRUV LQÀXHQFLQJ the adoption of broadband. According to Stanton (2004, p. 9), the “demographics for broadband access do not replicate the demographics found for computer ownership and home Internet access.” :KLOHUDFLDODQGVRFLRHFRQRPLFYDULDEOHVLQÀXence computer ownership and Internet access, they are less important for the purchase of broadband DFFHVV%URDGEDQGDFFHVVLVPRUHVWURQJO\LQÀXenced by geographical variables (Stanton, 2004), and the broadband disparity arises between urban and rural areas. A study carried out in the county of Møre & Romsdal, Norway (Molka-Danielsen, Ohren, & Jæger, 2003), found that there are clear differences between rural and urban areas in the development of broadband infrastructure. “The aggregation of demand of users in rural communities is often not great enough to justify the investments” (MolkaDanielsen et al., 2003). The cost of developing broadband service is too high relative to the demand. But even if rural areas are lagging behind, the situation is not exclusively negative. In an OECD study of broadband access in rural and remote areas (OECD, 2004), it was found that the results in rural areas are “sometimes surprising and counter-intuitive…There is a rapidly increasing amount of new private-sector entry occurring in the provision of broadband access in rural areas” (OECD, 2004, p. 4). These providers are delivering high-quality service at a competitive price. This chapter will further examine the development of broadband infrastructure in rural areas.
RESEARCH METHOD This study of rural broadband infrastructure development is a study of factors contributing to and LQÀXHQFLQJWKHGHYHORSPHQWRIWKHLQIUDVWUXFWXUH An interpretive approach is used to reach a better understanding of the development (Walsham, 1993). Interpretive research is concerned with the development of a deep and contextual understand-
ing of a phenomenon. The focus is on the subjective processes surrounding the social construction of the network, using qualitative research methods (Walsham, 1993). The aim of such interpretive analysis is to understand rather than to predict. The interpretive perspective helps to focus on the formal and informal parts of the process that took place. The study explores a longitudinal process which covers the years from 1998 until today, and which involves activities at different levels, both regional and local. At the regional level I started out by following the process as a participant observer (Baskerville & Wood-Harper, 1996; Cole, 1991). As the process stabilized my involvement in the development has decreased, and my main source data have been documents, status reports, project reports, and applications for continued funding. Data has been gathered as part of the process through the interaction, participation, and observation of meetings and activities. In addition three of my colleagues have been active in the process as project managers and change agents; dialogue and discussions with them have been important LQ¿OOLQJJDSV At the local level a combination of interviews with broadband providers and a survey of the various offers of broadband services in the region have been used to gather data. The survey was published in a series of status reports and updated based on developments. Early in the process the frequency of status reports was high, while lately the frequency has become lower. Additional sources of information have included the Web sites of the local initiatives, project reports, and interviews with some key actors from some of the local initiatives to complete the picture. The analysis has been carried out using a historical perspective (Granovetter & McGuire, 1998, p. 149). This approach opens up for a degree of uncertainty as the events are interpreted and the interpreter bases the interpretation on the knowledge he or she has about the development; this again can open up for different interpretations. In this case, the interpretation is carried out as a FULWLFDO UHÀHFWLRQ E\ P\VHOI DV RQH DFWRU LQ WKH network. The aim is to analyze the development
179
Improving Broadband Access in Rural Areas
of the broadband infrastructure exploring the GHYHORSPHQWSURFHVVDQGWKHIDFWRUVLQÀXHQFLQJ the development.
tures. Broadband is shared between all its users, it supports a number of different applications, and its users are aware when it does not function as expected.
KEY THEORETICAL PERSPECTIVES
Infrastructure Cultivation
Information Infrastructure
Infrastructures are heterogeneous networks (Shapiro & Varian, 1999) of technical and organizational components assembled together. The heterogeneity of the network makes them complex to plan and design. Traditionally, telecommunications infrastructures, top-down planned and rolled out, are managed internally without much interaction with actors who own neighboring infrastructures. They themselves have full control over the infrastructure, the installed base, and how it may be extended. As a result both the technical and organizational complexity of the development may be considered to be relatively low. In a deregulated market there are new challenges. Depending on the approach chosen, the complexity of the development can vary from ORZWRKLJK8UEDQDUHDVZLWKDVXI¿FLHQWPDUNHW demand will still be served by providers who can roll out the infrastructure without much coordination with other actors. In rural areas where the market demand is low, alternative strategies are needed to achieve the same level of service; this involves more coordination between owners of the installed base and results in more complexity through integrating different technical and organizational systems. According to Star and Ruhleder (1996), factors such as the conventions of a community, practice, and working routines both shape and are shaped by the infrastructure. Consequently, such factors must be taken into consideration in the design process. An infrastructure from this perspective is never developed from scratch; it builds on something and continuously interacts with it (Hanseth, 2002), and when changed or improved, the new version KDVWR¿WLQZLWKWKHH[LVWLQJLQIUDVWUXFWXUH7KLV µVRPHWKLQJ¶LVRIWHQWHUPHGWKHLQVWDOOHGEDVH7KH characteristics of the installed base in question will vary depending or whether a top-down or a bottom-up approach to development is taken. In a
$QLQIUDVWUXFWXUHLVGH¿QHGDV³DVHWRILQWHUFRQnected structural elements that provide the framework for supporting the entire structure. It usually DSSOLHVRQO\WRVWUXFWXUHVWKDWDUHDUWL¿FLDO´ :LNLpedia, 2006). An infrastructure may be looked upon as a network that connects an integrated and heterogeneous set of components, such as equipment, systems, applications, processes, and people. But at the same time these components are rather stable structures, with a certain inertia. An information infrastructure is an infrastructure that supports the information society: the equipment, systems, applications, and support systems needed for operating in the information society. The term information infrastructure was introduced as part of the National Information Infrastructure initiative launched by the Clinton Administration in 1994. In the Clinton–Gore initiative, the Internet was described as an infrastructure shared by its users. With the introduction of the Internet as a tool for the general public and for particular purposes, there was a need for a construct which emphasizes the distinction between Internet- type applications and ordinary information systems. Information infrastructures are shared, according to Hanseth (2000), in that they are a foundation for the community they support. They are also open and enabling, in that they support a large variety of applications. In addition they are evolving, standardized, and heterogeneous WKDWLV both their organizational and technical components are connected using standards or gateways. The chapter will focus on broadband as an information infrastructure. Broadband is both an infrastructure and an information infrastructure because it consists of a set of interconnected components, equipment, systems, applications, processes, and people in more or less stable struc-
180
Improving Broadband Access in Rural Areas
top-down setting, the installed base is controlled by the developer. However, in a bottom-up setting, the installed base may be managed by different organizations and consist of different technical solutions; as a result the complexity of the development increases, and the installed base will also vary from community to community. Consequently, it is impossible to make a blueprint copy that may be used from one location to the next. Hanseth (2002) has developed a notion of cultivating an infrastructure based on the available installed base as a bottom-up strategy. The FRQFHSWRIµFXOWLYDWLRQ¶RULJLQDWHVIURPWKHVSKHUH of agriculture and means to “to grow and cultivate plants and the domestication of animals.” In the sphere of information systems. the word has been used in relation to strategies for developing technological systems. Dahlbom and Janlert (1996) use the concepts of construction and cultivation to describe two different processes to shape technolRJ\µ&RQVWUXFWLRQ¶SXWVHPSKDVLVRQWKHKXPDQ ability to shape the world according to set of goals. µ&XOWLYDWLRQ¶UHIHUVWRWKHSURFHVVRIVHOHFWLQJDQG assembling components, but also to a process of growth or development in the material itself when we create a system. According to Hanseth (2002), cultivation as a concept views technological systems as organisms that have a life of their own; this focuses on the limits of the rational and human control involved in the development. +DQVHWKDQG/\\WLQHQ GH¿QHWZRNH\ design strategies for information infrastructures. (DFKRIWKHVHKDVDGH¿QHGVHWRISULQFLSOHVDQG guidelines to make the development process simpler and to increase the success rate of the implementation. The strategy “bootstrap installed base” considers how to cultivate the infrastructure by IRFXVLQJRQXVHIXOQHVVIRUWKH¿UVWDGDSWHUVDQG the size of the user groups as a way of ensuring momentum and creating a critical mass of users. Another way of building momentum is to extend the available installed base. When building on the installed base, the new components need to be simple and inexpensive, but they still need to IXO¿OOWKHPDLQDLPRIFUHDWLQJDXVHIXOVHUYLFH 7KH\ QHHG WR ¿OO WKH FRQQHFWLRQ WR WKH H[LVWLQJ entity using an interface based on standards and
gateways, but it is also easier for potential users to connect to the infrastructure without making a total break from the known and existing entity. Improvements should be on a needs basis, rather than a “nice-to-have” basis. Throughout the whole process the alignment of interest needs to take place in parallel to the building process; part of this is also a process of developing a support community for the infrastructure. In this way the new infrastructure is built brick by brick as part of a bottom-up strategy. The second strategy “avoid technology lock-ins” focuses on simplicity, modularization, and use of gateways between different layers and components and to neighboring infrastructures. A lock-in situation is something that can happen over time as the infrastructure develops. By choosing simple and inexpensive components and designing IRUÀH[LELOLW\LWPD\EHSRVVLEOHWRDYRLGORFNLQ situations.
Exploring the Installed Base When dealing with bottom-up cultivation of infrastructure in a rural area, the complexity is high and there is a need to explore features of the available local installed base before and during the cultivation process. What is the installed base in broadband development? Installed base in general is the available infrastructurethe existing standards, and technical and non-technical components needed to develop an infrastructure or develop it furtherto cultivate it. An infrastructure is never developed from scratch, it is an extension to the existing installed base (Hanseth & Lyytinen, 2006); therefore there is always an existing installed base that can be cultivated. In a bottom-up process, it is necessary to start by examining the installed base in a local community to identify what is the starting point and identify actors that can be aligned and utilized in the cultivation process. The design strategies (Hanseth & Lyytinen, 2006) for information infrastructures with their principles and guidelines provide guidance for the designers and constructors who are responsible for the cultivation process. They bring us to a certain level where the instruction is to draw on and expand the installed base by using the existing transport
181
Improving Broadband Access in Rural Areas
infrastructure and building gateways to the existing infrastructure. In the case of local broadband development, the number of local actors controlling the existing installed base may be high, and the number of needs and motivations may also be high. As a result, the development is often complex with a number of actors who need to align and move toward a common aim. One of the challenges is to identify the installed base and consider how best to utilize it in the design process. But how can we identify the available installed base and how can WKHLQVWDOOHGEDVHEHXWLOL]HGLQWKHGHVLJQ"6FK|Q¶V ³UHÀHFWLYHFRQYHUVDWLRQZLWKWKHVLWXDWLRQ´ offers one such framework. In an effort to idenWLI\³WKHUHÀHFWLYHSUDFWLWLRQHU´KHH[SORUHVWKH teacher–student relationship in one of the design professionsarchitecture, which includes urban design, regional planning, and town planning. Design processes for infrastructures in general and information infrastructures are met with many of the same challenges that the architect meets when approaching the location for a new development. In search of more information about exploring the installed base, I was therefore inspired by Schön DQG KLV ³UHÀHFWLYH FRQYHUVDWLRQ ZLWK WKH situation,” which views the process using three dimensions: the design, the implications discovered in the process, and changing stance regarding the situation, going from unit level to considering the whole. This chapter uses the framework as described by Schön for architecture to create a ³UHÀHFWLYHFRQYHUVDWLRQ´LQUHODWLRQWREURDGEDQG infrastructure, and in particular the design dimensions, which may offer a framework for information infrastructure in general. When an architect is designing a new construction, a house, a campus, the utilization of a plot, WKHUHDUHDQXPEHURIH[LVWLQJIDFWRUVLQÀXHQFLQJ the designthe installed base. Any decision made LQWKHGHVLJQSKDVHPD\LQÀXHQFHODWHUGHFLVLRQV ,Q SHUIRUPLQJ WKLV UHÀHFWLYH FRQYHUVDWLRQ WKH designer takes different stances in relation to the object and location. Hanseth and Lyytinen (2006) claim that we need to draw upon and expand the existing installed base. However, they do not offer DIUDPHZRUNIRUDVVHVVLQJDQGUHÀHFWLQJXSRQWKH LQVWDOOHGEDVHDQGLWVLQÀXHQFHRQWKHGHVLJQRI
182
the new infrastructure. For the design dimension Schön (1991) describes 12 clusters of elements describing the features, relations, and actions related to the architectural design and the norms to evaluate problems, consequences, and implications. The descriptive clusters focus on the sitethat is, WKHIDFWRUVWKDWLQÀXHQFHWKHGHYHORSPHQWWKDWDUH VSHFL¿FWRWKHORFDWLRQDQGKRZFRPSRQHQWVPD\ be organized in relation to each other. They also focus on the use and function of the construction, including the elements that need to be in place. The descriptive clusters are part of a process of making design decisions, but also part of a process of “zooming in and out” from a unit level to the whole as part of a process of considering the design implications from a number of shifting viewpoints. In the next section we will give a description of the case material and then we will use the cases to explore the descriptive clusters as part of the process of cultivation of the installed base.
THE DEVELOPMENT OF BROADBAND IN SOGN OG FJORDANE COUNTY Sogn og Fjordane County is located in the western part of Norway and covers an area of 18,634 square kilometers, with an average of about six persons per square kilometer. The county is characterized by a rugged physical environment which includes glaciers, mountains, and fjords, which separate populated areas. At the time of the deregulation, access to broadband infrastructure in Sogn og Fjordane County was very limited. The regional university college, research institution, and hospital, all had access to broadband capacity. There were also D IHZ HQWHUSULVHV WKDW OHDVHG OLQHV IRU VSHFL¿F purposes and used ISDN for e-mail and Internet access. However, in other more urban areas, both local and national providers offered a number of broadband alternatives. The industries in Sogn og Fjordane are to a large extent based on natural resources and are often located close to these resources. As a result the
Improving Broadband Access in Rural Areas
Table 1. Overview of local broadband initiatives and suppliers in Sogn og Fjordane as of June 2006. The two initiatives marked with * are chosen for consideration in this chapter. Initiative
Aurland &
BB
Owners/ Involve-
Since
ment
2003
Ltd owned by
Existing cable-TV &
2 Municipality + 2
ADSL & wireless &
el. suppliers
¿EHU
Ltd company
Wireless & ADSL
Lærdal Breiband
Datapart
Enivest
2002
2000
Infrastructure
Users/Use
Operation
Coverage Municipality
Sogn Breiband
Internal opera-
Aurland & Lærdal
tion
Private & busi-
Internal opera-
nesses
tion
Ltd owned by 4
ADSL & wireless &
Private & busi-
Internal opera-
municipalities and 3
¿EHU
nesses
tion
ADSL & wireless
Private & busi-
Internal opera-
Luster & Sogndal
Nine municipalities
el. suppliers Firdanett *
2000
nesses
tion
Jølster Breiband
2002
Ltd company
Wireless
Private & busi-
Internal opera-
nesses
tion
Kapasitets-
2000
Ltd Company
Fiber
Businesses &
Outsourced
public sector
operation
Luster
2005
Ltd company
Services on top of
Private & busi-
Internal opera-
Aurland & Lærdal
ALB
nesses
tion
Wireless & ADSL
Private & busi-
Internal opera-
nesses
tion
businesses
Internal opera-
Ltd company owned by el. supplier
laget * Sogn Breiband
SicoData
ViaVest
Vik IT-Partnar
?
2000
2002
Ltd company
Ltd joint venture
Wireless & leased
between companies
lines
Ltd company
Wireless
Gloppen
Mainly Jølster
Sogndal, Leikanger,
Stryn
Stryn
tion Private & busi-
Internal opera-
Vik Balestrand
nesses
tion
in coop. w el. supplier
Zet.no
1998
Ltd company
:LUHOHVV ¿EHU
Private & businesses
tion
Årdalsnett
2001
Ltd company
Existing cable-TV &
Private & busi-
Internal opera-
12 local owners
wireless
nesses
tion
population is spread over large parts of the region with only a few agglomerated areas. For commercial providers the cost of cabling the whole region was considered too high and the demand was too small. The providers are willing to deliver access to individual organizations, but at a high cost. To rent a 2 Mbit/s connection is 40 times as expensive per bit as the rent for 622 Mbit/s (Teleplan, 2003b); accordingly it is a discount based on quantum. Lo-
Internal opera-
5 towns
Årdal & Lærdal
cal actors can reduce cost by collaborating if they share a high-capacity connection and develop an access network locally. To meet the challenge a total of 12 regional and local initiatives were undertaken which utilized existing relational and organizational networks and established new ones where it was needed. These initiatives ensure that large areas of the region now have access to broadband infrastructure.
183
Improving Broadband Access in Rural Areas
Several of the initiatives were developed as part of a public-private partnership and have received funding from Høykom and from the public sector. Nine of these initiatives have been included in a study carried out by Norsk Telecom (2004). Table 1 lists all the initiatives with a short description of their main characteristics. As a result of these initiatives, the region has a broadband DSL coverage of 79.36% (Telenor_Jara_AS, 2006). However, DPRQJWKHUHJLRQV¶PXQLFLSDOLWLHVWKHFRYHUDJH varies from 34.51% to 97.16%. Figure 1 shows the development of broadband access from 2002 to 2006 in Sogn og Fjordane County. Each of these initiatives covers only a part of Sogn og Fjordane, and all are the result of local initiatives. The company Enivest is the only entity that may be considered to be a provider with regional coverage. The national provider Telenor is currently present in all 26 municipalities in the region, providing DSL service (BBF, 2005). In addition to the local initiatives, there is a regional initiative called Broadband-forum Sogn og Fjordane (BBF); this is a network organization initiated in 2001. BBF focuses on the establishment, development, and utilization of broadband infrastructure in local communities. BBF initiates and participates in activities focusing on disseminating information by means of meetings and conferences. It maintains an overview of providers in each municipality, and monitors projects by focusing on utilization and service delivery. By participating in these activities, BBF supports the demand side in achieving the aim of having the best possible coverage in the region. The following description focuses on two of the local development initiatives in the municipalities Sogndal–Leikanger and Gloppen (marked with * in Table 1). These two localities had different starting points (different installed bases) and also went through different processes subsequently (the differences are further outlined in Table 3).
Sogndal–Leikanger Municipalities Sogndal Municipality is one of the main agglomerations in the region, with about 6,600 inhabitants. It has a diverse business structure, and is also the
184
location of the regional university college and a small research institute. These two institutions both had access to broadband prior to deregulation. As the university college developed, it grew to occupy various buildings throughout Sogndal; over the years a broadband infrastructure has been built which connects these various buildings. This infrastructure originally leased lines from the national telecom provider Telenor, but with the liberalization of the market, the university college and the municipality purchased the available infrastructure. Below is a chronological list of the main events involved in this development process in Sogndal–Leikanger: •
•
•
•
1996: Today: two large public organizations, the county municipality,3 and the county governor4 participated in international projects together and separately, in the development of national e-government services. 1999: Increasing need for better, faster access to the Internet. The need for high-capacity access to provide services to regional and QDWLRQDOXVHUV6HUYLFHVJHQHUDWHKHDY\WUDI¿F ERWKZLWKUHJDUGWRODUJH¿OHVDQGWKHQXPEHU of requests, for example, the portal norge.no, which has approximately 55,000 page views per day.5 The two public organizations, along with regional partners, saw an opportunity WRDFKLHYHVXI¿FLHQWDFFHVVDWORZHUFRVWE\ aggregating demand. 1999: Applied for funding from the national program Høykom. The project started in -XQH DQG ZDV ¿QDOL]HG LQ 'HFHPEHU 2001. The focus was on developing the local infrastructure to facilitate service delivery. 1999-2001: A network structure was set up to connect the two public organizations located in Leikanger to the national trunk network in Sogndal with a 2 Mbit/s and a 4 Mbit/s line. The 20-kilometer line to Leikanger from Sogndal was established by leasing lines from the national provider. Subsequently they UHQWHG XQWHUPLQDWHG ¿EHU WR REWDLQ PRUH control over capacity and both endpoints. In Sogndal there was a 4 Mbit/s line to the
Figure 1. The map on the left shows access points for broadband coverage in 2002, and the one on the right shows the coverage in 2006 (Vorland, 2006).
Improving Broadband Access in Rural Areas
185
Improving Broadband Access in Rural Areas
•
•
•
•
•
•
university college, and the connection to the national trunk network was 4 Mbit/s. 1999-2001: In Leikanger a backbone network was developed in two phases to include 16 QRGHVORFDOO\FRQVLVWLQJRI¿EHUDQGUDGLR links 1999-2001: The county municipality extended its network to include the administration of nine municipalities in the region. The virtual private network (VPN) had a capacity of 2 0ELWVDQGDFRPPRQ¿UHZDOO 1999-2001: In Sogndal two R&D organizations were connected with a 4 Mbit/s line. Eight businesses cooperated in a joint effort to create a backbone network consisting of ¿EHUDQGUDGLRWRH[WHQGWKHFRPPRQLQIUDstructure to other potential users. January 2001: A joint venture company µ.DSDVLWHWVODJHWL6RJQ$6¶ ./ ZDVVHWXS Its aim was to continue operation and development. Ten local companies own equal shares in KL. The combination of owners ensured access to the installed base and to knowledge about how to operate the infrastructure. Initially KL was operated using employees from the participating owner organizations. Later the operation was out-sourced to one of the organizations. 2002: The access network was extended to the neighboring municipality Luster. A QHZ¿EHUOLQHZDVQHHGHG7KLVZDVPDGH possible by a joint effort between two local energy companies and the main national WHOHFRPSURYLGHUXWLOL]LQJH[LVWLQJ¿EHUDQG DGGLQJDUHODWLYHO\VKRUWQHZ¿EHUFDEOH$V a result, Luster municipality acquired commercial broadband at least one year earlier than planned. Summer 2003: $QHZSLHFHRI¿EHUFDEOH infrastructure was installed to connect Kaupanger; this infrastructure is owned by a local energy company
7KHNH\WRWKHGHYHORSPHQW¶VVXFFHVVZDVWKH DELOLW\ WR PDNH XVH RI H[LVWLQJ ¿EHU LQIUDVWUXFture owned by the municipality and the college. It was also possible to keep the operational costs
186
low by using existing employees from the owner organizations. The collaboration between KL and the owners of the infrastructure was a win-win situation. Compared to traditional top-down diffusion of infrastructure development, this approach worked well: it was instrumental in keeping cost to a minimum while providing broadband services in a very short period of time. Under this setup KL provides the backbone infrastructure; the company does not provide any other services. Provision of other services has been left to the users of the infrastructure. Organizations that wish to connect to the infrastructure may do so at their own cost or through a radio line delivered by one of the service providers in the network. KL negotiates access to the national infrastructure at regular intervals. As the use of the infrastructure increased, higher capacity access to the national trunk network was needed. Currently the capacity is 20 Mbit/s. The experience so far is that working jointly in this way has led to cheaper and faster access for all of the companies connected through KL. Summing up this initiative, the main actors and WKH¿UVWDGDSWHUVLQWKH./LQLWLDWLYHZHUHIURP the regional public sector, together with a limited number of enterprises and the two local research and educational institutions. The strategy chosen was ad hoc. The main focus was on the needs of WKH¿UVWDGDSWHUVDQGWKHQHHGVRIWKH¿UVWDGDSWHUV and the aim of less expensive broadband capacity were the main motivating factors behind the well-aligned network. There was prior knowledge about operating broadband infrastructure, and the organizations involved had a large knowledge base to draw on. In addition, the infrastructure was cultivated based on the utilization of existing LQIUDVWUXFWXUHXQXVHG¿EHURSWLFOLQNVIURPIRUPHU LQVWDOODWLRQVDQGXQWHUPLQDWHG¿EHUFDEOHVZKLFK were supplemented with some new components to connect the new broadband access network. KL has no employees; initially the operation of KL was based on human resources drawn from the participating organizations, while consultants stepped in to manage administrative and operational services later on. Consequently, the focus remained on the QHHGVRIWKH¿UVWDGDSWHUVDQGWKHHFRQRPLFDVSHFWV of the operation were not a problem.
Improving Broadband Access in Rural Areas
Gloppen Municipality
•
Gloppen is a municipality with a population of 6,000. The development of the local broadband initiative Firdanett is the result of several interacting processes. Below is a chronological list of the main events involved in this development process in Gloppen: •
•
•
•
•
1996: The chamber of commerce was revitalized and with it a strategy for revitalizing the economic development of the community (Gloppen-Næringsorganisasjon, 2001). This started a process that led to a consensus about the current situation and the future challenges. The chamber of commerce worked in cooperation with the municipality. 1996: $*ORSSHQEDVHG,&7¿UPVHWXSD/$1 to provide services to a customer located in RQHEXLOGLQJ7KH¿UPKDGUHFHLYHGUHTXHVWV for an extended service to customers not located in the same building. Due to certain legal regulations, this was not possible. 1998: The telecommunications liberalization policy was implemented. The regulations were changed and the LAN was expanded. The LAN and its services formed the basis for the development of the local broadband infrastructure. 1998: µ*ORSSHQ,7QHWW¶ZDVHVWDEOLVKHGDVD joint venture between the chamber of commerce, the municipality, and the local energy company (Gloppen-kommune, 2000). 1999: An application for funding was made to the Høykom program. The project started LQ-XQH DQGZDV¿QLVKHGLQ'HFHPEHU 1999. As part of the project, the LAN was extended to become a full broadband access network. The project was important in helping to expand the network to provide better coverage throughout the municipalLW\ 7KH GHYHORSPHQW RI µ*ORSSHQ ,7 QHWW¶ was considered to be an exemplary model (Samferdsledepartementet, 2002) for publicprivate partnerships for the development of broadband infrastructure.
•
•
1999: After about one year of operation, it became clear that there was a disagreement among the owners with regard to priorities. As a result services connected with the infrastructure, such as backup, server space, and so forth, were transferred back to the ORFDO,&7¿UPZKLOHWKHUHPDLQLQJRZQHUV UH¿QDQFHGWKHFRPSDQ\ ZKLFKDOVRFKDQJHG its name to Firdanett at that time). The local energy company was the majority owner of Firdanett, while the chamber of commerce and the municipality were minority owners. Firdanett had two employees, one technician and one salesman. 2003: Firdanett provided broadband access to 55 companies and 85 households. At that time the national provider Telenor did not offer broadband to households and small businesses in Sandane. March 2004: Firdanett merged with another broadband company, ENIVEST, and as a result of another merger process the two companies came under the same ownership. ENIVEST had regional coverage and expanded its access services based on demand and public sector projects, which were important for development. Coverage in Gloppen was increased further after the merger.
The local broadband infrastructure consists of a mixture of the early LAN, supplemented by ¿EHUFDEOHVFRQQHFWLQJWKHPDLQEXLOGLQJVLQWKH central parts of Sandane and radio transmitters/ receivers to reach communities elsewhere in the municipalities. Firdanett has access to national infrastructure through a 6 Mbit/s licensed radio link to Hunvikfjellet that is about 22 kilometers (straight-line distance) which connects the local network to the national trunk network. In Gloppen the LAN was the only technical installed base that could be cultivated in the same way as in the KL FDVH7KHFRVWRISXWWLQJD¿EHURSWLFLQIUDVWUXFWXUH in place was high, and there was no short-term payback from the investment. Firdanett offers broadband Internet access and ,3WHOHSKRQ\DQGWKH,&7¿UPDQGRWKHUVHUYLFH
187
Improving Broadband Access in Rural Areas
Table 2. Mapping of Schön’s (1991) descriptive clusters with the clusters found useful for rural broadband development Descriptive cluster w/ explanation according to Schön (1991) 9) Character - the kind of, mode of, or style of construction 1) Use and function - what is the intended use or function of the construction 2) Siting - Features of the site, factors that influence the development
Cluster broadband development
Definition of cluster for broadband development
Character
Kind of infrastructure
Use and function
Description of intended function and use of the infrastructure
Siting
10) Precedent - describing a model for the construction; 3) Building elements and components - what needs to be in place and can be used in the construction; 6) Structure/Technology and processes used - the material needed in the construction and an understanding of the process 7) Scale - Magnitudes of elements in relation to one another 4) Organisation of space and the relation of spaces to one another 5) Form - Shape, geometry, organisation of space and paths through space 8) Cost of development;
Precedent
Descriptions of the site(s) to be served by the infrastructure. If relevant description of neighbouring infrastructure, of geography or structural features that can influence the infrastructure and of organisations or actors that can play a role in bootstrapping the infrastructure. Models for the implementation, reference types. List of possible components. Design principles for development.
11) Representation - Language and notions for describing the construction 12) Explanation - context of interaction.
188
Technology, building elements and components
Scale
Form/ Organisation of space
Cost of development: Representation:
Explanation
Reach and capacity of infrastructure, EDQGZLGWKFRYHUDJH, coverage geography The shape of the infrastructure, distribution, redundancy, M:M or 1:M How and were are different places that will be using the infrastructure located, how can they be connected? Budget available for the development and for operations. Representation used to describe infrastructure Protocol for frequency and involvement of users in the development process.
Improving Broadband Access in Rural Areas
providers offer their ICT services to the businesses connected to the network along with other service providers such as accountants. Summing up this initiative, the main actors in the Gloppen initiative were local enterprises and the local public sector. The strategy chosen was ad hoc and the main focus was on the needs of the business community. The main motivating factor was access to broadband infrastructure for a relatively low number of organizations; as a re-
sult there was an immediate need to increase the number of users quickly which made the developPHQWPRUHGLI¿FXOW$VDUHVXOWWKHORFDOQHWZRUN was not well aligned at the start; however, after the restructuring it became a well-functioning initiative. The initiative had IT competence but little experience in operating this kind of network, and there was no common knowledge base. BeFDXVHLWZDVQHFHVVDU\WRKDYHVXI¿FLHQWLQFRPH to employ stafftwo persons were employed to
7DEOH'HVFULSWLRQRIWKHWZRFDVHVXVLQJPRGL¿HGFOXVWHUV Descriptive cluster 1) Character of information infrastructure: 2) Use and function
3) Siting:
4) Precedent:
Firdanett
Kapasitetslaget
Internet access, service platform
Internet access
Phase 1. Access to Internet & service provision public sector and businesses in centre. Phase 2. extending to also cover private homes and businesses outside centre Function: 1)Provision of services locally on the network, 2) Gateway to services on the Internet Municipality of Gloppen, Phase 1. Centre of Sandane, public sector + businesses. Phase 2. Byrkjelo, Reed, Austrheim, Vereide, Rygg and Hyen, schools + general public. Installed base: LAN inside buildings, Telephone lines, Organisational: Municipality, energy company, chamber of commerce and some businesses interested in leading the development. Knowledge: Had sufficient operational skills, but very dependent on individuals and one company. Received external help to get funding. Needed to employ service personnel and a manager to operate the network. Salaries were the main costs of the company and put pressure on compan\¶s economy. Internet
Access to Internet to deliver services to external customers. Cover needs of initiators. Function: Gateway to services on the Internet and provision of services to users on the Internet. . Negotiating power increase/lower total cost? Municipalities Sogndal and Leikanger, specific organisations. Later also extending to other neighbouring communities, Luster, Kaupanger. Installed base: Municipality and university college own fibre network covering much of Sogndal Centre. Access point for national trunk network in Sogndal. Knowledge: Partners in the development had across organisations skills and knowledge about running and extending the networks. Did not need to employ personnel to manage the network - was in the start-up phase operated based on “donation” of time from employer to KL. This kept the operation costs very low.
Internet
189
Improving Broadband Access in Rural Areas
Table 3 (cont.) Descriptive cluster 5) Technology, building elements and components
6) Scale,
Firdanett
Kapasitetslaget
Licensed Radio link to Eid to connect to national trunk network. Optical fibre in Sandane centre, licensed radio link and wireless broadband to the subscriber to connect more of the surrounding and remoter communities.
Fibre owned by university college available, leased lines to Leikanger connecting county municipality (Fylkeskommunen) and county governor (Fylkesmannen) to network. Leikanger installed trunk network in 2 phases to include 16 nodes locally; consists of both fibre and radio link. Use gateways to connect fragmented infrastructures. Extending to Luster and Kaupanger in collaboration with national provider, and two local electricity companies. Capacity on the link to national trunk network, determining the cost of KL services. Dependent on number of connected users, and type of data. Transfer of music files requires high capacity, web search requires lower capacity. Access to trunk network needs to be scalable to fit number of users and the demand of the users connected to the infrastructure.
Reach: Phase 1. Centre of Sandane, businesses & public sector. Phase 2: Small communities in municipality Gloppen, businesses and private homes. Bandwidth (initial): 10/100 Mb to trunk network. In the network aiming for 2-11 Mbit capacity using radio links. 7) Form/ Centre of Gloppen Sandane is Organisation of located about 20 km from a space location to connect to national trunk network. 2 possible trunk network providers. Star shaped infrastructure, no redundancy. Unlicensed radio and fibre are alternatives for distribution to end user. 8) Cost of Cost of putting down fibre and development setting up radio link and radio. Cost of serving an IP phone switch. Funded among others through 2 projects. Phase 1: 1.8 million NOK. Phase 2: 1.66 million NOK 9) RepresentaNo particular representation tion chosen 10) Explanation Aim of development to ensure broadband access to Internet for public sector, businesses and individuals.
190
Sogndal is on the national trunk network. Leikanger is 20+ km from Sogndal. Luster 10-15 km, Kaupanger 10 km. Star shaped infrastructure - no redundancy
Cost of connecting fragments, cost of gateways. Project phase: 2.9 Million NOK
No particular representation chosen Aim of development to ensure broadband access to Internet for public sector, businesses and individuals.
Improving Broadband Access in Rural Areas
7DEOH5HGH¿QLWLRQRIGHVFULSWLYHFOXVWHUVIRUWKHLQIUDVWUXFWXUHGRPDLQ PRGL¿HGIURP6FK|Q Cluster
Important Questions to Ask
Character
Kind of infrastructure. What kind of infrastructure is intended?
Use and function
Description of intended function and use of the infrastructure. What is the intended function? What is the intended use? Who are the intended users? (public sector/ business/ private)
Siting
Descriptions of the site(s) to be served by the infrastructure. If relevant, description of neighboring LQIUDVWUXFWXUHRIJHRJUDSK\RUVWUXFWXUDOIHDWXUHVWKDWFDQLQÀXHQFHWKHLQIUDVWUXFWXUHDQGRIRUJDnizations or actors that can play a role in bootstrapping the infrastructure. Where is the installation? Location of sites to connect? What is the installed base? What organizational infrastructure can be XWLOL]HG",VWKHUHVXI¿FLHQWNQRZOHGJHDQGFRPSHWHQFHWRRSHUDWHWKHLQIUDVWUXFWXUH":KRKDVWKLV NQRZOHGJHDQGFRPSHWHQFH":KRDUHWKH¿UVWDGDSWHUV"$UHWKHUHRWKHUSRWHQWLDO¿UVWDGDSWHUVWKDWDUH instrumental to be included? Which facilities need to be connected to the new infrastructure? What are potential growth paths? Who are potential second adapters?
Precedent
Models for the implementation, reference types. Has anyone done something similar before? Is it possible to learn from them and their choices? Study trip?
Technology, building elements, and
List of possible components. Design principles for development. Which components is it possible to
components
XWLOL]H"+DUGZDUH ¿EHUOLFHQVHGDQGXQOLFHQVHGUDGLRFRSSHUSKRQHOLQHV'6/ 6WDQGDUGVVZLWFKHV organization. Structure and basic choice of technology and development process.
Scale
5HDFKDQGFDSDFLW\RILQIUDVWUXFWXUHEDQGZLGWKFRYHUDJHFRYHUDJHJHRJUDSK\What bandwidth is needed? What is the geographical coverage? What is the percent coverage?
Form/organization of space
The shape of the infrastructure, distribution, redundancy, M:M or 1:M. How and where are different places that will be using the infrastructure located, how can they be connected? Possible designs taking technology, building elements, components, scale, and costs into consideration. Should there be redundancy in the system? If this is the case, how and where are the critical applications/users?
Cost of development
Budget available for the development and for operations. Cost of alternatives? Can the development be phased? What human resources are needed? Is there access to these resources? How can the project EHIXQGHG SXEOLFIXQGLQJ¿UVWDGDSWHUVGHSHQGHQWRQFULWLFDOPDVVRIXVHUV "%XGJHWDYDLODEOHIRUWKH development and for operations?
Representation
Representation used to describe infrastructure.
Explanation
Protocol for frequency and involvement of users in the development process.
operate and market the infrastructurethis put pressure on the organization, as the focus became income generation as opposed to the needs of the ¿UVWDGDSWHUV
CAN DESCRIPTIVE CLUSTERS BE USED TO EXPLORE THE INSTALLED BASE? Both the cases described above are successful developments, but the conditions for the developments were quite different. Out of the 12 initiatives in the
191
Improving Broadband Access in Rural Areas
region, these two were chosen in order to represent some of the variety that was present. Common to all is the bottom-up approach, including the approach of making development as simple as possible using available infrastructure and low-cost additions to ¿OOWKHQHHGVRIWKH¿UVWDGDSWHUV,QHIIHFWWKHVH developments create a larger total market for suppliers than that constituted by the individual needs of the actorsthat is, they aggregated the demand. This involved a public-private partnership sharing costs between regional and local public sector and local businesses. Transaction costs are also lower for suppliers, who interact with the initiative as one large customer rather than many small customers. The variety in the development process illusWUDWHVWKHQHHGWRWDNHVSHFL¿FORFDOIDFWRUVLQWR DFFRXQW WKLVLQFOXGHVWKHQHHGVRIWKHSRWHQWLDO¿UVW adapters, the existing infrastructure (installed base) which is available to be built upon (both technical and organizational), and the local organization(s) which are part of the installed base. This is where I VXJJHVWWKDW6FK|Q¶V ³UHÀHFWLYHFRQYHUVDWLRQ with the situation” may be useful. In Table 2, the FOXVWHUVGH¿QHGE\6FK|QDUHOLVWHGDQGUHGH¿QHGIRUWKHORFDOEURDGEDQGFRQWH[W0RVWRIWKH descriptive clusters are also useful when working with rural broadband infrastructure, but some seem not to be applicable or to be variations of one of the other clusters. The original 12 descriptive clusters (Schön, 1991) have been examined, combined, and UHGH¿QHGLQWRGHVFULSWLYHFOXVWHUVZKLFKFRYHU areas relevant to the local broadband context. There are small variations between the original cluster 3 µEXLOGLQJHOHPHQWV¶DQGFOXVWHUµVWUXFWXUHWHFKQRORJ\¶DQGDOVREHWZHHQFOXVWHUµRUJDQL]DWLRQRI VSDFH¶DQGFOXVWHUµIRUP ¶WKHUHIRUHWKHSURSRVDO LVWRXVHFOXVWHUVDVRSSRVHGWRWKHGH¿QHG by Schön (1991). When developing rural broadband infrastructures, what is the installed base and what are the input factors to the bottom-up cultivation process? In Table 3 the two local initiatives are described XVLQJWKHUHGH¿QHGFOXVWHUV7KLVLVGRQHWRHQable the extraction of important factors relating to the cultivation process and to identify rhetorical µZKDWLI¶TXHVWLRQVWDNLQJWKHLQVWDOOHGEDVHDQG
192
the features of the location into the consideration in the cultivation process. The descriptions shed light on the installed base, the choices made, and the components used as part of the cultivation process.
CONCLUSION AND PRACTICAL IMPLICATIONS Using a bottom-up strategy is more complex than a top-down rollout, because there are many actors and components to consider. Hanseth and /\\WLQHQ¶V IUDPHZRUNDQG6FK|Q¶V descriptive clusters are intended to serve as a framework for both understanding and guiding bottom-up development. In the description of the cases, it has been used in retrospect to examine the development. Looking at the information brought forward in the description (Table 3) of the cases, it is possible to extract a number of questions that are useful to consider when starting a bottom-up process. In all of the cultivation processes, there will be uncertainties. To reduce the uncertainties it is necessary to know as much as possible about local conditions and potential actors that the project should be aligned with, as well as the installed base; which components need to be added to the installed base as part of the cultivation must also be carefully considered. During the cultivation process the descriptive clusters may be used to arrive at a description of the installed base and the considerations needed to cultivate it. To identify the opportunity space of the local broadband development, the descriptive clusters become a tool, a set of rhetorical questions for uncovering the installed base and the choices that need to be consideredthat is, the “what if” and “if…then…” considerations that need to be taken as part of the cultivation process. In Table 4 the local broadband clusters are analyzed to produce a set of questions addressing key contributing factors to a bottom-up cultivation of broadband infrastructure in rural areas. In this way the descriptive clusters may be used to uncover the starting point of a development. The questions and answers describe important inputs
Improving Broadband Access in Rural Areas
into the development process, and in using them it is possible to gain an overview of what kinds of resources are available, of the features installed base, and of the kinds of decisions that need to be taken. In addition, the use of a tool like the descriptive clusters also puts the focus on the local context and culture in the community. With a bottom-up strategy questions have to be answered in relation to the local context. When changing from one location to another, there are always things to be learned, but the bottom-up strategy makes it impossible to make blueprint implementations. What can be transferred are rather the relevant methods and experience of dealing with a type of problem situation.
REFERENCES Baskerville, R., & Wood-Harper, A.T. (1996). A critical perspective on action researcha method for information systems research. Journal of Information Technology, 11, 235-246. BBF. (2005). Breibandsstatus Sogn og Fjordane 22.01.05. Retrieved March 31, 2005, from http:// www.it-forum.no/pub/Pubdok/Breibandsstatus% 20Sogn%20og%20Fjordane200105.pdf Cole, R.E. (1991). Participant observer research. In W.F. Whyte (Ed.), Participatory action research. Sage. Dahlbom, B., & Janlert, L.E. (1996). Computer future. Manuscript as referenced in: Hanseth, O. (2002). From systems and tools to networks and infrastructures—from design to cultivation. Towards a theory of ICT solutions and its design methodology implications. Retrieved November 29, 2003, from http://www.if.uio.no/~oleha/Publications/ib_ISR_3rd_resubm2.html Gloppen-kommune. (2000). Gloppen-modellen—utbygging av digitalt allemannsnett i distrikta (Høykom rapport 1019/240). Retrieved October 29, 2004, from http://www. hoykom.no/hoykom/web_hoykom_ prosjekter.nsf/677cffd27235dae8c1256d7a0041af6f/
efcfd7951176e01d41256823003ce603/$FILE/ P1019%20Gloppen.doc Gloppen-Næringsorganisasjon. (2001). Vedtekter Gloppen Næringsorganisasjon 27.3.2001. Retrieved October 29, 2002, from http://www.gloppennaerings.org/vedtekter.doc Granovetter, M., & McGuire, P. (1998). The making of an industry: Electricity in the United States. In M. Callon (Ed.), The laws of the markets (pp. 147-173). Oxford: Blackwell. Grubesic, T.H. (2003). Inequities in the broadband revolution. Annals of Regional Science, (37), 263289. Grubesic, T.H., & Murray, A.T. (2002). Constructing the divide: Spatial disparities in broadband access. Papers in Regional Science, (81), 197-221. Hanseth, O. (2000). The economics of standards. In C. Ciborra (Ed.), From control to drift (pp. 56-70). Oxford, UK: Oxford University Press. Hanseth, O. (2002). From systems and tools to networks and infrastructures—from design to cultivation. Towards a theory of ICT solutions and its design methodology implications. Retrieved November 29, 2003, from http://www.if.uio.no/~oleha/ Publications/ib_ISR_3rd_resubm2.html Hanseth, O., & Lyytinen, K. (2006). Theorizing about the design of Information infrastructures: Design kernel theories and principles. Retrieved -XO\IURPKWWS ZZZL¿XLRQRaROHKD 3XEOLFDWLRQV,65LQIUDVWUXFWXUH¿QDOSGI Molka-Danielsen, J., Ohren, O., & Jæger, B. (2003, November). Broadband infrastructure in the County of Møre and Romsdal and rural regional development. Proceedings of the Annual NOKOBIT 2003 Conference, Oslo, Norway. Norsk_Telecom. (2004). Norske Bredåndsaktører (Rapport 2. Mars 2004 for NHD). Retrieved March 31, 2005, from http://www.enorge.org/File$UFKLYH$NW&%UUDSSRUW¿QDO pdf OECD. (2004). The development of broadband access in rural and remote areas. Retrieved
193
Improving Broadband Access in Rural Areas
September 10, 2004, from http://www.oecd.org/ dataoecd/38/40/31718094.pdf
IURP KWWS ZZZRGLQGHSQR¿ODUNLY2IIHQWOLJHBDYJLIWHUB¿QDOSGI
Post_og _teletilsynet. (2006a). Bredbåndsdekning—tekniske muligheter og begrensninger. Retrieved September 10, 2006, from http://www. npt.no/iKnowBase/Content/bredbandsdekning. pdf?documentID=47302
Vorland, A. (2006). Status og behov breibandutbygging i Sogn og Fjordane. Retrieved September 28, 2006, from http://www.it-forum.no/konferanse/pdf/BBF.pdf
Post_og_teletilsynet. (2006b). Det norske telemarkedet 1. halvår 2006.
Walsham, G. (1993). Interpreting information systems in organizations. New York: John Wiley & Sons.
Samferdsledepartementet. (2002). Breiband i kommunane—Kva er breiband, kvifor er det nyttig og korleis kan det nyttast? Retrieved October 29, 2004, from http://www.dep.no/sd/norsk/publ/ veiledninger/028021-120002/dok-nn.html
Wikipedia. (2006). Infrastructure. Retrieved March 29, 2006, from http://en.wikipedia.org/ wiki/Infrastructure
Schön, D.A. (1991). 7KH UHÀHFWLYH SUDFWLWLRQHU How professionals think in action. Aldershot: Ashgate/Arena.
KEY TERMS
Shapiro, C., & Varian, H.R. (1999). Information rules: A strategic guide to the network economy. Boston: Harvard Business School Press. Skogseid, I., & Hanseth, O. (2005, May). Local actors build broadband infrastructure. Proceedings of the European Conference of Information Systems 2005, Regensburg, Germany. 6WDQWRQ/- )DFWRUVLQÀXHQFLQJWKHDGRStion of residential broadband connections to the Internet. Proceedings of the 37th Hawaii International Conference on System Sciences, Hawaii. Star, S.L., & Ruhleder, K. (1996). Steps to an ecology of infrastructure. Design and access for large information spaces. Information System Research, 7(1), 111-134. Telenor_Jara_AS. (2006). Retrieved June 2, 2006, from http://www.jara.no/produktlosninger/bredbaand/priser_avtaler.html# Teleplan. (2003a). Bredbånd—Dekning og tilknytning. Retrieved March 31, 2005, from http://www. enorge.org/modules/module_111/news_item_view. asp?iNewsId=2369&iCategoryId=153 Teleplan. (2003b). Kartlegging av bruk av offentlige vederlag for fremføring. Retrieved March 31, 2005,
194
Broadband: High-speed data transmission. 7KHUHLVQRFOHDUGH¿QLWLRQRIZKDWFRQVWLWXWHVD EURDGEDQG WUDQVPLVVLRQ 6RPH GH¿QLWLRQV RQO\ consider speeds above 2 Mbit/s, while in this chapter speeds above 384 kbit/s are considered as broadband, while ISDN and modem transmissions IDOORXWVLGHWKHGH¿QLWLRQRIEURDGEDQG Cultivation: A strategy for developing technological systems. The term originates from the sphere of agriculture and means to “to grow and cultivate plants and the domestication of animals.” Dahlbom and Janlert (1996) use the two terms “construction” and “cultivation” to describe two different processes to shape technology. “Construction” emphasizes the human ability to shape WKHZRUOGDFFRUGLQJWRVHWRIJRDOVµ&XOWLYDWLRQ¶ refers to the process of selecting and assembling components, but also to a process of growth or development in the material itself that has characWHULVWLFVWKDWLQÀXHQFHWKHGHYHORSPHQW Høykom: A Norwegian program to increase the demand for broadband services. The Høykom (www.hoykom.no) program is at present in its third period of operation. The program was initiated in 1999 for a period of three years (1999-2001). Its current period of operation lasts from 2005-2007. +¡\NRPKDVSURYLGHG¿QDQFLDOVXSSRUWIRUPRUH WKDQ SURMHFWV RQ D VKDUHG ¿QDQFLQJ EDVLV
Improving Broadband Access in Rural Areas
Høykom supports projects initiated by the public sector with up to 50% of total project costs, totalLQJPRUHWKDQ¼PLOOLRQDQGKDVOHGWRDWRWDO SURMHFWYROXPHRIPRUHWKDQ¼PLOOLRQ Infrastructure: A “set of interconnected structural elements that provide the framework for supporting the entire structure. It usually applies RQO\WRVWUXFWXUHVWKDWDUHDUWL¿FLDO´ :LNLSHGLD 2006). An infrastructure may be viewed as a network that connects an integrated and heterogeneous set of componentssuch as equipment, systems, applications, processes, and peoplewhile at the same time these components form rather stable structures with certain inertia. Infrastructures are often taken for granted; their existence is only noticed when a segment breaks down, for example, during a power outage. There are many types of infrastructures: • • •
Transportation, such as roads and public transportation Public utilities, such as telephone and electricity networks 3XEOLF VHUYLFHV VXFK DV ¿UH GHSDUWPHQWV police, and waste management National services such as defense, postal services, and the national bank
Information Infrastructure: An infrastructure that supports the information society: the equipment, systems, applications, support systems, and so forth that are needed for operating in the information society. The term was introduced with the National Information Infrastructure initiative launched by the Clinton Administration in 1994. In the Clinton-Gore initiative the Internet is described as an information infrastructure shared by the users.
ENDNOTES 1
2
3
4
5
I use “rural areas” to describe areas that are sparsely populated, where the inhabitants live in small towns, villages, farms, or isolated houses. This is used in contrast to “urban areas,” which I use to describe areas such as densely populated towns and cities. +RZHYHUWKHUHLVDÀXLGERXQGDU\EHWZHHQ the two, and some who reside in small towns or villages may regard these as urban areas, though the district on the whole may have rural characteristics. The Norwegian Høykom (www.hoykom. no) program is at present in its third period of operation. The program was initiated in 1999 for a period of three years (1999-2001). Its current period of operation lasts from +¡\NRPKDVSURYLGHG¿QDQFLDO support for more than 400 projects on a VKDUHG¿QDQFLQJ EDVLV +¡\NRP VXVWDLQV projects initiated by the public sector with up to 50% of total project costs, totaling more WKDQ ¼ PLOOLRQ ZKLFK KDV OHG WR D WRWDO SURMHFWYROXPHRIPRUHWKDQ¼PLOOLRQ Norwegian: Fylkeskommunen: Sogn og Fjordane County Municipalitypolitical and administrative regional organization. Norwegian: Fylkesmannen: The county governor in Sogn og Fjordanethe regional state representative. Calculated based on data accessed September 20, 2006, from http://www.norway.no/omnorgeno/statistikk/norgeno.asp
195
196
Chapter XIII
Metropolitan Broadband Networks:
Design and Implementation Aspects, and Business Models Antonios Alexiou Research Academic Computer Technology Institute and University of Patras, Greece Christos Bouras Research Academic Computer Technology Institute and University of Patras, Greece John Papagiannopoulos University of Aegean, Greece Dimitrios Primpas Research Academic Computer Technology Institute and University of Patras, Greece
ABSTRACT This chapter presents the design principles that cover the implementation of broadband infrastructure in the region of Western Greece, by examining all the necessary parameters that arise while implementing such a critical developmental project. The broadband infrastructure that is deployed is either based on RSWLFDO¿EHU RQELJPXQLFLSDOLWLHV RURQZLUHOHVVV\VWHPV 2)'0EDVHGDQG:L)LFHOOV )XUWKHUPRUH we present as two case studies all issues of the designing of the Metropolitan Area Network of Patras, the third largest city of Greece and the Wireless Access Network of Messatida. The major target of the broadband networks is to interconnect the buildings of the public sector in the city and also deploy infraVWUXFWXUH ¿EHUVRUZLUHOHVVV\VWHPV WKDWZLOOFUHDWHFRQGLWLRQVRIFRPSHWLWLRQLQSURYLGLQJERWKDFFHVV and content services to the advantage of the end consumer. The usage of the broadband infrastructure by service providers will be based on the open availability of the infrastructure in a cost-effective way. )LQDOO\ZHSUHVHQWWKHPDLQFKDUDFWHULVWLFVRIDSURSRVHGEXVLQHVVSODQWKDWHQVXUHV¿QDQFLDOYLDELOLW\ of the broadband infrastructure and guarantees the administration, growth, and exploitation of infrastructure. Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Metropolitan Broadband Networks
INTRODUCTION This chapter presents the design principles that cover the implementation of broadband infrastructure in the region of Western Greece, by examining all the necessary parameters and studying all the issues that arise while implementing such a critical developmental project. In particular, we present the main principles that should be followed while developing such metropolitan area networks. Regarding the design guidelines, in this chapter we cover issues such as architecture of the broadband network, topology selection, requirements of the passive and active equipment, and requirements of WKH¿EHUDQGGXFWLQJLQIUDVWUXFWXUH)XUWKHUPRUH we present as a case study critical issues regarding the design of the metropolitan area network of Patras, the third largest city of Greece. The main target of the MAN of Patras is to interconnect the buildings of the public sector in the city. The organizations that are going to be connected in the MAN are organizations of the sectors of public administration, education, health, culture, and so forth. The usage of the broadband infrastructure by service providers will be based on the open DYDLODELOLW\RIWKH¿EHURSWLFVLQIUDVWUXFWXUHLQD cost-effective way. Finally, we present the main characteristics of a proposed business plan that HQVXUHV ¿QDQFLDO YLDELOLW\ RI WKH EURDGEDQG LQfrastructure and guarantees the administration, growth, and exploitation of the infrastructure. Several related projects that implement neutral broadband infrastructure in cities are running across the world. For example, Ireland, Sweden, and New Zealand run such programs, where the local authorities design and fund the major part of the projects aiming to increase the broadband SHQHWUDWLRQZLWKEHQH¿WVWRWKHHQGXVHUV 8723,$ 2006; CityLink, 2006; Localret, 2006). This chapter is structured as follows. We next describe broadband infrastructure in Europe and in Greece, then present the general guidelines for the design of municipal broadband networks. Following this, the methodology of work regarding the designing of the MANs is offered, and the overall architecture and topology of the MAN of Patras LVEULHÀ\GHVFULEHG$GGLWLRQDOO\WKHVDPHVHF-
tion describes the architecture of smaller wireless networks that are implemented in smaller municiSDOLWLHVZKHUHWKHRSWLFDO0$1VDUHQRWSUR¿WDEOH and presents a typical case study that is Messatida municipality. A presentation follows of the main characteristics of a business plan that ensures the ¿QDQFLDO YLDELOLW\ RI EURDGEDQG LQIUDVWUXFWXUH Finally some concluding remarks and planned next steps are presented.
BROADBAND STATUS IN GREECE The importance of broadband infrastructure ZRUOGZLGHLVFRQ¿UPHGE\WKHDFWLYLWLHVRIFHUWDLQ advanced countries in order for the appropriate broadband infrastructure to be developed and adopted so as to contribute to economic growth and to tackle any possible cases of “technological exclusion” of citizens (Firth & Mellor, 2005). The importance of broadband networks for the GHYHORSPHQWRIDFRXQWU\PD\DOVREHFRQ¿UPHG by the intensiveness of the activities of many countries that set as their main strategic objective the implementation of such infrastructures. In addition, the development of such networks has also been adopted in the common European policy for the implementation of the Information Society. In eEurope 2005 as well as in i2010, broadband access is an important priority of the European Union (EC, 2002; Europa, 2004). In the current situation, the proper infrastructure in Greece is owned only by the former public telecommunications provider (OTE), while the alternative providers seem only to have plans in expanding their network infrastructure within the big cities of Greece. The business plans of the alternative telecommunications companies and network carriers do not include the expansion of their network throughout Greece, since they are afraid that non-urban areas do not appear to have DQ\EXVLQHVVLQWHUHVW%URDGEDQGDFFHVVDVGH¿QHG by the “Strategic Text on Broadband Access” of the relevant task force, requires the proper broadband infrastructure and the competition between the Internet service providers. Since broadband infrastructure is now being developed, the penetra-
197
Metropolitan Broadband Networks
tion of broadband usage has not been increased. $OWKRXJKWKHVH¿QGLQJVDUHSHVVLPLVWLFUHJDUGLQJ the growth of broadband, the action line of the operational program “Information Society” appear WREHDVLJQL¿FDQWKRSHIRUWKHQHDUIXWXUH Unfortunately, the broadband penetration level has been very low in Greece. Greece has been the last country among the EU of 15, and remained last, in providing broadband access among the 25 countries of the EU (see Figure 1) (Europa, 2006). This is caused by the fact that there is no investment by the private sector, therefore there is no competition in the broadband market in the region. The main issues that occur during this research for the current state of broadband in Western Greece could be divided in two main categories: D WKHFOLHQWV¶VLGHDVIDUDVEURDGEDQGGHPDQG LVFRQFHUQHG DQG E WKHSURYLGHUV¶VLGHDVIDUDV broadband supply is concerned. Regarding the end users, the major issues are the following: • •
High cost of broadband access regarding home usage Lack of broadband services that will take advantage of the infrastructure
Figure 1. Broadband penetration rates in EU25
198
The telecommunications companies have raised a number of issues that discourage them from investing in broadband infrastructure and services. Synoptically these issues are: •
•
Lack of the regulatory framework that adjusts DQGGH¿QHVWKHPDUNHWRIEURDGEDQGVHUYLFHV LQRUGHUWRHQVXUHPDUNHWDQGFRPSHWLWLRQ¶V functioning 'LI¿FXOWLHVLQGHYHORSLQJlocal loop unbundling (LLU) Lack of preparation in supporting the demand of broadband services by the former public telecommunications carrier (OTE)
Since the start of 2005, the xDSL penetration has been rising. The current access network (copper cables) is being updated by installing xDSL technologies such as ADSL mainly and HDSL. According to OTE, 10,929 ADSL ports were installed in the region of Western Greece by the end of 2005 (and 270,000 throughout the country). In addition, wireless hotspots have been installed in public places, such as ports, entertainment parks,
Metropolitan Broadband Networks
and so forth, through a project of the Information Society. Another measure that is expected to boost ADSL use in the major cities of the country is the provision of ADSL Internet access at a diminished cost to university students. This measure was announced by the Greek government in June 2005, and it will begin at the start of the next academic year. Taking into consideration the current state in the region of Western Greece where the telecomPXQLFDWLRQVPDUNHWVWLOOGHSHQGVRQWKHFRUH¿EHU optics networks of the dominant provider (OTE) and all the above mentioned issues, it can be said that it is important for the Greek government authorities to update the telecommunication law of the country. This could secure the viability of the public and private telecommunication companies, and simultaneously the viability of the communityowned networks. To this end, the Greek government authorities, taking into consideration the issues mentioned in this section and the variability of the Greek telecommunications market, passed the new telecommunications law. This law aims to secure the telecommunications market and competition in Greece, and as a consequence, better services will be provided to the citizens. The new law provides WKHWHOHFRPPXQLFDWLRQV¿UPVZLWKWKHFRQGLWLRQVWR HQDEOHWKHUHGXFWLRQRISULFHV7KHVSHFL¿FFODXVHV clearly describe the functioning of the Regulatory Authority (EETT) in terms of independence and HI¿FLHQF\7KHLVVXHVFRQFHUQLQJWKHULJKWVRIZD\ DUHDOVREHLQJGH¿QHG HJWKHULJKWVRIZD\RI ¿EHU RSWLF WKURXJK XUEDQ SODQQLQJ UHVRXUFHV RI different proprietors such as municipalities, individuals, institutions, and organizations), collocation issues, and the demarcation between public and private networks.
ISSUES ON DESIGNING BROADBAND INFRASTRUCTURE This section presents the main principles that should be followed while designing such metropolitan area networks. In Western Greece (our case) the networks belong to cities, where in most of the cases the municipalities do not have the proper know-how
of designing such networks. Therefore, a specialized technical consultant (the Research Academic Computer Technology Institute) has been selected in order to provide its know-how. In particular, the consultant works on conducting the studies of the QHWZRUNV¶GHVLJQDQGDUFKLWHFWXUHVXSHUYLVLQJWKH technical part of the construction of the networks and also designing the business plan that will be DSSOLHG DQG ZLOO HQVXUH WKH QHWZRUN¶V YLDELOLW\ The business plan proposes the scheme that will be responsible for the operation of the networks, according to the criteria and the guidelines of the EU (EC, 2002; Europa, 2004). The expected result of these projects is the creation of broadband infrastructure in Greece, which in turn will create conditions of competition in Greece in advance of the end user. A common strategic design and implementation in a regional OHYHO LV KLJKO\ GHVLUDEOH $W WKH ¿UVW VWDJH WKH main target of these networks, as metropolitan area networks, is to interconnect the buildings of the public sector of the cities in which they will be developed. The organizations that are going to be connected are those belonging on the education, health, and culture sectors, and so forth. During the second stage, this infrastructure will be available on network and content providers in a cost-effective way. The latter will create conditions of competition in broadband infrastructure and VHUYLFHVWKDWZLOOEHQH¿WHQGXVHUV7KHSODQIRU deployment of the infrastructure encourages the public exploitation through the leasing of pairs of ¿EHURSWLFV7KLVIDFWSURYLGHVUHYHQXHLQDFRVW effective way in order to cover all operating and maintenance costs. The projects mainly consist of manholes, ducts, FKDQQHOV¿EHURSWLFVSRLQWVRILQWHUFRQQHFWLRQ together with the installation of the passive and active equipment in order to provide the basic broadband access in the public sector buildings. In particular, a broadband infrastructure based on RSWLF¿EHUVZLOOEHLQVWDOOHGDFURVVWKHFLWLHVDQG several public collocation points will be created. The projects also fund some additional access infrastructure from public sector buildings to the collocation and aggregation points. This access LQIUDVWUXFWXUHDORQJZLWKDVPDOOSDUWRIWKHFLWLHV¶
199
Metropolitan Broadband Networks
broadband infrastructure will be used to provide network connectivity to the public sector. The UHPDLQLQJEURDGEDQGLQIUDVWUXFWXUH RSWLF¿EHUV and collocation points) will be available for the content and service providers to use, and it will provide broadband services to the citizens. The major characteristics of the mentioned broadband infrastructure according to the European Commission (EC, 2003) should be the following: 1.
2.
3.
Open access: The funded projects must be consistent with the new regulatory framework of electronic telecommunications and the rules of competition (public funding and antitrust). The appliance of these rules is a FRPPLWPHQWLQRUGHUWRKDYHFOHDUO\GH¿QHG open access. In particular, the construction of the networks, as already mentioned, should be limited in the construction of infrastructure and equipment that will be open to any telecommunication carrier and service provider (ED, 2003; Magnago, 2004). Neutral operator: The main principles in the deployment of the broadband infrastructure assign the network operator the obligation to retain the neutral character of the infrastructure. The network should be an open access installation to all the organizations that provide electronic networks and services with absolutely no discriminations against them (Economides, 1996). Infrastructure owner: The municipality constructs the broadband infrastructure and REYLRXVO\ KDV VSHFL¿F EHQH¿WV IURP WKHVH networks. Therefore, it is also responsible for planning the expansion of the current infrastructure in a controlled and rational way. Also its role is to solve all issues of the installed infrastructure and declare the rules of usage of the infrastructure by the providers.
METROPOLITAN AREA NETWORKS This section is dedicated to describing the methodology that we followed in order to design the 200
MAN of Patras. Furthermore, it presents the main characteristics of the MAN of Patras in terms of topology, architecture, and technology selection.
Designing Aspects The overall architecture of the MAN is shown in Figure 2 (ITU, 2006). The topology is based on a three-level model: main network, distribution network, and access network. In turn, there are three types of nodes in the system: main nodes, distribution nodes, and access nodes. The main network consists of a number of main nodes that are connected directly between each other. In the main network, there must be some direct redundancy between main nodes which are close together. This means that it must be possible from one main node to reach the main nodes next to it without passing through the active equipment of another node. The optical cables should be laid without a break between the main nodes, so as to achieve high operational dependability. 7KH PDLQ QHWZRUN¶V RSWLFDO FDEOHV WKDW FRQQHFW different main nodes are to be separately ducted. 7KHQXPEHURI¿EHUVEHWZHHQWKHPDLQQRGHVLQ the main network in a municipality ought to be not less than 72 per optical cable. This number results as follows: Each distribution node usually FROOHFWV WUDI¿F IURP DSSUR[LPDWHO\ HLJKW DFFHVV nodes, as designed in the technical studies for the development of the networks. Each access node JDWKHUV WZR SDLUV RI ¿EHUV RQH XSOLQN DQG RQH backup), thus each optical cable should have at OHDVW[ ¿EHUVSOXVIRUDOWHUQDWLYHURXWHV in the distribution network. The available cables XVXDOO\SURYLGHDQG ¿EHUVVRLVWKH lower acceptable optical cable. If main nodes in different municipalities are long distances apart, with a smaller number of distribution nodes, the QXPEHURI¿EHUVPD\SRVVLEO\EHVPDOOHULIWKLVLV MXVWL¿HGE\JUHDWGLIIHUHQFHVRI¿EHUFRVW As far as the distribution network is concerned, it consists of the distribution nodes. A distribution node shall connect to a main node and shall be planned to have a redundant connection to another main node. The optical cables should be laid without a break from each main node to any
Metropolitan Broadband Networks
Figure 2. General architecture
distribution node. Alternatively, an optical cable loop is laid with two or three distribution nodes where the need for each distribution node is hived off. Over long distances this will be a cheaper but PRUHYXOQHUDEOHRSWLRQ7KHQXPEHURI¿EHUVLQ the distribution network is affected by the following parameters: • •
Number of access nodes connecting with each distribution node Number of operators needing connections in the distribution network /HDVLQJRIGDUN¿EHUWRRWKHUDFWRUV
7KHQXPEHURI¿EHUVWRHDFKGLVWULEXWLRQQRGH ought not to be less than 72 per optical cable, as already described. The access network consists of the access nodes. A number of buildings are connected to an access QRGHWKURXJKD¿EHUFDEOHZLWKDWOHDVWIRXU¿EHUV WZRSDLUVRI¿EHUVRQHXSOLQNDQGRQHEDFNXS 5HJDUGLQJ WKH ¿EHU LQIUDVWUXFWXUH WKH DGditional optical cable will be installed, in order to handle situations of possible or anticipated penetration in the area, a large number of Internet operators active in the area, the positioning of ac-
tive equipment, and the degree of redundancy in the networks. As a general rule, if existing ducting is to be used, a careful assessment must be made of the best way to use it. If the number of existing optical pipes is small, an optical cable with many ¿EHUVZLOOKDYHWREHODLGVRDVWRPDNHPD[LPXP use of the ducting. For the part of the network that will provide FRQQHFWLYLW\WRWKHSXEOLFVHFWRU¶VEXLOGLQJVWKH ethernet technology has been selected (Metro Ethernet Forum, 2006). Ethernet switches are used RQ DFFHVV QRGHV DQG DJJUHJDWH WKH WUDI¿F IURP the buildings. Each building has a 100 Mbps or *ESVFRQQHFWLYLW\WKURXJK%DVH/;6)3V7KH distribution nodes do not have any active equipment, only passive. The main nodes have gigabit ethernet switches with advanced features. Those switches connect the ethernet switches of access nodes as well as the buildings that have dedicated ¿EHUFRQQHFWLRQV7KLVFKRLFHKDVEHHQGRQHGXHWR WKHIDFWWKDWWKHSXEOLFVHFWRU¶VEXLOGLQJVVKRXOGXVH YDULRXVVHUYLFHSURYLGHUV,QSDUWLFXODUWKHWUDI¿F from the connected schools in the MAN should be forwarded to the Greek School Network, the WUDI¿FIURPXQLYHUVLWLHVRQ*UHHN5HVHDUFKDQG Technology Network, and so forth. Therefore, the
201
Metropolitan Broadband Networks
GHVLJQLQJDSSURDFKLVWRFRQ¿JXUHWKHHTXLSPHQW in such a way that will provide Transparent LAN services, connecting the building to their service provider transparently and “friendly” (Kompella & Rekhter, 2005). The latter means that the approach will eliminate the limitations that add to the network and service providers.
Methodology of Work In order to handle the project work and secure the FRUUHFWQHVVDQGTXDOLW\RIWKHGHVLJQDVSHFL¿F methodology was followed. In particular the steps were the following: •
•
•
•
•
Several visits on municipalities took place, in RUGHUWR¿QDOL]HWKHORFDWLRQRIWKHEXLOGLQJ the points that the building will connect to WKHQHWZRUNDQGWKH¿QDOPDSSLQJRXWRIWKH ducts Design the network according to the three distinct levels (access, distribution, and main network) 'HVLJQ WKH UHTXLUHPHQWV RI WKH QHWZRUN¶V nodes and the equipment (passive and active) :ULWHWKH¿UVWYHUVLRQRIWKHDQDO\WLFDOGHVLJQ study of the network Write the specific requirements of the QHWZRUN¶V FRPSRQHQWV DFFRUGLQJ WR HDFK PXQLFLSDOLW\¶VQHHGV Perform indicative cost accounting of the requirements and the overall design in order to ensure that it is compliant to the allocated budget Finally, optimize the network design and equipment requirements
The above mentioned methodology was used HI¿FLHQWO\IRUHLJKWGLIIHUHQWPXQLFLSDOLW\QHWZRUNV in Western Greece. The case of MAN of Patras was the largest one and therefore the most complex.
The MAN of Patras The city of Patras is the largest municipality in the region of Western Greece, the third largest city of
202
Greece (its metropolitan area has a population of more than 200,000), an important commercial center, and a busy portthe second largest in Greece. 3DWUDV¶0$1FRQQHFWVSXEOLFEXLOGLQJVLQWKH city, among them three university institutes, six research centers, four hospitals, and 120 schools (primary and secondary). ,PPHGLDWHO\EHQH¿WLQJIURPWKLVQHWZRUNZLOO be all the employees of institutions of education, research, health, and public administration in the region of Patras, while in effect all the citizens RIDZLGHUUHJLRQRIWKHFLW\RI3DWUDVZLOOSUR¿W Additionally, all major Greek content and service providers can use (cost effectively) this infrastructure to provide broadband services to the citizens. 7KHLULQWHUHVWLVTXLWHODUJHDVWKHORFDOµ3DWUDV¶ market is very attractive. 7KH0$1RI3DWUDVFRQVLVWVRI¿YHULQJVZKLOH the total length of the ducts is 48 Km (see Figure 3). 7KHXVHG¿EHUFDEOHV YDULRXVW\SHV ¿EHUV DUHDSSUR[LPDWHO\.P$PRQJWKHULQJV a star topology is used for the connections of the buildings to the access nodes. Additionally, 100% redundancy has been designed for the distribution DQG PDLQ QHWZRUN 0RUH VSHFL¿FDOO\ WKH 3DWUDV MAN will consist of: • • • •
Four (4) main nodes Eight (8) distribution nodes Twenty-two (22) access nodes Nine (9) wireless access nodes
Regarding the cost of the total investment, the Patras infrastructure cost approaches the value of 0¼ZKLFKLVDEVROXWHO\FRPSDUDEOHWRWKHFRVWRI the telecommunication services of the public sector of municipality of Patras. As shown in Table 1, this FRVWLVDSSUR[LPDWHO\0¼DFFRUGLQJWRUHFHQW research for the municipality of Patras conducted by the Research Academic Computer Technology Institute. Consequently, the depreciation of the cost of the whole investment in the municipality of Patras will take no more than two years.
Figure 3: The general ducting schema in Patras MAN
Metropolitan Broadband Networks
203
Metropolitan Broadband Networks
Table 1. Telecommunications fees in the municipalLW\RI3DWUDV LQȀ¼ ORGANIZATION
Education–Research
Universities Technical Universities
SUM 2
880.41
1
73.37
Research Institutes
6
117.39
Elementary Schools
76
102.72
High Schools
44
108.58
Public occupational centers
2
5.57
Hospitals
4
622.16
Region
1
158.48
Prefecture
1
70.43
Municipalities
2
76.30
Libraries
1
3.22
Sum
140
2218.6
works can be implemented, investing a reasonable amount of money. Those networks are designed in a three-level hierarchy (see Figure 4), where there are: •
•
Government
Wireless Municipal Networks Additionally, the Information Society program funds similar smaller projects that are suitable for small municipalities (less than 10,000 citizens) that mainly are underpopulated. These projects have the same goals as the optical metropolitan area QHWZRUNVEXWGXHWRWKHORFDOJHRJUDSK\ GLI¿FXOW morphology), the underpopulation of these areas, and the small needs of the public sector, the deSOR\PHQWRIRSWLFDOLQIUDVWUXFWXUHLVXQSUR¿WDEOH Therefore, these special projects are carried out and are based on wireless technology. The main design goal is the existence of backhaul connections based on multi-carrier OFDM that can succeed with large transmit rates over long distances and also support non-line-of-site (NLOS) connectivity. These systems are accompanied by WiFi products that are used in the densely populated areas in order to connect various sites of public sector (local authorities, schools, hospitals, etc.). Following this approach, small wireless net-
204
Access nodes that provide connectivity through WiFi systems, or in special cases through multi-carrier OFDM (mainly :L0$; Retransmission nodes that are used to connect main and access nodes when their immediate connection is not feasible 0DLQQRGHVWKDWDJJUHJDWHWKHWUDI¿FIURPDOO connected sites and provide inter-municipal routing. Main nodes also support the interconnection with the global Internet through the appropriate providers in a federated way
Except for the basic connectivity between sites, those networks aim to provide inter-municipal VoIP calls for all public sector sites and connectivity with other federated networks that are higher in the hierarchy (for example, school networks, national administrative networks). Additionally, the expansion of those networks is a continuous goal that can be succeeded either with wireless or optical connections. Finally, these networks will comprise the basis for the e-government and all the e-services of the new digital world that the European countries try to establish. In the region of Western Greece, 19 small municipalities are going to implement such wireless metropolitan networks, aiming to limit their operational costs and increase the quality of the services provided to the citizens.
The Wireless Access Network of Messatida This section presents a case of a municipal wireless access networkthe case of Messatida. Messatida, in the western part of Peloponnesus and very close to Patras city. The population of the city is 11,873, and its area is 66.4 square kilometers. The proposed municipal wireless access network covers the wider region of Messatida and constitutes a network that could be easily implemented
Greek Government Network
Building 1
Bridge
Switch
Antenna WiFi PtP Km
1 - 25 Km
Bridge
-
1,5 K m
Km 1 ,5
0,1 -
0,1
Bridge
Bridge
Bridge
Switch
Building 5
Antenna WiFi PtP
Switch
Km ,5 -1
0,1 1,5 Km
0,1
Site Á
Building 4
Antenna WiFi PtP
Bridge
Antenna WiFi PMP
Switch
Antenna WiMAX (PtP)
Access Node
Site Â
25 Km
Bridge
WiFi PMP
Switch
Antenna WiMAX (PtP)
1-
Bridge
Bridge
5 -2
Access Node
Switch
Antenna WiMAX (PtP)
Switch
Bridge
Antenna WiMAX (PMP)
1
Retransmission Node
Bridge
Greek Government Network Router Router
Bridge
Antenna WiFi PMP
Main Node
0,1 - 1,5
1 - 25 Km
Main Site
Figure 4: The architecture of wireless broadband networks
Access Node
Bridge
Switch
Bridge
Antenna WiFi PMP
Building 3
Antenna WiMAX (PtP)
Antenna WiFi PtP
Switch
Building 2
Switch
Antenna WiFi PtP
,5 -1
Km
Km
0,1 - 1 ,5
0, 1
Site C
Bridge
Bridge
Antenna WiFi PtP
Switch
Building 7
Building 6
Switch
Antenna WiFi PtP
Metropolitan Broadband Networks
Km
205
Metropolitan Broadband Networks
since the covered distances are relatively small. 7KLVQHWZRUNZLOOEHQH¿WLQVWLWXWLRQVRIHGXFDWLRQ research, health, and public administration in the municipality of Messatida, and all citizens of the wider region of the city of Messatida will have access to this broadband network. The municipal wireless access network of Messatida consists of the backhaul network and the access network. The backhaul network consists of point-to-point and point-to-multipoint high capacity links (one and four respectively), with highpower transceivers able to cover long distances with a cost-effective way. The selected technology FRXOGEHHLWKHU:L0$;RUDQ\WHFKQRORJ\WKDW supports multicarrier OFDM. The whole wireless network of Messatida has one main node, six access nodes, and uses one retransmission node that is necessary due to the morphology. The total number of connected buildings is 22. As far as the access network is concerned, it consists of large-scale cell deployments LPSOHPHQWHGZLWKSRLQWWRPXOWLSRLQW:L0$; multi-carrier OFDM technology, as well as WiFi hotspots where the cell radius is small enough. In particular, two access connections are implemented XVLQJ:L0$;DQGVL[XVLQJ:L)LV\VWHPV
The outer goal should be the interconnection of those networks (not even for the needs of the public sector) through broadband infrastructure that will either be deployed or will be leased for a long time (at least 20 years). The solution for the deployment of interconnection infrastructure could be an easy solution, if we take advantage of existing projects for construction of highways and other roads. In this case, the deployment of optical infrastructure during the construction is worldwide a very cheap solution. Additionally, for solitary areas where the optical deployment is not feasible, alternative solutions such as wireless backhaul connections based on OFDM technology or satellite links can be studied. This prospect will offer global connectivLW\ WKURXJK LQIUDVWUXFWXUH FRPSOLDQW WR VSHFL¿F technology standards and neutrality. Over this infrastructure, each provider will offer its services to its clients. The latter also applies for public sector needs. Generally, it is a key point to study carefully this prospect of global national connectivity that will add value to these municipal broadband networks and will also expand their usage and viability.
Interconnection at a National Level
BUSINESS MODEL
A major issue for the viability and exploitation of those networks is the interconnection between them in order to make a full connected broadband country. It is a key point for future years, but also a big challenge due to the morphology of Greece. ,QD¿UVWVWDJHWKHLQWHUFRQQHFWLRQRIDOOWKH optical MANs and the wireless municipal networks from each “provider” of the public sector (school networks, Greek research and education networks, the Greek government network) independently is an obvious solution. In this case, each provider ZLOOHVWDEOLVKOLQNVWKDWZLOOURXWHLWVWUDI¿FIURP the main node of each municipal network. But this solution must be temporary, as it is not scalable and cost effective due to the fact that each provider leases lines (or buys access lines from OTE) that DUHH[SHQVLYHDQGWHFKQRORJ\VSHFL¿F
,Q RUGHU WR VHFXUH WKH ¿QDQFLDO YLDELOLW\ RI WKH broadband infrastructure of Western Greece, it has to be supported by a business plan. Such business models have been proposed by several researchers and are used on other broadband municipal networks. $EXVLQHVVPRGHOGH¿QHVWKHZD\DPHWURSROLWDQ SXEOLFQHWZRUNVKRXOGEHH[SORLWHG,WGH¿QHVWKH role of the municipality or the public authorities, the handling of competition issues, the involvement of private companies, and so forth. Its intention is to provide the citizens with a viable metropolitan ¿EHURSWLFQHWZRUNLQFOXGLQJWKHSURSHUUHVRXUFHV for the maintenance and the expansion of the network (Monath, Kristian, Cadro, Katsianis, & Varoutas, 2003; Ecomomides, 1996). The main goal is the provision of better and less expensive services to the public (Henderson, Gentle, & Ball,
206
Metropolitan Broadband Networks
2005). Figure 5 presents the three basic levels of a business model:
•
•
7KH¿UVWOHYHOUHIHUVWRWKHSK\VLFDOLQIUDVWUXFWXUHRIWKHQHWZRUN GXFWVGDUN¿EHUHWF DQG to the organization that provides and exploits this broadband network infrastructure The second level refers to the active equipment of the network (e.g., routers, switches, etc.) The third level refers to who has access to the network, its services, and the provided content
Figure 5. Basic business model’s levels
Access, Services, Content
Network Equipment (Backbone and Access )
Physical Infrastructure (Dark Fiber)
By assigning each of the above mentioned levels to different organizations, consortiums, or companies, different business model cases can be proposed: 1.
Equal Access Model: In this model the broadband infrastructure is built by the city or state and then it is leased or sold to an operating company or a consortium. The operating company adds the active equipment to the network and sells access to the operating broadband network to any service provider. The service providers pay the operating company a monthly fee per customer, while independent content providers are able to sell their content to the public or to business customers. This process is done through WKH RSHUDWLQJ FRPSDQ\¶V SRUWDO ZKLFK LV responsible for keeping billing records and having a direct billing relationship with the customer. The role of the public authorities (municipality or government) in this model is to stimulate competition at the level of content and services. The government in this model has already invested in passive infrastructure, thus the cost for a service or content provider to enter the market is much lower. Through the operating company the government ensures an equal confrontation of all the service and content providers. In this model the roles and responsibilities of all the involved organizaWLRQVPXVWEHFOHDUO\GH¿QHGEHFDXVHLWPD\
2.
-
TV / radio Internet Telephony Other services
- Routers - Switches - etc. - Ducts - Fiber optics
lead to complex structures and processes (Chlamtac, Gumaste, & Szabo, 2005). The equal access model is followed by a number of international and EU countries. Examples of this model can be found in The Netherlands (Enschede) with the Dutch/GerPDQ,QWHUQHW([FKDQJH 1',; LQ&DQDGD (Alberta) with Axia, and in the United States (Grant County) with the Zipp Network. Variations of this model can also be found in the case of the city of Amsterdam, where the physical infrastructure already existed, and in Spain (region of Catalonia) and Italy (Infratel), where local government also set up the active infrastructure of the network. Sole Private Provider Model: This model is applicable for cases where no service and content providers preexist, and the broadband market has not created a critical mass of customers that would be able to generate suf¿FLHQWUHYHQXHWRDODUJHQXPEHURILQYROYHG companies. In this model the operating company is also the service provider. In this way, the development of broadband infrastructure can become more easily a viable project, but the customers experience a less wide range RIVHUYLFHVDWDSULFHWKDWGRHVQRWEHQH¿WWKH competition of companies. For this reason, the monopoly of the operating company should be maintained temporarily, in order to lead WRDQHTXDODFFHVVPRGHOLQD¿[HGSHULRGRI time. This model requires a lower manage-
207
Metropolitan Broadband Networks
3.
ment complexity and is also more attractive to service providers as it becomes commercially viable much more quickly (Chlamtac et al., 2005). Examples of this model can be found in Sweden (Stokab–Stockholm) and The Netherlands with the Fiber Pilot program in Almere. Full Public Control Model: In this model all the layers are managed by one or more public organizations, including the development of passive and active infrastructure and the provision of services. The private sector can also get involved in a variation of this model, by the creation of public-owned companies with private capital investment. This model is suitable in cases where the involvement of an operating company is not feasible and no service providers are activated. Most of the time, this is the case of rural or remote communities, with no prior broadband infrastructure (Chlamtac et al., 2005). The main problem of this model, like the previous one, is that no competition is evolved between service and content providers. Moreover, the lack of technical expertise of the local government most of the time has as an effect of less innovation in the deployment of the network and its provided services. Finally, the cost of the total investment in building the network and providing the services is quite high for the local government. Examples of this model can be found in Italy with Terrecablate in Siena and Acantho in the region of Emilia-Romagna, and in Austria with the Wienstrom in Vienna.
Proposed Business Model The selection of the proper business model is essential for the viability and the success of the project. Regarding the metropolitan area networks, the use of the equal access business model based RQDQDWLRQDOOHYHOLVWKHRQHWKDWVHHPVWR¿WPRVW in the case of Greece. In this business model, the public sector is only the owner of the passive infrastructure (level one in the model presented in Figure 5) while in the upper two levels (Figure
208
5), competition is delivered between both telecommunications carriers that operate the networks and content and service providers. 7KH SXEOLF VHFWRU¶V RUJDQL]DWLRQ ZKLFK ZLOO RSHUDWHRQOHYHORQH ZLOOSURYLGHWKH¿EHURSWLFV infrastructure, in a cost-effective way, to the telecommunication companies. The mission of this organization will be to establish competition conditions between the providers, and maintain and expand the network, where this will be necessary. In particular, it should have certain responsibilities, such as the information of the citizens, the recording of the needs during the use of the network infrastructures, in order to be able to maintain the networks and conduct the studies for their expansion. The requested revenue for the expansion of the network will come by the leasing of the existing infrastructure to telecommunication and service providers. Thus, the service and content providers will focus on their role by providing their competiWLYHVHUYLFHVLQDFRVWHI¿FLHQWZD\7KHFRQVXPHU will be able to choose from among the competition a great variety of different services. All in all, this model has some certain advantages regarding the design and the deployment of an overall broadband network in Greece. In this way, the service and content providers will provide their services at a low cost and in a continuously developing way. The service providers will also gain from this situation, as they will have access to national infrastructure through one organization and network operating schema. Additionally, there will be competition in the network level between telecommunication companies and competition in the access, services, and content level among the service and content providers (see Figures 5 and 6).
CONCLUSION This chapter presented the basic design aspects for the broadband municipal infrastructure networks in the region of Western Greece. The proposed broadband networks will cover the needs of the eight major cities within the region of Western Greece and 61 other networks in other regions of Greece.
Metropolitan Broadband Networks
Figure 6. The application of the proposed business model in network layers Network’s Layers
Access, Services, Content
Competition between content providers Competition
Network’s active equipment (main and distribution network)
Competition between telecommunication providers
Network’s passive equipment
National Level non-profit Company
Cost Effective
Geographic Area
These networks will interconnect the organizations of education, research, health, culture, and the sum of the buildings of public administration via highspeed connections, and will also establish public FROORFDWLRQSRLQWVDQGH[FHHGLQJ¿EHUFDEOHVWKDW can be used cost effectively by service and content providers. The projects are in the construction phase, and in the second half of 2008, the network will be up and running. Additionally, the case of the city of Patras, the largest case on the designed MANs, is described. Finally, the chapter discusses the proposed business models for the operation of those broadband networks and the one that we believe better suits the Greek case. For future work, we already have plans to study and propose an appropriate business model to cover all technical and political aspects for the operation of those broadband infrastructure networks.
REFERENCES Chlamtac, I., Gumaste, A., & Szabo, C. (2005). Broadband services: Business models and tech-
nologies for community networks. New York: Wiley Interscience. CityLink. (2006). New ZealandWellington’s case. Retrieved June 14, 2006, from http://www. citylink.co.nz EC (European Commission). (2002). eEurope 2005: An information society for all. Retrieved January 13, 2006, from http://ec.europa.eu/information_society/eeurope/2002/news_library/documents/eeurope2005/eeurope2005_en.pdf EC. (2003). Guidelines on criteria and modalities of implementation of structural funds in support of electronic communications. Retrieved January 10, 2006, from http://ec.europa.eu/regional_policy/ consultation/telecom_en.htm Economides, N. (1996). The economics of networks. International Journal of Industrial Organization, 14(4), 673-699. Europa. (2004). Challenges for Europe’s Information Society beyond 2005: Starting point for a new EU strategy. Retrieved January 13, 2006,
209
Metropolitan Broadband Networks
from http://europa.eu/rapid/pressReleasesAction. do?reference=IP/04/1383
Metro Ethernet Forum. (2006). Homepage. Retrieved June 14, 2006, from http://www.metroethernetforum.org
Europa. (2006). Broadband access in the EU: Situation at 1 July 2006. Retrieved February 9, 2007, from http://forum.europa.eu.int/Public/irc/ infso/cocom1/library?l=/public_documents_200 6&vm=detailed&sb=Title
Monath, T., Kristian, N., Cadro, P., Katsianis, D., 9DURXWDV' (FRQRPLFVRI¿[HGEURDGband access network strategies. Communications Magazine, 41(9), 132-139.
)LUWK/ 0HOORU' %URDGEDQG %HQH¿WV and problems. Telecommunications Policy, 29(23), 223-236.
UTOPIA (Utah Telecommunication Open Infrastructure Agency). (2006). Retrieved June 14, 2006, from http://www.utopianet.org/
Henderson, A., Gentle, I., & Ball, E. (2005). WTO principles and telecommunications in developing nations: Challenges and consequences of accession. Telecommunications Policy, 29(2-3), 205-221. ITU. (2006). Construction, installation and protection of cables and other elements of outside plant. Retrieved June 14, 2006, from http://www. itu.int/rec/T-REC-L/e Kompella, K., & Rekhter, Y. (2005, December 28). Virtual private LAN service. Internet Draft (draft-ietf-l2vpn-vpls-bgp-06). Localret. (2006). The case of Barcelona. Retrieved June 14, 2006, from http://www.localret.net/idiomes/english.htm Magnago, A. (2004). Open accessbusiness models and operational costs. Proceedings of the Broadband Europe International Conference 2004, Brugge, Belgium.
210
KEY TERMS Access Network: Part of the network architecture that consists of a number of access nodes. Business Model:$SODQWKDWHQVXUHVWKH¿QDQcial viability of the broadband infrastructure. Distribution Network: Part of the network architecture that consists of a number of distribution nodes. Main Network: Part of the network architecture that consists of a number of main nodes. Optical Fiber Network: A type of network PDLQO\EDVHGRQ¿EHURSWLFVWHFKQRORJ\ Wireless Access Network: Type of network mainly based on wireless technology.
211
Chapter XIV
Small World:
The Irish Broadband Experience Diana Wilson Trinity College, Dublin Kevin O'Reilly eircom, Ireland Dave Murray eNua IT Consultancy, Ireland
ABSTRACT In this chapter we consider from a marketing perspective the political, cultural/social, and economic factors, both micro and macro, affecting the supply/demand nexus of broadband services for the Irish consumer. We chart the development of broadband and its current situation of rollout and uptake, examine the reasons for its continuing poor performance, and offer recommendations on how Ireland may close the gap and perhaps even move ahead. We collated data from a variety of resources, journals and press and trade publications. We attended a ministerial conference on the state of broadband to which many representatives of the telecommunications industry had been invited. We surveyed people with and ZLWKRXWWKHIDFLOLW\DQGLQWHUYLHZHGNH\SOD\HUVLQWKH¿HOG:HFRQFOXGHGWKDWDOWKRXJKWKHPDUNHWLV beginning to grow strongly, it is from a low base, and as a result the country still lags behind many of its European counterparts. There is still a lack of competition which is having an adverse effect on both supply and demand of broadband for the Irish domestic consumer. In a small country, the incumbent VWLOOKROGVFRQWURORIWKHPRVWSUR¿WDEOHDUHDVRIWKHWHFKQRORJ\SDUWLFXODUO\ZKROHVDOHDQGVHOOLQJWR other operators. Also, the Irish are still not convinced that broadband is the ‘killer app’ they need. This may change in the future as the technology delivers more of the content-rich multimedia fare that the Irish already enjoy in other formats.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Small World
INTRODUCTION: THE MACRO AND MICROECONOMIC BACKGROUND After more than 15 years, the Celtic Tiger still roars. The economy doubled in size in the 1990s, achieving the fastest growth in the OECD zone in WKDWSHULRG7KHUHZDVQROHWXSLQWKH¿UVWKDOIRI the following decade where it recorded the highest rate despite being hit by the worldwide slump in the information and communication technology (ICT) sector (OECD, 2006). Many countries, including some in the EU15 but especially those in the new EU27 and those awaiting accession, have looked to Ireland to learn the secrets of its success. There are both long- and short-term, macro and micro IDFWRUVIRU,UHODQG¶VHFRQRPLFVXFFHVVRIZKLFKD brief synopsis follows. Following a period of harmful protectionism (from the foundation of the State, but particularly from 1932-1938), successive Irish governmental SROLFLHVERWK¿VFDO LQWKHIRUPRIORZFRUSRUDWH taxes) and social (in the form of ongoing partnership agreements between trades unions, employers, and governments which have guaranteed workplace VWDELOLW\ KDYH RSHQHG ,UHODQG¶V PDUNHWV ZLGHU and wider. This opening of the markets was further enhanced by membership of the European Union in 1973 and being in receipt of numerous grants from there. This money was, by and large, well spent (Dorgan, 2006). The country, through its agencies, the Irish Development Authority and Enterprise Ireland, has for decades actively sought and encouraged foreign direct investment while also promoting the successful integration RI LQGLJHQRXV ¿UPV LQWR WKHVH JOREDO QHWZRUNV ϶5LDLQ 3DUDGR[LFDOO\DVDUHVXOWRIWKH protectionist policy mentioned above, Ireland was a predominantly agricultural nation and therefore did not have the large, heavy industrial base of many of its European counterparts. Such industries often require burdensome subsidies or protracted GLVPDQWOLQJ,UHODQGZDVOLWHUDOO\DµJUHHQ¿HOG¶ VLWH (FRQRPLVW 2¶+HDUQ Its geographical location on the edge of Europe, and its perceived pro-European attitude at a time when the United Kingdom was displaying distinctly anti-European tendencies, combined with a highly
212
HIIHFWLYH H[SORLWDWLRQ RI µ7tUJKUi¶1 attracted the EHQLJQ LQÀXHQFH RI WKH ,ULVK 'LDVSRUD DQG WKH economic interests of many multinationals, both American and European. Furthermore, in recent WLPHVµ7tUJKUi¶KDVEURXJKWPDQ\,ULVKQDWLYHVDQG descendants home. With them they brought back a world outlook and experience, known locally as having a BABeen Abroadand strong technical and advanced managerial skillsets. They joined the indigenous, well-educated, computer-literate, \RXQJ ÀH[LEOH (QJOLVKVSHDNLQJ SRSXODWLRQ which until recently was regarded as low cost (Wilson & Murray, 2000. All these factors make a formidable combination, which bears out Paul .UXJPDQ¶V FLWHGLQ϶5LDLQS SLWK\ VXPPDU\RI,UHODQG¶VDGYDQWDJHVEHLQJGXH³«SDUW to luck, in part to policies….” But some commentators do ask how, and for how long, can this growth be sustained (Economist, 2004; Gottheil, 2003; OECD, 2006; Thornton, (2004)? Many solutions are given: sustaining pruGHQW¿VFDOSROLFLHV PDLQWDLQLQJVWDEOHLQGXVWULDO relations; nurturing indigenous entrepreneurship; increasing investment in education, research, and development; and importantly, yet little discussed in the literature of the Celtic Tiger, protecting the peace process (McAleese, 2000).2 Time and time again, furthermore, the crucial role of advanced telecommunications infrastructure in Ireland as an important driver for economic growth is cited (DCMNR, 2006; Forfás, 2005; Ottens, 2005; Sunday Business Post, 2006c). Yet for all our roaring, we remain, at best, just below European averages or, at worst, slipping towards the bottom of the leagues, for PC penetration, connection to the Internet, broadband rollout and uptake for domestic and commercial consumers, local loop unbundling (LLU), and value for money indices (CSO, 2005; Forfás, 2005; Sunday Business Post, 2006c; TIF, 2002). In this chapter we consider from a marketing perspective the political, cultural/social, and economic factors, both micro and macro, affecting the supply/demand nexus of broadband services for the Irish consumer. We chart the development of broadband and its current situation3 of rollout and uptake, examine the reasons for its continuing poor
Small World
performance, and offer recommendations on how Ireland may close the gap and perhaps even move ahead. We collated data from a variety of resources, journals and press, attended a ministerial conference on the state of broadband, surveyed people with and without broadband, and interviewed key SOD\HUVLQWKHEURDGEDQG¿HOG
PLVVLRQZLWKWKHDEROLWLRQRI7(¶VVWDWHPRQRSRO\ One reason for bringing the date forward may have EHHQGXHWR0LFURVRIW¶VGHFLVLRQDWWKHEHJLQQLQJ of the same year not to locate its European mirror in Dublin, but London instead (Thompson, 1999). Among many technical reasons given by Microsoft, WKHFRPSDQ\VSHFL¿FDOO\PHQWLRQHGWKHIROORZLQJ at that time:
THE DEVELOPMENT OF BROADBAND IN IRELAND: THE TELECOMMUNICATIONS BACKGROUND
•
In the late 1970s, the Irish telecommunications infrastructure, managed by a state company, Telecom Éireann (TE), was at a breaking point; when (and it was a long when) a domestic or commercial subscriber eventually got a phone, the equipment was of poor quality, the lines unreliable, and the bills among the highest in Europe (Forfás, 1998). The detrimental impact of this situation was making itself felt on the economy as foreign investment looked elsewhere in Europe to spend its money (Burnham, 2003). The issue was raised LQWKH'DtO4: “We must recognise that we have the worst telephone service in Western Europe. The lack of a modern telecommunications service is proving to be a serious impediment to the economic development and social progress of our country…The Minister will agree it is becoming more and more GLI¿FXOWWRDWWUDFWLQGXVWU\WRWKHVHDUHDV«7KHUH are jobs at stake. There are businesses at stake….” 2¶'RQQHOO Thus, Ireland began a program of massive investment in telecommunications infrastructure, with particular regard to establishing state-of-theart, international communication links (Forfás, 2004). But to maintain such a position requires ongoing investment and to achieve this, Ireland began to open its markets and began to semiprivatize its state companies. Full liberalization (although the process had begun in 1995) of the telecommunications market in Ireland took place on December 1, 1998, one year ahead of the date agreed on with the European Com-
• • •
Ireland was uncompetitive for large-scale digital communications There was an absence of large-scale ISPs with good international links The Irish market was not perceived as being deregulated, with TE holding a monopoly E-commerce, part of the MS European Internet strategy, would require a vast amount of bandwidth which Ireland did not have
On the same date as liberalization, TE announced a £100 million investment in broadband, VSHFL¿FDOO\¿EHURSWLFFDSDELOLW\DQGWZRZHHNV later changed its name to eircom (Thompson, 1999). It was also obliged to dispose of its shareholding in Cablelink, the cable service company (Massey & Shortall, 1999). But commercial broadband did not commence until 2002, some two to three years behind other countries (DCMNR, 2006; TIF, 2003). :K\WKHµVWUROORXW¶" The delay can be attributed to factors that also affected other countries, including the dot.com bust of 2000 and the general downturn in the telecommunications sector resulting in a slowdown of investment (TIF, 2003). But there are factors that are peculiarly Irish....
It’s a Long Way from Here to There… Ireland is a young country; its own violent history (which for some is still unresolved) is within living memory. Fresh from a Civil War, Ireland adopted a SURWHFWLRQLVWSROLF\ZLWKWKHDLPRIVHOIVXI¿FLHQF\ Many industries, including telecommunications, were brought under state control, and monopolies were created for the common good; some remain in this situation to this day. In the 1960s, this policy was reviewed and reassessed under Sean Lemass,
213
Small World
who began to open Ireland to external markets. 7KHUHIRUHWKHQHFHVVLW\IRUDQHI¿FLHQWUHOLDEOH and affordable telecommunications system was crucial. Serious attempts to deal with increasing demand and capacity, rather than just geographical reach (which has left Ireland with a high ratio of local exchanges in rural areas with shared lines) began. The 1980s saw an extensive period of sustained investment (which remains, to date, the most expensive ever) to ensure that Ireland would never fall so far behind demand again (Flynn & Preston, %XWWKLVLQYHVWPHQWZDVQRWMXVWFRQ¿QHG to telecommunications. Similar investments had to be carried out in other sectors of Irish society: education, health, transport. From “the poorest RIWKHULFK´WREHFRPH³(XURSH¶VVKLQLQJOLJKW´ (Economist, 2004), Ireland had a lot of catching up to do.
See You in Court Competition law enforcement in Ireland has historically been weak. The State, as a policy instrument, has a tradition of creating national monopolies and protecting private ones, and allowing concentration of markets, both strong and weak (Ingleton, 2003). Irish companies have been known to use the courts as both a delaying tactic in the face of UHJXODWRU\HQIRUFHPHQWRUDVµUHQWVHHNLQJ¶5 For example, following complaints received in 1997 and 1998 about TE, the Competition Authority (CA) of Ireland decided there was evidence of a possible abuse of dominant position, particularly with regard to local loop unbundling, and warned the company that unless this was resolved they would instigate proceedings (CA, 1998). This they did in 1999. The case was discontinued in 2002. On an almost annual basis, since that time, eircom (the incumbent) has either been taken to court by various parties, or has taken various parties to court, despite the proposed provisions of the Communications Regulations Act of 2000, which sought to reduce the resort to the courts (OECD, 2001). To counteract the growing perception that the Commission for Communications Regulations (ComReg) was powerless to move against the
214
LQFXPEHQWDQGDIWHUDGHOD\RI¿YH\HDUVZKLFK is given as a reason for the lack of competition in WKH ,ULVK WHOHFRPV PDUNHW DQG WKH QDWLRQ¶V SRRU broadband infrastructure, the Communications 5HJXODWLRQ $PHQGPHQW %LOORIZDV¿QDOO\ SXEOLVKHG 7KH ELOO KDV LQFUHDVHG WKDW DJHQF\¶V LQYHVWLJDWRU\ SRZHUV DQG FDSDFLW\ WR ¿QH QRQ compliant behavior. It is also intended to eradicate the time-consuming and expensive trot to the court taken by providers against the regulatory body.
What Did You Do in School Today? Well, not computing. Of course there are computers in Irish schools, and they are integrated into subjects other than computing itself, but it is patchy. However, computer science or ICT-related subjects are still not examinable subjects in the Irish School Curriculum. The Broadband for Schools program was launched in 2004 to provide broadband to all of the 4,100 primary and secondary schools by the end of 2005 (Forfás, 2005). The deadline was missed, but as of July 2006, some 90% of schools were connected (Campbell, 2006). However, ongoing professional development of all teachers is lacking (Forfás, 2005). Korea enacted its Classroom Advancement Plan in 1999, and the UK and U.S. have policy documents on the transformative potential of ICTs in education. In Ireland, there is no government policy for equipping schools with even the minimum of equipment and no central procurement or distribution of technology by government to schools (Murphy cited in Kennedy, 2005). Schools fund themselves, DQGWKHLQHYLWDEOHµGLJLWDOGLYLGH¶UHVXOWV
These Little Telcoms Went to Market… eircom, the incumbent, has had a turbulent history since deregulation in 1998. It has changed ownership four times in less than seven years (Sunday %XVLQHVV3RVWD ,WKDVEHHQÀRDWHGRQDQG delisted from the stock market twice (it remains GHOLVWHGWRGD\ ,WV¿UVWÀRDWDWLRQLVSHUFHLYHGE\ some to be an example of a stock market bubble which left many Irish small investors with a jaun-
Small World
diced view of both the company and the government.6 This feeling persists. Currently the company is owned by Babcock and Brown, an Australian investment company, with employees holding 35% RIWKH¿UPLQDQ(627 HPSOR\HHVKDUHRZQHUship trust). It is undergoing major restructuring. It has nearly halved its workforce. The company UHFHQWO\ DQQRXQFHG D ¼ ELOOLRQ LQYHVWPHQW LQ telecommunications projects over the next three years, to include enhanced broadband services and LQLWLDWLRQRIUROORXWRI¿EHUWRWKHNHUE There are now more than 60 authorized operators, most of whom are active, in the fully liberalized, deregulated Irish market offering ¿[HGOLQH PRELOH LQFOXGLQJ * ¿[HG ZLUHOHVV access, wireless, and satellite communications platforms (ComReg, 2007). Telecommunications KDVEHFRPHDVLJQL¿FDQWVRXUFHRIHPSOR\PHQW employing on average 13,000 workers and revenue JHQHUDWLRQRI¼ ELOOLRQ9HQGRUVRIIHUDYDULHW\ of options, and pricing tariffs depend on contention rates, upload and download capacity, caps, and contracts. Helpful government and lobbying JURXSV¶:HEVLWHVJXLGHFRQVXPHUVLQWKHLUFKRLFH these include: http://www.callcosts.ie/home/default.asp and http://www.mulley.net/2006/02/09/ broadband-choices-in-ireland/. eircom continues WREHWKHODUJHVWRSHUDWRULQWKH¿[HGOLQHPDUNHW (ComReg, 2007). Ireland has young, mobile, well-educated, English-speaking, and technically literate workforce.
We have 103% mobile phone penetration and are the second highest number of users of voice and text services in Europe (ComReg, 2006b). We also have among the highest mobile phone costs in Europe (OECD, 2006). Figure 1 shows the overall call volume (ComReg, 2007). The Irish telecommunications market is small, and that may have made it unattractive for some players (OECD, 2006). Furthermore, as one of the 2(&'¶VPRUHRSHQHFRQRPLHV 2(&' WKHUH are multinationals entering the sector and they may EH VXVFHSWLEOH WR JOREDO ÀXFWXDWLRQV '&015 2003). The government and the regulator (ComReg) are trying to foster competition through a number of initiatives to foster supply, particularly through LVVXLQJOLFHQVHVWR¿[HGZLUHOHVVRSHUDWRUVDQG encourage demand using media campaigns. However, most licensees have delayed or failed to start up, ceased trading after a while, or chosen niche (rural) markets, thereby leaving little effective competition for the incumbent. Smart Telecom, a major competitor, went to the wall on October 2, 2006, discommoding some 40,000 phone customers, and eircom began the process of disconnecting broadband customers (RTÉ Business, 2006). This KDVVKDNHQFRQVXPHUFRQ¿GHQFHLQWKHFRPSHWLtiveness of the market, as eircom was able to offer VZLWFKEDFNWR6PDUW¶VFXVWRPHUV%7,UHODQGKDV SXEOLVKHGLWVDFFRXQWVZLWKUHSRUWHGORVVHVRI¼ billion. Magnet Networks has reported that it will
Figure 1. Overall call volume (ComReg, 2007)
215
Small World
Figure 2. Broadband subscribers by platform (ComReg, 2007)
take its emphasis off LLU and concentrate on more SUR¿WDEOHPDUNHWV .HQQHG\ Time and again, the issue of the slow progress of LLU is raised, with commentators stating that unbundling is crucial to broadband uptake (ECTA, 2006a; OECD, 2006) With only 4% of DSL fully unbundled, which is below the EU25 average of 17% (ComReg, 2006a), most of the alternative DSL providers are either resellers or offering bitstream products resulting in limited product differentiation (Forfás, 2005). While the incumbent is criticized by the press, the regulator, the Oireachtas Committee, and lobbying groups for dragging its feet on the LLU issue, it did pay for the infrastructure when it was privatized and therefore does own the loop, and in the words of its newly appointed chairman, Pierre Danon: “The price of the loop needs to be at a reasonable price in order that eircom can make money at selling the unbundled local loop and the cost is the cost.” (Irish Times, 2006) Paradoxically, for a country renowned for its strong ICT-producing sector, Ireland does not adopt such systems in non-ICT-related sectors. We spend VLJQL¿FDQWO\OHVVRQ,&7WKDQYLUWXDOO\DOORWKHU European countries (Forfás, 2005).
216
THE STATE OF BROADBAND IN IRELAND TODAY After a slow start, Ireland is moving quickly to demand and supply broadband services and products, just in time to meet the global slowdown in broadband uptake as established markets reach saturation levels and others see no reason to climb on board *XDUGLDQ 7KHUH KDV EHHQ D VLJQL¿FDQW slowdown in broadband uptake across Europe from 23% to 14% growth rates. ECTA (2007) reports also a corresponding growth in telecommunications monopolies in key European countriesa correlation they believe is not coincidental. In June 2006, the Parliamentary Joint Committee on Broadband, comprised of Teachta Dála7 and senators from the Lower and Upper Houses, called in CEOs from eircom, SmartTelecom, BT Ireland, eNet, and ComReg. They were subjected WRYHU\KHDY\TXHVWLRQLQJZLWKUHJDUGWR,UHODQG¶V SRRUSHUIRUPDQFH 'tRVSyLUHDFKWDL3DUOLDPHQWH 2006). ComReg released its Quarterly Key Data on the Irish Communications Market in March 2007. Figure 2 shows the number of broadband subscribers by platform. $FFRUGLQJWRWKH&HQWUDO6WDWLVWLFV2I¿FH &62 2005), broadband uptake by consumers with an Internet connection between 2003 and 2005 rose from just under 2% to almost 16%. However, dialup connections dropped from 94% to 81% in the same timeframe. This would seem to indicate that
Small World
Figure 3. Market share of total broadband (ComReg, 2006c)
for the majority of Internet users, broadband has been a matter of migration from one platform to another rather than substantial numbers of new uptakes. Finally, Figure 3 shows the market share of the EURDGEDQGPDUNHW$VFDQEHVHHQWKHLQFXPEHQW¶V DSL share has been dropping consistently. In February, 2007, ECTA, the industry benchmark used by the EC, national regulators, and the OECD released its quarterly scorecard for broadband across Europe (see Figure 4). Ireland is not doing well. Public perception of broadband supply and demand is not good. There are almost daily accounts in the popular and practitioner press of ,UHODQG¶VSRRUSHUIRUPDQFHDQGYRFLIHURXVSXEOLF OREE\LQJ IRUD VXFK DV ,UHODQG2IÀLQH %RDUGVLH and ComWreck. These performance indicators can range from technical issues, such as bad contention rates and low capacity, to more social factors such as static PC penetration levels, confusion of services and applications, and irrelevance to consumer needs.
IF YOU BUILD IT: FACTORS AFFECTING UPTAKE The Minister for the Department of Communications, Marine and Natural Resources (DCMNR), Noel Dempsey, has stated publicly that there is a lack of demand for broadband (Kennedy, 2006a). This was met with public and private derision, particularly with statistics about lack of supply DQGDYDLODELOLW\(PEDUUDVVLQJO\RQH¿IWKRIWKH PLQLVWHU¶V FROOHDJXHV LQ WKH 'DtO ZKR ZHUH invited by eircom to have broadband installed in WKHLUKRPHVRUFRQVWLWXHQF\RI¿FHVGLVFRYHUHGWKDW their lines failed the test (Sunday Times, 2006).8 But there is a grain of truth on both sides. We were intrigued by statements made not only by the minister, but also by ComReg (Kennedy, 2006b), eircom, and the Irish Business and Employers Confederation (IBEC), that one of the reasons for the slow uptake for broadband was lack of customer, both domestic and SME, demand. eircom cited countrywide coverage at 85% (although there is an ongoing dispute about line quality and fail tests) that they were unbundling EHWZHHQDQGORRSVDZHHN 'tRVSyLUHDFKWDL
217
Small World
Figure 4. Evolution of broadband penetration rates (ECTA, 2007)
Parliamente, 2006), government initiatives were underway to roll out in rural areas, and competitors were offering alternative broadband access, such as cable and wireless. Yet this contradicted press reports and results from our survey about the lack of availability of broadband infrastructure for the consumer being a major reason for lack of uptake. It would appear that Ireland was building it, and the Irish did want to come, but something was going wrong. This contradiction highlighted for us the complexity of supply/demand nexus and the effect it was having on the market; it is often said that connection pushes and content pulls. A further paradox is that Ireland buys and sells over the Net far above the European average (CSO, 2005; Ottens, 2004): 92% of all companies have access to the Internet (not necessarily via broadband) (Demunter, 2005), yet only 16% of Irish households with an Internet connection (48%) had broadband (CSO, 2005). For this reason, we were interested primarily in the domestic market.
218
Using a combination of the traditional four 3VRIPDUNHWLQJLQWHJUDWHGZLWK.RWOHU¶VIRXU&V of e-marketing, we analyzed the Irish broadband market as evidenced from the press, the literature, interviews with key personnel, and the results from our survey. We surveyed 305 people, from farmers to ICT professionals, the retired to factory workers, from all over the country to ascertain why they opted for broadband, why they did not, and the reasons they felt were hindering uptake; 12.6% had broadband at home only, 17.9% had broadband at work only, 61.3% had broadband both at home and at work, while 8.3% had neither. The four Ps, while they may be regarded as oldfashioned and no longer applicable in the age of the Net and often thought of as too product-centric, are still the standard in marketing textbooks and useful for analysis. The four Cs are an adaptation of the four Ps for the Net economy and are an attempt to counterbalance the product by placing the customer at the center of the marketing mix. By merging them, we aimed to reveal the interrelationships between supply and demand.
Small World
Traditional Marketing Based on 4 Ps
E-Marketing Based on 4 Cs
x
Product
x
Customer Needs and Wants
x
Price
x
Cost to the Customer
x
Place
x
Convenience
x
Promotion
x
Communication
Product and Customer Needs and Wants Well, it’s broadband, isn’t it? And I need it and want it, don’t I? Broadband is a complex suite of communications technologies, made even more complex by providers having to make investment and infrastructural decisions on maintaining, modifying, or scrapping equipment altogether. In Ireland, as with other countries, most broadband is rolled out on a combination of legacy telephony FRSSHUZLUHVDQG¿EHURSWLFFDEOHV7KLVFDQEH IXUWKHUFRPSOLFDWHGZLWKWKHLQWURGXFWLRQRI¿[HG wireless, wireless, television co-axial cable, and satellite delivery mechanisms. Then there is the decision on how the data, in all its multimedia forms, is going to move and how fast in whatever direction, and what other data, in all its forms, has to be accommodated on the stream. Currently,
WKHGHEDWHEHWZHHQ,ULVKµWHFKLHV¶LVGLIIHUHQWLDO between upload and download rates, contention issues, and lately, the pros and cons of dial-up vs. ISDN vs. xDSL. And all the customer wants to do is use it quickly, conveniently, reliably, securely, and at a reasonable price. A total of 78.1% (n=224) of our respondents gave “faster access to the Net” as a major reason for getting broadband, with 60.7% of them upgrading from dial-up. But why did they want it? Respondents were allowed to choose as many options as they wished from a given list; and 34% and 15% stated that leisure and family requirements were the drivers here, respectively. There was a voluntary option to specify what they wanted broadband for, and of the 39 respondents who completed this, 14 stated voice over Internet protocol as a reason, while 11 mentioned homework or telework needs. :KHQ&RP5HJ¶V7UHQGV6XUYH\6HULHV &RPReg, 2007) asked respondents their reasons (they
Figure 5. Reasons for not adopting broadband
219
Small World
could opt for as many as they wished) for not getting broadband in the home, they found the results in Figure 5.
Price and Cost to the Customer Ireland has the highest landline rentals and mobile phone bills in Europe (ComReg, 2006a). The costs to the customer of entry-level DSL services, which KDYHGHFUHDVHGVLJQL¿FDQWO\VLQFHDUHDPRQJ the most competitive in Europe (Forfás, 2005). However, when speed of connectivity is factored in rather than just price, Ireland drops almost down to the bottom of the EU15 (Sunday Business Post, 2006c). In an effort to guide the consumer, ComReg recently included comparison tables for broadband on its site9: 25% of 224 of our respondents felt that prices have become more competitive and that was a reason for uptake. Of the 207 out of 305 who had upgraded to broadband, 56% found the switch better value. When asked what would be an important consideration for broadband uptake, 63% of 295 respondents stated low costs. The cost of accessing broadband in rural areas can be as much as 20 times higher than in Dublin, as consumers have no alternative but to choose satellite. Latency problems with this platform have already caused one company to relocate its server to America (ISC, 2003). With the crash of Smart Telecom, government LQLWLDWLYHV ORVLQJ PRQH\ DQG LQFXPEHQW¶V GHEWV RI ¼ ELOOLRQ EURDGEDQG VWLOO LV QRW SUR¿WDEOH (Sunday Business Post, 2006b). Furthermore, if the private sector is unwilling to supply in areas it perceives as high cost for return and people are unwilling to subscribe due to concomitant high costs, WKHµEURDGEDQGJDS¶ZLOOJURZ ,6&
Place and Convenience On a geographical level, the penetration rate for EURDGEDQG LQ UXUDO DUHDV LV VLJQL¿FDQWO\ ORZHU than the national average and is among the lowest in the EU15. In this country, more than 40% of the population resides in rural areas. (Forfás, 2005; EC, 2005). Ireland has a high ratio of local exchanges to telephone lines, many of which are
220
not suitable for upgrade, and the longer distances from buildings to exchanges and from these to the backbone render them unattractive to the commercial market. Indeed a larger share of the population may be beyond the range of standard DSL or cable services. These factors increase the investment costs of a ubiquitous DSL system (TIF, 2003). Government initiatives, such as MANs and County and Group Broadband Schemes for communities of less that 1,500 people, are seeking to address the gap in the market that broadband suppliers will not enter. Both were launched in 2004, and at present, there is a lack of reliable data. Yet time and again, non-availability to location is cited in report after report. Of those respondents who did not have broadband in the home, 63.5% (n=85) of our respondents stipulated that they could not get broadband at their current address. A further 20% said that as they had it at work, they did not want it at home. The connection to landlines in new home builds is dropping as the Irish have taken mobile phones to heart (or rather to their ears). Most of us have two! While electricity and water are obviously SLSHG WR HDFK KRPH FDEOH 79 ¿EHU RSWLFV DQG telephone lines are not. Many of our respondents mentioned choosing a broadband service that did not require a landline or making the decision to install cable for the convenience of a single bill. However, number portability is an issue, and one LQ¿YHFXVWRPHUVKDYHWRZDLWWZRZHHNVIRUOLQH reconnection if they choose to switch from eircom (Sunday Business Post, 2006a). Long lock-ins to contracts were also mentioned. We asked respondents that if they could receive broadband in the home, and if their employer was amenable, would they like to telework. A total of 55.2% (n=183) said yes. Not only is this a strong indicator of the pervasiveness of the ICT sector ZLWKLQ,ULVKLQGXVWU\LWLVDVDGUHÀHFWLRQRQWKH VWDWHRI,ULVKWUDI¿FZLWKDYHUDJHFRPPXWHWLPHV within the Dublin area alone standing (pun intended) at one hour. With 61.3% of respondents (n=302) stating they had broadband both at home and at work, experiHQFLQJWKHEHQH¿WVRIWKHIDFLOLW\LQWKHZRUNSODFH may have been a deciding factor in upgrading or subscribing to it at home.
Small World
Promotion and Communication
FUTURE TRENDS
After some time, competition to the incumbent has entered the market and predictably, achieved a response from the latter. Cable, after delayed re-investment, is making some progress as a provider. Wireless and FWA access continue to grow (DCMNR, 2006): 40.4% (n=120) stated that they felt there was a lack of understanding/confusion as to exactly what broadband is and what it can offer the domestic consumer. There are more than 40 broadband providers, yet 50% of our respondents (n=292) thought there were only about 20. Broadband providers, products, services, and bundles are being touted in the media everyday, adopting different tactics on price, technical advantages, and lifestyle choices. But the Irish also have access to media from other countries, particularly the UK and Northern Ireland, and the differences in these factors are glaringmore bang for the EXFNZKLFKUHDI¿UPVSXEOLFRSLQLRQRIµWKHULS RII5HSXEOLF¶10 The incumbent is consistently blamed by consumers and the press for impeding the progress of broadband; slow unbundling, poor quality lines,11 the cash cow of dial-up, lock-in to other services such as Homewatch, and delay in switching accounts from themselves to other providers were the qualitative reasons given by 67% (n=199) of our respondents. This is not unique to Ireland however, as ECTA (2006c) has found that the incumbents in Europe retain 80% of the lines used to supply broadband. Many of the European incumbents are criticized for exactly the same reasons given in the Irish market. But this negative perception is having an impact with consumers. Many of our respondents gave ABe (Anyone But eircom) as an emotive reason for choosing an alternative provider. Research shows that broadband users were more likely to use the Net for a wider variety of purposes (ComReg, 2006b) than narrowband users; as the former sings the praises of more content-rich applications, it may persuade others to revisit their decision to get broadband.
7HOHFRPV¶JURZWKUDWHLVHVWLPDWHGWRLQFUHDVHE\ LQGULYHQE\¿[HGGDWDZKLFKZLOOWRS ¼PLOOLRQDQLQFUHDVHRIRQRQHRI the fastest growths in Europe (Daly, 2007). Fixed YRLFHDQGPRELOHYRLFHVHUYLFHVZLOOUHPDLQÀDW 7KHUH LV ZLGHVSUHDG FRQ¿GHQFH LQ WKH ZLOOingness of Irish consumers to purchase new technologies. But we do not buy for the sake of buying, rather the technology must be relevant to our lifestyles (DCMNR, 2006). We love talking, texting, gambling, gaming, Bebo and MySpace, video, and music. As mentioned above, we are a young demographic, but when the time comes, we will have our own children who will want to do all these things also (Demunter, 2005). These IDFWRUVDQGµSHVWHUSRZHU¶DUHOLNHO\WREHWKHELJ GULYHUVRIEURDGEDQGGHPDQGDQGµIDW¶EURDGEDQG at that. $FFRUGLQJWRD6RQ\EXVLQHVVEULH¿QJLQ0DUFK 2006, the country is the third largest purchaser, on a per capita basis, of PlayStations and PS2s. With the advent of PS3, which permits online gaming, the demand for broadband should rise accordingly. 7KHFRPSDQ\¶V&RPSXWHU(QWHUWDLQPHQW(XURSH President David Reeves said: “…the company is investigating the possibility of electronic distribution of content over broadband for its next generation system—describing the pursuit of this as the ‘ultimate goal’ for the PlayStation 3, and stating that Sony hopes to see always-on network access and viable broadband distribution in place when the PS3 becomes available in 2006.” (Fahy, 2005) Ireland has the second highest cinema attendance in Europe, and the country is set to be the ¿UVW LQ (XURSH WR LQVWDOO 'LJLWDO &LQHPD LQ picture houses (BBC, 2005). This will require the necessary bandwidth for download of simultaneous releases in Dublin and the rest of the country. ,WZRXOGDOVRVHHPWKDW¿OPVZLOOGULYH,QWHUQHW protocol TV, video on demand, and so forth; the proliferation of multimedia content and the Irish GHVLUHIRULWFRXSOHGZLWK¿OHVKDULQJLVOLNHO\WR be a huge driver for demand. Expanding the number and variation of devices that can go online should
221
Small World
increase broadband users, and this in turn will drive more broadband usage (ISC, 2003). As they continue their love affair with the PRELOHSKRQH¿[HGOLQHSHQHWUDWLRQKDVOHYHOHG off at 79% of Irish households. But as the major broadband access mode, and likely to continue to be so for some time, is DSL, subscribers will reTXLUHD¿[HGOLQHWRWKHKRPH '&015 * licenses have been issued (4G is on the way), and there have been successful promotion campaigns of this technology, particularly 3 Ireland offering free edited highlights of the recent Ryder Cup and Barclays Premiership football. But broadband mobile pricing looks set to incur a premium rate and therefore is not likely to be a viable alternative platform; thus its impact is likely to be minimum over the next few years (Forfás, 2004). VoIP is making headway into companies as a mechanism to drive down costs, and given our high domestic charges, the desire for VoIP should grow for the home user, particularly as many of us have American relations. :LUHOHVVDFFHVV ¿[HGDQGQRQ¿[HG FRXOGEH a way to enable intra-platform competition. eircom LVWULDOLQJ:L0$;DWHFKQRORJ\WKDWVKRXOGJR some way to addressing the problems in line quality (RTE Business, 2006).
CONCLUSION The rollout of broadband is often cited as a requirePHQWIRUDFRXQWU\¶VHFRQRPLFJURZWK,QPDQ\ ways, it should be no surprise that Ireland has lagged in the uptake in broadband; it has had a remarkable amount of catching up to do in all sectors to build the related infrastructures to support such massive and profound change (Economist, 2004). It could be said that broadband was not the priority for Ireland that it was for other countries; jobs, education, health, roads and transport, energy, as well as telecommunications were banging on the 0LQLVWHURI)LQDQFH¶VGRRU After a slow start, the market is growing robustly, but the issues are sustaining this growth and closing the gap. The small size of the market may make it unattractive (OECD, 2006), yet con-
222
versely, the increasing prominence of multinational ¿UPVHQWHULQJWKH,ULVKWHOHFRPVVHFWRUPDNHLW VXVFHSWLEOH WR ÀXFWXDWLRQV LQ WKH JOREDO PDUNHW (DCMNR, 2003). This tension can make the Irish sector dynamic and volatile. eircom has been asset stripped in the form of its cable arm (as part of privatization) and then its mobile phone division (as part of making money for shareholders).12 It now KROGVGHEWVRI¼ELOOLRQ,QQRYDWLRQLQEURDGEDQG both in terms of inter- and intraplatform, may not be a priority for this company. For example, the company only offers asymmetric services at the moment, when it is envisaged that symmetric services will be required in the near future. Smart Telecom, a major competitor, has gone to the wall. This has impacted on consumer con¿GHQFH LQ WKH FDSDELOLW\ RI VPDOOHU FRPSDQLHV to supply reliable broadband. LLU has been the VXSSRVHG VWXPEOLQJ EORFN RQ ZKLFK HLUFRP¶V ¿[HGOLQH EURDGEDQG FRPSHWLWRUV KDYH IDOWHUHG In other markets, full LLU has allowed operators to innovate in products and pricing. Sean Murphy, director of policy at Chambers Ireland, says LLU is academic because copper cannot handle the bandwidth required for the future (Kelly, 2006). This was also echoed by Peter Cochrane at the Facilitating Broadband Leadership Conference mentioned above. It is estimated that there is an immediately addressable broadband market of 800,000 subscribers (DCMNR, 2006). Data from government departments, ComReg, and the Information Society VXJJHVWWKDW,UHODQGKDVMXVWFRPSOHWHGWKHµHDUO\ DGRSWHUVWDJH¶LQWKH6WDQGDUG7HFKQRORJ\$GRStion Curve. A variety of reasonsincluding late start; late liberalization; slow progress with LLU; lack of competition; innovation in both inter- and intra-platforms, prices, products, and services; nonintervention from the government and a light-touch approach on behalf of the regulator; and lack of perceived relevanceindicate that Ireland may be H[SHULHQFLQJWKHµDGRSWLRQJDS¶ZKHUHE\WKHH[WHQW to which the technology will be fully adopted in society is questioned (ISC, 2003). But there is still evidence of market failure, both for supply and demand. It is acknowledged to be a GLI¿FXOWDQGXQGHUGHYHORSHGPDUNHW,QDQHIIRUW
Small World
to tackle markets which private companies may ¿QGXQDSSHDOLQJWKHJRYHUQPHQWKDVOHGLQLWLDtives in MANs and group broadband schemes to roll out broadband to rural areas. The government would fund 55% of the start-up costs. To date, RQO\DERXW¼KDVEHHQGUDZQGRZQZLWK participants stating too much bureaucracy, footGUDJJLQJTXLEEOLQJDQGJHQHUDOODFNRIµMRLQHGXS¶ government as the obstacles to achieving uptake of 90,000 subscribers (6,000 are participating currently) (Kennedy, 2006b). This lack of supply to the regions outside of Dublin is a barrier both to the consumer and to business. It would be in WKH FRXQWU\¶V LQWHUHVW LI WKHUH ZDV FRQYHUJHQFH between various government departments under an overall information and communications technology strategy. 7KHUH LV RQH ¿QDO IDFWRU ZKLFK PD\ UHDOO\ push the Irish into getting broadband: Ireland is the highest consumers of alcohol in Europe.13 The government recently introduced strict new anti-alcohol abuse measures, and that great social, cultural, political, and religious center of Irish life, the pub, is slowly emptying. The price of a pint is truly shocking....
REFERENCES BBC. (2005, March). Irish cinema set to go digital. Retrieved November 10, 2006, from http://newsvote.bbc.co.uk/mpapps/pagetools/print/news.bbc. co.uk/1/hi/northern_ireland/4365875.stml Burnham, J.B. (2003). Why Ireland boomed. The Independent Review, VII(4), 537-556. CA (Competition Authority). (1998). Competition Authority annual report 1998. Retrieved November 10, 2006, from http://www.tca.ie/NewsPublications/AnnualReports/AnnualReports.aspx Campbell, I. (2006, July 31). Schools round broadband corner. Retrieved November 10, 2006, from http://www.siliconrepublic.com/news/news. nv?storyid=single6810
ComReg (Commission for Communications Regulation). (2006a, June). Irish communications market: Quarterly key data report. Dublin. ComReg. (2006b, October). Trends survey series—wave 2, 2006. Dublin. ComReg. (2006c, December). Irish communications market: Quarterly key data report. Dublin. ComReg. (2007, March). ComReg trends survey series Q4 2006. Dublin. &62 &HQWUDO6WDWLVWLFV2I¿FH )HEUXDU\ Information society and telecommunications 2005. Dublin. CSO. (2006, December). Information society and telecommunications 2006. Dublin. Daly, S. (2007, February). IT and telecoms spend in Ireland. Retrieved November 10, 2006, from http:// www.ireach.ie/failid/Alert_21st_feb_07.pdf DCMNR (Department of Communications, Marine and Natural Resources). (2003). Ireland’s broadband strategy. Retrieved November 10, 2006, from http://www.micr.cz/images/dokumenty/Ireland_Broadband_Strategy.pdf#search=%22irelan G¶VEURDGEDQGVWUDWHJ\2 DCMNR. (2006, January). A review of demand in the Irish broadband market. Retrieved November 10, 2006, from http://www.dcmnr.gov.ie/Communications/Communications+Development/ Broadband+demand+report.htm Demunter, C. (2005). The digital divide in Europe, Eurostat. Retrieved November 10, 2006, from http://epp.eurostat.ec.europa.eu/cache/ ITY_OFFPUB/KS-NP-05-038/EN/KS-NP-05038-EN.PDF 'tRVSyLUHDFKWDL3DUOLDPHQWH -XQH Reports on broadband: Presentation (vol. 74). Dorgan, S. (2006, June). How Ireland became the Celtic Tiger. Retrieved November 10, 2006, from http://www.heritage.org/Research/WorldwideFreedom/bg1945.cfm
223
Small World
EC. (2005, July). Digital divide forum report: Broadband access and public support in underserved areas. Retrieved November 10, 2006, from http://europa.eu.int/information_society/eeurope/ LGRFVLPSOHPHQWDWLRQGGIBUHSRUWB¿QDOSGI search=%22EC%2C%20Digital%20Divide%20F orum%20Report%3A%20Broadband%20Access %20and%20Public%20Support%20in%20UnderServed%20Areas%2C%20July%202005%22 Economist. (2004). Survey: Tiger, tiger, burning bright. The Economist, (October 16). ECTA. (2006a, May). Broadband study highlights two-speed Europe. Retrieved November 10, 2006, from http://www.ectaportal.com/en/basic245. html ECTA. (2006b, July). Broadband penetration in EU: The haves and the have nots. Retrieved November 10, 2006, from http://www.ectaportal. com/en/upload/File/Broadband%20Scorecards/ Q106/FINAL%20BB%20ScQ106%20Press%20r elease%20Sept%2006.pdf
Forfás. (2005, November). Benchmarking Ireland’s broadband performance. Retrieved November 10, 2006, from http://www.forfas.ie/publications/forfas051205/webopt/forfas051205_broadband_webopt.pdf *RWWKHLO) ,UHODQG :KDW¶V&HOWLFDERXW the Celtic Tiger? Quarterly Review of Economics and Finance, 43(5), 720-737. Guardian. (2006). Global broadband take-up slows to lowest level ever. The Guardian, (October 3). Ingleton, J. (2003, June). De-monopolising Ireland. Proceedings of the EU Competition Law and Policy Workshop, Fiesole, Italy. Irish Times. (2006). The smart money is unlikely WREHWRQWKLVSKRQH¿UPThe Irish Times, (October 3). ISC (Information Society Commission). (2003). Ireland’s broadband future. Retrieved November 10, 2006, from http://www.isc.ie/downloads/InfoSocBroadband.pdf
ECTA. (2006c). 2006 review. Retrieved November 10, 2006, from http://www.ectaportal.com/en/basic522.html
Kelly, M. (2006). The future for broadband in Ireland. Retrieved November 10, 2006, from http:// www.enn.ie/frontpage/news-9674046.html
ECTA. (2007, February). Broadband take-up dramatically slows across Europe. Retrieved November 10, 2006, from http://www.ectaportal. com/en/upload/File/Broadband%20Scorecards/ Q306/FINAL%20European%20PR%20Sc%20Q 306_2_.pdf
Kennedy, J. (2005). Delays to school broadband scheme. Retrieved November 10, 2006, from http://www.siliconrepublic.com/news/news. nv?storyid=single5162
Flynn, R., & Preston, P. (1999). The long-run diffusion and techno-economic performance of national telephone networks: A case study of Ireland, 1922-1998. Telecommunications Policy, 23(5), 437-457. Forfás. (1998). Broadband telecommunications investment in Ireland. Retrieved November 10, 2006, from http://www.forfas.ie/publications/ bband/index.html Forfás. (2004, November). Broadband telecommunications benchmarking study. Retrieved November 10, 2006, from http://www.forfas. ie/publications/broadbandbenchmarking041126/ broadbandbenchmarking041126.pdf 224
Kennedy, J. (2006a, June 15). Dempsey concedes broadband demand problem. Retrieved June 15, 2006, from http://www.siliconrepublic.com/news/ news.nv?storyid=single6596 Kennedy, J. (2006b, November 6). Group broadband scheme labeled an ‘abject failure’. Retrieved November 10, 2006, from http://www.siliconrepublic.com/news/news.nv?storyid=single7307 Kennedy, J. (2007, February 15). Ireland is a dif¿FXOWPDUNHWIRUWHOHFRPVVD\V%7 Retrieved November 10, 2006, from http://www.siliconrepublic. com/news/news.nv?storyid=single7802 Massey, P., & Shortall, T. (1999, July). Competition and regulation in the public utility industries. Dis-
Small World
cussion Paper Number 7, Competition Authority. Retrieved November 10, 2006, from http://www. irlgov.ie/compauth/Dis_doc7.htm
Sony. (2006, March). Business EULH¿QJRetrieved November 10, 2006, from http://www.gamespot. com/news/6146007.html
McAleese, D. (2000, July-August). The Celtic Tiger: Origins and prospects. Options Politiques.
6XQGD\ %XVLQHVV 3RVW D HLUFRP¶V IXWXUH back on hold. The Sunday Business Post, (February 26).
2'75 2I¿FHRIWKH'LUHFWRURI7HOHFRPPXQLFDtions Regulation). (1999, December). Liberalisation in the Irish telecommunications market. Dublin. 2¶'RQQHOO 7' 0D\ 6SHDNLQJ RQ WKH Telecommunications Capital Bill. Díospóireachtaí Parliamente, 328. OECD. (2001). Regulatory reform in Ireland: Regulatory reform in the telecommunications industry. Retrieved November 10, 2006, from http://www. oecd.org/dataoecd/30/54/2511026.pdf#search=%2 2telecommunication%20in%20ireland%22 OECD. (2006). Economic survey of Ireland 2006. Retrieved November 10, 2006, from http://www. oecd.org/document/48/0,2340,en_2649_201185_ 36157872_1_1_1_1,00.html 2¶+HDUQ' *OREDOL]DWLRQµQHZWLJHUV¶DQG the end of the developmental state? The case of the Celtic Tiger. Politics and Society, 28(1), 67-92. ϶5LDLQ 6 7KH ÀH[LEOH GHYHORSPHQWDO state: Globalization, information technology and WKH µ&HOWLF 7LJHU¶ Politics and Society, 28(2), 157-193. Ottens, M. (2004). Internet usage by individuals and enterprises. Retrieved November 10, 2006, from KWWS ZZZPLFUF]¿OHV VWDWLVWLN\B vyu__v_n__ict_EU.pdf#search=%22internet%2 0usage%20by%20individuals%20and%20enterp rises%202004%20eurostat%22 Palcic, D., & Reeves, E. (2004, October). An economic analysis of privatisation in Ireland, 19912003. Proceedings of the Statistical and Social Inquiry Society of Ireland, Dublin. RTÉ Business. (2006, October 19). eircom to use WiMAX for broadband. Retrieved November 10, 2006, from http://www.rte.ie/business/2006/1019/ eircom.html
Sunday Business Post. (2006b). Eight tough calls for Babcock and Brown. The Sunday Business Post, (September 3). Sunday Business Post. (2006c). In the news. The Sunday Business Post, (October 22). Sunday Times. (2006). TDs discover the broadband revolution has failed them. The Sunday Times, (June 11). Thompson, H. (1999, May). Irish government sells off state telecom company. Retrieved November 10, 2006, from http://www.wsws.org/articles/1999/ may1999/tele-m03.shtml 7KRUQWRQ 5 &DJLQJ WKH WLJHU ,UHODQG¶V economy roars on. Harvard International Review, 26(3). TIF. (2003). Getting Ireland online. Retrieved November 10, 2006, from http://www.iia.ie/downloads/TSG_Interim_Report.pdf#search=%22gett ing%20ireland%20online%22 Wilson, D., & Murray, M. (2000, November). Pushing groupware, pulling teams: A re-evaluation of groupware in the light of m-business. Proceedings of 10th BIT Conference, Manchester.
KEY TERMS DCMNR: The Department of Communications, Marine and Natural Resources, affectionately known as the Department of Fish and Chips. Enterprise Ireland (EI): Government agency responsible for promoting indigenous enterprises. Forfás:,UHODQG¶VQDWLRQDOSROLF\DQGDGYLVRU\ board for enterprise, trade, science, technology, and innovation.
225
Small World
Industrial Development Agency (IDA): Government agency responsible for attracting foreign investment.
5
6
ENDNOTES
1
2
3
4
226
7tUJKUiOLWHUDOO\ µORYH RI FRXQWU\¶ $V D result of the forced emigration of many Irish from Famine times onwards, and the gaining of independence, albeit of 26 counties, from the British, the Irish Diaspora hold a strong ORQJLQJIRUWKHµDXOGFRXQWU\¶ At the time of writing (February 2007), the Assembly of Northern Ireland remains suspended pending elections to be held in March. Established in 1998, the Assembly has gathered intermittently and been suspended four times as a result of disagreements between the main Protestant parties and Sinn Féin, the largest Nationalist Party. ³,UHODQG KDV µZRUVW EURDGEDQG LQ (XURSH¶ says eircom owner,” according to the Sunday Business Post, August 6, 2006. The Lower House of the Oireachtas, the National Parliament
7
8
9 10
11
12
13
µ5HQWVHHNLQJ¶LVWKHWHUPJLYHQE\HFRQRmists to socially wasteful activities such as litigating to retain a monopoly. It may even have been a factor in the then Minister for Public Enterprise losing her seat in the General Election which followed. Teachta Dála are the equivalent of Members of Parliament. 7KLVVLWXDWLRQZDVUHFWL¿HGRYHUWKHVXFFHHGing months by eircom. http://www.callcosts.ie/home/default.asp The Rip-Off Republic was a impactful television program presented by Eddie Hobbs, ¿QDQFLDO DGYLVRU DQG FRQVXPHU FKDPSLRQ ZKLFK GUHZ LQ UHFRUG YLHZLQJ ¿JXUHV ,W made for depressing viewing. The bar set for eircom for the quality of landlines may be too high. A 5% reduction would still offer broadband suitable for most consumers needs. eircom subsequently purchased Meteor, the mobile phone company, in 2005. http://www.euromonitor.com/Alcohol-related_illnesses_cause_for_concern_in_Ireland
227
Chapter XV
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion: A German Case Study Axel Schulz De Montfort University, UK Bernd Carsten Stahl De Montfort University, UK Simon Rogerson De Montfort University, UK
ABSTRACT There is considerable interest worldwide in broadband diffusion, with research focusing on aspects such as the provision of broadband in remote areas and the socio-economic factors that determine the OLNHOLKRRGRIDGRSWLRQ7KLVFKDSWHULGHQWL¿HVWKHSROLFLHVDQGLQLWLDWLYHVXVHGWRHQFRXUDJHEURDGEDQG awareness, availability, and adoption. Using the case study of a local broadband initiative in remote and rural Germany, the chapter asks the question of who can and should be responsible for broadband provision, and how such responsibility ascriptions are realized.
INTRODUCTION National and international institutions across Europe and beyond encourage broadband adoption and coverage in rural areas. But often, on a local level, the broadband adoption does not take place. Broadband Internet is among the most important technologies in the telecommunication infrastruc-
ture market. Telephony and data transmission of all kinds will be provided and handled by broadband networks. Broadband is thus held to be a very important tool in order to take part in the knowledge and service society. But what if broadband is not available? This chapter will present a case study from East Germany where local authorities failed to take the demand for broadband seriously.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
Using the theoretical lens of responsibility, this chapter will also present an explanation why the demand/supply model fails when free markets do not emerge/exist. The chapter will concentrate on the question of responsibility for broadband. In order to explore what responsibility can mean in the context of broadband, we will start out with a discussion of the importance of broadband and technical infrastructures necessary for providing it. After EULHÀ\RXWOLQLQJWKHVRFLDODQGHWKLFDOUHOHYDQFHRI broadband, we will describe the German broadband market. This will set the scene for the case study of AltmarkDSL.de, a case where market forces did not lead to the provision of broadband in a remote and rural area in Germany. In this case, broadband was eventually provided because of pressure exerted by the local public. We will use this case study to discuss the question of responsibility for and in broadband, and to explore the link between ethical, political, social, and other factors. The main aim of this discussion is to show that broadband can be understood as a social innovation in which WHFKQRORJ\LVRQO\RQHLQÀXHQFLQJIDFWRUDPRQJ many.
SETTING THE SCENE: WHY BROADBAND IS IMPORTANT Broadband is a technology that facilitates participation in the global information society. It enables the transmission of large amounts of data in short time. In the near future TV and telephony may be delivered by broadband infrastructures only. Other examples of its importance include the KHDOWKRUQRWIRUSUR¿WVHFWRU+RVSLWDOVZLOOVHQG radiograms to foreign destinations via broadband connections. This will require doctors to be linked to a broadband connection. There is a range of other possible applications that cover all aspects of human life. What is important to note is that the political sphere has already recognized the importance of broadband. Germany is one example. The German coalition parties formulated in November 2005 their coalition contract using the following words: “In
228
order to advance the industry and research position of Germany, the coalition parties will provide incentives for the set up of modern broadband telecommunication infrastructures.”1 Furthermore, the European Union debates if broadband should get the status of a Universal Service Obligation (USO). “Universal service is a safety net for achieving social inclusiveness that is to ensure that basic communications services would always be available at a determined quality and an affordable price, even if the market would not provide them under normal commercial circumstances.” 7KH(8GH¿QHVXQLYHUVDOVHUYLFHDVD³PLQLPXP VHWRIVHUYLFHVRIVSHFL¿HGTXDOLW\WRZKLFKDOO end-users have access, at an affordable price in the light of national conditions, without distorting competition.”2 Thus, on an international and national level, the importance of broadband seems to be quite well recognized.
What Is Broadband? The speed and the quality of the signal by which data is transmitted around the globe via the Internet KDVEHFRPHDVLJQL¿FDQWHFRQRPLFDOIDFWRU7KH buzzword for the modern and fast transmission of data via Internet connection is broadband. Broadband transmission became the ordinary form of Internet connection. Many Internet applications do need in fact the availability of a broadband connection in order to be usable. The Internet is no longer a medium were scientists and young people share their thoughts, music, and videos. Instead, broadband is becoming an infrastructure for the many new industries and applications, a necessity for human life. Radiograms are nowadays transmitted via the Internet. The new form of telephony, called voice-over-Internet-protocol, is likely to become the new standard of how telephone calls are transmitted. In general one can say that the mass of data which will be transmitted by many new applications via the Internet will increase in the coming years. That is why it is important that people have access to new information infrastructures like broadband. Technically an Internet connection can be deVFULEHGDVµEURDGEDQG¶LIWKHEDQGZLGWKIRUGDWD
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
transmission is at least 0.5 Mbit/s fast (Lindskog & Johansson, 2005, p. 49f). However, the technical GH¿QLWLRQRIEURDGEDQGPD\YDU\IURPFRXQWU\ to country. In the United States, every Internet connection that exceeds the speed of 200 kbit/s is labeled as broadband. Other resources (Savage & :DOGPDQQ VXJJHVWHGWKHLPSUHFLVHGH¿QLtion of broadband as 50 times faster than dial-up connections (56 kbit/s). “Broadband is thus not a system or a technology, but rather refers to speed or capacity (bandwidth)” (Lindskog & JohansVRQ S 7KXV D FOHDUFXW GH¿QLWLRQ RI “broadband” is not available. For the purposes RI WKLV FKDSWHU ZH ZLOO GH¿QH EURDGEDQG DV DQ Internet access technology with a bandwidth of 2 Mbit/s downstream and 0.5 Mbit/s upstream. This should be acceptable for most of the Internet applications like video streaming, VoIP, browsing, and so forth. Broadband Internet connections can be realized via a range of different technologies. ADSL is in Europe one of the most common access technologies to broadband infrastructures. In the United States, cable TV infrastructures are commonly used to deliver broadband to households. Other technologies for the access to broadband infrastructures are WKHZLUHOHVVDFFHVVLQIUDVWUXFWXUHVOLNH:L0$; WLAN, or HSDPA. Potentially broadband can also EHGHOLYHUHGYLDVDWHOOLWHV7KHGLI¿FXOW\KHUHLV that the speed and bandwidth relies heavily on the number of user who use the satellite connection at the same time. Because of this disadvantage, broadband via satellites was not really successful in the broadband access market.
The Social and Ethical Consequences of Broadband Since the main argument of the chapter will aim at the question of responsibility for and in broadband provision, we need to introduce the idea of broadband as something more than just a technology. The discussion of the reasons why politicians are pushing for broader broadband provision has already pointed to some of the social and ethical consequences that broadband is believed to have or to develop. If we have already entered or are about
to enter the information era and live in a knowledge society, then it is clear that access to information is of central concern. This has to do with the fact that information work is a fast growing sector of meaningful employment, whereas most other sectors (agriculture, production) are in decline, at least in industrialized countries. Access to information is thus of importance for employment, but also for healthcare, social interaction, and entertainment, and it affects most areas of life. The often implicit argument then is that in order to be able to participate in the knowledge society, one will need technical access to information. In current terms this means that some form of broadband access is required to participate. Lack of such technology then means lack of inclusion of society. This leads us to the issue of digital divides (cf. Trauth, Howcroft, Butler, Fitzgerald, & DeGross, 2006; Rookby & Weckert, 2007). The digital divide and the discussion about “information-rich” and “information-poor” is an established topic of applied ethics. Normally the discussion refers to the large gap between the Third and the First Worlds in respect to the level of available telecommunication infrastructures and computers as such: the level of how people are able to use computers, the level of competency, and the level of how the issue is recognized by public policies. Recently this discussion got a new dimension. Politicians and scientists had to realize that the digital divide is not only a matter of fact EHWZHHQWKHWZRµZRUOGV¶WKHGLJLWDOGLYLGHLVDOVR a matter of fact within industrial countries. The digital divide is a good concept to understand why broadband technology can be viewed in terms of social and ethical responsibility. It allows a discussion of which rights citizens should have, KRZIDUHFRQRPLFFRQVLGHUDWLRQVVKRXOGLQÀXHQFH political decisions, and it is based on a percepWLRQWKDWIDLUQHVVPDWWHUVLQWRGD\¶VVRFLHWLHV$Q important consideration here is that the economic consequences of technology feed directly into its ethical evaluation. Precisely because broadband is VHHQDVHFRQRPLFDOO\EHQH¿FLDODFFHVVWRLWFDQ be interpreted as an ethical question.
229
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
THE GERMAN SITUATION The current state in many rural areas across Europe regarding broadband diffusion is quite bad. Many regions cannot be equipped by ordinary $'6/RU¿EHUVLQFHLWLVWRRFRVWO\IRU,QWHUQHW service providers (ISPs) to improve the existing infrastructure. The obvious solution to this problem is to equip rural and remote areas via a wireless technology (Sari, 2000; Vaughan-Nichols, 2004). In Germany this poses particular problems for ISPs investing in broadband infrastructure. One problem is the competitive character of possible broadband technologies of different ISPs. The other problem is the technology “pool” of one single ISP, for example, German Telekom, which LQÀXHQFHVLQYHVWPHQWGHFLVLRQVWUHPHQGRXVO\,Q the year 2000 more than 99.3 billion Deutsche Mark4 DURXQG ELOOLRQ ¼ ZHUH SDLG WR WKH German government by big mobile ISPs for the UMTS frequency spectrum. UMTS and the next generation of UMTS, HSDPA, enable providers to offer wireless broadband and even mobile wireless broadband services. But since these providers paid such a substantial amount of money for the licenses, they will need a fast return of investment. Thus, the investments in UMTS networks were
undertaken in large and urban cities/regions with a high population density (see Figure 1). In addition, other wireless technologies capable of delivering EURDGEDQG HJ:L0$; DSSHDUHG7KLVFUHDWHVD YHU\GLI¿FXOWVLWXDWLRQIRUWKHODUJHZLUHOHVV,63VLQ Germany. Large companies are not willing to risk substantial investments in rural and remote areas since none of them knows which market risks they KDYHWRIDFHE\FRPSHWLQJWHFKQRORJLHVOLNH¿EHU cable TV, or new and future wireless technologies OLNH:L0$;E\RWKHUSURYLGHUVRUE\WKHPVHOYHV *HUPDQ7HOHNRPIRULQVWDQFHRZQVWKH¿EHULQfrastructure in Germany, including the “last mile.” On the other hand, the company also owns UMTS licenses. The decision to set up a total-coverage UMTS network will probably cause a market loss LQWKH¿EHUPDUNHW $QRWKHUTXLWHUHFHQWH[DPSOHRIKRZGLI¿FXOW the German situation seems to be is the “Broadband Wireless Access” auction by the German Federal Network Agency (FNA). In December 2006 the FNA, which is responsible for the regulation of the telecommunication market in Germany, sold for a WRWDODPRXQWRIPLOOLRQ¼WKHVRFDOOHG³%URDGEDQG:LUHOHVV$FFHVV´ ³:L0$;´ OLFHQVHVLQWKH frequency spectrum of 3.4-3.6 GHz. The ISPs that now own these new licenses are not the same ones
Figure 1. UMTS coverage in Germany by the largest mobile telecommunication providers: Vodafone (left) and T-Mobile (right)3
230
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
as the ISPs that own the UMTS spectrum since none of those providers participated in the BWA auction. Compared with the results of the 2000 UMTS auction, there is a huge difference in the amount of money for frequency spectrums, which can basically be used to offer the same products and services. Products and services will constitute the overall market share for ISPs. At the end, this will be the ultimate indicator and value for such licenses. Wireless broadband services across Europe can also be delivered by ordinary WLAN. There are several commercial and non-commercial projects within the EU where ordinary WLAN is used to cover regions with broadband Internet services. Additionally the EU decided in 2003 that the frequency spectrum for WLAN has to be made available free of charge. This, of course, makes the decision for ODUJH,63VHYHQPRUHGLI¿FXOWZKHQWKH\KDYHWR decide which technology for what kind of region LVFRPPHUFLDOO\MXVWL¿DEOH 7KLVLVYHU\EULHÀ\WKHVLWXDWLRQRQDPDFUR level from a mere business perspective. The main focus of this chapter is which consequences such a situation has for rural and remote regions, and how social and political responsibility can be ascribed to certain actors.
Literature Review In recent years the diffusion of broadband has EHFRPHDPDMRUWRSLFLQYDULRXV¿HOGVRIUHVHDUFK DQGGLVFLSOLQHV7KH¿UVWHGLWLRQRIWKHELEOHRIGLIfusion research called “Diffusion of Innovations” E\(YHUHWW05RJHUVZDV¿UVWSXEOLVKHGLQ (Rogers, 2003). This book outlined a theory how new technologies or innovations diffuse within a society within a certain place and by a certain time. If one accepts that broadband is one of those innovations, then one can inquire about its diffusion. Some studies chose a normative approach. In such studies broadband is seen as an entity which can help to improve a situation within a society, for example, bridging the digital divide (Prieger, 2003; Oakes, 2004; Mbebe & Sewsunker, 2004; Zhang & Wolff, 2004; Gubbins, 2003a). Other studies use an analytical approach. Here the diffu-
sion process as such in a particular country society is described and analyzed in respect of adoption (Tookey, Whalley, & Howick, 2006; Adams, 2006; Papacharissi & Zaks, 2006). In most of the studies, recommendations for particular political actions are suggested such as how to overcome the demand/supply problem in the case of broadband GLIIXVLRQ /LQGVNRJ -RKDQVVRQ +ROOL¿HOG & Donnermeyer, 2003). But also success stories of broadband diffusion (Bradley, Eng, & Toms, 2004) and the role and status of broadband are UHSRUWHG HJ;DYLHU 7KLVFKDSWHULGHQWL¿HVWKHSROLFLHVDQGLQLWLDWLYHVXVHGWRHQFRXUDJH broadband awareness, availability, and adoption by presenting a case study.
Methodology: Action Research The methodological approach for our research ZDVµDFWLRQUHVHDUFK¶³$FWLRQUHVHDUFKZDVRULJLQDOO\SURSRVHGE\/HZLQ DQGLQÀXHQFHGE\ work at the Tavistock Institute (Rapoport, 1970; Trist, 1976). It uses intervention into problematic VRFLDOVLWXDWLRQVDVDPHDQVWRGHYHORSVFLHQWL¿F knowledge. Different action research approaches have been developed, one of the best known being 6XVPDQDQG(YHUHG¶VDFWLRQUHVHDUFKF\FOHFRQVLVWing of diagnosing, action planning, action taking, evaluating, and specifying learning (Davison et al., 2004; Susman & Evered 1978)” (quoted in Iversen, Mathiassen, & Nielsen, 2004, p. 397). Action research summarizes different methods of the social sciences under one umbrella. “Action research challenges the claims of a positivistic view of knowledge which holds that in order to be credible, research must remain objective and value-free” (Brydon-Miller, Greenwood, & Maguire, 2003, p. 11). We chose this method since one of the authors was heavily involved in the public-awareness initiative called AltmarkDSL, which emphasized the importance of broadband for rural and remote areas. Thus, the research and UHÀHFWLRQV RI WKH DXWKRU ZKR LV DOVR D OHDGLQJ ¿JXUHRIWKHORFDOLQLWLDWLYH$OWPDUN'6/KHDYLO\LQÀXHQFHGWKHIXUWKHUDFWLRQVRIWKHLQLWLDWLYH and thereby the actions of the other actors, like politicians, local authorities, and so forth. “There
231
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
is universal agreement among action researchers that it must be collaborative. The ultimate aim of the researcher is to empower those with whom he/she works to improve their condition. At the same time, the researcher typically enters the collaboration with particular political ideals or goals; the collaboration is typically intended to further this agenda” (Gergen, 1999, p. 100). We hold that action research seems to be the most appropriate approach to the research setting/object in our case. One of the implicit assumptions of an action UHVHDUFKSURMHFWLVWKDWWLPHDQGWKHUHÀHFWLRQVRI WKHDXWKRUZLOOLQÀXHQFHWKHHPSLULFDOVSKHUHRU the empirical setting. Thus, the researcher will determine which research subjects will enter the research scene and why. In the course of time this will become a dialectical process. Some defenders of action research even hold that the research as such must give an impact on the social and political situations in society (cf. Brydon-Miller et al., 2003). This is, of course, a fairly bold statement since a value-laden research approach in the social sciences was an object of rejection by many authors in the past. However, we hold that in order to generate valid data about the research subject, the UHVHDUFKHUPXVWEHSDUWRIWKH¿HOG,IWKLVKDSSHQV LQWHQWLRQDOO\DQGXSRQUHÀHFWLRQWKHQWKHZKROH research will generate very valuable empirical data DQGNQRZOHGJHDERXWDFHUWDLQ¿HOG In our case study we tried to identify which moral agents could potentially carry a responsibility for the diffusion and adoption process of broadband. Thus, at the beginning of the research process, it was not clear which moral agents would be the subject of research. When the impact of the local initiative AltmarkDSL started to develop, more DQGPRUHPRUDODJHQWVHQWHUHGWKH¿HOG,QRUGHU to understand the background of the initiative and why we chose action research as our method, we will now present the case study that outlines the SUREOHPVDQGGLI¿FXOWLHVRIEURDGEDQGGLIIXVLRQ and adoption in a rural and remote area in SaxonyAnhalt, Germany.
232
Case Study from Germany: AltmarkDSL.de Before we turn the focus on how one can ascribe UHVSRQVLELOLW\WRDQDFWRUZH¿UVWKDYHWRLGHQWLI\ which actors we have to deal with. The diffusion of broadband can be studied from a bottom-up or topdown perspective. From the top-down perspective the research question will mainly rely or focus on economic analysis and the easy-to-identify actors like governments, ISPs, trade agencies, laws, and so forth. The bottom-up approach is much harder to describe. Of course, there are the potential costumers, local authorities, social networks, and the local infrastructure which we can ascribe responsibility to. But the situation might differ from region to region and from nation to nation. This makes it impossible to really tell who is in charge or responsible since potentially every citizen can be held responsible. In our study we focused on a grassroots movement called AltmarkDSL. AltmarkDSL is a group of people who tried to create public awareness in an rural and remote region in middle East Germany. This region is called “Altmark.” Altmark is left behind in many ways. A low population density (48 people per square kilometer) and a high unemployment rate (approximately 22%) make this region one of the poorest in Germany. However, it is a beautiful place to live with low industry density, beautiful landscapes, and no motorways. In order to close the gap to more developed UHJLRQVDQGLQRUGHUWRSUR¿WIURPWKHSRWHQWLDOV and options of a service and knowledge society, this kind of region is in special need of broadband connections since new forms of work and employment can be based on broadband infrastructures. Broadband could play an important role in addressing the infrastructural problems of this area. As indicated above, the EU and the federal government are in support of further broadband development. At the same time, there are few economic incentives for commercial ISPs to provide access to the population. Additionally, there is the very disadvantageous situation of the German telecommunication sector for such remote and rural regions.
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
This was the point of departure for some local individuals in September 2004 to form an initiative which “publicized” the importance and potential of broadband for this region. At the same time they presented the possible option for total broadband FRYHUDJH :L0$; :L0$; ZDV DW WKDW WLPH a wireless broadband technology that was not standardized. However, the German FNA granted ³H[SHULPHQWDOOLFHQVHV´WRSRWHQWLDO:L0$;SURviders. The Swedish city of Skeleftea was one of WKH¿UVWFLWLHVLQ(XURSHWRXVHWKLVQHZWHFKQRORJ\ to cover a whole region with wireless broadband Internet connections. So, technically this technology should have also been working for the German UHJLRQ$OWPDUN:L0$;LVFDSDEOHRIGHOLYHULQJ high-speed Internet connections over a distance of up to 20 kilometers with a speed of up to 10 Mbit/s. The local initiative published ideas and concepts of how this region could be covered with wireless broadband connections. Members of parliament, local authorities, mayors, business development agencies, local companies, and interested citizens were asked to participate in advertising the idea of broadband in rural Altmark. The results were interesting. Companies and politicians participated little in that campaign. The local business development agency decided not to support that campaign. Their argument was that this is solely a matter of business and had to be resolved by business. The same position was declared by the District Administration, which has the highest political power in a region like Altmark. These reactions showed several facts. First, the importance of broadband was underestimated. The national (“Broadband-Atlas”) and international campaigns (“eEurope 2005”) for broadband diffusion and adoption did not reach the local levelin this particular case. The demand/supply failure was misunderstood or not recognized. The broadband market for the total coverage of broadband in Altmark did not emerge, because the demand was, generally speaking, too small. Only larger cities, which were supplied with DSL-broadband from German Telekom, had the chance to get a hold of broadband. On the other hand, as a consequence this exacerbated the situation for rural villages.
Approximately half a year after the campaign was started, it became clear that even citizens in such small villages still trusted the competence and promises of the largest ISP in the DSL market in Germany, German Telekom, to coversooner or laterthe region via cable DSL. Thus, the probroadband campaign had to face the trust of citizens in large ISPs. That made the potential market for alternative wireless broadband providers even worse. For this reason the AltmarkDSL initiative started a second campaign for a small company which offered to cover seven small villages via WLAN, another wireless technology capable of delivering broadband Internet. At the end, this campaign was successful. Today there are six villages which have broadband access of up to 2 Mbit/s via WLAN technology. Given our involvement and familiarity with the AltmarkDSL case, which stems from our active participation in the project, the question of theoretical interest we wish to discuss here is who is or can be held responsible for broadband provision. We will use the case study to explore this question in some depth. One guiding question here is whether a theory of responsibility will allow us to develop principles or practices that will facilitate the evaluation of comparable situations with the aim of overcoming some of the practical problems we have faced.
Responsibility for (or in?) Broadband The central question of this chapter is who should be held responsible for the diffusion of innovations, in this case: the diffusion of broadband. In the social sciences there had been extensive research on the “diffusion of innovation” problem (Rogers, 2003). 'LIIXVLRQ FDQ EH GH¿QHG ³DV SURFHVV E\ ZKLFK an innovation is communicated through certain channels over time among the members of a social system” (Rogers, 2003, p. 11), which is also the GH¿QLWLRQRIGLIIXVLRQ³7KHLQQRYDWLRQGHFLVLRQ process is essentially an information seeking and information-processing activity in which an individual is motivated to reduce uncertainty about the advantages and disadvantages of the innovation” (Rogers, 2003, p. 14). What will the advantages and
233
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
disadvantages be in my situation? “Observability is the degree to which the results of an innovation are visible to others. The easier it is for individuals to see the results of an innovation, the more likely they are to adopt” (Rogers, 2003, p. 16). This is where the chapter will introduce the concept of responsibility in order to investigate the question how such a dilemma can be understood and possibly overcome. Structurally, responsibility is the ascription of an object to a subject with the aim of setting sanctions. The whole social construct of responsibility is meant to improve social life. Responsibility ascriptions are complex social interactions that need to consider a range of angles and dimensions. We will structure this discussion by investigating each of the major dimensions of UHVSRQVLELOLW\XVLQJWKHZRUNRQUHÀHFWLYHUHVSRQsibility by Stahl (2004). We explicitly use the term “responsibility” and not “accountability” since our methodological approach does not allow the ascription of accountability to certain moral agents. This assumption relies on the Aristotelian distinctions of responsibility. Aristotle made a distinction between the merit-view of responsibility; that means a moral agent can be praised or blamed because of the actions he did or did not do, and the consequentialist view, where an action is judged to be good or bad dependent on the outcome. Accountability would presuppose that the subject of responsibility can EHSUHGH¿QHG$VDOUHDG\VWDWHGLQWKLVFKDSWHU this was impossible in our case study and with our methodological approach. On the other hand it was under no circumstances clear which moral agents could be held accountable for which actions. The synergy and network effects made it simply impossible to ascribe a certain accountability to a certain agent. That is why we chose the concept of responsibility as our point of departure for the analysis of the case study.
Object and Subject of Responsibility As indicated above, there are three major dimensions that need to be considered in each responsibility ascription: the subject, the object, and the authority or normative background. The subject is the one who is held responsible, the object is
234
what the subject is responsible for, and the authority provides the norms on which responsibility is ascribed and enforces sanctions. In our case the easiest of these three dimension seems to be the object. It is the provision of broadband access to the population of the Altmark region. All is not as easy as it seems, however. While the general aim of providing broadband access seems to be ZHOOHVWDEOLVKHGDQGQRWFRQWHQWLRXVWKHVSHFL¿F underlying reasons for requiring the object seem less clear. Broadband provision can be desired for economic, political, ethical, or other reasons. Depending on the underlying reason for the desire for broadband, different subjects, different norms, and different sanctions would be applicable. The object of responsibility ascription changed dramatically, depending on whether broadband was seen as a purely technical piece of infrastructure, a means for regional development, a vehicle for economic growth, or a source of higher standard RIOLYLQJDQGPRUHIXO¿OOHGLQGLYLGXDODQGFROOHFtive activities. The District Administration or the local initiative AltmarkDSL may serve as an example for a subject of responsibility which neglected the meaning of broadband as a “vehicle” to develop the region economically. Both subjects recognized their responsibility in two totally different directions. The District Administration rejected the responsibility by pointing the market. Their argument was that market forces have to resolve the broadband issue (object of responsibility). As PHQWLRQHG HDUOLHU WKH GLI¿FXOW VLWXDWLRQ LQ WKH broadband market in Germany leads the market forces not to invest in rural areas. This was already described by a number of studies (Lindskog & Johansson, 2005; Oakes, 2004; Prieger, 2003). Thus, the supply model fails because the market does not emerge. However, the demand for broadband is still prevailing because the people are in need of broadband, and at least some articulated this by participating in AltmarkDSL or in other circumstances. Further, subjects are the individuals or entities that are held responsible. In our case there are several contenders for the role of subjects. Politicians, ISPs, markets, the people, and many more
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
moral agents can be held responsible for successful diffusion and adoption of broadband in the rural area Altmark. As we saw in the case study, the local initiative AltmarkDSL publicized the need for broadband for the rural and remote area Altmark. AltmarkDSL is a voluntary and independent organization which articulated an ethical concern. This ethical concern is highlighted in much of the digital divide literature, and in national and international political initiatives which emphasized the importance of broadband. From July 2004 onwards AltmarkDSL tried to raise the awareness for the “broadband problem” to local politicians and MPs, who represent the region in the parliament of the county and the German Bundestag. This approach failed. None of the local politicians responded to this initiative. Additionally, AltmarkDSL tried to deliberate the issue with local entities that are responsible for the economical development (IGZ) of the rural region Altmark. Even here the response was a mere rejection of the issue. It is interesting to note that the IGZ noted that broadband coverage is a solely business issue. The ISPs must decide if they want to invest in the broadband infrastructure in the rural area. The ISPs on the other hand, as described earlier in this chapter, have little incentives to invest in a rural region with a low population density. Thus, it seems that the main responsibility lies in the hands of local authorities like politicians or the IGZ. The crucial point here is, these authorities rejected more or less the accountability for the moral, political, and social responsibility for this particular problem because they saw broadband as a mere “product,” like ice cream or a car, and not as an important multiplicative tool for the positive GHYHORSPHQWRIWKHUHJLRQ+HUHZHKDYHLGHQWL¿HG the crucial fact which generates an ethical issue in broadband. If authorities do not recognize the true value of an entity (here: broadband) or are not willing to evaluate the consequences the diffusion of broadband will have for the rural region, then they are blameworthy. So, the diffusion of broadband becomes an ethical issue.
Authority and Norms The subject is held responsible on the basis of accepted norms, and these are considered and sanctioned by the authority of the ascription. The authority in our case is, on the one hand, the general political agreement on the national level that broadband constitutes an important infrastructure for the people in order to take part in a modern IT society. The fact that broadband was discussed as an issue of Universal Service Obligation underlines this importance. On the other hand, Internet service providers and business as such emphasize the importance of broadband. Thus, politics and business declare in unison the high status of broadband. Our point here is that this seems only to “work” on a macro level. However, this accepted view of the importance of broadband on a macro level does not lead to provision-positive politics on a micro level. That would be the general expectation of the public and AltmarkDSL. The people themselves, as the authority, articulated within the initiative AltmarkDSL their desire for broadband. So, implicitly AltmarkDSL articulated an ethical claim (“broadband for the region”) which made at least every local authority blameworthy because these institutions do have the responsibility to respond to that claim.
Typical Responsibility Ascriptions Just describing all possible permutations of responVLELOLW\DVFULSWLRQVZRXOGJREH\RQGWKHFRQ¿QHV of this chapter. They are also not all relevant or useful. What the above enumeration of some of them should have done, however, is show that there is no simple linear way of ascribing responsibiliWLHV7KHPRVWSURPLQHQW¿JXUHVRIUHVSRQVLELOLW\ ascription in our case study were the following: the local District Administration, MPs, IGZ, ISPs, and AltmarkDSL. AltmarkDSL was during the period of time very well recognized as a successful grassroots movement which fought for needed technology. It ZDVWKH¿UVWHQWLW\ PRUDODJHQW ZKLFKDUWLFXODWHG WKHQHHGIRUEURDGEDQGDQGSRLQWHGWRWKHGLI¿FXOW
235
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
market situation in rural and remote areas for broadband provision. The crucial point here is that the local District Administration, MPs, ISPs, as well as the IGZ did not have to fear any bad consequences. Ascription of moral responsibility to these agents did not really cause bad consequences for them. The empirical evidences proved that all ethical economical and social concerns did not provide enough reasons and forces to stimulate actions of one of those moral agents.
CONCLUSION In our case the overall objective seems clear, namely the provision of broadband access to the rural population of Altmark. All other aspects are much less clear. These include the subject (who is responsible?) as well as the sanctions (rewards or punishments). Responsibility can be of different types (legal, moral, religious) and there seems to be a mix of different types in our case. It is not clear on what grounds responsibility should be ascribed, nor is it obvious how achievement is to be measured and how the sanctions can be implemented. We are thus looking at a complex net of responsibilities, some of which established, others still emerging, that currently do not facilitate a solution. One possible way to address this SUREOHPLVWKHFRQFHSWRIUHÀHFWLYHUHVSRQVLELOLW\ 6WDKO 5HÀHFWLYHUHVSRQVLELOLW\VWDQGVIRU the attempt to render responsibility ascriptions realizable. One important aspect of this is that the process of ascription must be undertaken in a participative manner to include the voices of those who are affected. In our case study there is a local imitative, AltmarkDSL, which conforms to the ideas of UHÀHFWLYHUHVSRQVLELOLW\EHFDXVHLWLVDERWWRPXS population-driven attempt to provide broadband. In the chapter we explored how this initiative articulated and may have changed existing perceptions of responsibility and assumptions about relevant realities (“broadband diffusion”).
REFERENCES $GDPV 3 ,VRODWLQJ µZK\¶ $XVWUDOLDQ households adopt broadband. Telecommunication Journal of Australia, 56(3-4), 27-36. Aronsson, M., Tholén, D., Josephson P.-E., Li, H., & Kong, S. (2003). Broadband services for residential and commercial tenants: A categorisation of current and future services and a survey on tenants needs in Sweden. Building and Environment, 38(2), 347-358. Bandias, S., & Vemuri, S.R. (2005). Telecommunications infrastructure facilitating sustainable development of rural and remote communities in Northern Australia. Telecommunications Policy, 29(2-3), 237-249. Bittlingmaer, G., & Hazlett, T.W. (2002). Open access: The ideal and the real. Telecommunications Policy, 26(5-6), 295-310. Boscher, C., Hill, N., Leiné, P., & Candido, A. (2003). Providing always-on broadband access to under-served areas. Alcatel Telecommunications Review, (3), 127-132. Bradley, L., Eng, P., & Toms, M. (2004). Rural broadband networks and economic development: The South Dundas story. Proceedings of the Fourth IASTED International Multiconference on Wireless and Optical Communications (art. no. 424-057, pp. 573-578). Brydon-Miller, M., Greenwood, D., & Maguire, P. (2003). Why action research? Action Research, 1, 9-28. Elnegaard, N.K., Olsen, B.T., & Stordahl, K. (2003). Broadband deployment in rural and non-competitive areas: The European perspective. Journal of the Communications Network, 2(3), 51-54. Filder, A.J.H., Hernandez, G., Lalovic, M., Pell, T., & Rose, I.G. (2002). Satellitea new opportunity for broadband applications. BT Technology Journal, 20(1), 29-37. Gergen, K.J. (1999). An invitation to social construction (p. 100). London: Sage.
236
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
Ghosh, A., Wolter, D.R., Andrews, J.G., & Chen, R. (2005). Broadband wireless access with :L0$; &XUUHQW SHUIRUPDQFH EHQFKmarks, and future potential. IEEE Communications Magazine, 43(2), 129-136. *ULPHV6 +RZZHOODUH(XURSH¶VUXUDO businesses connected to the digital economy? European Planning Studies, 13(7), 1063-1081. Grubesic, T.H. (2004). The geodemographic correlates of broadband access and availability in the United States. Telematics and Informatics, 21(4), 335-358. Grubesic, T.H. (2003). Inequities in the broadband revolution. Annals of Regional Science, 37(2), 263-290. Grubesic, T.H., & Murray, A.T. (2004). Waiting for broadband: Local competition and the spatial distribution of advanced telecommunication services in the United States. Growth and Change, 35(2), 139-165. Gubbins, E. (2003a). Broadband for everyone. Telephony, 244(4), 18. Gubbins, E. (2003b). Rural broadband funding going fast in future budgets. Telephony, 244(11), 12. Gutierrez, V. (2002). Broadband and the Internet. Arthroscopy, 18(6), 654-657.
provement: An action research approach. MIS Quarterly, 28(3), 395-433. Kaigo, M. (2005). Can the WSIS declaration principle and plan of action work in Japan? Digital VWUDWL¿FDWLRQRI-DSDQHVHVRFLHW\Telematics and Informatics, 22(4), 333-347. Lindskog, H., & Johansson, M. (2005). Broadband: A municipal information platform: Swedish experience. International Journal of Technology Management, 31(1-2), 47-63. Madden, G., Savage, S.J., Coble-Neal, G., & Bloxham, P. (2000). Advanced communications policy and adoption in rural Western Australia. Telecommunications Policy, 24(4), 291-304. Mason, R., & Rennie, F. (2004). Broadband: A solution for rural e-learning? International Review of Research in Open and Distance Learning, 5(1). Mbebe, Z., & Sewsunker, R. (2004). xDSL as a broadband solution for the developing countries. IEEE AFRICON Conference, 1, (pp. 155-160). Murray, B. (2003). Delivering education in remote areas by Internet via satellite broadband. Telecommunication Journal of Australia, 53(1), 59-63, 71. Oakes, S. (2004). The Internet and regional inequality. Geography Review, 18(2), 24-29.
Henten, A., & Kristensen, T.M. (2000). Information society visions in the Nordic countries. Telematics and Informatics, 17(1), 77-103.
Ohata, K., Kobayashi, K., Nakahira, K., & Ueba, M. (2005). Broadband and scalable mobile satellite communication system for future access networks. Acta Astronautica, 57(2-8), 239-249.
+ROOL¿HOG &$ 'RQQHUPH\HU -) &UHDWLQJGHPDQG ,QÀXHQFLQJLQIRUPDWLRQWHFKQROogy diffusion in rural communities. Government Information Quarterly, 20(2), 135-150.
Papacharissi, Z., & Zaks, A. (2006). Is broadband the future? An analysis of broadband technology potential and diffusion. Telecommunications Policy, 30(1), 64-75.
Ilcev, S.D. (2004). Inmarsat satellite-based global business solutions for remote and rural communications. IEEE AFRICON Conference, 1, (pp. 31-37).
Park, S., & Yoon, S.-H., (2005). Separating earlyadopters from the majority: The case of Broadband Internet access in Korea. Technological Forecasting and Social Change, 72(3), 301-325.
Iversen, J.H., Mathiassen, L., & Nielsen, P.A. (2004). Managing risk in software process im-
Pastrover, S. (2003). The prospects for broadband deployment in rural America. Government Information Quarterly, 20(2), 95-106.
237
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
P a t e l , D. , & M a n n i g s , R . (2 0 0 2) . 5HÀH[personalised wireless interaction in a broadband environment. BT Technology Journal, 20(1), 38-46. Philpott, M. (2003). Broadband technologies and rural areas. Journal of the Communications Network, 2(1), 9-16. Prendergast, D., Guillet, M., & Caron, B. (2004). Broadband Internet multimedia service return channel using DTV for rural and remote access. Proceedings of Fifth World Wireless Congress (pp. 166-171). Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363. Rao, B., & Parikh, M.A. (2003). Wireless broadband drivers and their social implications. Technology in Society, 25(4), 477-489. Rogers, E.M. (2003). Diffusion of innovations. New York: The Free Press. Rookby, E., & Weckert, J. (Eds.). (2007). Information technology and social justice. Hershey, PA: Idea Group. Savage, S.J., & Waldman, D. (2005). Broadband Internet access, awareness, and use: Analysis of United States household data. Telecommunications Policy, 29(8), 615-633. Sari, H. (2000). Trends and challenges in broadband wireless access. Proceedings of the Symposium on Communications and Vehicular Technology (pp. 210-214). Sinha, A., Mitchell, K., & Medhi, D. (2005). NetZRUNJDPHWUDI¿F $EURDGEDQGDFFHVVSHUVSHFWLYH Computer Networks, 49(1), 71-83. Stahl, B.C. (2004). Responsible management of information systems. Hershey, PA: Idea Group. Strover, S., Chapman, G., & Waters, J. (2004). Beyond community networking and CTCs: Access, development, and public policy. Telecommunications Policy, 28(7-8), 465-485.
238
Tookey, A., Whalley, J., & Howick, S. (2006). Broadband diffusion in remote and rural Scotland. Telecommunications Policy, 30(8-9), 481-495. Trauth, E., Howcroft, D., Butler, T., Fitzgerald, B., & DeGross, J. (Eds.). (2006). Social inclusion: Societal & organizational implications for information systems. Boston: Springer. Van Gorp, A.F., Maitland, C.F., & Hanekop, H. (2006). The broadband Internet access market: The changing role of ISPs. Telecommunications Policy, 30(2), 96-111. Vaughan-Nichols, S. (2004). Achieving wireless EURDGEDQG ZLWK :L0$; Computer, 37(6), 1013. Venkatachalam, S., & McDowell, S. (2003). What is broadband? Where is “rural”? Government Information Quarterly, 20(2), 151-166. Wang, R.Y., Sobti, S., Garg, N., Ziskind, E., Lai, J., & Krishnamurthy, A. (2004). Turning the postal system into a generic digital communication mechanism. Computer Communication Review, 34(4), 159-166. Warner, T.D., Monaghan-Geernaert, P., Battaglia, J., Brems, C., Johnson, M.E., & Roberts, L.W. (2005). Ethical considerations in rural health care: A pilot study of clinicians in Alaska and New Mexico. Community Mental Health Journal, 41(1), 21-33. ;DYLHU 3 6KRXOG EURDGEDQG EH SDUW RI Universal Service Obligations? Info, 5(1), 8-25. Zhang, M., & Wolff, R.S. (2004). Crossing the digital divide: Cost-effective broadband wireless access for rural and remote areas. IEEE Communications Magazine, 42(2), 99-105. Zhang, M., & Wolff, R.S. (2004). Using WiFi for cost-effective broadband wireless access in rural and remote areas. Proceedings of the 2004 IEEE Wireless Communications and Networking Conference (WCNC 2004) (vol. 3, pp. 1347-1352).
Social, Political, and Ethical Responsibility in Broadband Adoption and Diffusion
KEY TERMS
ENDNOTES 1
Action Research: A research method origiQDOO\SURSRVHGE\/HZLQDQGLQÀXHQFHGE\ZRUN at the Tavistock Institute. It uses intervention into problematic social situations as a means to develop VFLHQWL¿FNQRZOHGJH Broadband: 7KHUHLVQRFOHDUFXWGH¿QLWLRQRI “broadband.” Broadband can be used synonymous for every technology for high-speed transmission RI GDWD 6RPH GH¿QLWLRQV RQO\ FRQVLGHU VSHHGV DERYH0ELWVDVEURDGEDQG2IWHQWKHGH¿QLWLRQV DOVRLQFOXGHWKHFDSDFLW\RIµDOZD\VRQ¶DVW\SLFDO for broadband. Case Study: A form of qualitative descriptive research. It refers to the collection and presentation of detailed information about a particular group RULQGLYLGXDODQGWKHVSHFL¿FFRQWH[W7KHFRQFOXsions drawn from a case study research project are only applicable and valid for this particular case and context. The emphasis of case study research projects is placed on exploration and description. Diffusion of Innovation: (YHUHWW5RJHUVGH¿QHV diffusion as the process by which an innovation is communicated through certain channels over time among the members of a social system.
2
3
4
German: Die Koalitionsparteien werden zur Sicherung der Zukunft des Industrieund Forschungsstandorts Deutschland Anreize für den Aufbau bzw. Ausbau moderner und breitbandiger Telekommunikationsnetze schaffen. Taken form the “Coalition Contract” of the Social Democratic Party of Germany and the Christian Democratic Union of Germany from November 11, 2005. Retrieved June 19, 2006, from http://koalitionsvertrag. spd.de/servlet/PB/show/1645854/111105_Koalitionsvertrag.pdf Retrieved June 6, 2006, from http://europa.eu.int/rapid/pressReleasesAction. do?reference=MEMO/05/ 165&format=HT ML&aged=0&language=en&guiLanguage =en Vodafone map retrieved February 20, 2007, from http://netmap.vodafone.de; T-Mobile map retrieved February 20, 2007, from http://t-map.t-mobile.de http://www.bundesnetzagentur.de/media/archive/214.pdf (p. 26)
Ethics: According to http://onlineethics.org, “the term ethics is used in several different ways. It means the study of morals. It is also the name for that branch of philosophy concerned with the nature of morals and moral evaluation, e.g., what is ULJKWDQGZURQJYLUWXRXVRUYLFLRXVDQGEHQH¿FLDO or harmful (to others).” Responsibility: Structurally responsibility is the ascription of an object to a subject with the aim of setting sanctions. The whole social construct of responsibility is meant to improve social life. Responsibility ascriptions are complex social interactions that need to consider a range of angles and dimension.
239
Division V
North America
241
Chapter XVI
Competition in Broadband Provision and the Digital Divide Wei-Min Hu Peking University, China James E. Prieger Pepperdine University, USA
ABSTRACT This chapter examines the supply of DSL broadband by the incumbent local exchange company (LEC) LQ¿YH86VWDWHVLQWKHHDUOLHU\HDUVRIGHSOR\PHQW2XUHPSLULFDODQDO\VLVVKRZVWKDWLQFRPHRWKHU demographics, and cost factors are important determinants of entry and availability. After controlling for other factors, the racial characteristics of the area do not affect DSL provision. Active competition in broadband from competitive LECs reduces deployment of DSL by the incumbent, but potential competition from competitive LECs has the opposite effect. Competition from cable companies also negatively LQÀXHQFHVWKHLQFXPEHQW¶VGHFLVLRQWRVXSSO\'6/2XUREMHFWLYHLQJDXJLQJWKHLPSRUWDQFHRIWKHYDULRXV factors is to highlight the important drivers of broadband provision for policymakers.
INTRODUCTION The worldwide explosion in the growth of broadband infrastructure is allowing countries to enjoy productivity gains in industries that heavily use communications. Consumers also are deriving LQFUHDVHGEHQH¿WVIURPWKHDYDLODELOLW\RIEURDGband connection to the Internet, as the technology speeds up some applications (e.g., downloading music) and creates entirely new possibilities (e.g., telemedicine).1 Households typically connect to
the Internet through digital subscriber line (DSL) service provided on their telephone line or through cable modem service. Although broadband technology of these and other types is diffusing rapidly, some policymakers in the U.S. and abroad are concerned that not all regions within countries are receiving broadband access at the same time. The phenomenon of unequally diffusing information and communication technology is known as the “digital divide.” The digital divide in the United States takes the form of a well-documented gap
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Competition in Broadband Provision and the Digital Divide
in computer and Internet usage between richer and poorer households, majority and minority groups, and urban and rural areas (Newberger, 2001; NTIA, 2002; Mills & Whitacre, 2003; Fairlie, 2004). Broadband Internet access has become one of the latest aspects of the digital divide to be discussed, examined, and lamented in arenas of public policy. In the early years of broadband Internet adoption, the U.S. Department of Commerce (NTIA, LVVXHG DQ LQÀXHQWLDO UHSRUW VKRZLQJ WKDW black and Hispanic households had less access to broadband Internet access than did white households. The report also found that households in rural areas were less likely than urban households to subscribe to broadband access, and that lowerincome households lagged the subscription rates RI PRUH DIÀXHQW KRXVHKROGV 7KH 86 )HGHUDO Communications Commission (FCC) opined that low subscription to broadband services in rural, low-income, and minority areas was due at least in part to a lack of availability (FCC, 2000a). Since availability—the supply of broadband infrastructure by communications providers—is necessary for the household to subscribe to broadband service, fundamental analysis of the broadband digital divide must begin on the supply side. What are the determinants of the supply side of the digital divide? Both the demographic characteristics of the territory and competition among broadband service providers play a role in entry in the broadband market. The racial composition DQGLQFRPHRIDPDUNHWPD\LQÀXHQFHEURDGEDQG supply directly through overt discrimination (sometimes called “redlining” in the U.S.) on the part of the providers, or more likely, according to the SUR¿W PD[LPL]DWLRQ K\SRWKHVLV 3ULHJHU 2003), through reducing the demand in an area DQGWKHUHIRUHWKHSUR¿W SHUFHLYHGE\DSRWHQWLDO supplier. Competition in local communications also affects broadband access. For example, incumbent telephone and cable companies may be more likely to provide broadband access to “meet the competition” when they face broadband competitors (the competitive stimulus hypothesis). On the other hand, if potential entrants deem a market large enough to support one broadband provider
242
but not two, then the availability of cable modem service (for example) will discourage deployment of DSL (the carrying capacity, or market size, hypothesis modeled by Bresnahan & Reiss, 1987). If competition is an important driver of broadband availability, then policy efforts to close the digital divide should encourage competition. Although the data we examine are from the U.S., the public policy concerns of the digital divide and the issue of the role of competition in broadband rollout are universal. In this chapter, we examine the supply of DSL broadband by the incumbent local exchange comSDQ\ /(& LQ¿YH0LGZHVWHUQVWDWHVLQWKH86 Our objective is to gauge the importance in the DSL entry decision of the various markets and competiWLYHIDFWRUVGLVFXVVHGDERYH7KH¿QGLQJVLQWKLV study show that there is no evidence that race per se matters in the supply of DSL once the impact of the income of the area is removed. However, household income does matter, with lower-income DUHDVUHFHLYLQJOHVVDFFHVV:HDOVR¿QGWKDWWKH education levels, average length of commute, size RI¿UPVLQWKHDUHDDQGFRVWVDIIHFWWKHGHSOR\PHQW of DSL. The effect of competition in the area is in accord with the carrying capacity hypothesis.
LITERATURE REVIEW A growing body of literature exists on the determinants of broadband availability in an area. Our study has the advantage of examining a precise geographical market for DSL deploymentthe wire center serving area. This allows us to gather exact demographic characteristics of the markets, as would be considered by the decision makers at the potential broadband providers. Previous studies in the economics literature looking at deployment throughout the U.S. typically examine postal service code areas2 )ODPP ;LDR Orazem, 2005; Grubesic & Murray, 2004; Prieger 2003) or counties (Gillett & Lehr, 1999), neither of which conform to telecommunications geography. Gabel and Kwan (2001) also use the wire center serving area as the unit of observation. Studies considering broadband diffusion within a state
Competition in Broadband Provision and the Digital Divide
or metropolitan area have used other geographical units. For example, Grubesic (2003) uses the ¿QHVWOHYHORIJHRJUDSK\DYDLODEOHIRUHDFKW\SH of provider (townships for cable modem providers and wire centers for DSL providers) in his study of broadband availability in Ohio. The studies on the determinants of broadband DYDLODELOLW\JHQHUDOO\FRQ¿UPWKDWUXUDODQGORZ income areas are less likely to have broadband available, but the set of variables included in the HVWLPDWLRQVFDQJUHDWO\LQÀXHQFHWKHFRQFOXVLRQV For example, Prieger (2003) uses a nationally comprehensive broadband survey by the FCC to examine the income and racial aspects of the supply side of the digital divide. The author shows that if the only area characteristics included in an estimation for the availability of broadband are the racial composition, urban density, and identity of the incumbent telephone company, then areas with more minorities appear to be less likely to have broadband. However, one cannot draw a causal link between race and supply-side overt discrimination, because once Prieger (2003) controls for income, WKHVWDWLVWLFDOVLJQL¿FDQFHRIWKHUDFLDOFRPSRVLWLRQ of the area disappears. Furthermore, when Prieger (2003) introduces additional demographic and business characteristics of the area to the estimation, the author found that there is no evidence of unequal availability based on race or income, except for Native Americans. Rural location and demand characteristics such as age, education, commuting time, gender, and size of businesses in the area were determinants of broadband availability that were more important. There are many more studies on household-level demand for broadband Internet access than there are on supply (see other chapters of this book for citations). A few studies (Burnstein & Aron, 2003; Wallsten, 2005) commingle supply and demand, and examine the factors determining the state-level penetration rate (the ratio of subscribers to total households in a state). These studies consider the impact of competition (among other factors) on the penetration rate. Burnstein and Aron (2003) ¿QG WKDW LQWHUPRGDO FRPSHWLWLRQ EHWZHHQ FDEOH FRPSDQLHVDQGWHOHFRPPXQLFDWLRQV¿UPVLVSRVLtively correlated with broadband penetration. In a
VLPLODUYHLQ:DOOVWHQ ¿QGVWKDWVWDWHVZLWK more competition in local telephony (as proxied by the number of telephone lines sold by competitive local exchange carriers) have higher broadband penetration rates. One disadvantage of using statelevel penetration rates is that the relatively small number of observations makes identifying driving factors problematic. The most ambitious of the broadband entry studies, those mentioned above using the FCC data from postal service code areas, are subject to criticism because broadband access is not necessarily ubiquitous in an area.3 Thus, using characteristics of the postal code areas subjects the broadband deployment estimations to measurement error bias, since the characteristics of the actual geographical market may differ.4 By examining the actual unit of GHSOR\PHQWRI'6/WKHFHQWUDORI¿FHDUHDLQRXU study we are able to test conclusions from earlier studies in a setting free from the measurement error bias that stems from geographical imprecision. For H[DPSOHZH¿QGWKDWWKHUHLVQRHYLGHQFHWKDWUDFH matters in the supply of DSL, other things equal, ZKLFKFRQ¿UPVWKH¿QGLQJRI3ULHJHU :H thus lend credence to the methods used in the previous studies.
THE DECISION TO DEPLOY BROADBAND INFRASTRUCTURE There are several steps involved when a household connects to the Internet via broadband access. &RQVLGHUWKHFDVHRI'6/¿UVWWKHPRGHRIDFcess examined in this chapter. Data moves across various networks from the Internet backbone XQWLOLWUHDFKHVWKHORFDOH[FKDQJHFDUULHU¶VFHQWUDO RI¿FH7KHFHQWUDORI¿FHLVWKHFDUULHU¶VSRLQWRI presence in the local exchanges, and contains the telecommunications switching and DSL multiSOH[LQJHTXLSPHQW)URPWKHFHQWUDORI¿FHGDWD is transmitted over a DSL connection residing on WKHWHOHSKRQHOLQHEHWZHHQWKHFHQWUDORI¿FHDQG WKHVXEVFULEHU¶VSUHPLVHV,QFDEOHGDWDQHWZRUNV GDWD ÀRZV IURP WKH ,QWHUQHW WKURXJK WKH FDEOH FRPSDQ\¶VKHDGHQG DFDEOHVHUYLFHSURYLGHU¶VYHUVLRQRIWKHORFDOH[FKDQJHFDUULHU¶VFHQWUDORI¿FH
243
Competition in Broadband Provision and the Digital Divide
and on to regional high-capacity data networks. 7RUHDFKWKHFDEOHPRGHPVXEVFULEHU¶VSUHPLVHV GDWDWUDYHOVWKURXJKORFDO¿EHURSWLFQHWZRUNVDQG ¿QDOO\RYHUFRD[LDOFDEOH:LUHOHVVDQGVDWHOOLWH carriers also offer broadband capability, although VXFK¿UPVW\SLFDOO\IRFXVRQWKHEXVLQHVVPDUNHW and have small market shares.5 Thus, for residential subscribers, cable modem service and DSL are the broadband options of choice in the U.S., with cable modems enjoying about a three-to-one advantage around 2000, the vintage of the data analyzed here.6 The FCC found as early as 2000 that the Internet backbone and the networks between the backbone DQGWKHWHOHSKRQHFHQWUDORI¿FHVDQGWKHFDEOHFRPpany headends were generally adequate to provide broadband access (FCC, 2000a). The so-called “last PLOH´WRWKHVXEVFULEHUV¶SUHPLVHVLVWKHOLPLWLQJ factor on the supply side of the market. $EURDGEDQGFDUULHU¶VGHFLVLRQWRHQWHUDPDUNHW depends on the expected demand, costs, and entry E\RWKHU¿UPV7KHHQWU\GHFLVLRQGHSHQGVRQWKH H[SHFWHGSUR¿WV( S(t,w,z(t),H)), where SLVWKHSUR¿W ÀRZDWWLPHt, w is a vector of demographics of the area and other observed variables, z represents the FRPSHWLWRUV¶HQWU\GHFLVLRQV LWVHOIDIXQFWLRQRI time), and HLVDUDQGRPYDULDEOHUHÀHFWLQJXQFHUWDLQW\DERXWIXWXUHSUR¿WDELOLW\)ORZSUR¿WVGHSHQG explicitly on time because, for example, the cost of telecommunications equipment falls over time. (QWU\DOVRGHSHQGVRQWKH¿[HGFRVWLQFXUUHGDW time of deployment, C(t). Given a discount rate r, DULVNQHXWUDO¿UPKDVDQRSWLPDODGRSWLRQWLPH t*IRUDSDUWLFXODUPDUNHWGH¿QHGE\ f
t* arg max ³ e rs E ª¬ t
t
s, w, z (s), º¼ ds e rt C (t )
Entry time t* may be treated as a random variable from the standpoint of the econometrician, who not observe all the relevant elements of the GHFLVLRQ LQFOXGLQJWKH¿UP¶VVXEMHFWLYHSUREDELOLW\ distribution for H WKDWWKH¿UPWDNHVLQWRDFFRXQW Then the probability that DSL has been deployed in area i as of the time T of our data is: Pr(ti * < T |x)
244
(1)
where x is the subset of (w,z) observed in the data. We use the probit model as our econometric speci¿FDWLRQRIHTXDWLRQZKHUHWKHGHSHQGHQWYDULDEOH yi is a binary indicator for DSL availability in the market. Using the probit model is tantamount to assuming that the distribution of random variable T, conditional on x, is standard normal.7 Thus the HPSLULFDOVSHFL¿FDWLRQLV yi = 1 if ti * < T, 0 otherwise Pr(yi = 1|x) = )(E'xi)
(2)
where ELVDYHFWRURIFRHI¿FLHQWVWREHHVWLPDWHG and ) is the standard normal cumulative density function.
DATA In October 1999, SBC and Ameritech merged. Both companies were dominant telecommunications service providers,8 albeit in different locations. As such, their merger required FCC approval in addition to the usual scrutiny by the Department of Justice. The FCC approved the transaction subject to certain conditions, including the promotion of advanced services such as broadband Internet access by the company. Failure to meet the conditions was to trigger penalty payments of more than $2 billion in payments. In particular, SBC was required to locate at least 10% of its advanced service facilities in low-income areas in the Ameritech region.9 6WDWHUHJXODWRUVLQ$PHULWHFK¶VRSHUDWLQJUHJLRQ (Illinois, Indiana, Michigan, Ohio, and Wisconsin) DOVRSXVKHGWKHPHUJHG¿UPWRDFFHOHUDWHEURDGEDQG GHSOR\PHQW 7KH UHJXODWRUV¶ FRQFHUQ ZDV that Ameritech had made slow progress deploying DSL compared to other major telephone companies in other areas of the U.S. Precise statistics at the company level on DSL deployment are closely held data, and so comparisons must rely on estimates, but one industry source calculated that at the end of 1999, Ameritech had only 2% as many DSL lines in service as the next smallest dominant telephone company in the U.S.10 7R DOORZ UHJXODWRUV WR JDXJH $PHULWHFK¶V progress in implementing DSL, as part of the
Competition in Broadband Provision and the Digital Divide
7DEOH6XPPDU\VWDWLVWLFVIRUWKHFHQWUDORI¿FHOHYHOGDWD Variable
Mean
DSL
0.051
Std. Dev.
Min
Max
0.220
0.000
1.000
Race and Ethnicity % Black
0.033
0.109
0.000
0.984
% Native American
0.006
0.033
0.000
0.962
% Asian
0.007
0.017
0.000
0.315
% Other
0.023
0.070
0.000
1.000
% Hispanic
0.021
0.042
0.000
0.695
Income (log)
10.652
0.245
9.613
11.911
% in Poverty
-2.628
0.615
-5.778
-0.724
Households (log)
7.369
1.507
-0.693
11.562
Number of Firms (log)
4.537
1.638
0.000
8.661
% Less than H.S.
0.172
0.070
0.000
0.704
% Some College
0.269
0.052
0.000
0.465
% College Degree
0.104
0.061
0.000
0.447
% Graduate Degree
0.053
0.044
0.000
0.429
% Work at Home
0.046
0.033
0.000
0.356
% Commute 20-40 minutes
0.348
0.109
0.020
0.713
% Commute 40-60 minutes
0.104
0.059
0.000
0.556
% Commute > 60 minutes
0.069
0.045
0.000
0.499
Income and Poverty
Size of Market
(GXFDWLRQ3UR¿OH
&RPPXWLQJ3UR¿OH
Other Demographics % Female
0.500
0.028
0.062
0.728
38.022
3.783
19.943
73.752
Ave. Workers/Firm
2.253
0.671
-0.847
5.676
% Small Firms
0.900
0.072
0.000
1.000
Pop. Density (log)
4.479
1.680
-2.857
10.215
Rural
0.846
0.361
0.000
1.000
4.059
1.114
0.525
6.908
3.571
0.297
1.747
4.126
Median Age Business Market
Cost Variables
Phone Density
†
Structure Age (log) Competition Cable Modem Presence
0.338
0.474
0.000
1.000
CLEC DSL Presence
0.083
0.275
0.000
1.000
245
Competition in Broadband Provision and the Digital Divide
Table 1. (cont.) CLEC Presence
0.871
0.335
0.000
1.000
Illinois
0.266
0.442
0.000
1.000
Indiana
0.147
0.354
0.000
1.000
States
Michigan
0.188
0.391
0.000
1.000
Ohio
0.226
0.418
0.000
1.000
Wisconsin
0.173
0.379
0.000
1.000
†
Phone density is transformed with a reverse log transformation: -ln(1-x).
regulatory contract the company made available a list of their DSL subscribers by nine-digit postal code (known as ZIP+4 Codes) as of March 2000. The data are a snapshot of DSL deployment as of shortly after the merger and established a baseline to which regulators could compare future deployment. ZIP+4 areas are typically very small geographic areas, comprising a few blocks worth of addresses at most. The list contains every ZIP+4 &RGHLQ$PHULWHFK¶VVXEVFULEHUGDWDEDVHDQGWKH deployment date, and allows us to see which central RI¿FHVKDG'6/GHSOR\HG In addition to the DSL deployment list, the other GDWDIRUWKHVWXG\FRPHIURP¿YHRWKHUVRXUFHV D GIS database of ZIP+4 Codes and locations, a GIS telecommunications wire center database, the FCC local competition database, and two Census data sources for household demographic and business information. A complete list of variables and summary statistics for the data are in Table 1. Given the nascent state of the market in the region in 2000, the data are from the early years of broadband rollout in the area. The advantage to examining data from the early years is that there was still much geographic variation in the DYDLODELOLW\RIEURDGEDQGZKLFKDOORZVLGHQWL¿cation of the impact of explanatory variables on GHSOR\PHQW7RGD\¶VQHDUO\XELTXLWRXVFRYHUDJH in U.S. urban and suburban areas does not allow researchers to apply fruitfully similar methodology to current data.
246
DSL Availability Data There are more than 170,000 entries in the Ameritech DSL ZIP+4 list. This implies that there are at least that many DSL subscribers, which means that either the external industry source mentioned above grossly undercounted DSL lines, or Ameritech enjoyed a phenomenal growth rate over the three months between the two data collection dates. The pattern of deployment can be seen by plotting the centroids of the ZIP+4 areas with DSL. Figure 1 VKRZV DOO WKH =,3 GHSOR\PHQWV LQ WKH ¿YHVWDWH region.11 The striking picture shows that DSL is available in only a small fraction of the total geography. However, it is important to remember that we do not observe DSL deployed by incumbent phone companies other than Ameritech, and that Ameritech is not the incumbent carrier in many rural areas in these states. In Illinois, all deployment is in the extended Chicago area. There is no DSL in Indiana. In Michigan, most deployment is in the Detroit area. In Ohio all DSL is around Cleveland. The few DSL customers in Wisconsin live in the Milwaukee area. These are the most populous areas in the region. DSL, as deployed by Ameritech at least, clearly lags behind other forms of broadband infrastructure LQWKHUHJLRQDVVXJJHVWHGE\WKHRI¿FLDOVWDWLVWLFV cited above. Figure 2 compares the Ameritech DSL ORFDWLRQVZLWKWKHFRPSUHKHQVLYH)&&GDWDRQ¿YH digit ZIP Codes with broadband access for Illinois. Many parts of the state have broadband access from non-Ameritech sources. Only in the Chicago area
Competition in Broadband Provision and the Digital Divide
Figure 1. DSL subscribers in the Ameritech region
does the Ameritech DSL deployment come close to matching deployment from all sources. The local exchange carrier implements DSL LQWKHFHQWUDORI¿FHDQGVRWKHUHOHYDQWGHFLVLRQ XQLWLVDFHQWUDORI¿FHVHUYLQJDUHD$OWKRXJKLWLV theoretically possible that Ameritech could choose not to offer DSL to certain neighborhoods within a serving area, as a marketing decision the company offers it to all neighborhoods in the area.12 We deWHUPLQHGZKLFKFHQWUDORI¿FHVZHUH'6/FDSDEOH by matching the ZIP+4 centroids on the DSL list WRWKHFHQWUDORI¿FHDUHDV13 Our dependent variable in the estimations is an observation on a central RI¿FHDUHDHTXDOWRRQHLI'6/LVGHSOR\HGRU]HUR if not. The sample includes all Ameritech central RI¿FHVLQWKH¿YHVWDWHUHJLRQZKLFK\LHOGV observations.14
Market Characteristics Data We model the deployment decision as a probit function of demand variables, cost factors, and the competitive environment.
Demand Variables Socioeconomic statistics at the block or block group level from the 2000 U.S. Bureau of the Census GHFHQQLDOFHQVXVFDSWXUHIDFWRUVLQÀXHQFLQJFDUULHUV¶H[SHFWHGVXEVFULEHUGHPDQGIRUEURDGEDQG :H DJJUHJDWH WKH YDULDEOHV WR WKH FHQWUDO RI¿FH area for the entry estimations.15 These variables SUR[\WKHH[SHFWHGSRVWHQWU\SULFHDQGSUR¿WLQ the DSL deployment estimations. The variables include racial breakdown of the area, Hispanic ethnicity, income, market size and composition, rural,
247
Competition in Broadband Provision and the Digital Divide
Figure 2. Comparison with FCC data for Illinois (FCC, 2000b)
DJHDQGHGXFDWLRQSUR¿OHJHQGHUUDWLRFRPPXWH time, and telephone penetration. An additional set of business market-related variables (the average QXPEHU RI ZRUNHUV SHU ¿UP DQG WKH SHUFHQWDJH RI¿UPVWKDWKDYHIHZHUWKDQHPSOR\HHV DUH taken from Census sources for establishment and employment counts at the postal code level.16
Costs Various studies and industry sources suggest that relevant cost considerations for broadband deployPHQWDUH¿[HGFRVWVVXEVFULEHUGHQVLW\DQGWKH vintage of the telecommunications infrastructure.17 Controls for these costs include subscriber density and a proxy for the vintage of the local telecommunications networks (median age of the housing
248
structures) in the area. For a more complete explanation of broadband costs and these variables, refer to Prieger (2003).
Local Telecommunications and Broadband Competition Local telecommunications competition had started to spring up at the time of the passage of the Telecommunications Act of 1996. Anecdotal evidence from the industry suggests that incumbent local exchange carriers are more likely to offer advanced services in areas in which they face competition, although the endogeneity of entry prevents any conclusion of causality from this observation.18 Some facilities-based competitive local exchange carriers (CLECs) offer DSL. CLECs can offer DSL
Competition in Broadband Provision and the Digital Divide
7DEOH3URELWHVWLPDWLRQVIRUWKHDYDLODELOLW\RIEURDGEDQGVHUYLFHZLWKLQDFHQWUDORI¿FHDUHD 5DFH and income Estimation 1
Estimation 2
Race
Race and Income
&RHI¿FLHQW
Variable
Marginal
P-Value of
Effect
Coef.
&RHI¿FLHQW
Marginal
P-Value of
Effect
Coef.
Race and Ethnicity % Asian
13.030***
0.621
0.000
7.408**
0.155
0.013
% Black
-0.701**
-0.033
0.034
1.199***
0.025
0.008
-127.046***
-6.057
0.002
-0.410
0.510
-13.682**
-0.652
0.017
2.192
0.046
0.725
0.410
0.008
-0.219
-0.005
0.951
% Native American % Other % Hispanic
8.591***
-19.546
Income and Poverty Income (log)
2.227***
% in Poverty
0.047
0.000
-0.269
-0.006
0.253
Size of Market Households (log)
0.216**
0.010
0.018
0.157
0.003
0.174
Pop. Density (log)
0.692***
0.033
0.000
0.852***
0.018
0.000
0.000
-33.406***
Intercept
-7.364***
Log Likelihood N Pseudo R
2
0.000
-264.793
-220.818
1,120
1,119
0.490
0.575
VLJQL¿FDQW DW WKH OHYHO
VLJQL¿FDQW DW WKH OHYHO
VLJQL¿FDQW DW WKH OHYHO 1RWHV 7KHGHSHQGHQWYDULDEOHLVLIWKHUHLVDWOHDVWRQHEURDGEDQGFXVWRPHULQWKHFHQWUDORI¿FHVHUYLQJDUHDLIQRW7KHVDPSOHLQFOXGHVDOO$PHULWHFKFHQWUDORI¿FHDUHDV0DUJLQDOHIIHFWLVWKHPDUJLQDO HIIHFWRQWKHPHDQHYDOXDWHGDWWKHVDPSOHPHDQRI[$VWHULVNVDQG3YDOXHVDUHIRUWKHFRHI¿FLHQWQRW the marginal effect.
LQ$PHULWHFK¶VVHUYLQJDUHDHYHQLI$PHULWHFKGRHV QRWLWVHOIGHSOR\'6/LQWKHFHQWUDORI¿FHE\FROORFDWLQJWKHLURZQHTXLSPHQWLQWKHLQFXPEHQW¶V FHQWUDORI¿FHDQGUHQWLQJWKHVXEVFULEHUOLQHWRWKH FXVWRPHU¶VSUHPLVHV The FCC makes available a list of postal codes in which there is local competition. In some speci¿FDWLRQVZHLQFOXGHDQLQGLFDWRUYDULDEOHCLEC Presence, for the presence of at least one competing local exchange company in the area. The FCC data do not indicate whether the competitor offers DSL, however, and so CLEC Presence indicates
only the potential for competition. We collected information on actual competition in DSL from CLECs from the New Paradigm Resources Group (2001). The data from NPRG lists which CLECs operate in which cities and whether they offer DSL.19 For a subset of ZIP Codes, we also have data on deployment of cable modem service, which ZH PDWFKHG WR RXU FHQWUDO RI¿FH VHUYLQJ DUHDV The cable modem data cover a random sample of =,3&RGHVDFURVVWKH¿YHVWDWHV20 except for Ohio, where coverage is complete.21
249
Competition in Broadband Provision and the Digital Divide
Table 3. Probit estimations for the availability of broadband service within a FHQWUDORI¿FHDUHD $OOYDULDEOHV Estimation 3
Estimation 4: All Variables and State Fixed
All Variables
Effects Marginal
P-Value of
Marginal
P-Value of
&RHI¿FLHQW
Effect
Coef.
&RHI¿FLHQW
Effect
Coef.
% Asian
12.345**
0.036
0.016
7.145
0.050
0.193
% Black
0.143
0.000
0.866
0.172
0.001
0.859
% Native American
3.515
0.010
0.943
-92.174
-0.639
0.255
% Other
-1.875
-0.005
0.833
-8.318
-0.058
0.401
% Hispanic
-3.316
-0.010
0.518
2.115
0.015
0.723
Income (log)
1.893**
0.006
0.039
0.887
0.006
0.367
% in Poverty
0.056
0.000
0.892
0.304
0.002
0.485
Households (log)
0.296
0.001
0.116
0.275
0.002
0.172
Number of Firms (log)
0.268*
0.001
0.086
0.357**
0.002
0.026
% Less than H.S.
22.491***
0.066
0.000
13.795**
0.096
0.019
% Some College
12.621***
0.037
0.006
1.346
0.009
0.805
% College Degree
3.314
0.010
0.355
1.455
0.010
0.705
% Graduate Degree
8.004*
0.023
0.056
2.654
0.018
0.567
17.864
0.052
0.11
27.176**
0.188
0.018
Variable Race and Ethnicity
Income and Poverty
Size of Market
(GXFDWLRQ3UR¿OH
&RPPXWLQJ3UR¿OH % Work at Home % Commute 20-40 minutes
2.670*
0.008
0.095
3.918**
0.027
0.019
% Commute 40-60 minutes
17.069***
0.050
0.000
16.393***
0.114
0.000
% Commute > 60 minutes
-9.176***
-0.027
0.006
-7.704**
-0.053
0.043
Other Demographics % Female
-6.211
-0.018
0.198
-10.517**
-0.073
0.038
Median Age
0.538*
0.002
0.070
0.401
0.003
0.208
Median Age Squared
-0.007*
0.000
0.096
-0.004
0.000
0.296
Business Market Ave. Workers/Firm
1.135***
0.003
0.002
1.189***
0.008
0.002
% Small Firms
7.399*
0.022
0.051
7.194*
0.050
0.06
0.708***
0.002
0.001
0.784***
0.005
0.001
Rural
-0.804*
-0.004
0.063
-0.939**
-0.010
0.041
Phone Density†
0.611***
0.002
0.002
0.681***
0.005
0.001
Structure Age < Median (log)
0.965**
0.003
0.035
0.353
0.002
0.482
Structure Age > Median (log)
-1.762
-0.005
0.105
-1.813
-0.013
0.123
Cost Variables Pop. Density (log)
250
Competition in Broadband Provision and the Digital Divide
Table 3. (cont.) States Michigan
0.792**
0.009
0.035
Ohio
-0.613
-0.003
0.151
Wisconsin
-0.001
0.000
0.999
Competition CLEC Presence
-0.021
0.000
0.963
0.086
0.001
0.862
CLEC DSL Presence
-0.517**
-0.001
0.025
-0.544**
-0.003
0.035
-64.317***
0.036
0.000
-42.590***
0.050
0.003
Intercept Log Likelihood N Pseudo R
2
-152.242
-133.255
1,119
957
0.707
0.725
VLJQL¿FDQW DW WKH OHYHO
VLJQL¿FDQW DW WKH OHYHO
VLJQL¿FDQW DW WKH OHYHO † Phone density has a long left tail and is transformed with a reverse log transformation: -ln(1-x). Omitted racial variable is White. Omitted education variable is High School Degree. Omitted commute variable is 1-20 minutes. Marginal effects for dummy variables are for a zero to one change in x. Observations from Indiana are dropped from estimation 4 because no DSL is deployed in the state. See notes to previous table.
EMPIRICAL FINDINGS The estimations for the determinants of DSL entry, based in equation 2, are reported in Tables 2-4.
The Digital Divide Table 2 contains two estimations that show an apparent supply-side digital divide based on race and income. Recall that the unit of observation is DQ$PHULWHFKFHQWUDORI¿FHDUHD,QDOOWDEOHVZH UHSRUWWKHSURELWFRHI¿FLHQWVWKHPDUJLQDOHIIHFWV of the variables, and the p-value for the hypothesis WKDWWKHFRHI¿FLHQWLV]HUR7KHPDUJLQDOHIIHFWRI a variable expresses the change in the estimated probability that broadband is deployed from a unit change in the variable (evaluated at sample averages of the data). In estimation 1, we control for nothing except race and market size. Areas with high concentrations of blacks, Native Americans, and “other” UDFHV DVGH¿QHGE\WKH&HQVXVFDWHJRULHV DUHVLJQL¿FDQWO\OHVVOLNHO\WRKDYH'6/DYDLODEOH$UHDV with many Asians and (surprisingly) Hispanics are
VLJQL¿FDQWO\PRUHOLNHO\WRKDYHDFFHVVWR'6/7KLV simple portrait of the digital divide is incomplete, however. An obvious question is whether race is a causal determinant of broadband deployment or is merely correlated with other demographic FKDUDFWHULVWLFV XSRQ ZKLFK ¿UPV EDVH WKHLU HQtry decision, such as income, education, and the business market. In estimation 2, we add income variables. Access to DSL is correlated with higher income and fewer households in poverty, although WKHODWWHUFRHI¿FLHQWLVQRWVLJQL¿FDQW,PSRUWDQW WRQRWHLVWKDWWKHVLJQL¿FDQWQHJDWLYHHIIHFWVIRU blacks, Native Americans, and other races disapSHDU7KHFRHI¿FLHQWIRUEODFNVVZLWFKHVVLJQDQG LVQRZVLJQL¿FDQWO\SRVLWLYH7KXVLQFRPHDSSHDUV to be a more important factor contributing to the digital divide on the supply side than race per se.
Demographic and Economic Factors In estimations 3 and 4, we add a host of other demographic and economic variables. The distinction between the two is the addition of state ¿[HGHIIHFWVLQHVWLPDWLRQ7KH¿[HGHIIHFWVDGG
251
Competition in Broadband Provision and the Digital Divide
DVWDWHVSHFL¿FLQWHUFHSWWHUPWRWKHLQGH[E'xi in HTXDWLRQ7KHDGGLWLRQRIVWDWH¿[HGHIIHFWVUHmoves the impact of any factor, observed or not, that varies among states but not within a state. Examples of such factors include the state regulatory environment, unique or historical reasons why costs of provision differ among states, and average demand and supply conditions. Adding state effects necessitates dropping observations from Indiana, where no DSL was in place. Except for some of the education variables, there are no VWDWLVWLFDOO\VLJQL¿FDQWODUJHGLIIHUHQFHVEHWZHHQ WKHFRHI¿FLHQWVLQWKHWZRHVWLPDWLRQV22 The education level of the area has a curious correlation with DSL access. The excluded educational category in the estimation is high school HGXFDWLRQDQGDOOFRHI¿FLHQWVDUHSRVLWLYHUHODWLYH to this category. The largest positive effect is for “less than high school” education. Perhaps inner urban areas with low average educational attainment but other desirable characteristics (from the ¿UP¶VSRLQWRIYLHZ VXFKDVEXVLQHVVGHPDQGDUH driving this result. More expected is that the fraction of people with some college education or higher positively affects the probability of DSL.23 Commuting has the expected impact. There are large positive effects for the “work at home” variable, perhaps due to demand for telecommuting. For those who commute, entry rises with the length of the commute up to one hour, which may UHÀHFWDFRPELQDWLRQRIXUEDQDUHDHIIHFWV XUEDQ areas have more demand for DSL) and stronger demand for telecommuting. The industry press has noted the link between longer commuting times and demand for broadband access (NPRG, 2006). The longest commuting category (over an hour) causes the probability of access to fall, which LVSUREDEO\DUHÀHFWLRQRIORZUXUDOGHPDQGIRU broadband (perceived or real on the part of the ¿UP $QLQWHUHVWLQJ¿QGLQJLVWKDWDUHDVZLWKPRUH women are less likely to have access to DSL. Median age of the area has an inverted U relationship to access, with the maximum occurring around age WR7KHLPSDFWRIDJHPLJKWUHÀHFWORZHU demand for Internet access among the elderly and the very young.24
252
Regarding business demand, access rises with WKHDYHUDJHQXPEHURIZRUNHUVSHU¿UPEXWDOVR ZLWKWKHIUDFWLRQRI¿UPVWKDWKDYHIHZHUWKDQ HPSOR\HHV7KHODWWHUSUREDEO\UHÀHFWVWKDW'6/LV an especially attractive broadband access option for small businesses, given its low price in comparison to high-speed dedicated telecommunications lines (so-called “special access” lines) and the relative lack of availability of cable modem service to business addresses.25 In a separate estimation (results not reported), we added variables for the percent of the businesses in the area in various industrial categories. While some of the categories had VLJQL¿FDQWHIIHFWVWKH\GLGQRWFKDQJHWKHPDLQ results of interest.26 The cost variables mostly have the expected signs. Denser areas, measured by population or telephone penetration, have higher probability RIDFFHVVWR'6/5XUDODUHDVGH¿QHGLQDFFRUG ZLWKWKH&HQVXVGH¿QLWLRQRIDGHQVLW\OHVVWKDQ 500 persons per square mile, have lower access. The marginal effects indicate that Ameritech was 4 to 10 percentage points (depending on which estimation is used) less likely to deploy DSL in its UXUDOFHQWUDORI¿FHVWKDQLQLWVXUEDQRQHV2OGHU infrastructure, as proxied by structure age, has a negative impact—at least for levels of the variable above the median. Younger structure ages show the RSSRVLWHVLJQEXWQHLWKHUFRHI¿FLHQWLVVLJQL¿FDQW in estimation 4.
The Impact of Competition The carrying capacity, or market size, hypothesis states that competition from CLECs and cable modem providers should make the provision of DSL less attractive to the phone company, other things equal. Some markets may be large enough to support one entrant only, and if the cable company or a CLEC already offers broadband, then the phone company may delay DSL provision until the market grows. An alternative hypothesis is that competition provides competitive stimulus to telecommunications investment (Willig, 2003; Aghion, Harris, Howitt, & Vickers, 2001).27 To test ZKLFKRIWKHVHK\SRWKHVHVEHWWHU¿WVWKHGDWDZH examine how competition affects the probability
Competition in Broadband Provision and the Digital Divide
of DSL deployment. Since our cable modem data are not available for the full sample, in estimations 3 and 4 we include only the CLEC variables for potential and actual competition. 7KH FRHI¿FLHQW IRU &/(& SUHVHQFH LV QRW VLJQL¿FDQW EXW ZKHQ WKH &/(& DFWXDOO\ RIIHUV '6/WKHLPSDFWLVQHJDWLYHDQGVLJQL¿FDQW7KH marginal effect of -0.001 for CLEC DSL Presence from estimation 3 implies that when a CLEC is actively offering DSL service in the central of¿FHDUHD$PHULWHFKLVSHUFHQWDJHSRLQWVOHVV OLNHO\WRGHSOR\'6/LWVHOI:KHQWKHVWDWH¿[HG effects are included in estimation 4, the marginal effect of DSL competition rises in magnitude to -0.26 percentage points. Thus, while the potential broadband competition that a CLEC represents neither spurs not hinders deployment, actual competition in DSL is associated with less deployment by the incumbent. The magnitude of the effect, however, is small. To explore the impact of the availability of cable modem service on DSL deployment, we turn to RXU VXEVHW RI FHQWUDO RI¿FHV IRU ZKLFK ZH KDYH both CLEC and cable modem data. The carrying capacity hypothesis, which predicts a negative sign RQFRHI¿FLHQWVIRU&/(&DQGFDEOHPRGHPSUHVHQFHLQWKHFHQWUDORI¿FHDUHDGHSHQGVFUXFLDOO\RQ other things in the market being equal. If one fails WRFRQWURODGHTXDWHO\IRUWKHSRWHQWLDOSUR¿WDELOLW\ of a market (through cost and demand variables), then the unobserved differences among markets may make it look like the presence of competitors is associated with greater DSL entry. If areas H[SHFWHGWREHPRUHSUR¿WDEOHDSULRULDWWUDFWERWK DSL and cable modem entry, then not controlling DGHTXDWHO\IRUH[SHFWHGSUR¿WDELOLW\ZRXOGOHDGWR a false rejection of the carrying capacity hypothesis in favor of the competitive stimulus hypothesis. The correlation between DSL and cable modem entry would not be causal in nature. To illustrate this point, we estimate a model FRQWUROOLQJ RQO\ IRU VWDWH ¿[HG HIIHFWV DQG WKH competition variables (estimation 5 in Table 4). The presence of both cable modem and competitive '6/RSWLRQVIRUVXEVFULEHUVLQWKHFHQWUDORI¿FH area is associated with a higher likelihood of deployment by the incumbent DSL provider. Before
taking this as evidence for the competitive stimulus hypothesis, however, we control for other market IDFWRUVDIIHFWLQJH[SHFWHGSUR¿WDELOLW\LQHVWLPDtion 6 (Table 4). The results show that negative FRHI¿FLHQWVIRUFRPSHWLWLYH'6/DQGFDEOHPRGHP service appear/result in accord with the evidence for the carrying capacity hypothesis found in estimations 3 and 4. We cannot include as many controls for market factors as in estimations 3 and 4, due to the smaller sample size in estimation 6. 7KH FRHI¿FLHQWV RQ WKH YDULDEOHV Cable Modem Presence and CLEC DSL Presence are negative EXWQRWVWDWLVWLFDOO\VLJQL¿FDQWKRZHYHUGXHWR the small sample size). The marginal effect for CLEC DSL Presence, 0.52 percentage points, is even higher than that estimated in the full sample. Given these results, together with those from the larger estimations, it appears that support is greater for the carrying capacity hypothesis than the competitive stimulus hypothesis. The variable measuring potential competition, CLEC Presence, is positive but not statistically VLJQL¿FDQW 3ULHJHU IRXQG D SRVLWLYH FRUrelation between CLEC presence and broadband provision, but could not separate provision from the incumbent local exchange company from other sources. A positive association with DSL deployPHQW E\ WKH LQFXPEHQW PD\ UHÀHFW D GHVLUH E\ Ameritech to beat the competition to market (and perhaps prevent the CLEC from entering the DSL market at all).28
FUTURE TRENDS AND CONCLUSION Our empirical results for the determinants of the deployment of DSL broadband Internet connection indicate that the racial and ethnic composition of an area do not matter independent of related factors VXFKDVLQFRPHDQGHGXFDWLRQ7KH¿QGLQJVDUH LQDFFRUGZLWKWKHYLHZWKDWWKH¿UPVDUHVHHNLQJ WRPD[LPL]HSUR¿W DVUHTXLUHGE\WKHLU¿GXFLDU\ responsibility to their stockholders). Any apparent redlining in the region studied is completely H[SODLQHGE\GHPDQGVLGHIDFWRUVWKDWDIIHFWSUR¿W Thus, to the extent that society desires to close what
253
Competition in Broadband Provision and the Digital Divide
7DEOH3URELWHVWLPDWLRQVIRUWKHDYDLODELOLW\RIEURDGEDQGVHUYLFHZLWKLQDFHQWUDORI¿FHDUHD &RPpetition variables Estimation 5
Estimation 6 Marginal
P-Value of
&RHI¿FLHQW
Effect
Coef.
0.000
-0.204
-0.003
0.442
0.151
0.027
0.813
0.007
0.551
-0.008
0.867
-0.440
-0.005
0.187
Income (log)
1.491
0.023
0.108
% in Poverty
-0.145
-0.002
0.775
Households (log)
-0.380
-0.006
0.168
Number of Firms (log)
0.530***
0.008
0.009
Median Age
-0.998**
-0.015
0.028
Median Age Squared
0.014**
0.000
0.028
Variable
&RHI¿FLHQW
Marginal
P-Value of
Effect
Coef.
Competition Cable Modem Presence CLEC Presence CLEC DSL Presence
0.898*** 0.719** -0.029
0.270
Income and Poverty
Size of Market
Other Demographics
Business Market Ave. Workers/Firm
0.448
0.007
0.464
% Small Firms
4.469
0.069
0.385
1.264***
0.019
0.000
Rural
-0.624
-0.009
0.537
Phone Density†
0.226
0.003
0.444
0.454
0.007
Cost Variables Pop. Density (log)
Structure Age Intercept State Fixed Effects Log Likelihood N Pseudo R2
-1.800***
0.000
-15.522
0.471 0.212
No
Yes
-175.104
-71.056
374
360
0.116
0.634
VLJQL¿FDQWDWWKHOHYHO
VLJQL¿FDQWDWWKHOHYHO
VLJQL¿FDQWDWWKHOHYHO † Phone density is transformed with a reverse log transformation: -ln(1-x). See notes to previous table.
254
Competition in Broadband Provision and the Digital Divide
remains of the broadband digital divide in the U.S., either attention must be given to the demand side of the market or means should be found (through WDUJHWHGVXEVLGLHVSHUKDSV WRLQFUHDVHWKH¿UPV¶ SUR¿WDELOLW\RIHQWHULQJOHVVGHVLUDEOHDUHDV,QWKH latter case, however, subsidies may create their own GLVWRUWLRQVRIHFRQRPLFHI¿FLHQF\DQGWKHFRVWV DQGEHQH¿WVRISROLF\LQWHUYHQWLRQLQWKHPDUNHW must be carefully weighed. The experience in the U.S. with the diffusion of broadband since the early years examined here shows that the question of broadband deployment is not “if” but “when.” At the time of the data ZHH[DPLQHGWKH)&& E IRXQGLQLWV¿UVW comprehensive survey that there was at least one customer for high-speed service in 59% of all the postal code areas in the United States. By the end of 2005, the same statistic rose to 99% of postal code areas (FCC, 2006).29 The speed at which the U.S. attained nearly ubiquitous broadband access is all the more remarkable given that (unlike some other countries) there are no general federal subsidy programs for either broadband infrastructure or household subscription.30 Even though some form of broadband access is available in most areas of the U.S. today, the lessons learned from looking at the early years of adoption are still important. While broadband availability is necessary to close the digital divide, more is needed. Cooper (2004) emphasizes that bringing down the price of service is essential if lower-income households are to be brought online. Competition among providers holds the greatest promise for bringing down prices in the United States, given the reluctance of the FCC to add broadband explicitly to the list of services supported under federal universal service programs. As the market grows, it is reasonable to expect that the IDFWRUVLQÀXHQFLQJIXWXUHHQWUDQWVLQDQDUHDZLOO follow patterns similar to that of the incumbent DSL entrant uncovered in this chapter. Thus, understanding what drives initial DSL deployment lends insight into what drives increasing competition.31 Furthermore, many countries around the world are in situations today similar to where the U.S. was ¿YH\HDUVDJR,QWKH86KDGDERXWD
subscription rate to broadband, which is higher than in Turkey, the Slovak Republic, Mexico, or Greece in 2006, just to mention OECD countries.32 ,IWKHSUR¿WDELOLW\RIJHRJUDSKLFDOPDUNHWVUHODtive to each other in the U.S. is similar to that in other countries, then we should expect broadband WR EH DYDLODEOH ¿UVW LQ GHQVHU XUEDQ DUHDV ZLWK longer commutes, a more educated populace, and wealthier households. Furthermore, intermodal competition (and thus lower prices) may be slower in coming than initial diffusion, as areas with an existing broadband provider look less attractive to enter (other things equal) than do virgin territories.
ACKNOWLEDGMENT We gratefully acknowledge Sean Gill for research assistance and the many seminar participants who commented on earlier versions of this project, including those at the Center for Research in Regulated Industries 18th Annual Western Conference (San Diego), the FCC, the International Industrial Organization Conference (Georgia Tech), the NSFConacyt Internet Use in the Americas Workshop (Mexico City), the Pew Charitable Trusts Workshop on Measuring Broadband (Washington, DC), the Public Utility Research Center and London Business School Annual Telecommunications Conference (University of Florida), the Economic Impacts of Technology Conference (UC Santa Cruz), and the Western Economic Association International 80th Annual Conference (San Francisco). The above-mentioned individuals and institutions are responsible for neither the opinions nor any errors contained herein.
REFERENCES Aghion, P., Harris, C., Howitt, P., & Vickers, J. (2001). Competition, imitation and growth with step-by-step innovation. Review of Economic Studies, 68, 467-492.
255
Competition in Broadband Provision and the Digital Divide
Bresnahan, T.F., & Reiss, P.C. (1987). Do entry conditions vary across markets? Brookings Papers in Economic Activities, 18, 833-882.
FCC. (2000b). High-speed services for Internet access: Subscribership as of June 30, 2000. Washington, DC.
Burnstein, D.E., & Aron, D.J. (2003). Broadband adoption in the United States: An empirical analysis. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (pp. 119-128). Haupaugge, NY: Nova Science Press.
FCC. (2006). High-speed services for Internet access: Subscribership as of December 31, 2005. Washington, DC.
Cameron, A.C., & Trivedi, P.K. (2005). Microeconometrics. New York: Cambridge University Press. CED. (1999, December). DSL technology looms ever larger (p. 100). Cooper, M. (2004). Expanding the digital divide & falling behind on broadband. Washington, DC: Consumer Federation of America. Crandall, R.W., & Jackson, C.L. (2003). The $500 ELOOLRQRSSRUWXQLW\ 7KHSRWHQWLDOHFRQRPLFEHQH¿W of widespread diffusion of broadband Internet access. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (ch. 8). Haupaugge, NY: Nova Science Press. Dixit, A. (1980). The role of investment in entry deterrence. Economic Journal, 90(357), 95-106. Duffy-Deno, K.T. (2000, September). Demand for high-speed access to the Internet among Internet households. Proceedings of ICFC 2000, Seattle, WA. Retrieved October 20, 2006, from http://www. icfc.ilstu.edu/icfcpapers00/duffy-deno.PDF Fairlie, R.W. (2004). Race and the digital divide. Contributions to Economic Analysis & Policy, 3(1), Article 15. Faulhaber, G.R., & Hogendorn, C. (2000). The market structure of broadband telecommunications. Journal of Industrial Economics, 48(3), 305-329. FCC (Federal Communications Commission). (2000a). Deployment of advanced telecommunications capability: Second report (FCC 00-290). Washington, DC.
256
Flamm, K. (2005, February). The role of economics, demographics, and state policy in broadband availability. Proceedings of the PURC/London Business School Conference on the Future of Broadband: Wired and Wireless, 2005, Gainesville, FF. Fox, S. (2004). Older Americans and the Internet. Washington, DC: Pew Internet and American Life Project. Gabel, D., & Kwan, F. (2001). Accessibility of broadband telecommunication services by various segments of the American population. In B.M. Compaine & S. Greenstein (Eds.), Communications policy in transition: The Internet and beyond (pp. 295-320). Cambridge, MA: MIT Press. Gillett, S.E., & Lehr, W. (1999, September). Availability of broadband Internet access: Empirical evidence. Proceedings of the 27th Annual Telecommunications Policy Research Conference, Alexandria, VA. Grubesic, T.H. (2003). Inequities in the broadband revolution. Annals of Regional Science, 37, 263-289. Grubesic, T.H. & Murray, A.T. (2004). Waiting for broadband: Local competition and the spatial distribution of advanced telecommunication services in the United States. Growth and Change, 35(2), 139-165. Mills, B.F., & Whitacre, B.E. (2003). Understanding the non-metropolitan–metropolitan digital divide. Growth and Change, 34, 219-243. NTIA (National Telecommunications and Information Administration). (2002). A nation online: How Americans are expanding their use of the Internet. Washington, DC.
Competition in Broadband Provision and the Digital Divide
NTIA. (2000). Falling through the Net: Toward digital inclusion: A report on Americans’ access to technology tools. Washington, DC. Retrieved October 20, 2006, from http://www.ntia.doc.gov/ ntiahome/fttn00/contents00.html Newberger, E.C. (2001) Home computers and Internet use in the United States. Special Study P23-207, U.S. Department of Commerce, Washington, DC.
KEY TERMS &HQWUDO2I¿FH See Wire Center. Competitive Local Exchange Carrier (CLEC): A public telephone company that provides local telecommunications service in competition with the incumbent local exchange company.
New Paradigm Resources Group. (2001). CLEC report 2001 (vols. 1-2, 13th ed.). Chicago, IL.
Digital Divide: The gap in computer and Internet usage between richer and poorer households, majority and minority groups, and households in urban and rural areas.
New Paradigm Resources Group. (2006). The telecommuter boon: The opportunity is there, but FDQFDEOHFR¶VJHWLWULJKW"Competitive Telecom Advisor, (October 24). Chicago, IL.
Last Mile: The part of the local telecommuQLFDWLRQVQHWZRUNWKDWFRQQHFWVWKHFHQWUDORI¿FH to the subscriber premises. The last mile typically consists of copper telephone wire.
Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363.
Local Exchange Company (LEC): A public telephone company in the U.S. that provides local telecommunications service.
Rappoport, P.N., Kridel, D.J., Taylor, L.D., Alleman, J.H., & Duffy-Deno, K.T. (2003). Residential demand for access to the Internet. In G. Madden (Ed.), Emerging telecommunications networks: The international handbook of telecommunications economics (vol. 2, pp. 55-72). Cheltenham, UK: Edward Elgar. Wallsten, S. (2005). Broadband penetration: An empirical analysis of state and federal policies. Working Paper 05-12, AEI–Brookings Joint Center for Regulatory Studies, Washington, DC. Willig, R.D. (2003). Investment is appropriately stimulated by TELRIC. Unpublished manuscript. Retrieved October 20, 2006, from http://psc.ky.gov/ pscecf/2003-00379/5200700_efs/04132004/MCI_ 67B07%B(;BBSGI ;LDR0 2UD]HP3) Do entry conditions vary over time? Entry and competition in the broadband market: 1999-2003. Unpublished manuscript.
Redlining: The practice of deliberately avoiding selling products or services (in context here, broadband Internet service) in particular neighborhoods. The term comes from the credit industry, and originally meant the refusal of a bank or other lender to extend credit to customers located in a high-risk geographical area, usually a declining inner-city neighborhood. These institutions supposedly drew red lines on maps marking off the high-risk areas. Telephony: A system of telecommunications employing telephonic equipment to transmit sound between points. In common parlance, the “telephone system.” Wire Center: The location where a central of¿FHVZLWFKFRQQHFWVWRWKHORRSIDFLOLWLHVWKDWFRYHU part or all of a local exchange. It is also the physical location where the loop distribution plant in the local telecommunications system can be accessed, and where the equipment enabling DSL service for subscribers is deployed. Used synonymously ZLWK ³FHQWUDO RI¿FH´ LQ WKLV FKDSWHU DOWKRXJK elsewhere sometimes the wire center refers to the SK\VLFDOEXLOGLQJDQGWKHFHQWUDORI¿FHUHIHUVWR the switches within the wire center.
257
Competition in Broadband Provision and the Digital Divide
ENDNOTES 1
2
3
4
5
6
7
8
258
Crandall and Jackson (2003) estimate that WKHYDOXHRIWKHEHQH¿WVWRFRQVXPHUVLQWKH U.S. from broadband Internet access (net of subscription cost) reaches the hundreds of billions of dollars yearly. Postal codes are known as “ZIP Codes” in the U.S. The FCC data list a ZIP Code if there is at least one broadband subscriber whose premises are in the ZIP Code area. Researchers then treat the listed codes as having broadband available and all ZIP Codes not on the list as not having broadband available. In large ZIP Code areas, not all parts of a ZIP Code on WKH)&&¶VOLVWPD\EHLQWKHDFWXDOEURDGEDQG service areas, however. Prieger (2003) and Flamm (2005) discuss the issue at length. Measurement error bias is also known as error-in-variables bias. See Cameron and Trivedi (2005, ch. 26) for a textbook treatment of the issue. Local telephone companies also lease high-speed dedicated circuit access lines to residences and businesses, but their high prices generally restrict them to high-volume business use. More recently, some telephone FRPSDQLHVKDYHDOVREHJXQWRGHSOR\¿EHU optic cable in the last mile network, but most of such investment occurred after the period of the data examined here. $FFRUGLQJWRRI¿FLDOVWDWLVWLFVIRUUHVLGHQFHV and small businesses in the U.S., there were 2,179,749 coaxial cable broadband lines in 2000, compared to 771,311 DSL lines (FCC, 2000b). The remainder of the 3,120,350 broadband lines consisted of dedicated lines (T-1) rented from phone companies, satellite and ¿[HGZLUHOHVVOLQHVDQG¿EHURSWLFOLQHV See Cameron and Trivedi (2005, ch. 14) for a textbook treatment of the probit model. SBC and Ameritech were both Regional Bell Operating Companies (RBOCs), formerly known as the Baby Bells after the breakup of AT&T. The merged entity kept the name
9
10
11
12
13
14
15
16
SBC, and later adopted the moniker AT&T DIWHUPHUJLQJZLWKWKDW¿UPLQ 6%&¶V6(&¿OLQJVIURPWKHWLPHGLVFXVVWKHVH conditions. $PHULWHFK¶VHVWLPDWHG'6/OLQHVDUH RI86:HVW¶VPLOOLRQOLQHV &(' 1999). The same source estimates 17 million lines for Bell Atlantic, 5 million lines for Bell 6RXWKDQG PLOOLRQOLQHVIRU6%&3DFL¿F Bell. 7KHVWDWHDEEUHYLDWLRQVLQWKH¿JXUHDUH,OOLnois (IL), Indiana (IN), Ohio (OH), Michigan (MI), and Wisconsin (WI). Due to technological restrictions at the time, transmission speeds degrade beyond 2.2 miles. Therefore, homes in the area may not have access if they are too far from the central RI¿FH7KHGLVWDQFHOLPLWDWLRQFDQEHVHHQ in Figure 2 from the circular clusters of DSL points, each of which surrounds a central RI¿FH We matched the ZIP+4 centroids to central of¿FHVXVLQJ$UF0DS*,6VRIWZDUHDQG*'7¶V WireCenter Premium telecommunications geography product. When there were between one and nine DSL =,3VLQWKHFHQWUDORI¿FHDUHD RXWRIRYHU FHQWUDORI¿FHV ZHUHPRYHGWKHFHQWUDO RI¿FHIURPWKHHVWLPDWLRQWREHFRQVHUYDWLYH because these might be “stray” ZIP+4s that do not actually belong in the DSL list. The average number of ZIP+4s in a DSL-capable FHQWUDORI¿FHLV Blocks are the smallest unit of U.S. Census geography, and typically contain about 10-30 households. Some statistics, such as household income, are available at the block group level but not at the block level. Block groups typically contain about 600 households. We aggregated the statistics at the Census block RUEORFNJURXSOHYHOWRWKHFHQWUDORI¿FHVHUYing area using the GIS software. The business data are from U.S. Bureau of the Census, ZIP Code Business Patterns CDROM, 2000 data. The ZIP Code areas were PDWFKHGWRWKHFHQWUDORI¿FHDUHDVXVLQJ*,6 software.
Competition in Broadband Provision and the Digital Divide
17
18
19
20
21
22
23
24
25
26
See Faulhaber and Hogendorn (2000) and Gabel and Kwan (2001), for example. The endogeneity issue is that there are characteristics of markets, such as high household income, that make entry by both competitors and the incumbent more likely. In such cases, WKHFRUUHODWLRQEHWZHHQWKHFRPSHWLWRUV¶DQG WKH LQFXPEHQW¶V GHFLVLRQV GR QRW SURYLGH SURRIWKDWWKHFRPSHWLWRUV¶SUHVHQFHcauses DSL deployment by the incumbent. The data from NPRG are not as geographically precise as our other data, because the CLEC locations are listed by city or suburb name. We matched the city names to the FHQWUDORI¿FHDUHDVE\KDQG These data were collected by Kevin DuffyDeno and are described more fully in DuffyDeno (2000) and Rappoport, Kridel, Taylor, Alleman, and Duffy-Deno (2003). We gratefully acknowledge permission to use these data. The cable modem data for Ohio are from Grubesic (2003). We gratefully acknowledge permission to use these data. The apparently large differences in some of WKHFRHI¿FLHQWVGRQRWWUDQVODWHLQWRVLJQL¿cant differences in the marginal effects. The unexpected result for the less than high VFKRROFRHI¿FLHQWLVQRWDQDUWLIDFWRIFROOLQearity; the variable is not highly correlated with any variables expected to increase the likelihood of deployment. Fox (2004) documents low demand among WKHHOGHUO\DQG¿QGVWKDWLQRQO\ of Americans age 65 or older reported having access to the Internet. Low demand for household broadband subscription among \RXQJHUDGXOWVPD\UHÀHFWERWKODFNRILQFRPH and the availability of broadband at school or work as a substitute for access at home. Cable companies built their networks to offer primarily residential service. The main changes were that the statistical VLJQL¿FDQFHRIWKH$VLDQDQGLQFRPHYDULables declined, compared to estimation 3. The LQGXVWU\FDWHJRU\RI³¿QDQFHDQGLQVXUDQFH´ DQGRI³UHDOHVWDWH´KDGVLJQL¿FDQWQHJDWLYH
27
28
29
30
31
32
effects, and the category of “professional, VFLHQWL¿FDQGWHFKQLFDOVHUYLFHV´KDGSRVLWLYH effects on deployment. The former results are probably indicative of the other broadband RSWLRQVWDNHQE\¿UPVLQWKHVHFRPPXQLFDtions-intensive industries. The incumbent marketed DSL at the time toward home and small-business users, since the phone company had little incentive to cannibalize demand for the high-speed (and high-cost) dedicated circuits it leased to businesses. Willig (2003) characterizes the competitive stimulus hypothesis as follows: “Under this view, the previous lack of competition in monopoly local telephone markets may have dissuaded the I[ncumbent ]LECs from making certain investments, and the competitive stimulus from CLEC entry under the 1996 Act may have encouraged greater investment by both the ILECs and the CLECs.” The author is not referring to the DSL investment in particular, but the same notion applies to investment generally (Aghion et al., 2001). Note that when the competition is potential EXWQRWDFWXDOLWLVGLI¿FXOWWRGLVWLQJXLVKWKH competitive stimulus hypothesis from predatory conduct (i.e., investment undertaken VSHFL¿FDOO\WRGHWHUHQWU\ 'L[LW Cable modem or DSL service was available in 87% of the ZIP Codes. Most of the remaining ZIP Codes were covered by satellite. There are a few relatively small targeted federal infrastructure subsidy programs. Two of these are the eRate program subsidizing broadband infrastructure for schools (Flamm, 2005) and the USDA Rural Development broadband program for rural LECs (Wallsten, 2005). See chapter 18 by Prieger and Lee in this volume for an exploration of how policies related to competition affect broadband deployment. Source of statistics: OECD data for June 2006, released October 13, 2006, available at www.oecd.org/sti/ict/broadband.
259
260
Chapter XVII
Governmental and Cultural Factors in Broadband Adoption Elizabeth Fife University of Southern California, USA Laura Hosman University of Southern California, USA Francis Pereira University of Southern California, USA
ABSTRACT 7KRXJKWKHSRWHQWLDOEHQH¿WVRIEURDGEDQG,QWHUQHWDGRSWLRQDUHJUHDWWKHOHYHOVRIWDNHXSYDU\JUHDWO\ around the world. Some governments have adopted aggressive policies to deploy broadband networks and to encourage the use of these applications, while others have not. In the former cases, governments DUHPRWLYDWHGWRSURPRWHEURDGEDQGDGRSWLRQLQRUGHUWRUHDOL]HERWKHFRQRPLFDQGVRFLDOEHQH¿WV7KLV chapter argues that the high level of broadband adoption rates witnessed in certain Asian economies is attributable in part to the aggressive policies pursued by these governments. Independent of these policies however, social factors can also have an impact on whether broadband-related technology will EHDGRSWHG(YHQLIHFRQRPLFDQGVRFLDOEHQH¿WVH[LVWWKHUHIRUHDVLQWKHFDVHRIWHOHPHGLFLQHLQWKH United States, cultural and social factors may in fact hinder the deployment of such applications and retard the growth rate of broadband access.
INTRODUCTION The continued worldwide growth of broadband ,QWHUQHW XVH KDV H[SRVHG WKH EHQH¿WV RI DQ ³H VRFLHW\´WRDQLQFUHDVLQJQXPEHURIWKHZRUOG¶V population, resulting in improved access to healthcare through telemedicine, the diminishing of the social divide through tele-education, and increased
economic competitiveness through telework. Although many countries around the world have modern telecommunications networks that support broadband access to the home and business, broadband penetration rates vary across these countries. One of the key factors explaining this difference in penetration rates is the government initiatives in the various countries, particularly in
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
the more developed Asian economies, to establish national “information superhighways.” Governments are motivated to develop national EURDGEDQGQHWZRUNVIRUVHYHUDOUHDVRQV(I¿FLHQW information-age infrastructures can enhance productivity by providing intelligent networks that can accommodate the converging voice, data, and electronic commerce applications (Frieden, 2005). These infrastructures provide a competitive advantage in the knowledge-based industries that include data processing, insurance, management, customer relationship management, and logistics and distribution. Such a competitive advantage in the area of information and telecommunications technology (ICT), when combined with a stable economy and favorable regulatory system, has been shown to translate to higher levels of foreign direct investment (FDI). The deployment of such networks, either directly through the use of tax dollars or indirectly by use of appropriate poliFLHVVXFKDVVSHFWUXPDOORFDWLRQE\¿DWIXUWKHU DOORZQDWLRQDOJRYHUQPHQWVWRH[SORLWWKHEHQH¿WV RI HJRYHUQPHQW LQFOXGLQJ LQFUHDVHG HI¿FLHQF\ and transparency. Governmental involvement and promotion of broadband can encourage end user adoption, by effectively lowering the usage cost of applications for business and consumers alike, thus improving working and living standards, and enhancing productivity. 6SHFL¿F DSSOLFDWLRQV WKDW DUH PDGH SRVVLEOH through broadband capabilities, such as telemedicine, may soon enable governments—particularly developing countries where transaction costs are high because of logistical problems (Sein & Harindranath, 2004)—to extend universal healthcare services to remote areas and provide an economically viable means to increase the quality of medical services. Similarly, tele-education provides governments with the potential to increase the level of educational services to resource-strapped inner-cities or rural areas and the means to extend the learning experience into the home. In addition, teleworking allows employees to work from home while still being productive agents of society, DQGJLYHV¿UPVDPHDQVWRGHFUHDVHFRVWVLQWKH production process. Furthermore, from a societal
YLHZSRLQWWHOHZRUNLQJUHGXFHVWUDI¿FFRQJHVWLRQ and helps in pollution control. Given the numerous areas where broadband deployment can bring about both social and ecoQRPLFEHQH¿WVIRUVRFLHW\GHPDQGFDQEHH[SHFWHG to grow in the coming years. The question remains regarding the proper role of governmental intervention in this process, and what role the unique FXOWXUDOFKDUDFWHULVWLFVRIVSHFL¿FVRFLHWLHVEULQJ to bear. This chapter addresses these issues by presenting a number of case studies illustrating the roles both governments and culture have played in the broadband deployment process, followed by a GLVFXVVLRQRIRXU¿QGLQJV
THEORETICAL FRAMEWORK The relationship between telecommunications investment and growth has been demonstrated (Roller & Waverman, 2001). Indjikian and Siegel (2005) review work from the developed world that suggests evidence, both qualitative and quantitative, of a positive correlation between IT investment and economic performance. The implications for developing countries are encouraging, and it DSSHDUVWKDWWKHHPSLULFDO¿QGLQJVRQ,7DQGHFRnomic performance, including case analysis from China, India, Chile, and elsewhere, are contributing to improvements in productivity and growth. In the economic literature, endogenous growth theory (or new-growth theory) stipulates that it is technology and human capital, when endogenously present, that contribute to continuous economic growth and therefore play an essential role in a FRXQWU\¶V GHYHORSPHQW (DVWHUO\ .LQJ /HYLQH & Rebelo,1994; Barro, 1997). Logically, workers who are better educated and literate, better fed, healthier, and technologically capable can produce more than those who are hungry, illiterate, unhealthy, and technologically unskilled. %DUUR¶V VWDWLVWLFDO DQDO\VLV RI WKH GLIference in growth rates across a large number of countries reveals that a high initial level of huPDQFDSLWDOKDVDVLJQL¿FDQWO\SRVLWLYHHIIHFWRQ growth—a virtuous circle. We can predict that ICT penetration—the ability to use technology to
261
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
LQFUHDVHHI¿FDF\DQGHI¿FLHQF\²ZRXOGKDYHWKH same reinforcing effect. A key economic incentive for some governments to develop a broadband network is to create competitive advantage in attracting direct investments by multinational corporations (MNCs). Foreign direct investment increases domestic capital formation, augments host country stocks of technology and managerial expertise, improves access to export markets, and provides a comparaWLYHO\VWDEOHVRXUFHRIH[WHUQDO¿QDQFLQJ /HKPDQ 2002, p. 3). Furthermore, FDI strengthens a host FRXQWU\¶VDELOLW\WRUHDFKPDUNHWVDEURDGWKURXJK its international links. ,Q JHQHUDO WKH LQYHVWPHQW ÀRZV WR WKH (DVW Asian economies studies in this chapter have been directed at creating export-oriented industries (Fan & Dickie, 2000). The rapid growth of these economies is widely attributed to a strategy of building industrial capabilities through foreign direct investment, achieved principally by creating favorable conditions for MNCs to locate their activities in their respective economies (Matthews, 1999). As expected, many of the East Asian economies are countries with some of the highest broadband penetration rates, in order to create these favorable business and investment conditions.
CASE STUDIES Singapore Government’s Efforts to Promote a Digitized Knowledge Economy The government of Singapore has been particularly aggressive in promoting the deployment of new technologies and broadband access, as manifested in several initiatives, which are discussed below. The government of Singapore has pursued an international-oriented business focus for a number of years. One result of this plan is that the total SURGXFWLRQ RI 86 01& DI¿OLDWHV FRQVWLWXWHG VRPHRI6LQJDSRUH¶V*URVV'RPHVWLF3URGXFW (GDP). Altogether, there over 7,000 MNCs located in Singapore (Lim, 2006).
262
Relatively high labor costs in Singapore, coupled with high employee turnover rates—particularly in the IT sectora manpower shortage, and the high overall cost of doing business in Singapore is motivating the government to look to e-commerce solutions, as some have observed, to encourage ¿UPVWRDGRSWWHOHZRUNLQJ /LP 7HRS 560). 2QH RI WKH ¿UVW WZR 6LQJDSRUH JRYHUQPHQW initiatives was the deployment of TradeNet and the Singapore ONE program. TradeNet, which facilitates trade documentation online, processes 99% of all trade permit declarations and is used by over 2,400 companies. This service has reduced SURFHVVLQJWLPHIURPVHYHUDOGD\VWR¿YHPLQXWHV 7UDGH1HW3OXVDPRUHFRPSOHWHDQGHI¿FLHQWYHUsion that incorporates other online features, such as payment and insurance, is expected to further reduce processing time to between one and three minutes. Furthermore, it is estimated that it will generate annual savings of some S$2.8 billion (Shih, 1999, p. 2). The Singapore ONE (One Network for EvHU\RQH SURJUDP LV D KLJKVSHHG RSWLFDO ¿EHU QHWZRUNWKDWPDQLIHVWVWKHJRYHUQPHQW¶V,7 Masterplan, a blueprint to deploy IT in almost every government department with the end goal of transforming the state into an “intelligent island” (Economist, 2000). The IT2000 Masterplan was GHVLJQHGWRDOORZDOORI6LQJDSRUH¶VRYHU KRXVHKROGVWRFRQQHFWWRDK\EULG¿EHUFRD[LDO network, irrespective of whether they intended to subscribe to the various services (Cable and Satellite Asia, 1997). The Singapore ONE initiative strives to develop a nationwide broadband network to link businesses, homes, and schools to deliver interactive, multimedia applications and services (Santiago, 2000). It is estimated that by 2005, some 43% of all households had connected to this broadband network (Burton, 2005). In this respect, WKH6LQJDSRUHJRYHUQPHQW¶VH&LWL]HQ&HQWHUDQ integrated service delivery system, was designed to ensure that the public sector operated, and was seen to operate, as a single entity. This particular project is aimed at bringing together useful services and delivering them to Singaporeans in convenient and easily accessible packages. The Singapore
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
government hopes that eCitizen will enhance the ability of the public to be increasingly familiar and comfortable with IT, which has become a critical component in the knowledge economy (Economist, 2000). In June 2006, the Singapore government launched its latest 10-year Masterplan, dubbed iN2015 (Intelligent Nation 2015) which established ambitious goals for the country: the establishment RI6LQJDSRUHDV¿UVWLQWKHZRUOGLQKDUQHVVLQJ,&7 to add value to the economy, a two-fold increase in value-added of the ICT industry to S$26 billion, a three-fold increase in ICT export revenue to S$60 billion, the creation of 80,000 additional jobs, a 90% rate of home usage for broadband, and the achievement of 100% computer ownership for all homes with school-going children (Business Wire, 2006). The government hopes to accomplish these goals through programs and initiatives that LQFOXGHIRUH[DPSOHWKH&;2SURJUDPZKLFKZLOO facilitate strategic discourse among key decision makers, CEOs, CIOs, and CFOs, and the iEnterprise Challenge Program, which will focus on developing innovative products and solutions as reference projects, which will have a strong impact across various sectors while simultaneously creating new intellectual property (Lim, 2006). Another government initiative seeks to raise the percentage of small and medium-sized enterprises (SMEs) conducting e-commerce to 25% within WKUHH\HDUV¶WLPH$FFRUGLQJWRDVXUYH\FRQGXFWHG late in 1999 by the Singapore IT Federation, while 91% of the top 1,000 companies already had Internet access, only 4% of the 92,000 SMEs were conducting e-commerce (Raj, 2000, p. 7). Sixtysix percent of SMEs with sales below $1 million cited lack of in-house IT skills and infrastructure, inadequate knowledge of e-commerce, and low budgets as barriers to implementing e-commerce (Koh, 2000, p. 14). As such, the Economic Development Board (EDB) will assist businesses in their e-commerce strategy by supporting up to 50% of WKHLUH[WHUQDOFRQVXOWDQWV¶FRVWVLQIRUPXODWLQJD business plan or business collaboration venture, conducting feasibility studies, or providing assistance in implementation, up to a maximum limit of S$250,000 (Lee, 2000, p. 71). The government
also unveiled a three-year, $11.7 million plan to SURGWKHFLW\¶VJURZLQJORJLVWLFVLQGXVWU\LQWRGRLQJPRUHZLWKWKH,QWHUQHW6LQJDSRUH¶VWUDQVSRUW and logistics industry contributed more than 7% WRFRXQWU\¶VWKHJURVVGRPHVWLFSURGXFWODVW\HDU (Bangsberg, 2000, p. 10). 7KH6LQJDSRUHJRYHUQPHQW¶V¶*H%,=LPLWDWLYH is an attempt to create a one-stop, round-the-clock FHQWHUIRUWKHJRYHUQPHQW¶VEXVLQHVVGHDOLQJV7KH ¿UVW SKDVH ZDV ODXQFKHG LQ $SULO *H%,= HQDEOHV WKH ¿QDQFLDO V\VWHPV RI PLQLVWULHV DQG the procurement applications to work together. 7UDGLQJ SDUWQHUV FDQ ¿QG LQYLWDWLRQV WR WHQGHU and purchase orders on the site. Suppliers can also submit invoices, check payment status, post their catalogs, and bid for contracts. Currently, purchases are capped at S$30,000 (US$17,341), but the Ministry of Finance estimates that once new payment and security systems are introduced, 80% of all government procurement will transfer to GeBIZ. In 2005, some 41,000 small-value purchases worth some $36 million were transacted over GeBIZ (Business Times, 2006). As with other online B2B WUDGLQJQHWZRUNVWKHEHQH¿WVFRPHLQWKHIRUPRI more competitive bidding, easy access to suppliers around the world, time saved by online processing of orders, lower stocks, and automated collection of high-quality data (Economist, 2000).
Taiwan Launches National Information Infrastructure Plan The government of Taiwan has also taken major steps to promote Internet usage among the population, making it clear from the start that its program was designed to promote simultaneously economic development and quality of life for citizens. The enabling steps the government has taken have included the gradual liberalization, according to a timetable, of each aspect of the telecommunications industry. The end goal is to move from a statemonopolized market to a completely competitive RQH7KHJRYHUQPHQW¶VSURJUHVVLQWKLVHQGHDYRU has been remarkably successful. Its liberalization and development plan was exceedingly well enumerated, planned out, and funded, and the fruits
263
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
RIWKLVODERUDUHHYLGHQWLQWKHSURJUDP¶VOHYHORI success over a relatively short time span. The number of broadband subscribers in Taiwan has rapidly expanded since the government began promoting the service in 2001. At the end of 2000, Taiwan had 262,800 subscribers. This number increased to 3.04 million subscribers by December 2003. By June 2005, this number had risen to 3.87 PLOOLRQ VXEVFULEHUV ZKLFK SODFHG 7DLZDQ ¿IWK in the world in household broadband penetration (Point Topic, 2006). As another example of rapid progress, Taiwan was recently ranked seventh among 105 nations around the globe, in terms RI µQHWZRUN UHDGLQHVV¶ LQ WKH :RUOG (FRQRPLF )RUXP¶V*OREDO,QIRUPDWLRQ7HFKQRORJ\5HSRUW 2005-2006, up eight places from just the previous year (E-Taiwan Update, 2005). In June 2002, the government of Taiwan launched a major initiative for e-enablement—the e-Taiwan Program—as part of the Challenge 2008 Program. This program promotes digitization throughout Taiwan through a number of initiatives LQ RUGHU WR LPSURYH WKH QDWLRQ¶V ,7 SUR¿FLHQF\ and competitiveness, both domestically and on an international level. This program involves the participation of a number of government agencies and is wide ranging in focus. The original framework of the program was developed around ¿YHSLOODUV LQIUDVWUXFWXUHHVRFLHW\HLQGXVWU\ e-government, and e-transportation. The e-Taiwan Program has been regularly subjected to performance and progress reviews, and has been revised according to the changing conditions and requirements over time. In June 2004, for example, Premier Yu Shyi-kun recommended that e-Taiwan Program initiatives be expanded to include a greater number of disadvantaged groups. 7KXVWKHµHRSSRUWXQLW\¶SODQZDVERUQ7KLVSODQ has been incorporated into the e-Taiwan Program DQGZLWKWKHDLPRIEULQJLQJJUHDWHUEHQH¿WDQG convenience to the greater Taiwanese public, includes several initiatives such as teleworking, healthcare, and e-governance, intended to be more user oriented in its approach. Some of the goals of the e-Taiwan project are WRLPSURYHJRYHUQPHQWHI¿FLHQF\UDLVHLQGXVWU\ FRPSHWLWLYHQHVV LPSURYH FLWL]HQV¶ TXDOLW\ RI
264
life, and establish an equal opportunity e-society through the use of modern information and communications technologies. By doing so, the program will accelerate the development of a knowledge economy in Taiwan, with the aim of achieving the reality of a high-tech services island (E-Taiwan Program, 2005). Recognizing the importance of being able to benchmark progress, the government has estabOLVKHGVSHFL¿FPHDVXUDEOHWDUJHWVIRUWKHSURJUDP including: the creation of 20,000 jobs related to the program or to technology, to achieve a 75% satisfaction rate of e-government policies, to provide 600 online services, to enable e-commerce to represent 15% of total GDP, and to expand broadband use to 6 million subscribers (E-Taiwan Program, 2005). 'HPRQVWUDWLQJ WKH JRYHUQPHQW¶V FRQWLQXHG commitment to the e-Taiwan project as well as to expanding it, Dr. Feng-Ching Lin, Minister of 6WDWHDQQRXQFHGDWWKH:RUOG(FRQRPLF)RUXP¶V October 2006 meeting, the new basis for the JRYHUQPHQW¶VYLVLRQ RQHRI7DLZDQ¶VQHZJRDOV will be to focus on the development of a “ubiquitous network society” in which all citizens would be able to access the Internet anytime, from any location (FIND, 2006). One particularly noteworthy aspect of the eTaiwan Program has been its recognition that there are greater societal aspects to the digital revolution, and promoting telecommunications availability can serve the country in ways reaching far beyond just increased international economic competitiveness. The program has been consistently successful in focusing on assisting the disadvantaged and helping to bridge the digital divide. To this end, in 2005 the government established 33 Digital Opportunity Centers (DOCs) in remote and rural areas of the country. This has provided disadvantaged groups and citizens in remote areas DQHTXDORSSRUWXQLW\WREHQH¿WIURP,QWHUQHWDFcess. In its greater effort to help reduce the digital divide globally, the Taiwanese government has taken this successful program to the international level, assisting Peru, the Philippines, and Vietnam to build their own Digital Opportunity Centers, and has plans to assist Chile, Indonesia, and Papua
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
New Guinea with the same in the future (E-Taiwan Update, 2005).
Malaysia Seeks Foreign Direct Investment through Infrastructure Improvement In much the same way, Malaysia is striving to become a fully developed economy through its Vision 2020 program, designed to shift focus to high value-added products with more automated operations. Malaysia sees itself well suited for high-tech, knowledge-driven, and service-based outsourcing industries, believing that it will eventually lose its competitive advantage in labor-intensive manufacturing to highly populated countries such as China and India that have even lower labor rates. Malaysia aspires to be a global leader in the IT sector by producing leading-edge technology products to be developed and used both within the country and to be exported. As a result, various government policies and initiatives focus on IT-related industries (Simic & Lee, 1997). 7KHVXFFHVVRIWKHJRYHUQPHQW¶VLQLWLDWLYHVLV evidenced by the amount of foreign direct investment into the country. Not only has this investment been relatively stable, but it has shown a marked increase over the past year from $2.7 billion in 1998 to $4.5 billion in 2005 (UNCTAD, 2007). The United States continues to be the largest foreign investor in Malaysia, accounting for nearly 30% RI0DOD\VLD¶VWRWDO)',DQGWKHWZRQDWLRQVDUH currently exploring the possibility of a bilateral free trade agreement. The ASEAN region overall has also seen record levels of FDI over the past few years, higher than those seen before the Asian ¿QDQFLDOFULVLV 7DQ 0DOD\VLD¶V PRVW SURPLQHQW ,7 SURMHFW LV WKH Multimedia Super Corridor (MSC). Unveiled by the government in August 1996, the original MSC project encompasses a 15x40-kilometer area, stretching from the Kuala Lumpur City Center to the new Kuala Lumpur International Airport (KLIA). This region has been transformed into a multimedia research and manufacturing zone. It will be home to Putrajaya, the paperless Malaysian federal government, and Cyberjaya, a new informa-
tion-age industrial and residential city. Projected to run well over US$40 billion, the MSC will include tele-suburbs, technology parks, and a multimedia university (Tan, 2006). The MSC will be equipped with a state-of-the art telecommunications and network infrastructure capable of supporting interactive applications and serviced by high-speed connections to Japan, the United States, Europe, and the ASEAN region. The Malaysian government has designated VHYHQ³ÀDJVKLSDSSOLFDWLRQV´LWVHHNVWRGHYHORS LQSDUWQHUVKLSZLWKSULYDWH¿UPVWKDWKDYHJDLQHG MSC status. These applications are in the following areas: electronic government, smart schools, telemedicine, multi-purpose smart card development, borderless marketing, an R&D cluster for developing software and next-generation multimedia technologies, as well as a multimedia-based worldwide manufacturing web that will serve as DVXSSRUWKXEIRU¿UPVLQKLJKFRVWFRXQWULHVWR access lower-cost facilities throughout Malaysia and the region (Simic & Lee, 1997). When entirely completed, the MSC asynchroQRXV WUDQVIHU PRGHEDVHG ¿EHURSWLF QHWZRUN within the corridor will support automated information network (AIN) telephones and enhanced services, cable television, video-on-demand, data communications, MSC applications, and teleport services. Telecom Malaysia is also considering using ADSL technology to connect to homes and RI¿FHV7KHIDVWHVWJURZLQJVHUYLFHDWSUHVHQWLV dedicated ISDN networks (Legard, 1998). The infrastructure will also be linked to Telekom 0DOD\VLD¶V FRUSRUDWH LQIRUPDWLRQ VXSHUKLJKZD\ (COINS) network. COINS is an open multimedia network with extreme high-speed capacity (Kim, 1997). 7KH 06& SURMHFW¶V GHYHORSPHQW KDV JRQH forward over the past 10 years, yet it has also experienced some growing pains. For example, there has been a call for greater oversight of the JRYHUQPHQW¶V FRQWUDFWLQJ DQG ,&7 VSHQGLQJ LQ light of revelations that some of the companies contracted to provide services under two of the JRYHUQPHQW¶VÀDJVKLSDSSOLFDWLRQV²LQWKHDUHDV of telehealth and electronic government—had not
265
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
completed their assignments. Concessions needed to be extended or revoked as a result. In 2001, Malaysian Prime Minister Mahathir Mohamad reported that the impact of the MSC RQ0DOD\VLD¶VHFRQRP\ZDVQRWDVVXFFHVVIXODV the government had expected (ZDNet Asia, 2001). Nevertheless, the government has continued its work on developing and promoting the MSC. Responding to numerous calls by corporations (and the media) to increase the number of ICT-skilled Malaysian workers, the government recently announced a double tax deduction for employee costs in ICT training. In other words, corporations are allowed to doubly deduct any expenses incurred while their employees are being trained in ICT, at an approved institution or by government training institutes (Measures, 2006). Additionally, the government has streamlined the entry visa and work permit issuing process, also in response to reported frustration on behalf of corporate interests (Matthews, 2006). 7KHJRYHUQPHQWKDVODXQFKHGWKHSURMHFW¶VVHFond phase which will be extended across the nation. The project has been renamed MSC Malaysia, in an effort to “brand” the project and increase global awareness of its location. In addition to launching a second cyber center in the city of Malacca, south of Kuala Lumpur, the Prime Minister declared the DLPRIWKHSURMHFW ³WRH[SDQGWKH06&¶VFRUULGRU network throughout the country and to bridge the digital gap between those with access to information and communications technologies and those without” (Matthews, 2006). Although the original focus and objectives of the 06&SODQUHPDLQWKHVDPHLWLVWKHJRYHUQPHQW¶V intent to improve the methods, approaches, and strategies employed to achieve the goals of the MSC Malaysia plan. As part of this plan, additional cyber centers and cyber cities will be established in various parts of the country. In addition to the Malacca center mentioned above, the rollout of MSC Malaysia continues at Johor and is expected to be extended to Perak within the year. To date, the MSC project has created 48,200 jobs, with 1,522 international and local companies SDUWLFLSDWLQJ7KHSURMHFWKDVEHQH¿WHGIURPPDMRU investments by companies such as Dell, DHL, Sun
266
Microsystems, BMW, Shell, EDS, Unisys, and AMD (Matthews, 2006). MSC Malaysia is not the only area in which the Malaysian people have seen progress in ICT use and uptake over the past decade. Mobile penetration in Malaysia surpassed the 80% mark in early 2006, with subscribers numbering over 20 million. This was up from just 2 million subscribers in 1998. Malaysia has the second highest mobile penetration rate in South East Asia, after Singapore (CAN Newswire, 2006). The government awarded two 3G licenses in March 2006, and this is expected to boost activity and competition in this sub-sector. The number of PC users has also risen dramatically. Malaysia had 888,000 PC users in 1996, and this had risen to 4.2 million users by 2006. The Internet penetration rate rose from 1.8% of the population in 1998 to nearly 40% in 2006. The JRYHUQPHQW¶VRYHUDOOSODQLVWRPDNH0DOD\VLDD knowledge-based economy and community, where ICT is seen as the key enabling tool for all other sectors of the economy, including manufacturing, services, and agriculture (MSC, 2006).
The Role of Advanced Telecommunications Infrastructure in Hong Kong +RQJ.RQJ¶V6SHFLDO$GPLQLVWUDWLYH5HJLRQ 6$5 KDVRQHRIWKHZRUOG¶VPRVWVRSKLVWLFDWHGDQGDGvanced telecommunications networks, which is crucial for Hong Kong to maintain its status as a OHDGLQJEXVLQHVV¿QDQFLDODQGLQGXVWULDOFHQWHU Sizeable investments in communications technology have made the telecom market in Hong Kong KLJKO\HI¿FLHQWDQGFRVWHIIHFWLYHIRUWKHPDQXfacturing industry. An advanced network, together with a wide range of professional and high valueadded manufacturing support services, ample supply of skilled and low-cost workers from China DQGRWKHU$VLDQ3DFL¿FFRXQWULHVDOOFRXSOHGZLWK the cross-border production competence of Hong Kong companies, has positioned Hong Kong as a control center in global production and sourcing (Chan, Yung, & Burns, 2000, p. 288).
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
Hong Kong boasts what is widely considered WKH ZRUOG¶V IUHHVW HFRQRP\ DV ZHOO DV D KLJKO\ VNLOOHG ZRUNIRUFH ÀXHQW LQ (QJOLVK WKH OLQJXD franca of the business, legal, and technology industries. The government of Hong Kong has been one of the leading investors in its IT sector through its resource allocation, e-government program, its implementation of information systems and procurement arrangements, as well as through DGPLQLVWUDWLYHDGRSWLRQ7KHJRYHUQPHQW¶VLQYHVWment in this industry has averaged HK$4.6 billion per year (Commerce, Industry, and Technology Bureau, 2004). The city launched its bid to privatize and demonopolize local and long-distance telephone services in 1995 and 1998, respectively. By 2001, all telecom services and infrastructures were liberalized, which resulted in wider choices and a higher quality of services at more competitive prices. Today, Hong Kong has one of the most advanced and competitive telecom markets in the world, with six major mobile operators competLQJWRVHUYH+RQJ.RQJ¶VPDUNHWQRQHRIWKHP dominating (Yan, 2006). In fact, with a penetration level of 123 mobile phones per 100 inhabitants, it may not be an exaggeration to claim that Hong Kong has the highest level of telecom competition, as well as adoption of mobile technology, in the world (Yan, 2006). 7KHJRYHUQPHQW¶VSROLFLHVDQGWKHGULYHWRZDUG market liberalization created a great deal of presVXUHIRULQQRYDWLRQ$VDUHVXOW+RQJ.RQJ¶VIRUPHU WHOHFRPPRQRSROLVWEHFDPHWKH¿UVWSURYLGHULQWKH world to launch interactive television (video-ondemand) service commercially in 1998, and was DOVRWKH¿UVWWRRIIHU,QWHUQHWSURYLGHUWHOHYLVLRQ (IPTV) service. In 2006, they launched a mobile real-time TV broadcasting service as well (Yan, 7KHSHRSOHRI+RQJ.RQJKDYHEHQH¿WHGLQ QXPHURXVZD\VIURPWKHJRYHUQPHQW¶VSURPRWLRQ RIFRPSHWLWLYHSROLFLHV +RQJ.RQJZDVWKHZRUOG¶V ¿UVWFLW\WRIXOO\GLJLWL]HLWVWHOHFRPPXQLFDWLRQV QHWZRUNV,WZDVDOVRWKH¿UVWHFRQRP\LQWKHZRUOG WRLQFRUSRUDWH¿[HGOLQHQXPEHUSRUWDELOLW\DQG WKLUGLQWKHZRUOGZLWK¿[HGOLQHPRELOHQXPEHU portability.
$JDLQVWWKLVEDFNJURXQGWKHJRYHUQPHQW¶V¿UVW Digital 21 IT Strategy was published in 1998. It was revised and updated in 2001 and then again in 2004. The plan is a comprehensive strategy to bring Hong Kong into the 21st century as a global leader in e-business, new technologies, and the LQIRUPDWLRQHFRQRP\7KHJRYHUQPHQW¶VYLVLRQLV that it should be an effective facilitator to promote innovation and the development of industry, as well as to ensure that the entire community will EHQH¿WIURP,7GHYHORSPHQWE\WDNLQJPHDVXUHV to bridge the digital divide. 7ZR ÀDJVKLS SURJUDPV RI WKH JRYHUQPHQW¶V Digital 21 Plan are the Cyberport and the Science and Technology Park. Both of these industrial clusters provide high-technology-focused value DGGHGWR+RQJ.RQJ¶VRYHUDOOHFRQRP\ZLWKWKH larger concept of attracting leading IT companies and a critical mass of professional talent into Hong Kong. The Cyberport is a high-technology multimedia hub designed to attract the research and development sectors of major high-tech companies. The &\EHUSRUW¶V3KDVHRFFXSDQF\UDWHKDVDOUHDG\ reached 80%, attracting such major tenants as Microsoft and GE Information Services. The Cyberport is designed to be more than a workplace, as it incorporates both a world-class living and working environment, and attracts businesses from many industries, including entertainment, communications, advertising, sales, trading, and ¿QDQFH 0$,7 The Science and Technology Park focuses on four sectors: ICT, electronics, biotechnology, and precision engineering. It is equipped with broadband networking and wireless LAN, laboratories, and support centers. +RQJ.RQJ¶VHFRQRP\EHQH¿WVIURPLWVVWDWXV DVDPDMRUJOREDO¿QDQFLDOVHUYLFHVFHQWHUDVZHOO as a trade, commerce, transportation, and logistics hub. All of the sectors comprising these industries are major users of IT, which has helped provide the impetus for Hong Kong to be at the forefront of the IT development and innovation, as well as for the government to facilitate and augment this SURFHVV+RQJ.RQJ6$5¶VEDQNLQJVHFWRULVWKH second largest in Asia, after Japan, and in terms
267
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
of total bank assets to GDP it is only matched by Singapore. +RQJ.RQJ¶VWUDGHDQG¿QDQFLDOFRPPXQLW\ is extremely “IT oriented.” The Hong Kong Trade Development Council (HKTDC) promotes Hong .RQJ¶VJRRGVDQGVHUYLFHVE\VHUYLQJDVDQLQformation broker and matchmaker among traders, exporters, manufacturers, buyers, and service providers. With the advent of the World Wide Web, the HKTDC introduced a host of electronic advertising and publishing activities to take advantage RIWKH:HE¶VLQVWDQWDQHRXVORZFRVWLQIRUPDWLRQ distribution to a wide online audience. Besides using the Internet for Web browsing, there is an increasing trend by Hong Kong corporations to use the Internet as the backbone to develop intranets for mission-critical applications, including dynamic access to databases, and functional applications such as purchasing, trading, and inventory control. The rapid growth of the Internet and intranets is generating rapid expansion in LANs (IMF, 2002). Hong Kong has a high level of LAN penetrationapproximately 70% of all corporations. +RQJ.RQJ¶VHFRQRP\KDVZLWQHVVHGUHPDUNDEOH JURZWK LQ )', LQÀRZV RYHU WKH SDVW IHZ GHFDGHV*OREDO)',LQÀRZVURVHVXEVWDQWLDOO\LQ 2005, up 29% from 2004, as Hong Kong maintained its number two position (behind China) as East $VLD¶VOHDGLQJUHJLRQDO)',WDUJHW +RQJ .RQJ¶V FXOWXUH LV FKDUDFWHUL]HG E\ LQdustriousness and by individuals who desire to better themselves (MAIT, 2005). As was menWLRQHGLQ.RUHD¶VFDVHDERYHVXFKFKDUDFWHULVWLFV may promote the adoption and acceptance of new technologies, as these are seen as methods for LQFUHDVLQJ HI¿FLHQF\ WRROV IRU OHDUQLQJ DQG DV an opportunity to show that one owns and knows how to use the latest device. $JUHDWSURSRUWLRQRI+RQJ.RQJ¶VSRSXODWLRQ KDVUHDOL]HGWKHEHQH¿WVRIWKHJRYHUQPHQW¶VSUR technology efforts. Broadband is now available to over 95% of households and to more than 98% of business buildings. The PC penetration rates in 2004 stood at 71% for households and 58% for businesses, while Internet penetration rates held at 65% for households and 50% for businesses (MAIT,
268
2005). The mobile penetration rate of 123% was mentioned above, while 70% of Internet users had used e-government services. Despite this level of technological progress, the government still acknowledges that it must go further in order to bridge the digital divide within its borders. In order to ensure that the entire FRPPXQLW\ZLOOEHQH¿WIURP,7GHYHORSPHQWWKH government has begun collaborating with both industry and NGOs in order to introduce a wide UDQJHRIPHDVXUHVWRIXUWKHUHQKDQFHUHVLGHQWV¶ quality of life. The “IT Hong Kong” campaign was designed as a result, to promote IT adoption in the greater community. It consists of free courses providing IT training, free talks in libraries, district promotional activities, and infotainment programs for citizens. Other initiatives include providing public computers with Internet access at convenient locations, computer recycling for the needy, ¿QDQFLDODVVLVWDQFHWRSHRSOHZLWKGLVDELOLWLHVIRU the purchase of computer facilities to enable working from home, and the installation of devices in public computers so that they may be accessed by the blind and visually impaired (Commerce, Industry, and Technology Bureau, 2004).
Korea’s Initiative to Develop a Broadband Infrastructure The Korean government has taken a strongly proactive approach to promoting an information society for all of its people, businesses, and industries. In DOORILWVHQGHDYRUVWKHJRYHUQPHQW¶VRYHUDUFKLQJ objective has been to utilize ICT to enhance the quality of life in Korea. .RUHD¶VIRFXVRQSURPRWLQJLQIRUPDWLRQDQG communication technologies began in the mid1990s, with the establishment of the Ministry of Information and Communication (MIC) in 1994 and the announcement of its overall broadband strategy, the Korea Information Infrastructure .,, LQ .RUHD¶VEURDGEDQGPDUNHWLVFXUrently an open market, free of regulation and controls over licensing and pricing. Its incumbent telecom provider, Korea Telecom, has been completely privatized as of May 2002 (Lee, 2002).
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
One aim of the KII initiative was to stimulate private sector investment and competition in the broadband market, by lowering the regulatory barriers to entry. A second aspect of the initiative was for the government to provide a high-speed network that would serve as a public backbone, providing broadband services to 30,000 governPHQWDQGQRQSUR¿WRUJDQL]DWLRQVDVZHOODVWR research institutes and to around 10,000 schools. The government had another ambitious goal: to provide high-speed Internet to every household in Korea. Not only has this goal nearly become realLW\EXWWKHJRYHUQPHQW¶VWDUJHWHG,QWHUQHWVSHHG (threshold) is continually revised upward: the current target is to provide 20 Mbps connectivity to all homes by the end of 2006, and 50-100 Mbps to all homes by the end of 2010 (Lee, 2002). The Korean government is continually revising its strategies, due in part to rapid changes in technological innovation and market demands (Lee & Chan-Olmsted, 2004) and putting forth new initiatives in its wide-ranging telecommunications efforts. In December 2002, Korea announced its eKorea Vision 2006 plan. This initiative, with a focus RQLPSURYLQJWKHTXDOLW\RI.RUHDQFLWL]HQV¶OLYHV KDG¿YHREMHFWLYHV7KH¿UVWZDVWRPD[LPL]HWKH opportunity for and ability of all Korean citizens to utilize the information society and communication technologies in order to actively participate in the information society (or to be “prosumers”—both producers and consumers of information.) The government took action by providing increased opportunities for Internet access for all, and by establishing a lifelong education system available through online learning (Ministry of Information DQG &RPPXQLFDWLRQ &LWL]HQV¶ DELOLW\ WR utilize IT will lead to the creation of added value in all aspects of society. The second objective is to strengthen the FRXQWU\¶VOHYHORIJOREDOHFRQRPLFFRPSHWLWLYHness by promoting “informatization” within all Korean industries, as the IT industry strengthens the competitiveness and productivity of most all other industries. This objective also has an environmental component: the use and application of ICT will enable improvements in the ecological system and ensure the conservation of a clean and
natural environment by strengthening environmental policy functions, while its use in the realm RIWUDQVSRUWDWLRQZLOOUHOLHYHWUDI¿FSUREOHPVDQG reduce logistics costs (Ministry of Information and Communication, 2002). 7KHWKLUGREMHFWLYHLVWRUHDOL]HDPRUHHI¿FLHQW and transparent government structure through informatization, by offering more online public services in the areas of education, culture, and VRFLDO ZHOIDUH 7KH 0LQLVWU\¶V IRXUWK REMHFWLYH is to facilitate continued economic growth by promoting the IT industry in general, and both core and new technologies in particular. Finally, the Ministry also aims to become a leader in regional technological cooperation, by establishing a system that promotes ICT collaboration within WKH$VLD3DFL¿FUHJLRQ 0LQLVWU\RI,QIRUPDWLRQ and Communication, 2002). In 2004, the MIC produced a new and ever further-reaching policy plan: the IT 839 Strategy. One of the overarching economic goals announced in this document was to see Korean GDP per capita grow to US$20,000. The exploitation and utilization of ICT is seen as crucial to achieving this goal (Ministry of Information and Communication, 2004). The IT 839 Strategy also seeks to establish and promote new and cutting-edge technologies, including a broadband convergence network, Internet protocol version 6, and new services, such as WiBro, digital multimedia broadcasting, home network service, telematics service, and Internet telephony (VoIP). These services, available over a converged network, will ensure Internet access at any time in any place (Ministry of Information and Communication, 2004). In terms of competition, both DSL and cable have been unbundled (de-monopolized and made competitive). This has resulted in making Korea one of the most truly competitive nations in terms of the levels of domestic ICT competition. In turn, this high level of competition has made for comparatively low costs and high rates of subscription E\.RUHDQFRQVXPHUV,QGHHG.RUHDLVWKHZRUOG¶V leading nation in household broadband penetration, with 78.5% of households having access (Point Topic, 2005).
269
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
The importance of foreign direct investment in the Korean economy can be seen in the table below. With policy reforms to allow increased foreign participation in the Korean economy, including the purchase of troubled corporations, the value of FDI between 1997 and 2000 quadrupled and VKDUH LQFUHDVHG IURP RI EXVLQHVV ¿QDQFLQJ to 13%. In 1998, FDI was equal to some 14.8% of total gross domestic investment in the Korean HFRQRP\7KRXJKWKHUHKDVEHHQVRPHÀXFWXDWLRQ in investment levels over recent years, 2004 data show a strong resurgence in foreign interest in .RUHD¶VHFRQRP\ ,QDGGLWLRQWRWKH0,&¶VRYHUDOOSROLF\SODQV the Korean government has launched a number RIVSHFL¿FEURDGEDQGUHODWHGLQLWLDWLYHVPDQ\RI them with an eye to bridging the digital divide. Among these has been the Public Fund Program, wherein the government provided low-interest ORDQVWRSURYLGHUVWRHDVHWKH¿QDQFLDOEXUGHQRI bringing access networks to small and mediumsized cities. An additional initiative the government has taken to close the digital divide was to provide 3,000 free Internet access points in public places, VXFKDVSRVWRI¿FHVDQGFRPPXQLW\FHQWHUV7KH government has also provided PCs and Internet infrastructure to schools. Further, within the context of its “Plan for Promoting ICT Use and Distributing PCs to Children of Low Income Families,” the government provides discounted ICT access and a PC to those who otherwise could not afford it (ITU, 2003). Further, the government has launched IT literacy programs, targeted at social groups that traditionally would not have access to computers or the Internet, such as housewives, the elderly, and prisoners. More than eight million people were trained between 2000 and 2002. The government has also started offering high school education programs that are broadcast over the Internet, increasing the demand for PCs and Internet connectivity in family homes (ITU, 2003). Beyond government initiatives, Korean culture has played a large role in determining ICT adoption. Some cultural characteristics prevalent in Korea include a high valuation placed on family
270
relationships, social advancement (as well as on social status in general), and learning and education (Davies, 2006). These cultural values promote hard work and personal development, and education is seen as a key means for achieving economic and social success. Another cultural characteristic at play is the “copycat syndrome” or “keeping up with the Joneses,” by which if one household subscribes to broadband Internet service, other households in the neighborhood will subscribe to an equal or superior service in order to maintain their social standing (Davies, 2006, p. 7). Both of these social characteristics have promoted the adoption of ICT within the Korean society. One downside of the cultural propensity for ICT uptake has been the widespread addiction to gaming. Most see this addiction spreading in reaction to the high level of societal pressure to H[FHOLQRQH¶VSHUIRUPDQFH JDPLQJRIIHUVDQDWtractive—and addictive—form of escape. This phenomenon has been abetted by the increasing prevalence of PC Baangs (PC rooms). These are inexpensive Internet cafes that, over the past few years, have morphed mostly into gaming rooms. The number of PC Baangs increased from 3,000 in 1998 to 22,500 in 2002. These Baangs, and gaming in general, are becoming so popular that addiction to gaming is currently considered one of the biggest social problems facing Korea—indeed, Korea is widely FRQVLGHUHG WR EH WKH KRPH RI WKH ZRUOG¶V PRVW extreme gamer culture, as well as the epicenter of gaming addiction. The situation has grown so acute that 10 South Koreans—mostly teenagers and people in their twenties—died in 2005 from game addiction-related causesfrom sitting in a single, cramped position for too long. To combat this problem, the government launched a game addiction hotline in 2006, after having opened a treatment center in 2002. Numerous private hospitals and psychiatric clinics have also opened units to treat this addiction (Faiola, 2006). ,Q WHUPV RI WKH EHQH¿WV RI ,&7 KRZHYHU DV PHQWLRQHGDERYHRQHRIWKHJRYHUQPHQW¶VNH\REMHFWLYHVLVWRHQVXUHWKDW,&7ZLOOPDNHDVLJQL¿FDQW difference in the daily lives of its citizens. To this
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
end, the government has implemented numerous ICT-oriented public services, such as e-health, e-transport, and e-learning, with positive results 'DYLHVS 7KHJRYHUQPHQW¶VSURPRWLRQ of ICT development is also expected to stimulate job creation. 7KH .RUHDQ JRYHUQPHQW¶V RYHUDUFKLQJ ,&7 objectiveto create an information societyhas been marked by its strong and continually renewed commitment. The results have been remarkable: one decade after the announcement and implementation of its initial ICT plan, the country now ranks at or near the top of almost any global-level survey measuring broadband use, uptake, or readiness, or of ICT penetration.
SIGNIFICANCE OF SOCIAL AND CULTURAL FACTORS FOR BROADBAND ADOPTION As the case studies above demonstrate, governmental initiatives can be important drivers of broadband penetration, by transforming the economy to a digitized, information-based one and promoting economic growth through numerous channels. However, the adoption of such a system also requires a transformation of societal structures and institutions, as well as an existing receptiveness to using the services and applications that can be provided through broadband, in order to more fully UHDOL]H,&7¶VSRWHQWLDOEHQH¿WVIRUVRFLHW\ A variety of social and cultural factors explain the growth of Internet adoption. Some of the cultural characteristics present in Asian states, including industriousness, the drive to keep up with RQH¶VSHHUVDQGWKHJDPLQJDVHVFDSHIURPVWUHVV mentality, were discussed above. In the United States, the society has a high degree of receptiveness to technological innovation. The population density is comparatively low relative to much of the developed world, and suburban life continues to increase with single-family homes the standard. With the need to communicate across great distances being the rule, the Internet offers DEHQH¿WLQWHUPVRIPDLQWDLQLQJFRPPXQLFDWLRQ links between families and friends, and colleagues.
Since the United States is a highly mobile society, the prospect for mobile Internet adoption among Americans initially appeared promising. However, mobile Internet has not caught on as rapidly in the U.S., and the majority of Americans still do the JUHDWHUSDUWRIWKHLU,QWHUQHWVXU¿QJIURPKRPH The decentralized governance of the Internet and the market structure in the U.S. led to a reducWLRQLQWKHFRVWRI:HEFRQQHFWLRQVÀDWUDWHDFFHVV schemes, and the emergence of high-speed bandwidth systems via telephone lines, cable systems, as well as satellites. As a result, there has been a fast spread of Internet connections and use. However, take-up rates for broadband to the home also have not met expectations. As such, even when broadband connections are available, consumer demand remains an issue: residential consumers may not ¿QGEURDGEDQGXVHIXOWRWKHLUOLYHVRUWKHUHPD\ EHLQVXI¿FLHQWVHUYLFHVDQGFRQWHQWWRMXVWLI\WR subscribing to broadband when it is available. One major explanation for this discrepancy between expectations and actual adoption patterns in the U.S. is that, unlike the Asian economies described above, the U.S. government has not to date established any comprehensive national policy to drive broadband development and adoption rates. ,WKDVDOVREHHQDUJXHGWKDWWKH86JRYHUQPHQW¶V policies have been prohibitive to competition, and in fact have promoted monopolistic behavior on the part of industry incumbents (Koski & Majmudar, 2000). Another major difference between the information superhighway programs in the Asian countries and the rest of the world is the concomitant government initiatives (in collaboration with the private sector) to develop content, particularly multimedia and IT applications. In Malaysia, for example, the government has established numerous incentives WR HQFRXUDJH WKH GHYHORSPHQW RI VHYHQ ³ÀDJship applications,” including telemedicine, smart schools, and electronic government, for its “Multimedia Super Corridor.” In much the same vein and for similar objectives, Taiwan has established the “Internet Applications/Electronic Commerce Program.” Similarly, in Singapore, the government KDV LQDXJXUDWHG D ³3LRQHHUV¶ &OXE´ WR FXOWLYDWH innovative applications and content providers for
271
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
its Singapore One Network. Club members are UHZDUGHGZLWKEHQH¿WVVXFKDVSUHIHUHQWLDOUDWHV for broadband usage as well as reimbursements of up to 70% of the cost for being new applications or content online.
Social Factors Behind the Success of Japanese Wireless Internet The unexpected success of i-Mode, the Internetenabled phone service from NTT DoCoMo in Japan, illustrates the pivotal role that culture and social traits play in explaining the different adoption rates for new technologies. In this case social and cultural factors, as well as geographical and historical circumstances, in large part account for the unprecedented growth of i-Mode: since its launch in February 1999, i-Mode has attracted on average some 500,000 new users monthly, and as of 2004, counted some 42 million users. While the European market has not yet met hopes for wireless Internet access, Japan remains the only market that has exceeded expectations. However, it is still debatable whether the success of i-Mode is transferable to other markets with different cultural, JHRJUDSKLFDQGVRFLHWDOLQÀXHQFHV $ PDMRU IDFWRU EHKLQG L0RGH¶V LQLWLDO VXFcess was its complementarity with the lifestyle of young urban Japanese consumers and the focus on entertainment-related content. i-Mode is hugely popular in densely populated cities like Tokyo ZKHUHZDLWLQJLQOLQHVIUHTXHQWWUDI¿FMDPVDQG general delays are commonplace. Some 55% of i0RGHDFFHVVLQWKH¿UVWIHZPRQWKVRIZHUH to entertainment-related areas, with the most popular categories including downloads of ring tones, characters and other pictures, horoscopes, games, and information about music (U.S. Department of Commerce, 2001). Overall, the trade-off between “reach,” which explains accessibility and ubiquity of coverage, and “richness,” which is the depth of content as well as interactivity, customization, and bandwidth, was appropriate for the cultural background and lifestyle of the Japanese consumer (Baldi & Thaung, 2002, p. 8). In addition, the relative rarity of home comSXWHUV LQ -DSDQ LV DQRWKHU VLJQL¿FDQW IDFWRU EH-
272
KLQGL0RGH¶VVXFFHVV'HVSLWH-DSDQ¶VLPDJHDV techno-friendly, in fact, the number of households ZLWK,QWHUQHWDFFHVVRQO\SDVVHGRQHLQ¿YHWKLV year—one of the lowest rates among industrialized countries. In addition, expensive Internet connections and phone calls via wireline connections have led to a high acceptance of the wireless Internet compared to the U.S. or Western Europe. As a result, wired penetration of the Internet has remained extremely low in Japan relative to other markets where wireless access is viewed as an alternative rather than primary means of Internet access. (Funk, 2000) Thus, i-Mode provides a cheap alternative to accessing the Internet for Japanese consumers, providing information and services FDOLEUDWHGWRWKHVSHFL¿FQHHGVDQGFXOWXUHRIWKH society (Suda, 2001).
Telemedicine in the United States The United States presents a distinct contrast to -DSDQLQWKHDGRSWLRQRIQHZWHFKQRORJLHV6SHFL¿cally, in the United States the value-proposition for telemedicine is attractive. Healthcare in the United States has been experiencing substantial and increasing cost pressures which could be addressed in part by telemedicine. For example, a 1992 study prepared by the Arthur D. Little consulting company estimated that the implementation of telemedicine would result in savings of $36 billion annually (Shoor, 1994). The fact that the savings estimates today remain strikingly similar to those made in 1992 UHÀHFWV KRZ OLWWOH SURJUHVV KDV EHHQ PDGH RYHU the past decade in the adoption of telemedicine in the United States (Pharmalicensing Limited, 2006).These savings could be generated from reduced costs for serving patients, through savings in time and travel for doctors and patients, fewer unnecessary referrals, and the replacement of doctors with less medically trained personnel supported by telemedicine. There would also be savings from the provision of better health care, generating cost reductions from early diagnosis and treatment. 'HVSLWHWKHSRWHQWLDOHFRQRPLFEHQH¿WVDQGWKH fact that telemedicine technology has existed since
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
the 1920s, usage has not been widespread due to structural and social constraints.a These barriers include low compatibility with existing medical practices, complexity of telemedicine equipment and interfaces,b multiple technical standards (Choi, Krause, Seo, Capitan, & Chung, 2006), the absence of reimbursement by third-party agencies, and incompatibility of state laws regarding telemedicine and licensure issues (McGee, 2004, p. 53; Smith, 2005). Overall, the social system surrounding the adoption of telemedicine is very structured and complex. In addition, the lack of clear support from key institutions, such as the American Medical Association and most medical colleges and medical schools, save the American College of Radiology, presents another impediment. This ambivalence is linked to four major social impediments to the use of telemedicine. First, the cost of implementing a telemedicine infrastructure is a large obstacle, both in the United States and in Europe (Statura, 2006). Currently, a large majority of telemedicine initiatives are sponsored by organizations where reimbursement is not crucial, like research centers, the Armed Forces, or state-owned hospitals, and these initiatives are IUHTXHQWO\¿QDQFHGE\GHPRQVWUDWLRQJUDQWV2QO\ DVPDOOQXPEHURIIRUSUR¿WPHGLFDOFHQWHUVDUH involved in telemedicine, and many of these, like the Mayo Clinic, are employing closed telemedicine systems (Tangalos, 1994). Furthermore, medical organizations are reluctant to purchase equipment because of the risk that it will be quickly outdated (Charles, 2000, p. 66; Muirhead et al., 2000, p. 96). In the United States, where healthcare is most often DSULYDWL]HGIRUSUR¿WFRQFHUQPDQ\FRPSDQLHV DUHORDWKWRWDNHWKH¿UVWVWHSVKDUHLQIRUPDWLRQ ZLWK FRPSHWLWRUV RU SXW WKHPVHOYHV DW ¿QDQFLDO risk regarding investment in an unproven technological direction. Secondly, under the present individual state licensure system, the potential of telemedicine is limited to the somewhat arbitrary borders of a state. Physicians are required to have medical licenses in each state in which they practice. This clearly OLPLWVWKHSRWHQWLDOJHRJUDSKLFEHQH¿WVWHOHPHGLcine could provide (Charles, 2000).
7KLUGO\ WKHUH LV VLJQL¿FDQW XQFHUWDLQW\ UHgarding whether malpractice insurance policies cover services provided by telemedicine (Western *RYHUQRUV¶$VVRFLDWLRQ 7KHOHJDOSUREOHPV associated with telemedicine malpractice liability are especially intricate when services cross state borders (Wasley, 1992, p. 82). In a highly litigious society like the United States, physicians are reluctant to increase their exposure to potential malpractice lawsuits. Finally, like other communications technologies, there is a concern regarding the security of personal medical information stored in telemedicine systems (Smith, 2005). Sanders (1994) notes the possible use of encrypting algorithms and OHJDO SUHFHGHQW \HW WR EH GH¿QHG GHWHUPLQLQJ “reasonable and customary” efforts in protecting LQGLYLGXDOV¶LQIRUPDWLRQ S 7KXVLWLVFOHDUWKDWVLJQL¿FDQWQRQWHFKQRORJLcal barriers exist to the widespread adoption of telemedicine in the United States, including regulatory, cultural, and cost issues. There has been little to no governmental effort in the United States to assist in promoting the adoption of telemedicine. 7KHHI¿FLHQFLHVDQGEHQH¿WVWHOHPHGLFLQHFRXOG bring about have not been realized.
CONCLUSION ,QVXPDOWKRXJKLWLVGLI¿FXOWWRDFFRXQWFRPSOHWHO\ for the differences in the deployment and adoption rates of broadband access throughout the world, a few provisional conclusions can be made. First, due in large part to the high costs for the deployment of advanced networks, thus far the most extensive examples of implementation have occurred due to strong government involvement. The governments that have taken an aggressively pro-ICT approach are motivated to implement high-speed networks largely for clear economic reasons: to encourage foreign direct investment and to realize multiple EHQH¿WV IURP WUDQVIRUPLQJ WKHLU HFRQRPLHV WR technological, information-based ones. In the U.S. by contrast, where the role of FDI in the domestic economy is relatively less important, coordinated efforts from the government are lacking.
273
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
,QWHUPVRIWKHVRFLDOEHQH¿WV,&7FDQEULQJ VLJQL¿FDQW XVH RI KLJKVSHHG QHWZRUNV FDQ EH expected to occur when services and applications are available that improve the quality and ease of life. In a number of the cases above, pro-ICT governments have also focused on and actively promoted citizen use of broadband technology LQ RUGHU IRU VRFLHWDO DQG LQGLYLGXDO EHQH¿WV WR accrue, emphasizing the role that ICT can play in increasing the quality of life through such areas as tele-education, teleworking, and telemedicine. Our cases have demonstrated that where the government does not take an active role, broadband XSWDNHLVVORZHUDQG,&7V¶PDQ\SRWHQWLDOEHQH¿WV often are not realized. Finally, the importance of considering social and cultural factors is paramount to understanding the different ways that broadband will develop and grow throughout the world. Although applications, content, and devices must be appropriate to cultural context, a transformation of social structures and organizations is crucial to support greater ubiquity of broadband in daily life activities.
REFERENCES Baldi, S., & Thaung, H.P.P. (2002). The entertainLQJZD\WRPFRPPHUFH -DSDQ¶VDSSURDFKWRWKH mobile Internet—a model for Europe? Electronic Markets, 12(1), 6-13. Bangsburg, P.T. (2000). Singapore to meld Internet logistics. Journal of Commerce, (March 20), 10. Barro, R.J. (1997). Economic growth in a cross-section of countries. Quarterly Journal of Economics, 106, 407-443. Barro, J. (1997). Determinants of economic growth. Cambridge, MA: MIT Press.
Business Times. (2006). Measures to further ERRVW0DOD\VLD¶VVHUYLFHVVHFWRUBusiness Times, (August 31). Business Wire. (2006). Singapore iN2015 Masterplan offers digital future for everyone. Business Wire, (June 26). Cable and Satellite Asia. (1997). No couch potatoes? Cable and Satellite Asia, (July), 18. CAN Newswire. (2006, September 12). 2006 Asia telecom, broadband and mobile reports, Malaysia, Philippines. Retrieved January 28, 2007, from http://www.marketresearch.com Chan, J.W.K., Yung, K.L., & Burns, N.D. (2000). (QYLURQPHQWVWUDWHJ\¿W $VWXG\RI+RQJ.RQJ manufacturing logistics. Logistics Information Management, 13(5), 286-300. Charles, B. (2000). Telemedicine can lower costs and improve access. Healthcare Financial Management, 54(4), 66-69. Choi, Y.B., Krause, J., Seo, H., Capitan, K., & Chung, K. (2006). Telemedicine in the USA: Standardization through information management and technical application. IEEE Communications Magazine, 44(4). Commerce, Industry, and Technology Bureau. (2004, March). Digital 21 strategy: Sustainability and opportunities. Government of the Hong Kong Special Administrative Region, Hong Kong. Davies, J.R. (2006, July). Korea broadband market December 2005. eCommerce Innovation Centre, Cardiff University, Wales. Easterly, W., King, R., Levine, R., & Rebelo, S. (1994). Policy, technology adoption and growth. NBER Working Paper No. 4681, National Bureau of Economic Research, Cambridge, UK.
Burton, J. (2005). High speed links to the Internet encounter consumer slowdown. Financial Times London, (March 21).
Economist. (2000). Government and the Internet: A survey. The Economist, (June 24), 17.
Business Times. (2006). New GeBIZ features lets ¿UPVSRVWHOHFWURQLFFDWDORJXHBusiness Times, (March 25).
**Faiola, A. (2006, May 27). When escape seems just a mouse-click away: Stress-driven addiction to online games spikes in S. Korea. Washington Post.
274
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
)DQ; 'LFNLH3 7KHFRQWULEXWLRQRI foreign direct investment to growth and stability: A post crisis ASEAN-5 review. ASEAN Economic Bulletin, 17 (3), 312-323.
Lee, N.C. (2002, February). Broadband Internet 6HUYLFH .RUHD¶V ([SHULHQFH 5HWULHYHG -DQXary 28, 2007 from eng.mic.go.kr/eng/secureDN. tdf?seq=5&idx=2&board_id=E_04_03
Frieden, R. (2005). Lessons from broadband development in Canada, Japan, Korea and the United States. Telecommunications Policy, 29, 595-613.
Lee, S.S. (2000). EDB 50% grant to aid switch to e-commerce. The Straits Times, April 12, p.71.
Funk, J. (2000, September). The mobile internet market: Lessons from Japan’s i-Mode system. Paper presented at The E-Business Transformation: Sector Developments and Policy Implications, Washington, DC. Indjikian, R., Siegel, D. (2005). The Impact of Investment in IT on Economic Performance: Implications for Developing Countries, World Development, 33 (5), 681-700. ,QWHUQDWLRQDO 0RQHWDU\ )XQG 3HRSOH¶V Republic of China- Hong Kong Special Administrative Region: Selected Issues. IMF Country Report, 02/99. International Telecommunications Union (2003) Broadband Korea: Internet Case Study. Retrieved February 1, 2007 at http://www.itu.int/ITU-D/ict/ cs/korea/material/CS_KOR.pdf Kim, F.F. (1997). Government IT Infrastructure. International Market Insights, National Trade Data Base, U.S. Department of Commerce, December 16. Koh, J. (2000). E-biz yet to touch SMEs. Business times, July 17, p.14. Koski, H.A. & Majmudar S.K. (2000). Paragons of virtue? Competitor entry and the strategies of incumbents in the US Local Telecommunications industry. Berkeley Center for Research on Telecommunications Policy Working Paper 55. Accessed February 11, 2007 from http://groups. haas.berkeley.edu/imio/crtp/publications/workingpapers/wp55.PDF Lee, C. & Chan-Olmstead, S.M. (2004). Competitive advantage of broadband internet: A comparative study between South Korea and the United States. Telecommunications Policy, 28, 649-677.
Legard, D. (1998). Forex losses impact Telekom Malaysia. Computerworld, February 28. Lehmann, A. (2002). Foreign direct investment in emerging markets: Income, repatriations and ¿QDQFLDOYXOQHUDELOLWLHVInternational Monetary Fund Working Paper Series, WP/02/47, 1-31. Lim, R. (2006). Masterplan to transform InfoComm. The Business Times, Singapore, July 3. Lim, V. & Teo, T (2000). To work or not work at home: An empirical investigation of factors affecting attitudes toward teleworking. Journal of Managerial Psychology, 15 (6) 560-586. MAIT. (2005). Hong Kong’s ICT Industry. MAIT Country Intelligence June, Volume 48. Matthews, J. (1999). Silicon Island of the East: Creating a semiconductor industry in Singapore. California Management Review, 41 (2), 55-78. Matthews, J. (2006). Malaysia pushes MSC into 2nd phase. Nikkei Electronics Asia, August. McGee, M. (2004). E-health on the horizon. Information Week, May 17, p.53. Ministry of Commerce, Industry and Energy, ReSXEOLFRI.RUHD .H\¿JXUHV )',E\PDMRU country. (2005). Retrieved January 28, 2007 from http://english.mocie.go.kr/language/eng/news/ news_view2.jsp Ministry of Information and Communication. (2002, December 16). E-Korea vision 2006. Korea. Ministry of Information and Communication. (2004). IT 8-3-9 strategy: The road to $20,000 GDP per capita. Korea. MSC. (2006, April 7). MSC revitalised for global stage. Retrieved January 28, 2007, from http://sso.
275
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
msc.com.my/index.php?option=com_content&tas k=view&id=114&Itemid=57 Muirhead, G. et al. (2000). An update on telemedicine. Patient Care, 34(6), 96-110. Pharmalicensing Limited. (2006). 2006 healthcare business market research handbook. Retrieved February 11, 2007, from http://pharmalicensing. com/intelligence/reportsearching.php?action=to c&productID=1211311 Point Topic. (2006, July 17). 7DLZDQUDQNHG¿IWK broadband country in Point Topic’s latest report. 5HWULHYHG )HEUXDU\ DW KWWS ZZZ¿QG org.tw/eng/news.asp?msgid=247&subjectid=2& pos=0 Raj, C. (2000). Singapore to pump up SMEs doing e-commerce. Business Times, (April 11), 7. Roller, L., & Waverman, L. (2001). Telecommunications infrastructure and economic development: A simultaneous approach. The American Economic Review, 91(4), 909-923. Sanders, J.H. (1994). Telemedicine, 1994: Challenges to implementation. Written Testimony to the Telemedicine Hearing before the Subcommittee on Investigations and Oversight, Committee on Science, Space and Technology, U.S. House of Representatives, 103rd Congress, U.S. Government 3ULQWLQJ2I¿FH 6DQWLDJR7 6LQJDSRUH¶VHFRPPHUFHQHW strategy bears fruit. Electronic Engineering Times, (June 10). Sein, M.K., & Harindranath. (2004). Conceptualizing the ICT artefact: Toward understanding the role of ICT in national development. The Information Society, 20(1), 15-24.
Simic, T., & Lee, J. (1997, November 24). Malaysia computer market brief. Washington, DC: National Trade Data Base, International Trade Administration, U.S. Department of Commerce. 6PLWK' 7KHLQÀXHQFHRI¿QDQFLDOIDFtors on the deployment of telemedicine. Journal of Health Care Finance, 32(1), 16-27. Statura, M. (2006). Telemedicine/tele-health: A national development tool and economic engine. International Journal of Economic Development, 18(3), 679-681. Suda, K. (2001, November). Broadband access in Japan: Implications for the future universal network. Economic and Social Research Institute. 7DQ&6 6WHDG\ÀRZVRI)',VLQWR$6($1 The Star Online, (September 1). Retrieved January 28, 2007, from http://sso.msc.com.my/index. php?option=com_content&task=view&id=184& Itemid=57 Tangalos, E.G. (1994). Telemedicine: An information highway to save lives. Written Testimony to the Telemedicine Hearing before the Subcommittee on Investigations and Oversight, Committee on Science, Space and Technology, U.S. House of Representatives, 103rd Congress. U.S. Government 3ULQWLQJ2I¿FH UNCTAD. (2007). Country fact sheets: Malaysia. Retrieved December 5, 2007, from http://www. unctad.org/Templates/Page.asp?intItemID=2441 &lang=1 U.S. Department of Commerce. (2001). NTT /DXQFKHVZRUOG¶V¿UVW*FHOOXODUVHUYLFHInternational Trade Administration, International Market Insights, (October 10).
Shih, T.H. (1999). Singapore at home and abroad. Business Times, (April 8), 2.
Wasley, T.P. (1992). What has government done to our health care? Washington, DC: Cato Institute.
Shoor, R. (1994). Long-distance medicine: Telecommunication promises to deliver better care. Business and Health, (June). Retrieved February 11, 2007, from KWWS ¿QGDUWLFOHVFRPSDUWLFOHV mi_m0903/is_n6_v12/ai_15496262
:HVWHUQ *RYHUQRU¶V $VVRFLDWLRQ Telemedicine action report. Retrieved February 1, 2007, from http://www.westgov.org/wga/publicat/actrept.htm
276
*RYHUQPHQWDODQG&XOWXUDO)DFWRUVLQ%URDGEDQG$GRSWLRQ
;X< -XQH 7KHUHJXODWRU\HQYLURQPHQW for future mobile multimedia services: The case of Hong Kong SAR and China. Proceedings of the ITU New Initiatives Workshop on the Regulatory Environment for Future Mobile Multimedia Services. ZDNet Asia. (2001, July 8). Malaysia: MSC’s performance defended. Retrieved January 28, 2007, from http://www.zdnetasia.com/news/hardware/0,39042972,30102699,00.htm
KEY TERMS Culture: Patterns of human activity and the symbolic structures that give such activity sigQL¿FDQFH'LIIHUHQWGH¿QLWLRQVRI³FXOWXUH´UHÀHFW different theoretical bases for understanding, or criteria for evaluating, human activity. Consists of values, norms, institutions, and artifacts passed on from generation to generation by learning. Endogenous Growth Theory: An macroeconomic theory in which both technology and policy measures have an effect on long-term growth
rates. It arose as a critique of neoclassical growth models which did not explain the origin of growth, assuming it to be exogenous. Foreign Direct Investment (FDI): A long-term investment by a foreign investor in an enterprise resident in an economy other than that in which the foreign direct investor is based. The FDI relationship consists of a parent enterprise and a IRUHLJQDI¿OLDWHZKLFKWRJHWKHUIRUPDPXOWLQDWLRQDO corporation (MNC). In order to qualify as an FDI, the investment must afford the parent enterprise FRQWURORYHULWVIRUHLJQDI¿OLDWH Tele-Education: The use of information and communication technologies to provide education from a distance. Telemedicine/Telehealth: The delivery of health- or medical-related services and information via telecommunications technologies. Telework: A work arrangement in which emSOR\HHVHQMR\OLPLWHGÀH[LELOLW\LQZRUNLQJORFDWLRQ and hours. In other words, the daily commute to a central place of work is replaced by telecommunication links. It refers to substituting telecommunications for any form of work-related travel, thereby eliminating distance restrictions.
277
278
Chapter XVIII
Regulation and the Deployment of Broadband James E. Prieger Pepperdine University, USA Sunhwa Lee University of California-Davis, USA
ABSTRACT This study examines the impact of telecommunications regulatory policy on broadband service deployPHQW8VLQJ86GDWDFRYHULQJDOOIRUPVRIDFFHVVWHFKQRORJ\ FKLHÀ\'6/DQGFDEOHPRGHP DQGDOO areas served by major carriers, we investigate the impact of state and federal regulation on broadband availability. Alternative regulation increases the probability of broadband availability, particularly for price caps. Unbundled network element (UNE) rates, the prices incumbent carriers charge to competitors for access to the local exchange network, also matter. Areas with lower UNE rates have a slightly higher probability of broadband availability. The effects of UNE rates on broadband deployment are largest where incentive regulation is in place. Our objective in examining regulatory factors is to highlight the role of incentive regulation and local telecommunications competition policy—policies used or available around the world—in stimulating broadband service deployment.
INTRODUCTION Broadband Internet access services are available in the United States via digital subscriber line (DSL), cable modem, and wireless or satellite technology for residences and additionally, dedicated leased lines for businesses. New Economy enthusiasts have touted the promise of broadband to provide VLJQL¿FDQWEHQH¿WVWRFRQVXPHUVDQGEXVLQHVVHV since the inception of the technology. Crandall DQG-DFNVRQ HVWLPDWHWKDWWKHEHQH¿WVWR consumers in the United States from broadband
Internet access are valued at hundreds of billions RI GROODUV DQQXDOO\ %XVLQHVVHV DOVR FDQ EHQH¿W from broadband technology through lowering their cost of doing business.17KHORQJHUWHUPEHQH¿WV may be even higher, if broadband follows the pattern set by the information technology sector as a whole, which is a major contributor to GDP growth (Jorgenson, 2001). Broadband access technology spread quickly in the U.S., compared to other popular consumer technology innovations such as analog cellular telephony and videocassette recorders (VCRs)
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Regulation and the Deployment of Broadband
(Faulhaber, 2002). Notwithstanding, commentators (e.g., Hausman, 2002) note that the pace of broadband diffusion has not been as swift in the U.S. as in some other countries, and point to regulatory roadblocks as potential causes. Although broadband technology is diffusing rapidly, not all areas of the U.S. are receiving broadband access at the same time. This study examines whether differences in regulatory treatment across states contributes to the unequal availability of broadband. Federal and state regulation may give some ¿UPVOHVVLQFHQWLYHWRGHSOR\EURDGEDQGWKDQWKH\ would have absent regulation. At the federal level, the unbundling and resale provisions of the Telecommunications Act of 1996 meant that in 2000, the time of the data we analyze, telephone companies could face “instant competition” from competitive local exchange carriers (CLECs) and broadband UHVHOOHUVLIWKH\GHSOR\HG'6/&DEOH¿UPVRIIHUing cable modem service have no such obligations, and therefore may have more incentive to invest in the necessary infrastructure. Similarly, smaller and rural telephone companies are exempt from unbundling and resale rules,2 and may therefore (other things equal) have more incentive to deploy DSL than do larger local exchange carriers. Due to lack of variation in the federal regulations in the cross-sectional data we use in this study, we are not able to investigate the primary effects on broadband deployment from federal regulation. Other authors attempt to uncover the impact of federal telecommunications regulation on broadband from variation over time (Hazlett & Bazelon, 2005) or by comparing cable modem deployment to DSL (Hausman, 2002). Neither of these strategies is ideal, because many other factors may change over time, and cable modem and DSL service differ in other ways beside regulation.3 In this chapter, we instead investigate the effects on deployment of state telephone regulation, the rates charged to competitors for the incumbent WHOHSKRQHFRPSDQLHV¶QHWZRUNHOHPHQWVDQGWKH interaction between state telephone regulation and federal unbundling policy. State legislatures and regulatory commissions FDQLQÀXHQFHWKHGLIIXVLRQRIEURDGEDQGWKURXJK (at least) three channels. First, a legislature could
pursue policies to encourage infrastructure development through tax incentives, loan programs, or other means. As of 2000, the vintage of the data explored in this study, few states had such policies and we do not investigate this channel.4 Second, a state regulatory commission could regulate broadband pricing directly, through ratemaking cases, inclusion in the rate base for rate-of-return regulation, or the application of price caps. Often, however, broadband pricing is effectively unregulated.5 Third, and of primary interest for this study, the VWDWHUHJXODWRUPD\LQGLUHFWO\LQÀXHQFHEURDGEDQG deployment through its regulation of basic services and the prices it allows incumbents to charge competitors for access to the local network. The two main classes of telecommunications regulation are rate-base rate of return regulation (RORR) and alternative regulation, the latter of which includes price caps, rate freezes, and hybrid regulatory schemes combining price caps and rate freezes. These regulatory schemes are described in more detail in the section on regulation later in the chapter. The incentives facing local exchange carriers to deploy DSL may differ under RORR and alternative regulation. RORR may not be conducive to investment in general, since an expost “prudency review” by the regulator can strike failed investments from the rate base (Kolbe & Tye, 1990). If revenue from DSL is included in the rate of return calculation, then any underinvestment problems associated with RORR would also apply to broadband infrastructure.6 On the other hand, Lehman and Weisman (2000) show WKDWLQVWDWHVZLWKDOWHUQDWLYHUHJXODWLRQ VSHFL¿cally price caps), regulators have insisted on lower prices for unbundled network elements (UNEs).7 :HFRQ¿UPWKLVIDFWZLWKRXURZQGDWD%HFDXVH a local exchange company is required to offer UNEs to competitors that allow the offering of EURDGEDQGVHUYLFHWKHSUR¿WIURPGHSOR\LQJ'6/ may be smaller in states with alternative regulation because the UNE rates are (expected to be) lower. )LQDOO\LIVWDWHUHJXODWLRQLQÀXHQFHVWKHWHOHSKRQH FRPSDQLHV¶LQFHQWLYHVWRRIIHUEURDGEDQGVHUYLFH it will have another, indirect, effect through the cable companies. If regulation affects the decision WRRIIHU'6/LWZLOODOVRDIIHFWWKHSUR¿WDELOLW\RI
279
Regulation and the Deployment of Broadband
offering cable modem service, which is a close substitute competing for the same customers. Economic models of entry, old and new (Chamberlain, 1933; Bresnahan & Reiss, 1987), as well as common sense, suggest that markets may only be able to hold a certain number of competitors, and thus the entry decisions of potential broadband providers in a market are intertwined. Burnstein and Aron (2003) note that unbundling obligations laid on local exchange carriers, if they discourage DSL, may encourage cable modem deployment in DUHDVWKDWZRXOGRWKHUZLVHEHXQSUR¿WDEOH7KH total effect of the regulation on broadband deployment is ultimately an empirical matter. Since we have data on both regulation and UNE rates, we attempt to disentangle the regulatory policy effect from the UNE price effect. We also examine the interaction between the two. 7KH¿QGLQJVLQWKLVVWXG\VKRZWKDWDOWHUQDWLYH regulation at the state level generally increases the probability of broadband availability, particularly for price caps. UNE rates do appear to matter, but not as expected by opponents of unbundling. After excluding the highest and lowest outlying UNE UDWHV ZH ¿QG WKDW DUHDV ZLWK ORZHU 81( UDWHV have a slightly higher probability that broadband is available. Interacting UNE rates with the form RIUHJXODWLRQZH¿QGWKDWWKHHIIHFWVRI81(UDWHV on broadband are largest where there is alternative UHJXODWLRQ:HWKXVGRQRW¿QGHYLGHQFHVXSSRUWLQJ /HKPDQDQG:HLVPDQ¶V DUJXPHQWWKDWORZ 81(UDWHVUHWDUGWKHLQFHQWLYHVRILQFXPEHQW¿UPV to invest in infrastructure more under alternative regulation than under RORR, since under price caps WKH¿UPFDQQRWPDNHXSORVWSUR¿WE\JHQHUDOO\ raising prices, as it can under RORR. There is relatively little formal econometric investigation of the link between broadband investment and regulation. This chapter uses a nearly comprehensive broadband survey by the FCC that covers services offered by nearly all cable and telecommunications companies. In the next section, we review the extant literature on the determinants of broadband deployment. In subsequent sections, we provide an overview of broadband technology and federal and state broadband regulation. We then outline our
280
empirical strategy, discuss our data, and present the results.
LITERATURE REVIEW There are many studies on the diffusion of broadband Internet access services. Several (Flamm, ;LDR 2UD]HP *UXEHVLF 0XUUD\ 2004; Prieger 2003) make use of the same data source we do, the FCC broadband survey described in our data section later in the chapter. Studies using other data include Gillett and Lehr (1999), Gabel and Kwan (2001), Grubesic (2003), and the chapter by Hu and Prieger in this volume.8 Most of these studies focus on non-regulatory factors such as demand-side determinants and cost factors. Studies focusing directly on regulatory impacts on broadband supply include Howell (2002), several of the chapters contained in Crandall and Alleman (2002), and Wallsten (2006). The former examines cross-country evidence from the OECD nations and concludes that local line unbundling is less successful at encouraging broadband rollout than intermodal competition between cable and telephone companies. The studies in Crandall and Alleman (2002) and Wallsten (2006) look at the effect of local network unbundling on broadband access, and we discuss them (along with related literature) in the section on regulation below. In another study, Wallsten (2005) examines some UHJXODWRU\ IDFWRUV LQÀXHQFLQJ WKH VWDWHOHYHO penetration rate (the ratio of subscribers to total households in a state). Note that the penetration rate commingles supply and demand considerations (the authors do not attempt to disentangle the two). :DOOVWHQ ¿QGVWKDWVWDWHOHYHOSROLFLHVDLPHG at promoting broadband, such as universal service and tax subsidies, are largely ineffective. Flamm (2005) also found demand-side subsidies (for federal programs intended to help connect schools and libraries to the Internet) to be ineffective at increasing broadband supply measurably. Most JHUPDQHWRRXUVWXG\LV:DOOVWHQ¶V ¿QGLQJ that states with more unbundled platform elements sold have lower penetration rates. However, given WKDWKHDOVR¿QGVWKDWWKHPRUHOLQHVWKDWDUHUHVROG
Regulation and the Deployment of Broadband
in the state, the higher the broadband penetration rate, it is not clear how to interpret his results. A related literature exists on the effect of telecommunications regulation on investment and innovation more generally. Several authors have found alternative regulation to speed the introduction of new technology and services (Taylor, Zarkadas, & Zona, 1992; Greenstein, McMaster, & Spiller, 1995; Prieger, 2001). Similarly, Prieger ¿QGVWKDWGXULQJDSHULRGRIOLJKWHUUHJXlation in the 1990s, the Bell operating companies introduced new information services such as fax, voicemail, and audiotext services more rapidly and DWDJUHDWHUUDWH$LDQG6DSSLQJWRQ ¿QG that costs are lower under alternative regulation, compared with RORR.
BROADBAND TECHNOLOGY There are several steps involved in accessing the Internet through a connection to a broadband network. Figure 1 depicts a (stylized) collection of terrestrial
(i.e., non-satellite) broadband networks.9 In a DSL network, data moves from the Internet backbone (the left side of Figure 1) through various high-speed networks (the “middle mile”) until it reaches the ORFDOH[FKDQJHFDUULHU¶VFHQWUDORI¿FH7KHFHQWUDO RI¿FHLVWKHORFDOWHOHSKRQHFRPSDQ\¶V³SRLQWRI presence” in a neighborhood. The so-called “last PLOH´RIWKHQHWZRUN ZKLFKLVWKH¿UVWPLOHIURP WKHHQGXVHU¶VSRLQWRIYLHZ LVWKHIRFXVRIWKLV chapter. In a DSL network, the last mile consists of a DSL connection residing physically on the existing WHOHSKRQHOLQHEHWZHHQWKHFHQWUDORI¿FHDQGWKH XVHU¶VFRPSXWHU WKHWRSULJKWRI)LJXUH /RFDO exchange carriers also lease high-capacity access lines (known as special access linemiddle right of Figure 1) to end users, but their high prices (at least $450 per month) take them out of the massmarket category. Figure 1 also shows a cable data network. After receiving data from the Internet DWWKHFDEOHFRPSDQ\¶VKHDGHQG DFDEOHVHUYLFH SURYLGHU¶VYHUVLRQRIWKHORFDOH[FKDQJHFDUULHU¶V FHQWUDORI¿FH WKHFDEOHFRPSDQ\VHQGVLWRQWR regional high-capacity data networks (the middle
Figure 1. A stylized depiction of terrestrial broadband Internet access networks
281
Regulation and the Deployment of Broadband
milebottom of Figure 1). In the last mile of the cable modem network, data travels through local ¿EHURSWLFQHWZRUNVDQG¿QDOO\RYHUFRD[LDOFDEOH to the end user (bottom right of Figure 1). Over WLPH FDEOH FRPSDQLHV DUH SXVKLQJ ¿EHU FORVHU to the end user as they upgrade their networks. Wireless and satellite carriers also offer broadband FDSDELOLW\DOWKRXJKVXFK¿UPVW\SLFDOO\IRFXVRQ the business market and had small market share in 2000. Nearly all residential subscribers choose cable modem service and DSL for their broadband option.10 In 2000, cable modems enjoyed a threeto-one advantage in the U.S. Since that time, DSL has nearly caught up, due to the currently lower average prices of DSL compared to cable modem subscription.
REGULATION AFFECTING THE PROVISION OF BROADBAND 7KHPRVWJHUPDQHIHGHUDOUHJXODWLRQVLQÀXHQFLQJ the broadband diffusion and competition in the U.S. DUHWKH)&&¶VUHVDOHDQGXQEXQGOLQJUHTXLUHPHQWV for common carriers. As of 2000, CLECs wishing WRRIIHU'6/FRXOGPDNHXVHRIWKHLQFXPEHQW¶V network through several channels. First, any service that the incumbent offers to retail customers, such as DSL, had to be offered to CLECs at wholesale prices.11 The CLECs can then mark up the service and resell it to end users. Second, the CLEC can purchase the local loop between the central of¿FHDQGWKHVXEVFULEHU¶VSUHPLVHVWKH³ODVWPLOH´ of the network, on an unbundled basis from the incumbent.12 The CLEC can then use the line to offer voice and data services. Finally, the CLEC could purchase only the high-frequency portion of the local loop as a UNE to offer DSL. This last option is known as “line sharing,” because the incumbent uses the lower-frequency part of the same line to offer voice service.13 By any of these methods, a CLEC uses the “last mile” owned by the incumbent instead of duplicating this part of the network with their own facilities. Hausman (2002), Shelanski (2002), and others have argued that granting a CLEC these “free options” to offer service causes the incumbents
282
to invest less in broadband infrastructure (or in upgrading the narrowband local loop to enable line sharing).14 Hazlett and Bazelon (2005) supSRUW+DXVPDQ¶V FRQWHQWLRQE\SRLQWLQJWR evidence that the rate of DSL subscription growth jumped up after the federal rule mandating line sharing was lifted. Bittlingmayer and Hazlett UHYLHZHYLGHQFHIURPWKHVWRFNPDUNHW¶V reaction to potential regulatory changes and conclude that deregulation of common carrier obligations (resale and unbundling) for broadband services would stimulate investment in the technology. Other authors expect that by making entry easier for competitors, unbundling should increase broadband penetration (Garcia-Murillo & Gabel, )DXOKDEHU FRQFOXGHVWKDWWKH)&&¶V loop-unbundling requirement has some effect on investment incentives to deploy broadband infrastructure, but that the effect is small compared to other factors.15 Little direct empirical evidence is available to date on whether unbundling helps or hinders deployment, and much of it is inconclusive, based only on case studies or returns results that are only PDUJLQDOO\ VWDWLVWLFDOO\ VLJQL¿FDQW +DXVPDQ 2002; Garcia-Murillo & Gabel, 2003; Wallsten, $ QRWDEOH H[FHSWLRQ LV :DOOVWHQ¶V international comparison of broadband penetration, ZKLFK¿QGVVWDWLVWLFDOO\VLJQL¿FDQWHYLGHQFHWKDW sub-loop unbundling requirements are correlated with lower penetration levels in OECD countries.16 +RZHYHU:DOOVWHQ IDLOHGWR¿QGHYLGHQFH that full-loop unbundling hindered broadband penetration. From the above description, it is clear that UNE prices for both the local loop and for line sharing may affect the incentive to deploy broadband. Burnstein and Aron (2003) provide evidence that lower UNE prices retard broadband diffusion, measured as state subscription rates and the latter in terms of the fraction of postal areas within a state with at least one broadband provider.17 While the requirement to unbundle network elements comes from the federal level, the states determine the UNE rates. After the Telecommunications Act of 1996, the FCC instructed state regulatory commissions to establish rates in accord with the cost
Regulation and the Deployment of Broadband
of operating a hypothetical network built with the PRVWHI¿FLHQWWHFKQRORJ\DYDLODEOH,QSUDFWLFHWKH hypothetical cost that is supposed to determine rates depends on many contested parameters and state regulatory agencies have wide leeway in setting rates. UNE rates vary greatly among states, even after considering local cost conditions. We use this variation to explore the impact of UNE rates on broadband deployment. UNE prices for line sharing are not available for our study; the UNE prices described in the data section below are for the local loop and switching (i.e., platform) costs. We expect that states that price the platform low also price line sharing low, and so the UNE rates used in this study are a proxy for both. We introduced the main classes of telecommunications regulation in the introduction: rate of return regulation and alternative regulation. RORR was the dominant form of utility regulation in the United States for much of the 20th century. Under RORR, the regulator approves prices for WKH UHJXODWHG ¿UP WKDW DUH MXVW KLJK HQRXJK WR cover production costs and a “fair” rate of return on installed capital.18 The allowed rate of return and prices are set through regulatory rate cases, which in most states were infrequent, lengthy, and costly formal hearings (Joskow, 1972). Firms must UHWXUQUHDOL]HGSUR¿WVLQH[FHVVRIWKHDOORZHGUDWH of return to the ratepayers. Alternative regulation includes price cap plans, rate freezes, and hybrid regulatory schemes. Price FDSSODQVOLPLWWKHSULFHVD¿UPFDQFKDUJHIRUD market basket (or “index”) of services that the UHJXODWHG ¿UP RIIHUV %UHQQDQ 8QGHU D SXUHSULFHFDSSODQWKH¿UPFDQUDLVHSULFHVIRU individual goods as long as it concomitantly reduces prices of other goods to keep the overall price index from rising. Price cap plans thus give the regulated ¿UPPRUHSULFLQJÀH[LELOLW\WKDQ5255GRHVDQG SODFHQRGLUHFWOLPLWVRQSUR¿WV Rate freezes, also known as rate case moratoria, are often part of a broader social contract between WKHUHJXODWRUDQGWKH¿UP$Q\UDWHIUHH]HLQYROYHV the “freezing” of the price charged for a service of particular interest to the regulator—often the price of basic local telephone service. In exchange for the promise not to request a rate hearing for a
VSHFL¿FSHULRGIRUWKHUHJXODWRU¶VVHUYLFHRIVSHFLDO FRQFHUQWKH¿UPRIWHQUHFHLYHVFRQFHVVLRQV7KH concession may be additional freedom to price certain other classes of services (for example, advanced, business, or toll services), acceptance RIWKH¿UP¶VFRVWLQJPHWKRGRORJ\IRUIXWXUHUDWH cases, or merger approval. Hybrid regulatory schemes combine price caps, rate freezes, or other non-traditional regulatory approaches such as earnings sharing or “banded” RORR.19 The most common hybrid plan, and the one we focus on in our empirical work, is a rate freeze combined with price caps for the non-frozen services. Cable company regulation in the U.S. differs greatly from telephone regulation. Only the basic service tier of cable programming was subject to rate regulation at all in 2000. The local franchising authority (typically a city or county) determines rates, not the state or the FCC. Furthermore, the FCC treats cable modem service as an information service rather than as a cable or a telecommunications service. This regulatory distinction means that cable modem service, unlike DSL in 2000, is not subject to the common carrier requirements under the Communications Act. Thus cable companies, in contrast to local exchange carriers, do not have to offer discounted services for resale or unbundled access to their network. Although policymakers in the U.S. discussed laying so-called “open access” requirements on cable companies a few years ago, they were never mandated. A cable company deploying cable modem infrastructure does not face the prospect of having to offer use of the network to rivals. Some authors (e.g., Hausman, 2002) argue that is why cable modem service took such a large early lead over DSL in the U.S.
METHODOLOGY We model the decision by a broadband carrier to enter a geographic market as depending on the expected demand in the area, costs, entry by other ¿UPVUHJXODWLRQDQG81(UDWHV20 We base our discussion on the results from probit estimations, where the dependent variable is binary: yi = 1 if
283
Regulation and the Deployment of Broadband
broadband of any type is available in ZIP Code area i; yi = 0 otherwise. In the probit model, the probability of observing yi = 1 is modeled as: E(yi) = )(E'xi) where xi is a vector of regressors, ) is the standard Normal cumulative density function, and E LVDYHFWRURIFRHI¿FLHQWVWREHHVWLPDWHG21 The QRQOLQHDUVSHFL¿FDWLRQRIWKHPHDQLPSOLHVWKDW WKHFRHI¿FLHQWVWKHPVHOYHVDUHOHVVLQWHUSUHWDEOH than are the marginal effects:
the results for the telephone regulation variables are interpretable as effects on DSL providers only if the cable modem deployment decision is largely XQLQÀXHQFHGE\WHOHSKRQHUHJXODWLRQ7KDWLVLI broadband deployment by cable companies actively responds to the form of regulation that their telecommunications rivals labor under, as discussed above, then the marginal effects of a regulatory variable is a net effect, subsuming all impacts on all players in the market.
DATA dE(y)/dx = I(Ecx)E where I is standard Normal probability density function. The marginal effects are the increase in probability that y = 1 due to a unit change in x. We calculate marginal effects for all estimates at the sample averages of the regressors. Although economic theory suggests variables that we included as controls for demand and costs, the estimations are of reduced form and we do not assign a structural interpretation to the estimates.22 Of the potential impacts of regulation mentioned in the introduction, the available data will not allow us to explore the differing regulatory treatment of telecommunications companies and cable companies, because we cannot distinguish cable modem access from DSL in the data. We are also not able to explore the direct impact of the UNE requirements, because we limit our sample to areas served by telephone carriers subject to the requirement. Firms qualifying for the rural H[HPSWLRQ IURP XQEXQGOLQJ DUH WKH ¿UPV PRVW likely to fall below the reporting threshold for the broadband survey (as explained in the next section) and are different in other respects to the RWKHU¿UPV:HFDQH[DPLQHKRZHYHUWKHLPSDFWV of state regulation on broadband deployment. We restrict attention to major local exchange companies (those with over 50,000 lines). These companies are all subject to UNE requirements, and so we can explore the effect of UNE rates on broadband in various regulatory regimes. Note that since we cannot separate DSL from cable modem access,
284
The data for the study come from four major sources: the FCC, the U.S. Bureau of the Census, a telecommunications wire center database, and a compendium of state telecommunications regulation. A complete list of variables and summary statistics for the data are in Table 1.
Broadband Availability The dependent variable is broadband availability within a ZIP Code area, covering all ZIP Codes in the mainland states. Since 1998, the FCC has collected data on where broadband access is available LQWKH867KHGDWDZHXVHDUHIURPWKH)&&¶V VHPLDQQXDOEURDGEDQGVXUYH\UHÀHFWLQJFRQGLtions on June 30, 2000. We examine data from the nascent years of broadband deployment in the U.S. because there was still much geographic variation in the availability of broadband.23 Through its survey, the FCC requires facilitiesbased telecommunications carriers to provide basic information about its operations in a state.24 The survey is obligatory for the carriers providing any kind of broadband technology if they serve more than 250 broadband lines in a given state.25 For SXUSRVHV RI WKH VXUYH\ WKH )&& GH¿QHV EURDGband as transmission speed in excess of 200 kbps in at least one direction. The survey is intended to cover all possible purveyors of end user broadband accessincumbent and competitive local exchange carriers, wireless carriers, cable companies, and otherswhether they are otherwise subject to FCC regulation or not.
0.047 0.105
0.076 0.960 0.073 0.967 0.324 0.055
),5(¿UPV
¿UPVZ HPSOR\HHV
% graduate degree
% have phone
% high-school diploma
PDQXIDFWXULQJ¿UPV
0.127
0.029
Verizon
UNE rate ($/mo.)
SBC
rate freeze regulation
Qwest
proxy cost ($/mo.)
price cap/rate freeze regulation
price cap regulation
population density (ln)
median structure age (years)
median income (ln)
deregulation/other alternative reg
CLEC presence
BellSouth
0.243
19.705
0.252
0.041
0.100
41.273
0.418
0.418
4.328
32.293
10.565
0.025
0.632
0.140
2.113
DYHHPSOR\PHQW¿UP OQ
0.314
0.483
0.040
Mean
% workers with no commute
VHUYLFH¿UPV
% rural non-farm households
% rural farm households
Independent Variables (cont.)
Variable
0.429
4.131
0.434
0.198
0.300
35.037
0.493
0.493
2.244
13.353
0.380
0.158
0.482
0.347
0.958
0.037
0.121
0.418
0.058
Deviation
Standard
Note: Demographic data are from the 2000 U.S. Census of Population. Business market data are from the Census Bureau’s “2000 ZIP Code Business Patterns.” The sources for the other data are described in the text or appendix. All percentages are expressed as IUDFWLRQV),5(VWDQGVIRU¿QDQFHDQGUHDOHVWDWH&/(&LV&RPSHWLWLYH/RFDO([FKDQJH&DUULHU2WKHUDOWHUQDWLYHUHJXODWLRQLQFOXGHV WKHIHZLQFHQWLYHUHJXODWRU\VFKHPHVQRWRWKHUZLVHFODVVL¿HGDVSULFHFDSUDWHIUHH]HRUK\EULGSODQV
0.065
0.070
0.046
0.058
0.040
0.393 0.505
% college degree
0.049
0.043
0.045
0.049
0.060
0.123
0.069
1.596
1.783
0.429
Deviation
Standard
% female
0.162
0.072
% age 19-24 yrs
% age 50-64 yrs
0.265
DJH\UV
0.128
0.262
% 30-60 min commute
0.236
0.085
% > 1 hr commute
% age 30-49 yrs
7.536
% age 25-29 yrs
4.587
# households (ln)
0.756
Mean
¿UPV OQ
Independent Variables
Broadband availability (1=yes)
Dependent Variable
Variable
Regulation and the Deployment of Broadband
Table 1. Summary statistics of the data
285
Regulation and the Deployment of Broadband
The publicly available broadband data from the FCC have some limitations. First, the data only contains ZIP Code areas with broadband providers reported. ZIP Codes without any reported broadband provider are not reported, and so the universe of ZIP Codes must be constructed outside the dataset. The universe of ZIP Codes in our study is the subset of the ZCTAs (ZIP Code tabulation areas) in the 2000 Census of Population that we ZHUHDEOHWRPDWFKWRWHOHSKRQHFRPSDQLHV¶FHQWUDO RI¿FHVXVLQJDZLUHFHQWHUGDWDEDVH26 A major limitation of the FCC data is that they do not indicate the type of company offering a service. The regulatory regime variable in this study applies only to local exchange companies, that is, providers of DSL service. The effects of telephone regulation on cable modem companies are indirect: if regulation affects the decisions of local exchange companies to offer DSL, that may DIIHFW WKH SUR¿WDELOLW\ RI RIIHULQJ FDEOH PRGHP service, a close substitute. Since we cannot tell from which types of companies broadband is available, we are measuring the total impact of the telephone regulation on the market, including (with no ability to separate out) any indirect effects through rivals to DSL. The reporting threshold raises an interpretational issue for the rural location variables we include in the estimations below. Carriers serving fewer than 250 broadband lines do not have to submit information (though the FCC states that some did voluntarily). It is therefore possible that some areas are incorrectly recorded as lacking broadband access. The carriers not meeting the reporting threshold are the mostly likely to be in rural areas, and thus availability may be understated there. However, since rural carriers are generally not subject to the unbundling obligations and are not in our sample, we do not attempt to explore the rural/urban divide in any event.27
coupled with interim rate freezes, and deregulation grouped with a few miscellaneous other alternative regulatory schemes. The incumbent local exchange carrier in each ZIP Code is matched with the form of regulation it operates under within that state.29 The other regulatory variable is the UNE rate. 7KHVSHFL¿F81(UDWHVIRUHDFKORFDOH[FKDQJH carrier are set by the state public utility commissions. They may differ by company and by density ]RQH :H FDQQRW PDWFK VSHFL¿F 81( UDWHV WR each ZIP Code area, and so we instead use statewide average rates.307RUHÀHFWWKHGLIIHUHQWFRVW structure at each wire center, we also include the proxy cost of local telephone service at each wire center. See the appendix for information on the proxy cost data.
Demand and Cost Variables All demographic, socioeconomic, and business market variables are from the U.S. Bureau of the Census.31 The data from the Bureau of the Census are for ZIP Code tabulation areas (ZCTAs), which convert the collection of point addresses in a ZIP Code to a geographic area. Demand and cost variables are expected to DIIHFWD¿UP¶VHQWU\GHFLVLRQEHFDXVHWKH\SUR[\ IRUWKHH[SHFWHGSRVWHQWU\SULFHDQGSUR¿WLQWKH estimations (Bresnahan & Reiss, 1987). Demand variables include income, market size and composiWLRQUXUDODJHDQGHGXFDWLRQSUR¿OHJHQGHUUDWLR commute time, and telephone penetration. Relevant cost considerations for broadband GHSOR\PHQWDUH¿[HGFRVWVVXEVFULEHUGHQVLW\DQG the quality of the telecommunications infrastructure.32 Controls for these costs include population density and a proxy for the vintage of the local telecommunications and cable networks (median age of the housing structures) in the area. For a more complete explanation of broadband costs and these variables, refer to Prieger (2003).
State and Federal Regulation Local Telecommunications Competition Regulatory variables consist of the state regulatory regime and UNE rates. We divide regulatory UHJLPHV LQWR ¿YH FDWHJRULHV IROORZLQJ 155, (2000):28 RORR, price caps, rate freezes, price caps
286
Local telecommunications competition, enabled by the Telecommunications Act of 1996, grew steadily until the telecom bust around the time of the data.
0.205 -0.069 1.517 1.18 0.689 0.948 1.300 2.931
% aged younger than 19
% aged between 19 and 24
% aged between 25 and 34
% aged between 35 and 49
% aged between 50 and 64
% high school education
% college education
% graduate education
% rural farm households -0.465
-2.138
Median income (> median income)
-0.229
0.179
Median income (< median income)
% female
-0.082
ORJ QXPEHURI¿UPV
% rural non-farm households
0.210 0.368
log (number of households)
Demand and Cost Variables
36.38
16.14
11.77
8.56
14.64
18.83
-0.85
2.55
-2.84
-5.78
-26.54
2.23
-1.02
4.56
2.60
0.396***
0.188***
0.211***
0.411*
0.378***
0.430***
0.395
0.316
0.349
0.068***
0.286***
0.124
0.093
0.025***
0.024***
1.201
1.229
0.537
0.398
0.311
1.081
0.334
0.065
-0.313
-0.399
-0.169
0.330
-0.199
0.396
0.332
0.361
2.155
Verizon 0.031***
0.000
SBC-Ameritech-PacBell
CLEC Presence
0.107 0.968
-0.380
0.168
-0.088
0.141
Coef.
Qwest
5.87
0.087***
0.073***
0.049***
0.048
s.e.
BellSouth
0.420
-9.26
-0.522
Bell Operating Company
Deregulation
5.81
0.857
Rate Freeze
2.12
0.174
0.39
0.032
Price Caps
x 100
Marginal effect x 100
8.90
9.11
3.98
2.95
2.31
8.01
2.48
0.48
-2.32
-2.95
-1.25
2.45
-1.47
2.93
2.46
3.04
9.54
0.00
3.79
0.74
-3.89
1.08
-0.66
1.03
Marginal effect
and CLEC Presence
and CLEC Presence Coef.
Regulatory Variables, BOCs,
Regulatory Variables
Price Caps & Rate Freeze
State-Level Regulation
Variable
ESTIMATION 2
ESTIMATION 1
0.446***
0.210***
0.236**
0.467
0.420
0.483**
0.433
0.359
0.392
0.074***
0.308
0.139**
0.102*
0.027***
0.027***
0.034***
0.067***
0.043
0.063***
0.050**
0.095***
0.086*
0.059
0.056**
s.e.
2.870
1.444
0.839
0.656
1.274
1.559
-0.128
0.310
-0.355
-0.503
-2.257
0.243
-0.088
0.378
0.242
-0.609
0.788
0.237
0.094
Coef. x 100
35.93
18.07
10.51
8.22
15.95
19.44
-1.60
3.88
-4.44
-6.30
-28.25
3.04
-1.10
4.73
3.03
-11.46
5.65
2.90
1.17
Marginal effect
without CLEC Presence
Regulation Variables
ESTIMATION 3
0.391***
0.186***
0.209***
0.407
0.375***
0.426***
0.390
0.313
0.346
0.067***
0.286***
0.123**
0.092
0.024***
0.024***
0.086***
0.073***
0.049***
0.047**
s.e.
Regulation and the Deployment of Broadband
Table 2. Probit estimations for the availability of broadband service within a ZIP Code area (cont. next page)
287
288 -2.36
-0.580 0.157
),5(¿UPV
VHUYLFH¿UPV
median structure age
s.e.
0.437 -5281.70
Pseudo R-square
Log-likelihood
0.945***
0.001***
0.026**
0.017**
0.056*
0.022***
0.320
0.102
0.243**
0.188**
0.341***
0.199***
0.119***
0.399***
-2.859
0.006
-0.046
-0.006
0.172
-0.099
-0.356
0.225
0.200
0.582
0.726
1.117
0.710
0.492
Coef.
-4420.92
0.529
18,978
0.04
-0.34
-0.05
1.27
-0.73
-2.64
1.67
1.48
4.31
5.38
8.27
5.26
3.64
x 100
Marginal effect
1.033***
0.002***
0.030
0.019
0.060***
0.024***
0.367
0.116*
0.261
0.213***
0.350**
0.221***
0.133***
0.437
s.e.
-5.049
0.014
-0.059
0.042
0.110
-0.120
-0.213
0.183
-0.547
0.492
2.607
1.183
0.557
1.836
Coef.
-5374.01
0.427
18,978
0.17
-0.74
0.52
1.37
-1.51
-2.67
2.30
-6.84
6.16
25.87
14.81
6.97
22.98
x 100
Marginal effect
without CLEC Presence
Regulation Variables
ESTIMATION 3
0.933***
0.001***
0.026**
0.017**
0.055**
0.022***
0.315
0.102*
0.241**
0.186***
0.339***
0.198***
0.118***
0.398***
s.e.
VLJ Q LI L F D Q W D W WK H OH YH O
VLJ QLI L F D Q W D W WK H OH YH O
VLJ Q LI L F D Q W D W WK H OH YH O Notes: Dependent variable is 1 if there is at least one broadband customer in the ZIP Code area, 0 if not. The excluded regulation indicator is Rate of Return Regulation. The excluded BOC indicator is all non-Bell LECs. Marginal effects are expressed in percentage points. The sample includes only ZIP Codes matched to large ILECs (those subject to the UNE provisions of the Telecommunications Act of 1996); these include the BOCs and other LECs such as Sprint, Valor, Cincinnati Bell, Citizens, Concord, and so forth. Marginal effect is the marginal effect on the mean evaluated at the sample mean of x; for indicator variables these are discrete changes. S.e. is WKHVWDQGDUGHUURUIRUWKHFRHI¿FLHQWQRWWKHPDUJLQDOHIIHFW&/(&LV&RPSHWLQJ/RFDO([FKDQJH&RPSDQ\
18,978
0.15
-0.80
0.52
1.19
-1.45
-7.20
5.70
22.59
13.63
Observations
-4.774
-0.064 0.012
population density (> median)
Constant
0.096 0.042
DYHHPSOR\PHQW¿UP OQ!PHGLDQ
population density (< median)
-0.190
0.459
PDQXIDFWXULQJ¿UPV
-0.117
1.820
% households with phone
¿UPVZLWK HPSOR\HHV
1.098
% over 1 hr commuting time
DYHHPSOR\PHQW¿UP OQ PHGLDQ
1.95
0.559
6.94
1.906
% 30-60 min commuting time
23.66
x 100
Marginal effect
and CLEC Presence
and CLEC Presence Coef.
Regulatory Variables, BOCs,
Regulatory Variables
% workers with no commute
Demand and Cost Variables (continued)
Variable
ESTIMATION 2
ESTIMATION 1
Regulation and the Deployment of Broadband
Table 2. (cont.)
Regulation and the Deployment of Broadband
Competition in local telephony may spur incumbent carriers to offer broadband. Furthermore, some facilities-based competitors offer DSL themselves, although Faulhaber (2002) disparages the effectiveQHVVRIWKHGDWDRI&/(&V¶EXVLQHVVSODQV The FCC makes available a list of ZIP Codes in ZKLFKWKHUHLVORFDOFRPSHWLWLRQ,QVRPHVSHFL¿FDtions, we include an indicator for the presence of at least one competing local exchange company in the area.
Telecommunications Carriers’ Operating Areas To control for institutional differences among the major incumbent local exchange carriers, we include indicator variables for territory served by the Bell Operating Companies at the time: BellSouth, Qwest (formerly U.S. West), SBC (formerly 6RXWKZHVWHUQ%HOO7HOHSKRQH3DFL¿F%HOO1HYDGD Bell, and Ameritech), and Verizon (formerly Bell $WODQWLF DQG 1<1(; 7KH H[FOXGHG LQGLFDWRU is all non-Bell carriers large enough to remain in the data set.33
RESULTS The results from the estimations are in Tables 2 and 3. The variables of interest are the regulatory variables, and we discuss their impact on broadband GHSOR\PHQW¿UVW:HQH[WWXUQWRWKHRWKHUGHPRgraphic, business market, and cost variables.
Regulatory Variables ,QWKH¿UVWVSHFL¿FDWLRQLQ7DEOHWKHYDULDEOHVRI interest are indicators for the forms of regulation: price caps, price caps coupled with rate freezes (“hybrid” plans), rate freezes, and the “not elseZKHUHFODVVL¿HG´FDWHJRU\RIGHUHJXODWLRQRURWKHU alternative regulatory schemes. We collectively refer to these four categories as “alternative regulation,” opposed to traditional RORR. The marginal effects of alternative regulation are positive (except for the deregulation/other alternative regulation category), showing that alternative regulation is
correlated with more broadband deployment, compared to the excluded category RORR. The coef¿FLHQWIRUSULFHFDSVLVQRWVWDWLVWLFDOO\VLJQL¿FDQW however.34 Given the miscellaneous nature of the deregulation/other alternative regulation category and the fact that it contains few observations, we will not discuss it further below. Before concluding that alternative regulation of the incumbent telephone companies provides a favorable environment for deploying broadband, we PXVW¿UVWQRWHWKDWLWPD\EHGLI¿FXOWWRGLVHQWDQJOH WKHLPSDFWRIUHJXODWLRQIURPFRPSDQ\VSHFL¿F factors. The data we have are not the outcome of a controlled experiment, and regulation is not randomly assigned to companies. BellSouth and SBC did not face RORR at all, having embraced price caps or hybrid plans in most of their territory. Qwest and Verizon have more variation in regulation across their operating states, facing both RORR and most forms of alternative regulation. GTE and Sprint operate under both RORR and alternative regulation, but do not face rate freezes. Thus, without controlling for the company incumbent in the area, the interpretation of the regulaWLRQFRHI¿FLHQWVLVSUREOHPDWLF)RUH[DPSOHWKH RORR effect is indistinguishable from a “neither BellSouth nor SBC” effect. 7RUHPRYHWKH¿UPVSHFL¿FIDFWRUVIURPWKH effects of regulation, we include indicator variables for the four large Bell Operating Companies (BellSouth, Qwest, SBC, and Verizon) in estimation 2 (see Table 2). The excluded indicator variable is “other LECs.” All Bell Operating Company (BOC) indicators (for BellSouth, Qwest, SBC, and Veri]RQ DUHVLJQL¿FDQWDQGSRVLWLYHFRPSDUHGWRWKH excluded category of other carriers subject to the unbundling requirements, which are all smaller than the big four. The result is expected, since the BOCs rolled out broadband faster than other local H[FKDQJHFDUULHUVGLGDQGVLQFHWKH%2&V¶WHUULtory includes the most-populated areas of the U.S., making it also attractive for cable modem entry. There is a marked change in some regulatory estimates when we include the BOC indicators. In particular, the size of the positive effect of price FDSVLQFUHDVHVDQGJDLQVVWDWLVWLFDOVLJQL¿FDQFHWKH HIIHFWRIK\EULGSODQVORVHVVLJQL¿FDQFHDQGWKH
289
Regulation and the Deployment of Broadband
positive effect of rate freezes is greatly attenuated (although still positive). The marginal effects, all with reference to the excluded category of RORR, DUHIRUSULFHFDSV EXWLQVLJQL¿FDQW IRU hybrid plans, and 1.1 for rate freezes (all in perFHQWDJHSRLQWV :KLOHWKH¿JXUHVDUHVPDOOHYHQ the one percentage point increase associated with price caps represents almost 200 additional postal areas nationwide. 7KXVLWDSSHDUVWKDWWKHSUR¿WOLPLWLQJDVSHFWV of RORR outweigh the potentially negative effects of alternative regulation in the broadband deployment decision, except perhaps for hybrid plans. That regulation consisting of a rate freeze encourages the growth of broadband infrastructure is not surprising, because prices for advanced services typically are not frozen. In most of the estimations in Tables 2 and 5, an indicator is included among the controls for the presence of at least one local telephone competitor (CLEC). The variable is potentially endogenous, however, given that there may be common unobVHUYHGIDFWRUVMRLQWO\LQÀXHQFLQJD&/(&¶VHQWU\ decision and the broadband deployment decision RIWKH¿UPVLQWKHPDUNHW3UHYLRXVDXWKRUVGHDO with the endogeneity issue differently. Prieger ¿QGVWKDW&/(&SUHVHQFHLVHQGRJHQRXV when jointly estimating the CLEC and broadband entry decision. Flamm (2005) also discussed the potential endogeneity of CLEC entry in the broadband equation, and presents estimations including and excluding the variable. Here, we follow the latter approach and redo estimation 1 excluding the CLEC variable (estimation 3 in Table 2). Comparing estimations 1 and 3 in Table 2 shows that the results with and without the CLEC variable are consistent, in the sense that there are no sign FKDQJHVRIWKHFRHI¿FLHQWVEHWZHHQWKHHVWLPDWLRQV The sizes of the marginal effects is also generally similar in the two estimations, although the marginal effect of price caps is about three times higher when the CLEC variable is excluded in estimation 3. Therefore, in the estimations that follow we leave WKH&/(&YDULDEOHLQWKHVSHFL¿FDWLRQ The causal interpretation of the results for the regulatory variables is problematic if regulators implemented alternative regulation only after
290
H[WUDFWLQJ DJUHHPHQWV IURP WKH ¿UPV WR GHSOR\ broadband. For example, many incentive regulation SODQVLQFOXGHFRPPLWPHQWVE\WKH¿UPWRGHSOR\ advanced telecommunications infrastructure. In VXFKFDVHVUHJXODWRUVDOORZWKH¿UPVWRVZLWFK out from RORR in exchange for promises to upgrade their infrastructure. Thus, while alternative regulation may be positively correlated with broadband deployment, it may not be because of incentive effects stemming purely from the price regulation. $UHODWHGFRQFHUQLVZKHWKHURXU¿QGLQJVDUH merely a demonstration effect (Sappington & Weisman, 1996). A demonstration effect is created when regulated telephone companies perform actions regulators desire, such as introducing DSL, to “demonstrate” the success of a favorable regulatory regime—in this case, alternative regulation. In such cases, the desirable outcome may be short-lived after the regulatory regime is FRQ¿UPHG+RZHYHULQWKHFRQWH[WRIEURDGEDQG deployment, if incentive regulation elicits a demonstration effect, then so much the better (from the point of view of encouraging broadband deployPHQW 2QFH'6/LVLQSODFHLQDFHQWUDORI¿FHWKH telephone company will not uninstall it or cut off its broadband customers. Thus, any demonstration effect can properly be counted among the virtues of alternative regulation. Finally, the regulatory scheme in place in a state may itself be endogenous (Ai & Sappington, 2002).35 There may be underlying factors leading to both adoption of alternative regulation by state public utility commissions and broadband deployPHQWE\¿UPV7KHLQFOXVLRQRIVWDWH¿[HGHIIHFWV (i.e., indicator variables for the states) can mitigate WKLVFRQFHUQVLQFHWKH¿[HGHIIHFWVXEVXPHVDOO state-level unobserved factors (as long as they do QRWFKDQJHRYHUWLPH 7KHDGGLWLRQRIVWDWH¿[HG effects to estimation 1 (results not reported) only strengthens the positive association between alternative regulation and broadband deployment, KRZHYHU :KHQ ¿[HG HIIHFWV DUH LQFOXGHG WKH regulatory marginal effects have the same signs but even larger magnitudes (and all are statistically VLJQL¿FDQW 36:HGRQRWLQFOXGHVWDWH¿[HGHIIHFWV in the main estimations because in many states
-0.247
Deregulation
UNE rates (deciles 2-9)
Proxy Cost
0.70
0.531 -4402.44
Pseudo R-square
Log-likelihood
0.000***
0.019**
0.008***
0.072*
0.034***
0.098**
0.137
0.059
0.057*
s.e.
0.767
5.34
1.60
-4398.51
0.531
18,976
Yes
-0.75
-0.15
-0.103
-0.02 -0.020
-0.25
0.01
0.39
-0.20
-0.003
-0.034
0.001
0.054
-0.027
0.219
3.66 2.94
2.62
-0.22
0.356
x 100
Marginal effect
2.073
0.582
-0.030
Coef.
0.077
0.031
0.017
0.017**
0.000***
0.029*
0.015*
0.081***
0.034***
1.743
0.83
0.351
0.339**
s.e.
VLJQL¿FDQWDWWKHOHYHO
VLJQL¿FDQWDWWKHOHYHO
VLJQL¿FDQWDWWKHOHYHO Notes: Dependent variable is 1 if there is at least one broadband customer in the ZIP Code, 0 if not. Controls include all the variables in Table 1. See notes to previous table.
18,976
Yes
0.01
0.27
-0.30
0.93
2.93
-2.25
0.78
-0.71
Observations
All other variables from Estimation 2
UNE rates * Deregulation
UNE rates * Rate Freeze
UNE rates * Price Caps & Rate Freeze
UNE rates * Price Caps
UNE Rates and Regulation Interaction
0.037 0.001
UNE rates (10 decile)
th
0.126 -0.041
UNE rates (1st decile)
UNE Rates and Proxy Cost
0.353
0.117
Rate Freeze
CLEC Presence
0.096 -0.095
Price Caps
x 100
Marginal effect
and Interaction Terms
and UNE Rates
Coef.
Regulatory Variables, UNE Rates,
Regulatory Variables
Price Caps & Rate Freeze
State-Level Regulation
Variable
ESTIMATION 5
ESTIMATION 4
Regulation and the Deployment of Broadband
Table 3. Probit estimations for the availability of broadband service within a ZIP Code area: The impact of UNE rates
291
Regulation and the Deployment of Broadband
Table 4 . Summary statistics of UNE rates by the type of regulatory regime Number of Obs. (Percent) Rate of Return
Price Caps
Price Caps with Rate Freeze
Rate Freeze Deregulation and Other Alternative Regulation
Standard Mean
1,999 (9.76%) 8,557 (41.78%) 8,561 (41.80%) 841 (4.11%) 522 (2.55%)
Deviation
Min
Max
21.97
4.08
12.89
33.42
18.43
2.74
13.87
33.42
19.10
3.23
12.89
33.42
31.62
3.62
14.87
33.42
22.57
1.27
18.00
26.40
Notes: UNE rates ($) are statewide weighted averages, from Gregg (2002). The unit of observation is a ZIP Code.
Table 5. OLS regression of UNE rates on the type of regulatory regime and costs State Average UNE Rates ($)
&RHI¿FLHQW
Standard Error
Price Caps
-6.610
1.863***
Price Caps with Rate Freeze
-5.932
1.650***
Rate Freeze
1.332
2.834
Deregulation/Other Alternative Regulation
-2.816
2.022
Proxy Cost
0.222
.0853**
Constant
19.295
2.380***
2
R = 0.514 Adjusted R2 = 0.425 F statistic = 5.97, p-value = 0.00 Number of Observations = 49
DQG
VLJQL¿FDQWDWDQGUHVSHFWLYHO\ Notes: The estimation method is OLS with heteroskedasticity-robust standard errors. There is one observation per state and Washington, DC, excluding Alaska and Hawaii. UNE rates are statewide weighted averages, from Gregg (2002). Proxy Cost is the state=level weighted average of proxy cost for local telecommunications service in the local exchange area, as calculated for individual wire centers by the FCC’s Hybrid Cost Proxy Model (HCPM) in January 2000. The weights for the state-level SUR[\FRVWYDULDEOHDUHFDOFXODWHGIURPWKHWRWDOQXPEHURIKRXVHKROGVDQG¿UPVZLWKLQWKH=,3&RGHDUHD
292
Regulation and the Deployment of Broadband
there is no variation in regulatory regime, which PHDQVWKDWWKHUHJXODWRU\FRHI¿FLHQWVDUHEHLQJ VWDWLVWLFDOO\LGHQWL¿HGEDVHGRQO\RQWKHVXEVHWRI states with more than one regulatory scheme in SODFH7KHUHVXOWVPD\QRWUHÀHFWWKHWUXHLPSDFWRI regulation in the full national sample, therefore.
Unbundling and UNE Rates We now turn to the impact of the unbundling of the LQFXPEHQWV¶ORFDOWHOHFRPPXQLFDWLRQVQHWZRUNV In estimations 4 and 5 in Table 3, we add the statewide average UNE rate and the local proxy cost of providing telephone service. If we could control perfectly for local loop and switching costs, DQGKDG81(UDWHVVSHFL¿FWRHDFKREVHUYDWLRQ then movements in the UNE rates would also be PRYHPHQWVLQWKHLQFXPEHQWV¶QHWUHYHQXHIURP sale of unbundled platform elements. Given that ZHKDYHSUR[\FRVWVDWWKHFHQWUDORI¿FHOHYHOEXW UNE rates only as the statewide average, our UNE rate variable only imperfectly measures the net revenue from unbundling, which may reduce the precision of the estimates. A further potential problem in the interpretation RIWKHFRHI¿FLHQWRQWKH81(UDWHYDULDEOHLQWKH estimation stems from the relationship between UNE rates and regulation in the data. An ideal experimental design37 for determining the causal incentive effect of regulatory policy on broadband diffusion would randomly assign UNE rates and UHJXODWRU\VFKHPHVWR¿UPV,QUHDOLW\/HKPDQ and Weisman (2000) show that in states with price caps, regulators have set lower UNE rates for the local loop and platform. Correlation and co-determination of UNE rates and the form of regulation do not necessarily introduce bias into the estimation. However, if there are unobserved factors that LQÀXHQFH81(UDWHVUHJXODWLRQDQGEURDGEDQG diffusion jointly, then the exogeneity assumption required of the independent variables in the probit HVWLPDWLRQIDLOVDQGWKHFRHI¿FLHQWHVWLPDWHVZLOOEH inconsistent. This is the endogeneity problem discussed in the previous subsection. In this instance we cannot adopt the solution of adding state-level ¿[HGHIIHFWVVLQFHWKH81(UDWHVDUHVWDWHDYHUDJHV and do not vary within the state.
:H UHSOLFDWH /HKPDQ DQG :HLVPDQ¶V result with our data in Tables 4 and 5. Table 4 shows the average UNE rates by form of regulation. Where the incumbent carrier is price capped, UNE rates are lowest, about $3.50 lower than RORR observations. Observations associated with rate freezes have the highest UNE rates—so high, in fact, that we suspect that cost differences are driving at least part of that result. To control for costs, we regress UNE rates on the form of regulation and average proxy costs in the state (see Table 5). The proxy costs do affect rates: every dollar that average proxy cost goes up leads to UNE rates going up 22 cents, DQGUDWHIUHH]HVWDWHVQRORQJHUVKRZDQ\VLJQL¿FDQW difference from RORR states. After controlling for cost, the difference between price capped (with or without rate freezes) and RORR states is even JUHDWHU DQGVWDWLVWLFDOO\VLJQL¿FDQW WKDQLQGLFDWHG in Table 4. The R2 statistic shows that 51% of the variation across states in UNE rates is explained by variation in the regulatory regime. Return now to our main estimation for the determinants of broadband diffusion. The UNE UDWHYDULDEOHLQHVWLPDWLRQVDQG 7DEOH LV¿W with a piecewise linear spline with knots separating the lowest and highest decile groups.38 The reported UNE parameters in Table 3 are the slopes and marginal effects for the relevant range of the UNE rates. In estimation 4, the marginal effect for the mid-range of the UNE rates is negative and VLJQL¿FDQW7KXVIRUDUHDVZLWKWKHPLGGOHRI UNE rates, higher UNE rates are associated with less broadband availability—the opposite outcome than that feared by commentators arguing that “cheap” unbundling schemes retard investment. The size of the effect is modest: if UNE rates were to rise by one dollar, the probability of broadband deployment falls by 0.3 percentage points. For the lowest and highest decile groups of UNE rates, however, the marginal effects are positive. While the effect for the lowest group is only marJLQDOO\VWDWLVWLFDOO\VLJQL¿FDQW WKHREVHUYDWLRQVDW this extreme come from only three states), the effect IRUWKHKLJKHVWJURXSLVVLJQL¿FDQWDWWKHOHYHO In the latter case, then, it may be that when UNE rates rise very high, the incumbents feel insulated from the prospect of competition. If the entry of
293
Regulation and the Deployment of Broadband
competitors via UNE and resale is less likely, then the return on infrastructure deployment rises for the incumbent. While it is not usually interesting to focus on outliers, in the present context the top decile group of UNE rates could be highly relevant for the following reason. Some policy advocates, DQGWKH¿UPVVXEMHFWWRXQEXQGOLQJUHTXLUHPHQWV argue that the methodology used to set UNE rates (known as TELRIC, for Total Element Long Run Incremental Cost) results in uneconomically low UNE rates. A federal court agreed in 2004 and UHMHFWHG WKH )&&¶V 81( SODWIRUP UXOHV DQG WKH TELRIC methodology. With the cheapest UNE rates (those for the platform) now unavailable, the impact of UNE rates at the high end in our data becomes more relevant. Comparing estimation 4 to estimation 2 (the most comparable, since it also includes the BOC indicator variables), we see that the marginal effects of the regulatory variables are reduced in PDJQLWXGH H[FHSW IRU WKH LQVLJQL¿FDQW K\EULG category). Given the association between type of regulation and UNE rates evidenced in Tables 4 and 5, it is not surprising that the inclusion of UNE rates directly in the estimation changes the strength of the regulatory effects. For example, as UNE rates rise, it is less likely that the observation is in a price cap area, and so the larger positive marginal effect for price caps in estimation 2 PLJKWUHÀHFWLQSDUWWKHLPSDFWRIORZ81(UDWHV in accord with the negative effect found for UNE rates in estimation 4. In estimation 5 in Table 3, we interact the UNE rate with the forms of regulation. The main UNE FRHI¿FLHQWVKDYHWKHVDPHVLJQVDVLQHVWLPDWLRQ 4. More interesting is that all four interaction coHI¿FLHQWVDUHQHJDWLYH$OWKRXJKRQO\WKHHIIHFW IRUSULFHFDSVLVLQGLYLGXDOO\VLJQL¿FDQWWKHIRXU FRHI¿FLHQWVDUHMRLQWO\VLJQL¿FDQWDWWKHOHYHO This means that the negative marginal effect of UNE rates shown for the bulk of the observations (from estimation 4) is even larger in magnitude in areas with alternative regulation. The main UNE FRHI¿FLHQWV JLYH WKH LPSDFW RQ 5255 VWDWHV since RORR is the excluded regulatory variable. 7KLV¿QGLQJLVWKHRSSRVLWHRIWKDWVXJJHVWHGE\ /HKPDQ DQG :HLVPDQ¶V DUJXPHQW WKDW
294
XQGHUDOWHUQDWLYHUHJXODWLRQWKH¿UPKDVQRZD\ WRPDNHXSORVWSUR¿WE\UDLVLQJUHWDLOSULFHVDVLW can under RORR. Their argument suggests that DQ\SUR¿WGDPSHQLQJHIIHFWVRIWKH81(UHJLPH for incumbents would be felt most strongly under alternative regulation. Note, however, that since we are not looking at investment by the incumbent local exchange carriers only, we do not have a clean test of that hypothesis.
Other Demographic, Business Market, and Cost Variables 7KHRWKHUYDULDEOHVLQWKHVSHFL¿FDWLRQDUHLQFOXGHG to control for underlying factors that may affect WKH¿UPV¶GHFLVLRQVWRGHSOR\EURDGEDQG,IWKHVH factors are correlated with the regulatory variables of interest, omitting them would bias the regulatory estimates. Although the other variables are QRWRISULPDU\LQWHUHVWZHEULHÀ\GLVFXVVWKHP here. Unless otherwise noted, the impacts of the YDULDEOHVDUHVLPLODUDFURVVVSHFL¿FDWLRQVDQGZH discuss the results from estimation 1. 7KHPDUNHWVL]HFRHI¿FLHQWVIRUKRXVHKROGVDQG ¿UPVDUHSRVLWLYHDQGVLJQL¿FDQW7KHPDUJLQDOHIfect of 0.026 for log number of households means that if the number of households nearly tripled, there would be a 2.6 percentage point increase in the probability of broadband access.39 The marginal effect of the size of the business market (the log QXPEHURI¿UPV LVDOPRVWWZLFHDVVWURQJ 7KHUHLVQRVWDWLVWLFDOO\VLJQL¿FDQWHYLGHQFHIRU unequal access to broadband based on the income of the area in estimation 1, although there is in estimations 2 and 3. Exploratory analysis revealed that income has a nonmonotonic relationship to broadband access.40 The relationship between median income and broadband deployment is 8VKDSHG LQ DOO VSHFL¿FDWLRQV GHSOR\PHQW GHcreases with income for areas below the national median and increases with income for wealthier DUHDV ,Q WHUPV RI VLJQL¿FDQFH WKH HYLGHQFH LV stronger for the latter than the former. The latter LV SUREDEO\ WKH UHVSRQVH RI ¿UPV WR WKH JUHDWHU demand for broadband in wealthier areas, since higher household income increases demand for broadband (Rappoport, Kridel, Taylor, Alleman,
Regulation and the Deployment of Broadband
'XII\'HQR 7KH¿QGLQJWKDWEURDGEDQG deployment increases as incomes fall (when below median income) may result from unobserved cost differences among areas. Controlling for rural location, as we do, lower-income areas may be in denser urban areas that have lower average costs to serve, or may have stronger business demand. %RWKUXUDOJHRJUDSKLFFRHI¿FLHQWVDUHQHJDWLYH 7KH)&&DQGRWKHUVWXGLHV¿QGWKDWUXUDODUHDVODJ urban areas in broadband availability (FCC, 2000b; Prieger 2003). However, recall most rural areas of the country are not included in our data. 7KHDJHFRHI¿FLHQWVLQWKHWR\HDUROG UDQJHDUHVLJQL¿FDQWDQGHDFKRIWKHVHDJHJURXSV¶ effects is positive compared to the excluded senior group (aged 65 and older). The education effects are all positive, compared to the excluded group lacking a high school diploma. In addition, the higher the educational level, the stronger the efIHFWLV7KHFRPPXWLQJGLVWDQFHFRHI¿FLHQWVDUH all positive compared to the excluded category of a 1-30-minute commute. A larger fraction of workers who do not commute (and who thus have higher demand for broadband for purposes of telecommuting) increases access likelihood. For workers with commutes, longer commuting times are associated with higher broadband deployment. The fraction of households with a telephone also affects broadband deployment. Since DSL (but not other modes of broadband) requires a phone line to function, it is not surprising that the marginal effect of the fraction of households with phone access in the area is positive. Regarding business demand, a higher fracWLRQRI¿UPVLQWKHPDQXIDFWXULQJDQG LQVRPH VSHFL¿FDWLRQV VHUYLFH LQGXVWULHV OHDGV WR PRUH broadband access. The opposite is true for the fracWLRQRI¿UPVLQWKH¿QDQFHDQGUHDOHVWDWH ),5( category, which is surprising given the relatively KHDY\XVHRIWHOHFRPPXQLFDWLRQVE\WKH¿QDQFH LQGXVWU\:KHQWKHUHDUHPRUHVPDOO¿UPVLQWKH DUHDDVPHDVXUHGE\WKHIUDFWLRQRI¿UPVZLWKOHVV than 50 employees, the probability of broadband GHSOR\PHQWLVORZHUEXWWKHHIIHFWLVQRWVLJQL¿FDQW statistically or in magnitude. There is a U-shaped association between the average employment per ¿UPDQGEURDGEDQGDFFHVV:KLOHWKHQHJDWLYHFR-
HI¿FLHQWRQDYHUDJHHPSOR\PHQWEHORZWKHPHGLDQ is unexpected, the positive impact on broadband DFFHVVRIDYHUDJHHPSOR\PHQWIRUODUJHU¿UPVLV QDWXUDO0DQ\RIWKHODUJHVW¿UPVXVHKLJKVSHHG dedicated access lines rented from the local phone company, which are included in the data. 7KHFRVWYDULDEOHVKDYHVLJQL¿FDQWHIIHFWVRQ the probability of broadband access. Population density has the expected positive sign in the upper median group, although it has a negative (but small in magnitude) effect in the lower median group. The age of housing structures in the area, as a proxy for the vintage of the network infrastructure of the telephone and cable systems, has an unexpected positive effect on broadband deployment. It is hard to imagine a causal reason why older infrastructure would lead to more broadband deployment, and the variable is probably acting as a proxy for other factors not otherwise adequately controlled, such as urban location or long-established business areas.
IMPLICATIONS AND OUTLOOK This study explores the interplay between reguODWLRQ DQG EURDGEDQG ,QWHUQHW DFFHVV :H ¿QG that alternative telephone regulation in the U.S. increases the probability of broadband availability in the U.S., particularly for price caps. Higher UNE rates are correlated with lower probability of broadband availability in general, although not throughout the entire range of the data. The effect is strongest in areas with price caps. 2XUGDWDDUHVSHFL¿FWRWKH8QLWHG6WDWHV+RZever, the implications of the results are relevant to the world telecommunications market, since many of the regulatory schemes we discuss are in use around the globe. By 2002, 23 OECD countries had mandatory local loop unbundling regimes in place, and prices for UNEs varied greatly across countries (OECD, 2003). By 2005, regulators in at least 19 OECD countries used price caps to UHJXODWHWHOHFRPPXQLFDWLRQV¿UPV 2(&' 7KHEHQH¿FLDOLPSDFWRISULFHFDSVRQEURDGEDQG deployment in the U.S. that we document here implies that the international push toward price
295
Regulation and the Deployment of Broadband
cap regulation is a movement in the right direcWLRQ2XU¿QGLQJWKDWORZ81(UDWHVGRQRWDSSHDU to hinder broadband deployment may be more germane to other countries than it is for the U.S. today, given that low UNE rates have disappeared in the U.S.41 Broadband has diffused in the U.S. to the point where today there is at least one customer for high-speed service in 99% of all the ZIP Code areas in the United States (FCC, 2006).42 Despite the ubiquity of broadband access today in the U.S., lessons learned from the early years of deployment are still valuable. There will always be technoORJLFDOFKDQJHLQWKH¿HOGRIWHOHFRPPXQLFDWLRQ and information services. Incentive regulatory schemes that encourage infrastructure deployment and network modernization today are likely to encourage it tomorrow as well (as opposed to the older-style command and control regulation, which adapts slowly to new technological opportunities). Furthermore, since competition among providers holds the greatest promise for reducing prices in the U.S. and abroad (Gruber & Denni, 2005), the most relevant question today is not if there is broadband access but how many providers offer it. Good regulatory schemes should encourage not only initial entry but subsequent competition among providers as well.
ACKNOWLEDGMENT 7KH¿UVWDXWKRUJUDWHIXOO\DFNQRZOHGJHVDJUDQW from the Institute of Governmental Affairs at UC Davis that supported this work. We thank participants at the Center for Research in Regulated ,QGXVWULHV¶WK$QQXDO:HVWHUQ&RQIHUHQFHDQG the Transportation and Public Utilities Group Annual Meeting (San Diego) for helpful comments on versions of this work.
REFERENCES Ai, C., & Sappington, D.E.M. (2002). The impact of state incentive regulation on the U.S. telecommunications industry. Journal of Regulatory Economics, 22(2), 133-159. 296
Berry, S.T. (1992). Estimation of a model of entry in the airline industry. Econometrica, 60(4), 889-917. Bittlingmayer, G., & Hazlett, T.W. (2002). The ¿QDQFLDOHIIHFWVRIEURDGEDQGUHJXODWLRQ,Q5: Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 223-277). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Brennan, T. (1989). Regulation by capping prices. Journal of Regulatory Economics, 1, 133-147. Bresnahan, T.F., & Reiss, P.C. (1987). Do entry conditions vary across markets? Brookings Papers on Economic Activity 1987 (3, Special Issue on Microeconomics), 833-881. Burnstein, D.E., & Aron, D.J. (2003). Broadband adoption in the United States: An empirical analysis. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (pp. 119-128). New York: Nova Science. Cameron, A.C., & Trivedi, P.K. (2005). Microeconometrics. New York: Cambridge University Press. Chamberlain, E.H. (1933). The theory of monopolistic competition. Cambridge, MA: Harvard University Press. Crandall, R.W., & Alleman, J.H. (Eds.). (2002). Broadband: Should we regulate high-speed Internet access? Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Crandall, R.W., & Jackson, C.L. (2003). The $500 ELOOLRQRSSRUWXQLW\ 7KHSRWHQWLDOHFRQRPLFEHQH¿W of widespread diffusion of broadband Internet access. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (ch. 8). Haupaugge, NY: Nova Science. Crandall, R.W., Sidak, J.G., & Singer, H.J. (2002). The empirical case against asymmetric regulation of broadband Internet access. Berkeley Technology Law Journal, 17(1), 953-987.
Regulation and the Deployment of Broadband
Donald, S.G., & Sappington, D.E.M. (1997). Choosing among regulatory options in the United States telecommunications industry. Journal of Regulatory Economics, 12(3), 227-243. Faulhaber, G.R. (2002). Broadband deployment: Is policy in the way? In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate highspeed Internet access? (pp. 223-244). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Faulhaber, G.R., & Hogendorn, C. (2000). The market structure of broadband telecommunications. Journal of Industrial Economics, 48(3), 305-329. FCC (Federal Communications Commission). (2000a). Deployment of advanced telecommunications capability: Second report (FCC 00-290). Washington, DC. FCC. (2000b). High-speed services for Internet access: Subscribership as of June 30, 2000. Washington, DC. FCC. (2001). Instructions for form 477. Washington, DC. FCC. (2006). High-speed services for Internet access: Subscribership as of December 31, 2005. Washington, DC. Flamm, K. (2005, February). The role of economics, demographics, and state policy in broadband availability. Proceedings of the PURC/London Business School Conference on the Future of Broadband: Wired and Wireless, 2005, Gainesville, FL. Ford, G.S., & Spiwak, L.J. (2004). The positive effects of unbundling on broadband deployment (Phoenix Center policy paper no. 19). Retrieved February 1, 2007, from http://www.phoenix-center. org/pcpp/PCPP19Final.pdf Gabel, D., & Kwan, F. (2001). Accessibility of broadband telecommunication services by various segments of the American population. In B.M. Compaine & S. Greenstein (Eds.), Communications policy in transition: The Internet and beyond (pp. 295-320). Cambridge: MIT Press.
Garcia-Murillo, M., & Gabel, D. (2003). International broadband deployment: The impact of unbundling. Proceedings of the 31st Telecommunications Policy Research Conference, Arlington, VA. Gillett, S.E., & Lehr, W. (1999, September). Availability of broadband Internet access: Empirical evidence. Proceedings of the 27th Annual Telecommunications Policy Research Conference, Alexandria, VA. Greenstein, S., McMaster, S., & Spiller, P. (1995). The effect of incentive regulation on infrastrucWXUHPRGHUQL]DWLRQ /RFDOH[FKDQJHFRPSDQLHV¶ deployment of digital technology. Journal of Economics and Management Strategy, 4, 187-236. Gregg, B.J. (2002). A survey of unbundled network element prices in the United States. Public Service Commission of West Virginia. Gruber, H., & Denni, M. (2005). The diffusion of broadband telecommunications: The role of competition. Retrieved February 1, 2007, from http://ssrn.com/abstract=829504 Grubesic, T.H. (2003). Inequities in the broadband revolution. Annals of Regional Science, 37, 263-289. Grubesic, T.H., & Murray, A.T. (2004). Waiting for broadband: Local competition and the spatial distribution of advanced telecommunication services in the United States. Growth and Change, 35(2), 139-165. Hastie, T., & Tibshirani, R.J. (1990). Generalized additive models. London/New York: Chapman and Hall. Hausman, J. (2002). Internet-related services: The results of asymmetric regulation. In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 129-156). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies. Hazlett, T., & Bazelon, C. (2005). Regulated unbundling of telecommunications networks: A stepping stone to facilities-based competition? Proceedings
297
Regulation and the Deployment of Broadband
of the 33rd Telecommunications Policy Research Conference. Howell, B. (2002). Broadband uptake and infrastructure regulation: Evidence from the OECD countries. Working Paper BH01/01, New Zealand Institute for the Study of Competition and Regulation, New Zealand. Jorgenson, D.W. (2001). Information technology and the U.S. economy. American Economic Review, 91, 1-32. Joskow, P.L. (1972). The determination of the allowed rate of return in a formal regulatory hearing. Bell Journal of Economics and Management Science, 3, 632-644. Kolbe, A.L., & Tye, W.B. (1990). The Duquesne opinion: How much “hope” is there for investors LQUHJXODWHG¿UPV"Yale Journal on Regulation, 8, 113-157. Lee, C.H. (2001). State regulatory commission treatment of advanced services: Results of a survey. Retrieved January 11, 2001, from http://www. nrri.ohio-state.edu/programs/telcom/pdf/broadband_survey_3-01.pdf Lehman, D.E., & Weisman, D.L. (2000). The political economy of price cap regulation. Review of Industrial Organization, 16, 343-356. Litan, R.E., & Rivlin, A.M. (2001). Projecting the economic impact of the Internet. American Economic Review, 91, 313-317. National Regulatory Research Institute (2000). Forms of regulation for basic services provided by major LECS in the U.S. states. Retrieved January 11, 2001, from http://www.nrri.ohio-state.edu/ programs/telcom/pdf/Forms_of_regulation_2001. pdf OECD. (2003). Developments in local loop unbundling. DSTI/ICCP/TISP(2002)5/FINAL; Directorate for Science, Technology and Industry; Committee for Information, Computer and Communications Policy; Working Party on Telecommunication and Information Services Policies; OECD, Paris.
298
OECD. (2006). Telecommunication regulatory institutional structures and responsibilities. DSTI/ICCP/TISP(2005)6/FINAL; Directorate for Science, Technology and Industry; Committee for Information, Computer and Communications Policy; Working Party on Telecommunication and Information Services Policies; OECD, Paris. Pindyck, R.S. (2004). Mandatory unbundling and irreversible investment in telecom networks. Working Paper No. W10287, NBER. Prieger, J.E. (2001). Telecommunications regulation and new services: A case study at the state level. Journal of Regulatory Economics, 20, 285-305. Prieger, J.E. (2002). Regulation, innovation, and the introduction of new telecommunications services. Review of Economics and Statistics, 84, 704-715. Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363. Rappoport, P.N., Kridel, D.J., Taylor, L.D., Alleman, J.H., & Duffy-Deno, K.T. (2003). Residential demand for access to the Internet. In G. Madden (Ed.), Emerging telecommunications networks: The international handbook of telecommunications economics (vol. 2, pp. 55-72). Cheltenham, UK: Edward Elgar. Reiss, P., & Wolak, F. (2005). Structural econometric modeling: Rationales and examples from industrial organization. Stanford GSB Working Paper. Retrieved November 15, 2006, from http:// www.stanford.edu/~preiss/makeit.pdf Sappington, D.E.M., & Weisman, D.L. (1996). Potential pitfalls in empirical investigations of the effects of incentive regulation plans in the telecommunications industry. Information Economics and Policy, 8, 125-140. Shelanski, H.A. (2002). Competition and regulation in broadband communications. In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 157-196). Washington, DC: AEI–Brookings Joint Center for Regulatory Studies.
Regulation and the Deployment of Broadband
Taylor, W., Zarkadas, C., & Zona, J.D. (1992). Incentive regulation and the diffusion of new technology in telecommunications. Unpublished Manuscript, National Economic Research Associates, USA. Wallsten, S. (2005). Broadband penetration: An empirical analysis of state and federal policies. Working Paper 05-12, AEI–Brookings Joint Center for Regulatory Studies, USA. Wallsten, S. (2006). Broadband and unbundling regulations in OECD countries. Working Paper 06-16, AEI–Brookings Joint Center for Regulatory Studies, USA. ;LDR0 2UD]HP3) Do entry conditions vary over time? Entry and competition in the broadband market: 1999-2003. Unpublished Manuscript.
Wire Center: The location where a central of¿FHVZLWFKFRQQHFWVWRWKHORRSIDFLOLWLHVWKDWFRYHU part or all of a local exchange. It is also the physical location where the loop distribution plant in the local telecommunications system can be accessed, and where the equipment enabling DSL service for subscribers is deployed. Used synonymously with ³FHQWUDORI¿FH´LQWKLVFKDSWHUDOWKRXJKHOVHZKHUH sometimes the wire center refers to the physical EXLOGLQJDQGWKHFHQWUDORI¿FHUHIHUVWRWKHVZLWFKHV within the wire center. ZIP Code Tabulation Area (ZCTA): An approximate area representation of a United States Postal Service ZIP Code service area. A ZCTA may not exist for every valid ZIP Code. In most FDVHVWKHH[FOXGHG=,3&RGHLGHQWL¿HVDVLQJOH delivery location for which the Bureau of the &HQVXVKDGLQVXI¿FLHQWGDWDWRHVWDEOLVKDGLVWLQFW ZIP Code area.
KEY TERMS ENDNOTES &HQWUDO2I¿FH See Wire Center. 1
Competitive Local Exchange Carrier (CLEC): A public telephone company that provides local telecommunications service in competition with the incumbent local exchange company.
2 3
Last Mile: The part of the local telecommuQLFDWLRQVQHWZRUNWKDWFRQQHFWVWKHFHQWUDORI¿FH to the subscriber premises. The last mile typically consists of copper telephone wire. Local Exchange Company (LEC): A public telephone company in the United States that provides local telecommunications service. Unbundled Network Element (UNE): Allows competitive local exchange carriers to offer comSHWLWLYHO\ SULFHG DOWHUQDWLYHV WR WKH LQFXPEHQW¶V services rapidly. UNEs are purchased at wholesale rates from the incumbent telephone companies. UNEs include the local loop, switch port, switching, transport, signaling systems, and databases (such as operations support systems and directory assistance). UNE-Platform (UNE-P, generally no longer available) combined all elements necessary to offer local service.
4
5
Litan and Rivlin (2001) estimate that the Internet could potentially reduce business costs by $100 billion to $250 billion annually. 47 U.S.C. § 251(f). Cable modem service suffers from congestion effects in the last mile, whereas DSL does not. The two services are also bundled differently with other services that consumers value, such as cable television subscription or local phone service. There have also been price differences between the services. Three states had tax incentives for broadband and one state was considering a loan program (Lee, 2001). Only 14 state commissions out of 39 answering a survey (Lee, 2001) decided broadband rates in rate cases, and some of these only under special circumstances. Lee (2001) also reports that the majority of states do not apply price caps for broadband; he does not report how many do. The survey report does not say which states enact which policies, so these variables cannot be used in the present study.
299
Regulation and the Deployment of Broadband
6
7
8
9
10
11
12
13
14
300
By “underinvestment” we mean the reduced incentive to offer a new service under RORR EHFDXVHWKHUHJXODWLRQOLPLWVWKHSUR¿WWKDWFDQ be made from service. This is distinct from the better-known overcapitalization property of RORR (the Averch-Johnson effect). Their explanation is that price caps insulate the regulator from the consequences of low 81(VEHFDXVHWKHSULFHVWKH¿UPFDQFKDUJH are constrained. Under RORR, on the other KDQGLI81(SULFHVDUHWRRORZDQGWKHSUR¿WDELOLW\RIWKH¿UPVXIIHUVWKHQWKH¿UPZLOO be allowed to raise its other prices to meet the allowed rate of return. See Chapter 16 for a more detailed review of the literature. 7KH¿JXUHDQGWKLVVHFWLRQGUDZKHDYLO\RQ Prieger (2003). $FFRUGLQJWRRI¿FLDOVWDWLVWLFVIRUUHVLGHQFHV and small businesses in the U.S., there were 2.2 million coaxial cable broadband lines in 2000, compared to 0.7 million DSL lines (FCC, 2000b). The remainder of the 3.1 million broadband lines consisted of dedicated lines (mostly T-1) rented from phone FRPSDQLHVVDWHOOLWHDQG¿[HGZLUHOHVVOLQHV DQG¿EHURSWLFOLQHV2IWKHPLOOLRQPHdium- and large-business high-speed lines in VHUYLFHZHUH¿EHUZHUH'6/ were coaxial cable, and 53% were dedicated lines. Smaller companies are exempt from this obligation. CLECs at the time could buy the loop by itself (known as UNE-L, for “loop”) or a package including the loop and switching (known as UNE-P, for “platform”). A court decision in 2004 struck down the UNE-P option. The local exchange carrier is not required to unbundle its DSL packet-switching facilities, however. See Shelanski (2002) for a fuller discussion of regulatory issues regarding unbundling and DSL. The FCC eliminated the requirement to offer line sharing to CLECs in February 2003. The term “free option” does not mean that the CLEC does not pay rental charges on the
15
16
17
18
19
20
21
22
last mile facilities. It refers to the fact that while the incumbent bears the investment risk in deploying the facilities, the CLEC can choose to purchase only those inputs that have turned out to be successful (at less than the cost of building their own facilities) (Pindyck, 2004). This option value provided to the CLEC from the unbundling rules is not included in the rental rates. Faulhaber (2002) points to evidence that the incumbent telephone companies are limited in their provision of DSL by their operational FDSDELOLWLHVQRWLQYHVWPHQWSUR¿WDELOLW\ In sub-loop unbundling, non-incumbent carriers can interconnect with the local access QHWZRUNDWSRLQWVEHWZHHQWKHLQFXPEHQW¶V site and the end user (for example, at a remote terminal deployed between the incumbent WHOHSKRQHFRPSDQ\¶VFHQWUDORI¿FHDQGWKH end user). See Ford and Spiwak (2004) for a non-refereed empirical study reaching the opposite conclusion. 7KH YDOXH RI WKH ¿UP¶V LQVWDOOHG FDSLWDO LV called the “rate base.” Earnings sharing and banded RORR schemes DOORZWKHUHJXODWHG¿UPWRHDUQPRUHWKDQ the “fair” rate of return on capital, as long as SUR¿WVIDOOZLWKLQDQDXWKRUL]HGUDQJHFDOOHG the “deadband.” Earnings above the deadband are returned partially or completely to ratepayers. A less common variant of earnings sharing is revenue sharing. See Chapter 16 by Hu and Prieger in this YROXPHIRUDIRUPDOPRGHORID¿UP¶VGHFLsion to deploy broadband. See Cameron and Trivedi (2005, ch. 14) for a textbook treatment of the probit model. In structural econometrics modeling, economic theory is used to impose structure on the problem, and the resulting estimates may be used to reveal the underlying “deep” parameters facing the economic agents (for example, the parameters of the cost and revenue functions). See Reiss and Wolak (2005) for a discussion of structural econometric PRGHOLQJLQWKH¿HOGRILQGXVWULDORUJDQL]D-
Regulation and the Deployment of Broadband
23
24
25
26
27
28
29
tion within the economics literature. The FCC broadband data are not well suited to existing structural models of entry using the QXPEHURI¿UPVLQWKHPDUNHW HJ%UHVQDhan & Reiss, 1987; Berry, 1992) because the FCC data are heavily censored. If there are RQHWRWKUHH¿UPVLQWKH=,3&RGHDUHDWKH exact number is not given; the majority of the GDWDDUHVRFHQVRUHG6HH;LDRDQG2UD]HP (2005) for an attempt at structural modeling with the FCC broadband data. Such geographic variation allows us to identify the impact of our regulatory variables of interest on broadband deployment. Today, broadband is available to almost all U.S. households (at least in urban and suburban areas), and thus we could not learn much from using similar methodology with current data. The FCC considers a carrier to be “facilities based” if it provides broadband access over its own local loop, or over unbundled network elements (UNEs) or leased lines that it obtains from other carriers and equips as broadband (FCC, 2001). Restricting the survey to facilities-based carriers removes double counting of resold access lines. The FCC has since removed this reporting threshold. We matched the ZIP Code areas by centroid WRFHQWUDORI¿FHVXVLQJ$UF0DS*,6VRIWZDUH DQG *'7¶V :LUH&HQWHU 3UHPLXP WHOHFRPmunications geography product. These data are proprietary. Another commonly cited disadvantage of the FCC broadband data is censoring. If the number of the reported broadband providers is one, two, or three, the actual number LVFHQVRUHGGXHWRFRQ¿GHQWLDOLW\FRQFHUQV However, since we only make use of the fact that there is at least one provider in the area, this limitation does not affect our methodology. The regulatory variable is constructed from NRRI (2000) and our own survey of state regulatory agencies. The regulatory variable is not necessarily the same for all observations within a state,
30
31
32
33
34
35
36
because there may be more than one carrier in the state, and not all carriers are necessarily under the same regulatory regimes. The UNE data are from Gregg (2002). They are the basic monthly “UNE platform (UNEP) price,” which includes local loop, port cost, and 1,000 minutes of switching. The personal and household data are from the U.S. Bureau of the Census “Census 2000 Summary File 3” DVD. The business data are from the “Census ZIP Code Business Patterns 2000” CD-ROM. See Faulhaber and Hogendorn (2000), Gabel and Kwan (2001), and Prieger (2003) for discussion. The other carriers are Alltel, Concord Telephone, Contel, Century, Cincinnati Bell, Citizens Telecom, Frontier Communications, IowaTel, Quaker State Telephone, and Sprint. Bell Atlantic had recently acquired GTE to form Verizon at the time of our data. Given that the merger had just taken place, and that most infrastructure in place was deployed by GTE when they were under their own manDJHPHQWZHGRQRWLQFOXGH*7(LQ9HUL]RQ¶V territory for purposes of the company indicator variables. The standard errors in Tables 2 and 3 are calculated using the usual formula assuming homoskedastic errors. We did not use heteroskedasticity-robust standard errors, as one sometimes sees for probit models, because: (1) the maximum likelihood estimates of WKHFRHI¿FLHQWVWKHPVHOYHVDUHQRWUREXVWWR heteroskedasticity (and the standard errors on inconsistent estimates are not interesting), and (2) the robust standard errors are very close to those we report. For example, none RIWKHVLJQL¿FDQFHVWDUVIRUWKHUHJXODWLRQDQG CLEC variables in estimation 1 would change if we used the robust standard errors. See also Donald and Sappington (1997) on the political economy of telecommunications regulation. The one exception is the rate freeze effect, which is about the same in magnitude as it is in estimation 1.
301
Regulation and the Deployment of Broadband
37
38
39
40
302
Ideal for the econometrician, presumably not WKH¿UPV The lowest decile group encompasses the lowest 10% of observations, ranked by UNE rates. The highest decile group includes the top 10% of observations so ranked. Preliminary nonparametric analysis from Hastie DQG7LEVKLUDQL¶V JHQHUDOL]HGDGGLWLYH model (GAM) indicated that there was possible nonmonotonicity in the effect of UNE rates across the range of the data. We chose the location of the knots for the spline based on the visual evidence from the GAM. The marginal effect of a variable that is in logs corresponds to the effect of multiplying the variable in levels by e (#2.7). We assess the functional form of such variDEOHV XVLQJ +DVWLH DQG 7LEVKLUDQL¶V
41
42
generalized additive model, which allows YDULDEOHV WR HQWHU WKH VSHFL¿FDWLRQ RI WKH PHDQÀH[LEO\7KHYLVXDOHYLGHQFHIURPWKH results informally guided our choice of locations for the knots of the spline functions for the income, population density, and average employment variables. A court decision since the time of the data we study here effectively rendered uneconomical large-scale competitive entry in local telephone markets by competitors using UNEs. Incumbents are no longer required to offer platform UNE (UNE-P) to competitors. Cable modem or DSL service was available in 87% of the ZIP Codes. Most of the remaining ZIP Codes were covered by satellite.
Regulation and the Deployment of Broadband
APPENDIX In this appendix, we describe the variables collected on characteristics of the ZIP Code areas. The QXPEHURI¿UPVLVWKHFRXQWRIHVWDEOLVKPHQWVIURPWKH³=,3&RGH%XVLQHVV3DWWHUQV´,QFRPHLV measured as median household income. Households are categorized as urban, rural non-farm, or rural farm (urban is the excluded category in the estimations). The excluded age category is 65 years and older. The excluded education category is “less than high school degree.” The commuting variables are taken from the travel time to work of workers aged 16 years and up; the excluded category is “commute is less than 30 but greater than zero minutes.” The units for population density are persons per square mile. Median structure age refers to dwelling places. The other demographic variables are straightforward and are described in the text. The cost variable is the proxy cost for local telecommunications service in the local exchange area, as FDOFXODWHGE\WKH)&&¶VK\EULGFRVWSUR[\PRGHO +&30 LQ-DQXDU\7KH+&30LVDQHFRQRPLF HQJLQHHULQJPRGHOWKDWFDOFXODWHVWKHFRVWRISURYLGLQJORFDOWHOHFRPPXQLFDWLRQVVHUYLFHXVLQJHI¿FLHQW WHFKQRORJ\JLYHQDQDUHD¶VJHRJUDSKLFWHUUDLQDQGVXEVFULEHUGHQVLW\3UR[\FRVWVDUHQRWDYDLODEOHIURP the model for some of the wire centers (mostly for smaller carriers); in these cases, we used the proxy cost of the nearest wire center for which cost was available. We matched wire center locations to ZIP Code areas as described in the text. The UNE rates are described in the text.
303
Section II
Consumer-User Behavior
Division I
Australia
306
Chapter XIX
Factors Affecting Broadband Adoption for Mainstream Consumers Peter Adams Charles Sturt University, Australia
ABSTRACT This chapter explores whether past experiences with telecommunications providers and current ‘plans’ on offer serve as barriers between an individual consumers’ persuasion phase of the Innovation-Decision 3URFHVV 5RJHUV DQGWKHGHFLVLRQSKDVH:LWKEURDGEDQGLQDURXQGRI$XVWUDOLDQKRPHV it is important that telecommunications providers understand why the future mainstream segment of consumers will want to adopt broadband, and any barriers to this. This analysis suggests future studies are needed to investigate whether the telecommunications providers are collectively confusing potential broadband consumers in their attempts to differentiate a generic product in the market. It argues that future technology adoption studies need to consider including the complexity of the actual purchase decision when developing constructs for quantitative models. If we are to build a picture of why mainstream consumers adopt broadband, more than just the perceptions of using the technology itself need to be investigated.
INTRODUCTION There is no simple explanation as to why one household installs broadband Internet and another chooses not to. While economic variables like price change over time and demographic distributions of income among a population will often correlate with the diffusion of an innovation, these correlations do not help us explain why members of the population under study adopt. These economic and demographic descriptions become even less useful as we move through the traditional diffu-
sion curve and study the mainstream segment of the population. Key literature across four disciplines (information systems, sociology, psychology, and consumer behavior) underpins the discussion and study presented in this chapter. The majority of the technology adoption literature in the information systems area can trace its roots back to sociology and psychology studies. Work in the areas of innovation diffusion theory (Rogers, 2003), the theory of planned behavior (Ajzen, 1991), and the theory of reasoned action (Fishbein & Ajzen,
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Factors Affecting Broadband Adoption for Mainstream Consumers
1975) are the building blocks for most technology adoption studies in the information systems literature. The majority of technology adoption research has studied workplace adoption intention DQGQRWLVVXHVVSHFL¿FWRKRXVHKROGFRQVXPHUV while most existing broadband adoption studies have primarily concentrated on economic and demographic descriptions of adoption (Madden & Coble-Neal, 2005; Choudrie & Dwivedi, 2005a; Cameron, 2004; Roy Morgan, 2003; Madden, Simpson, & Savage, 2002; Hausman, Sidak, & Singer 2001). The participants in the study outlined here discussed what technologies they found useful in their lives from the perspective of the household consumer. The transcripts were analyzed for both HPHUJLQJ WKHPHV DQG LQ WKH FRQWH[W RI ¿QGLQJV from existing studies. This chapter argues that future technology adoption studies need to investigate in more breadth and depth the role complexity of the purchase decision plays in acting as a barrier to adoption. Qualitative longitudinal studies of a purposive sample of households would provide more depth to this issue, while quantitative studies should consider including the complexity of the actual purchase decision when developing broad consumer technology adoption models. Building a picture of why mainstream consumers adopt broadband requires more than just investigating either their perceptions pre-adoption of the technology itself, or post adoption of using the technology. The critical issue is: what is the tipping point (Gladwell, 2002; Adams, 2005) which takes them from the persuasion to the decision phase (Rogers, 2003) in purchasing new technology for their household?
LITERATURE OVERVIEW There is much common ground in the studies of LQQRYDWLRQDQGWHFKQRORJ\DGRSWLRQLQWKH¿HOGVRI information systems, consumer behavior, psychology, and sociology. This study approached the issue from an information systems perspective initially through the lens of the following technology adoption models: the technology adoption model (TAM) of Venkatesh and Davis (2000) and Davis (1989);
the XQL¿HG WKHRU\ RI WKH DGRSWLRQ DQG XVDJH RI technology (UTAUT) of Venkatesh, Morris, Davis, and Davis (2003); and most recently the model of adoption of technology in households (MATH) of Brown and Venkatesh (2005). The origins of these models can be traced back to the sociology work of Rogers (1962, 2003) with his innovation diffusion theory (IDT), and two theories from the psychology discipline: the theory of reasoned action (TRA) of Fishbein and Ajzen (1975) and the theory of planned behavior (TPB) of Ajzen (1991). Work across several disciplines (Rogers, 2003; Moore, 2002; Bass, 1969; Mahajan, Muller, & %DVV VKRZVRFLDOIDFWRUVDQGVRFLDOLQÀXHQFHV KDYH D VWURQJHU LQÀXHQFH RQ PDLQVWUHDP consumers than on the innovators or early adopters. 5RJHUV¶ ZRUNIURPDVRFLRORJ\SHUVSHFWLYH LVWKHEHVWNQRZQRIWKHVHDQGKHGHVFULEHV¿YH attributes of innovations which can be studied in trying to predict their rate of adoption. In addition 5RJHUVGHVFULEHVD¿YHVWDJHSURFHVVKHFDOOVWKH Innovation-Decision Process (2003, pp. 174-179). 7KHVHFRQGRIWKHVHVWDJHVLVµSHUVXDVLRQ¶ZKHUH WKHLQQRYDWLRQ¶VFKDUDFWHULVWLFVLQÀXHQFHWKHIRUPDtion of a favorable or unfavorable attitude towards WKHLQQRYDWLRQEHIRUHWKHQH[WµGHFLVLRQ¶SKDVHRI adoption or rejection (see Figure 1). The process from the persuasion to the deciVLRQVWDJHLQ5RJHUV¶ ,QQRYDWLRQ'HFLVLRQ 3URFHVV VHH)LJXUH LVWKHVSHFL¿FDUHDRILQWHUHVW for the research outlined here. In the popular press Gladwell (2002) has described this transition as the µWLSSLQJSRLQW¶+HRXWOLQHVKRZVRPHLGHDVDQG products spread like viruses or epidemics, while others fail to gain any traction. Moore (2002) GHVFULEHVWHFKQRORJ\µPDLQVWUHDP¶FRQVXPHUVDV ZDQWLQJ WKH µZKROH SURGXFW¶ ZKHUH WKH JHQHULF SURGXFW¿WVZLWKWKHLUH[SHFWDWLRQVEXWDOVRSURvides the add-ons (peripherals, support, interoperability), as well as seeing some future potential for satisfying their growing needs. In developing a model of consumer adoption of technology, the idea of matching both current and perceived future needs is important. A more general view of conVXPHULQÀXHQFHLVSURYLGHGE\&LDOGLQL D psychologist. He sees the six principle categories RILQÀXHQFHDV UHFLSURFDWLRQFRQVLVWHQF\VRFLDO
307
Factors Affecting Broadband Adoption for Mainstream Consumers
Figure 1. The Innovation-Decision Process (Rogers 2003, p. 170).
proof, liking, authority, and scarcity. The principal of social proof is most relevant to informing the development of a model explaining why consumers adopt technology. Cialdini (2001, p. 140) says people look at others they see similar to themselves to see what they are doing, particularly where the situation contains uncertainty. Adopting a new technology is a situation which can be seen as containing uncertainty for many mainstream consumers. This LGHDRIVRFLDOSURRIRUVRFLDOLQÀXHQFHLVDFRPmon theme throughout the literature (Venkatesh et al., 2003; Brown & Venkatesh, 2005; Moore & Benbasat, 1991). There is a range of work in the marketing literature devoted to the communication channels used to inform decision making in the Innovation-Diffusion Process. The Bass (1969) model has formed the basis of much of this work which is well summarized by Mahajan et al. (1990). The %DVV PRGHO VKRZV WKDW LQWHUSHUVRQDO LQÀXHQFHV EHFRPHPRUHVLJQL¿FDQWDVZHPRYHWKURXJKWKH adoption curve. The Bass model assumes potential DGRSWHUVDUHLQÀXHQFHGE\WZRPDLQFRPPXQLFD-
308
WLRQFKDQQHOV WKHPDVVPHGLD H[WHUQDOLQÀXHQFH DQG ZRUG RI PRXWK LQWHUQDO LQÀXHQFH :KLOH there is a large body of work utilizing the Bass model, it is aimed at describing the market at an aggregate level. The work described here is aimed VSHFL¿FDOO\DWH[SORULQJwhy households make the purchase decision. In the information systems discipline over the past 20 years, there has been a considerable body RI UHVHDUFK GHYRWHG WR H[SODLQLQJ LQGLYLGXDOV¶ intentions to adopt technology. The most mature of these streams of research is in the area of DGRSWLRQLQDQRUJDQL]DWLRQDOFRQWH[W7KLV¿HOG of research was brought together by Venkatesh et DO SURSRVLQJWKHXQL¿HGWKHRU\RIDGRStion and usage of technology. With their model explaining up to 70% of the variance in intention in an organizational setting, they speculated that “we may be approaching the practical limits of our ability to explain individual acceptance and usage decisions in organizations” (p. 471). While this is a well-established area of research, the results need to be considered in the context of the research be-
Factors Affecting Broadband Adoption for Mainstream Consumers
ing done in organizations and studying individual adoption intention. Brown and Venkatesh (2005) have focused their work on household adoption of technology, and they have empirically tested and extended their model of adoption of technology in households, ¿UVWSURSRVHGLQ %URZQ 9HQNDWHVK This model was developed and tested in the context of personal computer (PC) adoption in households and explains 74% of the variance in intention to DGRSWD3&IRUKRPHXVH%URZQDQG9HQNDWHVK¶V (2005) extended MATH model incorporates household lifecycle stages and income (Brown & Venkatesh, 2005, p. 404). The extended MATH model is designed to better capture the complexities of household decision making by taking into account characteristics such as marital status, age of adults, and the presence (and age) of children in the household. Brown and Venkatesh used work originally developed by Gilly and Enis (1982) GH¿QLQJKRXVHKROGOLIHF\FOHVWDJHV7KHVRFLR economic factor of income was incorporated from work developed by Wagner and Hanna (1983). There is a large body of current work by Choudrie and Dwivedi (2004, 2005a, 2006a, 2006b; Dwivedi, Choudrie, & Brinkman, 2006) related to the adoption of broadband by consumers. The
same group of authors have also published on usage issues, but only the relevant adoption-related papers have been cited here. While their work informs the topics explored in this study, in the context of this chapter it is important to note that these papers relate to data collected in two surveys developed and implemented during the period 2003 to early 2005. In the design of their survey instrument, Dwivedi et al. (2006) have used work by Venkatesh and Brown published in 2001 which developed the MATH model from qualitative data gathered using a longitudinal phone survey related to the adoption of personal computers in the home. Subsequent to Dwivedi et al. (2006) developing their survey, Brown and Venkatesh published a paper in 2005 which extended their original MATH model to include a number of changes and additions (Brown & Venkatesh, 2005, p. 404). They tested their extended MATH model by surveying 746 U.S. households. The contemporary nature of the broadband adoption issue, combined with the time lags associated with publishing academic work, has PHDQW WKH µHDUOLHU¶ %URZQ DQG 9HQNDWHVK paper is more relevant in the context of this study than the work published in 2006 by Dwivedi et al.
Figure 2. The non-cumulative diffusion of innovations by adopter category over time (Rogers, 2003, p. 281)
309
Factors Affecting Broadband Adoption for Mainstream Consumers
which had relied on the 2001 work by Venkatesh and Brown. Other contemporary research being done in the broadband adoption area includes work by Cameron (2003, 2004, 2005) which is primarily concerned with comparing the adoption of broadband in Australia between Canberra (the national capital with good infrastructure) and the Northern Rivers area of NSW (coastal hinterland with a distributed population and pockets of infrastructure). &DPHURQ¶VGDWDZDVFROOHFWHGLQ1RYHPEHU and found the key demographic factors relevant to broadband adoption to be related to a range of factors including: location, education level, occupation level, ownership of other technology in the household, and home ownership for example. While the data is fairly recent, in terms of helping to explain why mainstream consumers adopt, it is too early for the Australian context. At the end of 2003 approximately 5% of Australian households had broadband (ACCC, 2006), which is still at the VWDUWRI5RJHUV¶HDUO\DGRSWHUSKDVHDQG5RJHUV DQGRWKHUV 0RRUH VKRZWKHUHDUHVLJQL¿cant differences in the attributes and attitudes of consumers at this stage of the diffusion cycle. There is also a range of work which focuses on broadband adoption in Korea. Korea has been a popular country to study because up until 2005 it had the highest penetration rate of 25.5 subscribers per 100 inhabitants according to the OECD (2005). One of the lessons we can learn from the Korean experience is that there are many ways to stimulate broadband adoption. Some of these are: government initiatives (Choudrie, Papazafeiropoulou, & Lee, 2003), geographic and cultural issues (Choudrie & Lee, 2004), as well as the role of experience with broadband at the individual level (Oh, Ahn, & Kim 2003). In 2006 Korea was overtaken by Denmark, The Netherlands, and Iceland at the head of the broadband subscriber rankings. The Netherlands comes in second with 28.8 broadband subscribers per 100 inhabitants, with Australia ranking 17th with 17.4 per 100. Barr (2006) presents a useful insight into some of the innovative approaches being taken in The Netherlands where a government-backed partnership with private enterprise called KenQLVZLMNZHQWWRWKHµXVHUV¶DQGDVNHGWKHPZKDW
310
they would like to see delivered via broadband. As Barr outlines, this user-centered approach offers DVLJQL¿FDQWDOWHUQDWLYHLQVLJKWLQWRKRZWRPDNH broadband appealing to mainstream consumers. Most of this literature does not exclusively focus on household-level adoption, but does offer some useful insights and demonstrates that there are many lenses through which to view the broadband adoption issue. While these studies have informed the approach taken here, for this chapter the critical elements of the literature are those facets which focus on the actual decision to purchase, described by Gladwell (2002) as the “tipping point.” Visually, it is best GHVFULEHGE\5RJHUV¶ ,QQRYDWLRQ'HFLVLRQ Process (see Figure 1). The diffusion model views the whole Innovation-Decision Process as having ¿YH VWDJHV 5RJHUV S NQRZOHGJH persuasion, decision, implementation, and con¿UPDWLRQ:KLOHFRQVXPHUVPD\KDYHDQDZDUHness of or be persuaded to consider a technology, at some point they make an adoption or rejection decision. In the information systems (IS) discipline, technology adoption research has been dominated by quantitative approaches. Choudrie and Dwivedi (2005b) found that 74% of studies on technology adoption in the leading IS journals used the survey method. They also looked more broadly at the issue of technology adoption in the household context in a wider range of journals and found the survey method still dominated, having been used in 63% of studies. The next most common approach used in 25% of the studies involved a mixed method approach where a survey was used in combination with interviews or a time use diary. The dominance of quantitative approaches in researching technology adoption may be explained by researchers tending to use the same methods as others in their discipline and the desire to be able to generalize about the potential adopting population to inform broad management or marketing strategies. The study of broadband adoption by households presents some unique challenges. Broadband is an intangible commodity to a certain extent (Adams, 2004) and one that can serve a range of
Factors Affecting Broadband Adoption for Mainstream Consumers
KRXVHKROGPHPEHUVLIWKHµKRXVHKROG¶DGRSWVRQH shared broadband connection. This is in contrast to mobile phones or personal computers which may have multiple adopters of different phones or computers within a household. While the exploratory nature of the research discussed in this chapter is outlined in the next section, there is a range of views on how to study WKH µKRXVHKROG¶ DGRSWLRQ RI EURDGEDQG ,Q WKHLU work studying the family purchasing process of broadband Internet in Australia, Chang, Ahn, and Lee (2006) adopted a family buying decision approach proposed by Sheth, Mittal, and Newman (1999). Chang et al. (2006) chose not to use the work of Rogers (2003) because they saw it more suited to individual adoption decisions, not family purchasing decisions. This approach worked for them because they chose to limit their study to RQO\IDPLOLHVZLWKFKLOGUHQDWWKHµIXOOQHVW¶VWDJH of the family lifecycle (Sheth et al., cited in Chang et al., 2006, pp. 286-288). There are many unanswered questions about whether broadband adoption in households follows similar patterns to individual purchase decisions or follows family decision models such as that proposed by Sheth et al. (1999). While that question will not be answered here, the approach used considers the relevance of the traditional adoption models (TAM, UTAUT, and MATH) used in the information systems and sociology (IDT, TRA, and TPB) literature. The study reported here is designed to provide a basis for the development of a survey instrument which will study a representative sample of all types of Australian households. It is projected that over the 26-year period 20012026, the fastest-growing segment in the makeup of Australian households will be the single person household, and the average household size will decline from 2.6 in 2001, to between 2.2 and 2.3 people by 2026 (ABS, 2004). Interestingly two of the three data series used for this analysis projects that by 2026, couples without children will be more abundant than couples with children. The subset RIWKHµIXOOQHVW¶WUDGLWLRQDOIDPLO\KRXVHKROGDV studied by Chang et al. (2006) only offers insights into one type of household. The family buying decision or individual adoption approach to studying broadband adoption need
not be seen as mutually exclusive. In households with more than one individual, the methodology used just needs to ensure there is scope for capturing the individuals views, which when aggregated build the household view.
PHILOSOPHY AND RESEARCH METHOD Given that technology adoption in the household is consumer behavior, and many of the popular technology adoption models (TAM, UTAUT, and MATH) owe their origins to sociology theories (IDT, TRA, and TPB), it is surprising interpretivist methods are not used more regularly to inform knowledge in this area. In keeping with the exploratory objectives of this study, the focus group method was selected as an appropriate way to investigate themes in addiWLRQWRWKRVHLGHQWL¿HGLQWKHH[LVWLQJWHFKQRORJ\ adoption studies. Focus groups were chosen over household interviews to assist in promoting discussion outside the parameters of the semi-structured topics gleaned from the IS literature. Smithson (2000, p. 116) argues that focus groups develop a discourse within the context of their discussions, and opinions should not be viewed as necessarily those individuals had formed prior to these discussions. Of course, the corresponding weakness in the focus group approach is the researcher has less control of the discussions than he or she might in an interview situation (Morgan, 1988; Warr, 2005). The topic of household use of technology was well suited to the focus group method as participants were keen to share their views and experiences. A more contentious issue such as downloading of pornography by household members may not be suited to this approach. To ensure the groups had a common basis for discussing technology adoption in their household, the four groups used were recruited based on the following criteria: Group A: Were existing broadband users Group B: Were planning on adopting broadband in the coming six months
311
Factors Affecting Broadband Adoption for Mainstream Consumers
Group C: Did not plan on adopting broadband in the foreseeable future Group D: Lived in a single-person household The participants from single-person households FRXOG IDOO LQWR DQ\ RI WKH ¿UVW WKUHH FDWHJRULHV (Group A, B, or C) of broadband adoption, but were VSHFL¿FDOO\WDUJHWHGDVDGLVWLQFWLYHJURXSEHFDXVH this is the fastest-growing segment of Australian KRXVHKROGV 6FKLIIPDQ%HGQDOO2¶&DVV3DODGLQR & Kanuk, 2005; ABS, 2001). Other studies (Chang et al., 2006; ACMA, 2005; Cameron, 2004) have shown the presence of children in the household makes the adoption of technology more likely, so studying single households was designed to explore additional themes which may extend current thinking. The sample of participants was drawn from RQH RI $XVWUDOLD¶V ODUJHVW UHJLRQDO FLWLHV ZKLFK was chosen as the most convenient way to access the views of participants from a city, regional, and rural background. The participants were selfselecting based on responding to media activity promoting the research. No incentive was offered for participating in the research. Participants attended the appropriate focus groups based on the four categories outlined earlier. If they lived in a single-person household, this took precedence to ensure this group was of an adequate size. At the time the study was undertaken, Australia had approximately 25% of households with a broadband Internet connection. This meant participants in the planning to adopt group could be considered WRFRPHIURPWKHµHDUO\PDMRULW\¶ 5RJHUV or mainstream (Moore, 2002) segment of the population.
It is generally accepted that an ideal focus group size is in the order of six to ten participants. Kitzinger (1995) suggests four to eight, Krueger (1994, p. 6) seven to ten, and Williamson (2002, S VHYHQWRWZHOYH7KH¿QDOVL]HRIWKHIRXU focus groups is listed in Table 1. Not surprisingly those who did not plan on installing broadband in the foreseeable future ZHUHWKHKDUGHVWWRUHFUXLW%\GH¿QLWLRQLI\RX are not interested in something, then you are not motivated to volunteer to talk about it. The initial PHHWLQJ WLPH IRU WKH µQRW LQWHUHVWHG¶ JURXS ZDV SRVWSRQHGDQGPRUHVSHFL¿FSURPRWLRQDOPHGLD ZDVXQGHUWDNHQWRUHFUXLWWKRVHµQRWLQWHUHVWHG¶LQ broadband. Conversely, the group most directly relevant to the research was the easiest to recruit. Those who were planning on adopting broadband in the next six months were keen to talk with other like-minded individuals. And despite the fact that this group was the largest in number, the members were the easiest to manage from a participation point of view as they exchanged ideas and elaborated on themes others had raised. In each case, one or two participants who had FRQ¿UPHGGLGQRWVKRZXSRQWKHQLJKWIRUDYDULety of reasons, meaning the target of at least eight participants was not always met. Future studies may consider over-recruiting, allowing for this contingency.
Preliminary Findings The focus groups were transcribed and coded in two SKDVHV7KH¿UVWSKDVHWRRNDQRYHUWO\LQWHUSUHWLYLVW approach in allowing the themes to emerge from the transcripts and the codes to be developed as
Table 1. Focus group make-up
312
Group
No. of Participants
Theme of Grouping
A
6
existing broadband users
B
10
were planning on adopting broadband in the coming six months
C
4
did not plan on adopting broadband in the foreseeable future
D
5
live in a single-person household
Factors Affecting Broadband Adoption for Mainstream Consumers
the transcripts were read and analyzed. Given the guiding themes used to shape the initial topics in this research came from existing quantitative studies, it would be naive to assume these studies did QRWLQÀXHQFHVRPHRIWKHFRGLQJGHVFULSWLRQVXVHG With this bias recognized, 20 codes were developed in addition to constructs/themes which could be associated with the existing quantitative models studied during the initial literature review. 7KHVHFRQGSKDVHRIFRGLQJVSHFL¿FDOO\ORRNHG for themes in the transcripts which aligned with constructs from the existing quantitative models: UTAUT (Venkatesh et al., 2003) and MATH %URZQ 9HQNDWHVK DVZHOODV5RJHUV¶ (2003) innovation diffusion theory. While the purpose of this chapter is to outline the emergLQJWKHPHVWKHVHFRQGSKDVHRIFRGLQJLVEULHÀ\ RXWOLQHGODWHUWRSODFHWKH¿QGLQJVLQWKHFRQWH[W of the existing body of literature. The exploratory phase of analyzing the tranVFULSWVLGHQWL¿HGWKUHHEURDGWKHPHVIURPWKH codes to emerge from the data. These emerging themes were outside the scope of those uncovered in the initial literature review. The most consistent theme revolved around the complexity of making the purchase decision. This chapter will focus on the issues raised around this theme of the complexity of the purchase decision. Future papers will explore the other two key themes to emerge, those being: the need to feel in of control of technology, and a cynical view of the telecommunication providers. Before exploring the theme of complexity in making a purchase, it needs to be recognized that ”those who adopt early are those who stand to EHQH¿WPRVWRWKHUWKLQJVEHLQJHTXDO´ 0F'RQDOG Corkindale, & Sharp, 2003, p. 92). While McDonald et al. (2003) were studying the early adoption of energy-saving light globes, the logic applies to technology innovations like broadband. Participants often had very practical reasons for why they had or had not adopted broadband. P: I buy in when what I’ve got is not doing the job anymore. So I don’t have an iPod because my Discman still works okay. I don’t have a digital
camera because my SLR still works. But when the SLR breaks I’ll go and get a digital camera. And I only got a DVD because the video didn’t work anymore. Sometimes people make a spur-of-the-moment GHFLVLRQWRDGRSWZKHQWKHµIDFWV¶DUHSXWLQIURQW of them. R: There was a usage issue in that I had a very cheap dialup plan and my usage pattern has recently changed and so I was exceeding my download limits in the last couple of months. So I was getting stabbed extra for that. So when I rang up to change my dialup plan they said, well you may as well go broadband because it’s the same cost. And Bigpond has got free installation at the moment. While it is an issue to be explored in more depth in other studies, it is interesting to speculate ZKHWKHU WKLV SDUWLFLSDQW ZHQW WKURXJK 5RJHUV¶ (2003) Innovation-Decision Process in a very condensed period of time within this phone call; or is there a completely different cognitive process happening in this case compared to a consumer who might take six months to make his or her purchase decision? Interestingly, practical reasons, while common, were not raised as consistently as the complexity of making the purchase decision.
The Complexity of the Purchase Decision Even when participants had essentially decided they wanted to purchase some sort of technology for their homes, often they found the information DYDLODEOHGLI¿FXOWWRXQGHUVWDQG ) ,¿QGDOORIWKDWPDUNHWLQJVRFRQIXVLQJ M1: Oh it is. F: And you know, we’ve got the home line because we’ve had it for years and years and years, but trying to sift through plans just leaves me cold.
313
Factors Affecting Broadband Adoption for Mainstream Consumers
,QRWKHUFDVHVZKHUHWR¿QGWKHLQIRUPDWLRQ itself was an issue. M1: We’re exploring voice over and I read the DUWLFOHRQLWEXW,¶YHQHYHUEHHQDEOHWR¿QGDJDLQ a lot more information on it. I’ve asked around [hometown] and…. K: That’s one of the advantages of broadband. M1: Yeah, but I can do and have voice over. But to get the information…. Frustration at not being able to reach a point where they felt comfortable making a decision was a common theme among participants. J: There’s too many different variables. You can’t compare them. Like >SDUWLFLSDQW¶VQDPH@ was saying, apples and apples. There is so many different things that was, it’s too hard to compare. In some cases the complexity of the purchase decision was enough of a barrier to result in a decision not to purchase, despite having committed to WKHµLGHD¶RIEX\LQJDSDUWLFXODUWHFKQRORJ\ M2: Computers are a really hard one to try and choose I reckon. I’ve had a look before and I just can’t even decide. I just give it a miss. Interestingly the line between the purchase of broadband and the computer itself was blurred with a number of the participants. The Internet connectivity was seen as part of the computer package, because this was how they received their Web access and e-mail. While the purchase of a commodity like EURDGEDQG LQYROYHV D VPDOOHU LQLWLDO ¿QDQFLDO commitment than purchasing a PC, one participant discussed how they had committed to installing broadband into their home, despite not understanding it. The overriding factor in making the purchase was simplifying the purchase process. V: Don’t understand it, can’t, don’t understand it, can’t waste the time…It’s like buying, the technology is so complex that for me to waste time and work out what’s what, which is which, how you 314
do it. I went to >SHUVRQ¶VQDPH@ over there and we went to [organization name] and I’ve got a very expensive thing…Somebody said, why would you get that one? Why would you pay that much? I VDLG,GRQ¶WFDUHVRLWORRNHG¿QH,GLGQ¶WZRUU\ about it. Might save $20 a month or something like that, don’t understand it. Demographic descriptors cannot be ignored when looking for ways to broadly isolate common WKHPHVIRUVSHFL¿FJURXSVVXFKDVHDUO\DGRSWHUV RUUHWLUHHV'HVSLWHVHHLQJVRPHEHQH¿WVRIEURDGband, like not tying up the phone line with dial-up Internet and having access to Internet banking, this participant (a retiree) still had cost as a barrier to WKHKRXVHKROG¶V¿QDODGRSWLRQGHFLVLRQ B: I suppose from our point of view that two things have come up…me as a retiree, what do I need to use the broadband for, and maybe Internet banking, it’s certainly one…If I wanted to look up a particular topic or something, or maybe as a matter of interest to do some research or whatever, but it’s, I’m not missing at the moment. I can do telephone banking, which I do. And the other major thing is the cost and the linking in with the telephone bill using your computer on Internet, whether you’re on broadband or whatever, is people not being able to contact you via phone. This is consistent with work by Brown and Venkatesh (2005), who incorporate a range of work on the household lifecycle into their adoption model. Their work also recognizes fear of technological advances and declining cost as playing roles in the decision to adopt technology by household consumers. J: The frustration is waiting for the right moment to pick the…to buy it, because as soon as you’ve bought it and it’s superseded. While it is not proposed to cover in detail the second phase of coding the transcripts in relation to the existing quantitative models, viewing the transcripts through the lens of existing quantitative PRGHOVSURYHGDXVHIXOH[HUFLVHLQFRQ¿UPLQJWKHLU relevance in the context of broadband adoption.
Factors Affecting Broadband Adoption for Mainstream Consumers
While a number of models were coded, the focus was on the model of adoption of technology in households (Brown & Venkatesh, 2005), as it is the most consumer focused of the IS models. This research found support for some existing constructs such as perceived ease of use (Brown & Venkatesh, 2005) and compatibility (Rogers, 3HUFHLYHGHDVHRIXVHDVGH¿QHGE\%URZQ DQG9HQNDWHVK UHÀHFWVDQLQGLYLGXDO¶VSHUceptions of their relationship with the technology. &RPSDWLELOLW\LVGH¿QHGE\5RJHUV DVDQ innovation being perceived as consistent with existing values and current needs. While the existing work by Rogers (2003) has complexity as one of WKH¿YHNH\FRQVWUXFWVWKHFRPSOH[LW\KHUHIHUV WRLVVSHFL¿FWRWKHXVHDQGXQGHUVWDQGLQJRIWKH technology under study, rather than the complexity of the purchase decision which emerged as the dominant theme in this study.
Connecting the Emerging Themes with the Literature Further literature reviews exploring the themes emerging from the focus groups revealed synergies
with some ongoing work in the psychology area. 6SHFL¿FLVVXHVWRXFKHGRQGLYHUVHHOHPHQWVVXFK as branding (Schiffman et al., 2005) and authority (Cialdini, 2001). Some relevant and informative research in the area of consumer choice has been undertaken by Schwartz et al. (2002), and also 6WUHEHO2¶'RQQHOODQG0H\HUV Firstly Schwartz et al. (2002) suggest that some people can feel worse off as the number of options they face increases. This study calls these people maximizersthey want the best option and are likely to experience regret because either their choice was not the best in all respects, or they were unable to properly weigh up all the alternatives. This has particular resonance in light of the themes to emerge from the focus groups. Broadband is a commodity where providers are often selling the same product bought from a broadband wholesaler (usually Telstra in Australia). The telecommunication companies are operating on small margins and are marketing their broadband services in order to build volume. Some consumers seem overwhelmed E\WKHDPRXQWRIFKRLFHWKH\KDYHDQG¿QGLWKDUG to choose the best alternative. Personality factors may play a role in different individuals with simi-
Figure 3. Conceptual model of frustration and the purchase process in a high-technology decision environment (Strebel et al. 2004, p. 1065)
315
Factors Affecting Broadband Adoption for Mainstream Consumers
lar demographic characteristics, making alternate adoption decisions. This links to the work by Strebel et al. (2004) (see Figure 3). They found consumers experience frustration in purchasing high-technology products in two dimensions: Consumers display frustration with the pace of technological change and with processing decision-related information. The results indicate that the probability of committing to a technology is VLJQL¿FDQWO\ ORZHU WKH KLJKHU FRQVXPHUV¶ IUXVtration is with their information environment.” (Strebel et al., 2004, p. 1072) 7KHLUPRGHOFRPSOHPHQWV5RJHUV¶ Innovation-Decision Process and suggests consumers may have made a commitment to buy a technology in principal, but that does not mean they have decided when to actually buy it. While further work is needed to explore this idea, the work of Schwartz et al. (2002) and Strebel et al. (2004) could be used to investigate the validity of the common practice of measuring behavioral LQWHQWLRQ 6WUHEHO HW DO¶V ZRUN DQG WKH dominant theme of this study, suggests consumers go through a two-stage process in what has often been studied as behavioral intention in the
IS literature. A decision to commit to a technology EDVHGRQDUDQJHRIIDFWRUVLVD¿UVWSKDVHEXWWKHQ a second decision phase of if and when to purchase is moderated by other factors. It is already recognized that intentions are not always a good predictor of behavior, particularly in survey research (see Chandon, Morwitz, & Reinartz, 2005). The emerging themes from this work suggest a longitudinal study investigating the process of how consumers move from an initial commitment, to a technology, to the purchase decision may add to our understanding of how consumers process and translate complex information about technology in order to make a purchase (or rejection) decision. This exploratory work indicates study of the process of how consumers move from commitment to purchase may provide deeper insights into why households adopt technology like broadband Internet.
DISCUSSION AND LIMITATIONS While the focus group method was particularly well suited to exploring the concepts of household WHFKQRORJ\ DGRSWLRQ LQ JHQHUDO DQG VSHFL¿FDOO\ EURDGEDQG LQ WKLV FDVH WKH ¿QGLQJV FDQQRW EH
Figure 4. CAIT (Consumer Adoption of Interactive Technology) model (Adams, 2006)
Perceived usefulness
Lifestyle motivations (career, family, entertainment) Purchase complexity
Image
Information sources
Perceived complexity Technology orientation (pessimistic, optimistic)
316
Purchase intention
Commitment to a technology
Perceived cost
Factors Affecting Broadband Adoption for Mainstream Consumers
generalized from one study. For an area like broadband adoption, generalizations are useful tools for a range of stakeholders interested in the adoption of broadband Internet in Australian households. They can assist telecommunications providers in making deployment decisions, retailers with marketing activities, governments and regulators in developing policies, and other researchers in exploring related themes. To this end it is recommended that the emerging theme of purchase complexity be incorporated into a quantitative study representative of the Australian population, to test its validity to be generalized. Recent work by the author (Adams, 2005, 2006) proposes a conceptual model (see Figure 4) to be used as the basis for a quantitative study of broadband adoption. Building on this conceptual model to create a valid survey instrument requires more than just taking existing constructs from the instruments used in previous technology adoption studies. The VSHFL¿FFRQWH[WRIWKHKRXVHKROGDVWKHDGRSWLQJ unit under study and the unique attributes of broadband as a technology need careful consideration before automatically adopting existing constructs. For a technology like broadband that has entered the mainstream phase of the diffusion curve, it is important to consider a range of psychographic constructs in addition to demographic ones. Being able to generalize about sections of the population based on common attributes is important for not only future research directions, but also to both governments and the telecommunications industry who want to see increased penetration of broadband adoption. &RQYHUVHO\LWLVDOVRLPSRUWDQWIRUPRUHVSHFL¿F in-depth studies to be carried out with individual households to follow them through the decision process in introducing new technologies like broadband Internet into their homes. An appropriate method would be a longitudinal study involving a series of interviews in a purposive sample of households over a 12- to 18-month period, trackLQJ WKH KRXVHKROGV¶ YLHZV RQ YDULRXV HPHUJLQJ technologies, when and why they felt committed to a technology, and when and why this translated into a purchase (or rejection) decision.
Further work is needed to gain an understandLQJ RI ZKDW LVVXHV DUH VSHFL¿F WR EURDGEDQG LQ comparison to other communications technologies. While this work found the purchase decision DVLJQL¿FDQWEDUULHUWREURDGEDQGDGRSWLRQZLWK 94% (Australian Government, 2006) of Australians having mobile phones, this almost ubiquitous adoption suggests that marketing of plans by competing telecommunication providers does not always prove to be a barrier to adoption for all technologies.
CONCLUSION The large body of IS technology adoption literature has historically been dominated by studies of organization adoption, with consumer models like MATH only emerging recently. This study shows that while these quantitative models do provide a solid basis for studying some key themes in household technology adoption, they are not good at explaining why the technology was adopted. The complexity of the purchase process DSSHDUVWREHDNH\IDFWRULQLQÀXHQFLQJWKHDFWXDO adoption decision. The emerging themes from this exploratory Lifestyle motivations study suggest consumers build their commitment (career, family, entertainment) RU µEHKDYLRUDO LQWHQWLRQ¶ WR D WHFKQRORJ\ LQ RQH phase, and then the purchase decision is a distinct phase moderated by key factors such as the complexity involved in gathering and analyzing enough information to make the purchase decision. It is no GRXEWGLI¿FXOWIRUWKHDYHUDJHFRQVXPHUWRFRPPLW to purchasing an intangible commodity he or she does not fully understand. A: So my expectation, I’m wondering now…asking people would Commitment this be unreal,to is that a if we went to broadband, that we would have access if we technology ZLVKHG RIWKHWLPH$QGWKDWH[SHFWDWLRQLV not unreal? Perceived cost
REFERENCES ABS (Australian Bureau of Statistics). (1999). Household and family projections Australia 1996 to 2021, 3236.0. Canberra. 317
Factors Affecting Broadband Adoption for Mainstream Consumers
ABS. (2004). Household and family projections Australia 2001 to 2026, 3236.0. Canberra. ACMA (Australian Communications and Media Authority). (2005). Digital media in Australian homes. Sydney. Adams, P. (2004). Is broadband really an innovationor just the Internet faster? Proceedings of ANZCA04, Sydney. Retrieved from http:// www.conferences.arts.usyd.edu.au/viewabstract. php?id=142&cf=3 $GDPV3 ,QVHDUFKRIEURDGEDQG¶VWLSSLQJ pointa conceptual model. Proceedings of the Communications Research & Strategy Forum, Sydney. Retrieved from http://networkinsight. YHUYHFRQWHQWFRPYHUYHBUHVRXUFHV3$GDPVB¿OH pdf Adams, P. (2006). Isolating why Australian households adopt broadband. Telecommunications Journal of Australia, 56(3). Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes, 50(2). Australian Government. (2006). Snapshot of industry: The Australian wireless industry. Retrieved November 15, 2006, from http://www. investaustralia.gov.au/index.cfm?id=9A0C9842D0B7-180C-16403701E984FDFB Barr, T. (2006). User-centred broadband: The Kenniswijk experiment. Telecommunications Journal of Australia, 56(3). Bass, F.M. (1969). A new product growth model for consumer durables. Management Science, 13(5). Brown, S.A., & Venkatesh, V. (2005). Model of Adoption of Technology in Households: A baseline model test and extension incorporating household life cycle. MIS Quarterly, 29(3). Cameron, A. (2003). Net drag: The understated discrimination in trickle-down models of broadband rollout. Brisbane, Australia: Australia & New Zealand Communications Association.
318
Cameron, A. (2004). Factors affecting broadband adoption and demand: A comparative study between metropolitan and regional Australia. Telecommunications Journal of Australia, 54(2). Cameron, A. (2005). Examining broadband demand. Proceedings of the Communications Research & Strategy Forum, Sydney. Retrieved from http://networkinsight.vervecontent.com/ verve/_resources/Cameron_S4.pdf Chang, S., Ahn, S., & Lee, H. (2006). The family purchasing process of broadband Internet in Australia. International Journal of Technology Marketing, 1(3). Chandon, P., Morwitz, V.G., & Reinartz, W.J. (2005). Do intentions really predict behavior? Self-generated validity effects in survey research. Journal of Marketing, 69(2). Choudrie, J., & Dwivedi, Y.K. (2004). Towards a conceptual model of broadband diffusion. Journal of Computing and Information Technology, 12(4). Choudrie, J., & Dwivedi, Y.K. (2005a). The demographics of broadband residential consumers in a British local community: The London borough of Hillingdon. Journal of Computer Information Systems, 45(4). Choudrie, J., & Dwivedi, Y.K. (2005b). Investigating the research approaches for examining technology adoption issues. Journal of Research Practice, 1(1). Choudrie, J., & Dwivedi, Y.K. (2006a). A comparative study to examine the socio-economic characteristics of broadband adopters and non-adopters. Electronic Government, 3(3). Choudrie, J., & Dwivedi, Y.K. (2006b). InvestigatLQJIDFWRUVLQÀXHQFLQJDGRSWLRQRIEURDGEDQGLQ the household. Journal of Computer Information Systems, 46(4). Choudrie, J., & Lee, H. (2004). Broadband development in South Korea: Institutional and cultural factors. European Journal of Information Systems, 13(2).
Factors Affecting Broadband Adoption for Mainstream Consumers
Choudrie, J. , Papazafeiropoulou, A., & Lee, H. (2003). A web of stakeholders and strategies: A case of broadband diffusion in South Korea. Journal of Information Technology, 18(4).
Mahajan, V., Muller, E., & Bass, F.M. (1990). New product diffusion models in marketing: A review and directions for research. Journal of Marketing, 54(1).
Cialdini, R.B. (2001). ,QÀXHQFH—science and practice. Needham Heights, MA: Pearson.
McDonald, H., Corkindale, D., & Sharp, B. (2003). Behavioral versus demographic predictors of early adoption: A critical analysis and comparative test. Journal of Marketing Theory and Practice, 11(3).
Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3). Dwivedi, Y.K., Choudrie, J., & Brinkman, W.P. (2006). Development of a survey instrument to examine consumer adoption of broadband. Industrial Management & Data Systems, 106(5-6). Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley. Gilly, M.C., & Enis, B.M. (1982). Recycling the IDPLO\OLIHF\FOH $SURSRVDOIRUUHGH¿QLWLRQAdvances in Consumer Research, 9(1). Gladwell, M. (2002). The tipping point. New York: Time Warner.Hausman, J.A., Sidak, J.G., & Singer, H.J. (2001). Residential demand for broadband telecommunications and consumer access to unaf¿OLDWHG,QWHUQHWFRQWHQWSURYLGHUVYale Journal on Regulation, 18(1). Infoplease. (n.d.). U.S. household PC growth and penetration. Retrieved from http://www.infoplease.com/ipa/A0908456.html
Modahl, M. (2000). Now or never. New York: Harper Business. Moore, G.A. (2002). Crossing the chasm. New York: HarperCollins. Moore, G.C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, (2). Morgan, D. (1988). Focus groups as qualitative research (vol. 16). Newbury Park, CA: Sage. OECD. (2005). OECD broadband statistics—June 2005. Retrieved November 15, 2006, from http:// www.oecd.org/document/16/0,2340,en_2649_ 34225_35526608_1_1_1_1,00.html OECD. (2006). OECD broadband statistics—June 2006. Retrieved November 15, 2006, from http:// www.oecd.org/document/9/0,2340,en_2649_ 34223_37529673_1_1_1_1,00.html
K i t z i n g e r, J. (19 9 5 ) . Q u a l i t a t i v e researchintroducing focus groups. British Medical Journal, 311(7000).
Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitudes. Journal of Information Technology, 18(4).
Krueger, R. (1994). Focus groups: A practical guide for applied research. Thousand Oaks, CA: Sage.
Rogers, E. (1962). Diffusion of innovations. New York: The Free Press.
Madden, G., & Coble-Neal, G. (2005). Australian residential telecommunications consumption and substitution patterns. Review of Industrial Organization, 26(3).
Rogers, E. (2003). Diffusion of innovations. New York: The Free Press.
Madden, G., Simpson, M., & Savage, S. (2002). Broadband delivered entertainment services: Forecasting Australian subscription intentions. The Economic Record, 78(243).
Roy Morgan Research. (2003). Broadband in Australia: The broadband consumer in Australia. Melbourne. 6FKLIIPDQ/%HGQDOO'2¶&DVV$3DODGLQR A., & Kanuk, L. (2005). Consumer behaviour. Frenchs Forest, NSW, Australia: Pearson.
319
Factors Affecting Broadband Adoption for Mainstream Consumers
Schwartz, B., Ward, A., Monterosso, J., Lyubomirsky, S., White, K., & Lehman, D.R. (2002). 0D[LPL]LQJ YHUVXV VDWLV¿FLQJ +DSSLQHVV LV D matter of choice. Journal of Personality and Social Psychology, 83(5).
Williamson, K. (2002). Research methods for students, academics and professionals. Wagga Wagga, Australia: Center for Information Studies, Charles Sturt University.
Sheth, J.N., Mittal, B., & Newman, B.I. (1999). Customer behavior: Consumer behavior and beyond. Orlando: Dryden.
KEY TERMS
Smithson, J. (2000). Using and analysing focus groups: Limitations and possibilities. International Journal of Social Research Methodology, 3(2).
Image: “The degree to which use of an inQRYDWLRQLVSHUFHLYHGWRHQKDQFHRQH¶VLPDJHRU VWDWXVLQRQH¶VVRFLDOV\VWHP´ 0RRUH %HQEDVDW 1991, p. 195).
6WUHEHO-2¶'RQQHOO. 0\HUV-* Exploring the connection between frustration and consumer choice behavior in a dynamic decision environment. Psychology & Marketing, 21(12). Venkatesh, V., & Brown, S.A. (2001). A longitudinal investigation of personal computers in homes: Adoption determinants and emerging challenges. MIS Quarterly, 25(1). Venkatesh, V., & Davis, F.D. (2000). A theoretical extension of the Technology Acceptance Model: )RXUORQJLWXGLQDO¿HOGVWXGLHV Management Science, 46(2). Venkatesh, V., Morris, M.G., Davis, G.B., & Davis, F.D. (2003). User acceptance of information WHFKQRORJ\ 7RZDUGDXQL¿HGYLHZMIS Quarterly, 27(3). Wagner, J., & Hanna, S. (1983). The effectiveness of family life cycle variables in consumer expenditure research. Journal of Consumer Research, 10( 3). :DUU'- ³,WZDVIXQ«EXWZHGRQ¶WXVXally talk about these things”: Analyzing sociable interaction in focus groups. Qualitative Inquiry, 11(2).
320
Information Sources: The degree to which VRFLDOLQÀXHQFHDGYHUWLVLQJSXEOLVKHGUHYLHZV DQGH[SRVXUHWRDWHFKQRORJ\LQÀXHQFHWKHGHYHORSment of an individuals normative beliefs (Brown & Venkatesh, 2005; Rogers, 2003). Lifestyle Motivations: “The degree to which an innovation is perceived as consistent with the existing values, part experiences, and current needs of potential adopters” (Rogers, 2003, p. 15). Perceived Complexity: “The degree to which DQLQQRYDWLRQLVGLI¿FXOWWRXQGHUVWDQGDQGXVH´ (Rogers, 2003, p. 16). Perceived Cost: $Q LQGLYLGXDO¶V SHUFHSWLRQ of cost. This is a secondary attribute as it is how the consumer considers price relative to his or her disposable income that is important (Moore & Benbasat, 1991, p. 194). Perceived Usefulness: The degree to which a person believes using the innovation is an improvement over the technology it supersedes (Davis, 1989; Rogers, 2003). Technology Orientation: $Q LQGLYLGXDO¶V optimistic or pessimistic feeling about technology (Modahl, 2000).
Division II
Europe
322
Chapter XX
Developing a Dynamic View of Broadband Adoption Herbert Daly Brunel University, UK Adriana Ortiz TECNUN University of Navarra, Spain Yogesh K. Dwivedi Swansea University, UK Ray J. Paul Brunel University, UK Javier Santos TECNUN University of Navarra, Spain Jose M. Sarriegi TECNUN University of Navarra, Spain
ABSTRACT The widespread domestic use of broadband Internet technology has been recognized to have a posiWLYHLQÀXHQFHRQQDWLRQDOHFRQRPLHVDQGLPSURYHWKHOLIHRIFLWL]HQV'HVSLWHVXEVWDQWLDOLQYHVWPHQWWR develop the infrastructure, many countries have experienced slow adoption rates for broadband. This chapter develops a view of UK broadband adoption using dynamic modeling techniques based on an existing statistical study. The contrasting approaches to modeling are compared. Principles of a dynamic modeling system are introduced and an appropriate form for broadband adoption chosen. The process of building a dynamic model based on an existing static model of broadband adoption is presented. Finally, the new perspective of the dynamic model is explored using the causal loop analysis technique.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Developing a Dynamic View of Broadband Adoption
INTRODUCTION The growth of the domestic use of broadband Internet access in the United Kingdom (UK) is a topic of interest for both government and the private sector. At the European level the widespread use of broadband is seen as a key to promoting economic activity, outlined in the 2000 Lisbon Agenda, as well as supporting civil projects and rural development (COM, 2006). This focus of interest has led to research into the spread of broadband technology, including the domestic take-up of broadband services in various countries (Choudrie & Dwivedi, 2005; Heejin, 2¶.HHIH .\RXQJOLP 6WDQWRQ Typically these studies explore the current state of national broadband penetration and usage. A structured approach is applied, using data collection and statistical analysis to identify the critical factors affecting change. By contrast, other studies focus on predicting the effects of future demand for broadband including penetration and usage. For example BSG & ANALYSYS (2006), commissioned by the UK government advisory body the Broadband Stakeholder Group, attempts to predict domestic bandwidth requirements for the years 2008 and 2012. The method described for obtaining the SURMHFWLRQV LV IDLUO\ XQVWUXFWXUHG XVHU SUR¿OHV are estimated, and developments in network applications and bandwidth usage are projected on that basis. While both approaches clearly provide useful results, they are affected by the limitations that result from the methods they apply. Both provide snapshot views of the situations they describe, one in the present, one in the future. Neither provides much insight into the processes that eventually led to those states or where they might lead to in the future. Studies such as Choudrie and Dwivedi (2005), Stanton (2004), and Aron and Burnstein (2003) contain a great deal of structure but require a VXI¿FLHQW TXDQWLW\ DQG TXDOLW\ RI GDWD LQ RUGHU to apply statistical measures. The rigorous structured approach provides little room for speculation except by using hypothesis testing. The BSG &
ANALYSYS (2006) study, by contrast, provides no traceable structure to understand how the future predictions are made. Although undoubtedly formed by experienced professionals, it is speculative and includes caveats to that effect. A dynamic modeling approach provides a compromise between these two extremes. Using all available resources, a structure for the problem is developed and used to examine how change may affect it over time. This chapter aims to demonstrate what dynamic models offer researchers investigating the issue of broadband adoption. An example of an existing static model that may be used as the basis of a dynamic model is also described. The System Dynamics (SD) technique of causal loop analysis is used to investigate the resulting dynamic model and analyzes the implications of the structure.
STATIC AND DYNAMIC MODELING: COMPARING CONCEPTS Static models provide a snapshot view of the state of a problem, as it is currently or as it may be in the future. Techniques may be applied to examine the static state, for example revealing the driver mechanisms for future change; however, they can provide no indication whether these same mechanisms were active in the past or will remain so in the future. Dynamic models are able to represent issues of change over time in more detail. There are many methods available to build static or dynamic models, and the issue of comparing different styles of mathematical model can be far reaching. In order to provide a brief example of comparing approaches, we refer to Meadows (1985) which reviews a number of models created, using different approaches, to understand the effects of policy on social and environmental problems.
Reviewing Approaches The static or dynamic properties of a model are E\QRPHDQVGH¿QLWLYHLQGHWHUPLQLQJWKHPRGHO¶V form. Meadows (1985) considers forms to be separate schools or even paradigms of modeling. The
323
Developing a Dynamic View of Broadband Adoption
different schools produce models with different properties, and may provide different views and different values on model creation and evaluation. Table 1 gives a brief summary of some schools of modeling that are available. 7KHµEDVLV¶UHIHUVWRWKHPHFKDQLVPE\ZKLFK the models in each paradigm produce results. However even where the same method is used, different principles may apply. For example, Discrete Event Simulation (DES) typically uses stochastic simulation with variable time steps. System Dynamics uses deterministic simulation with ¿[HG WLPH VWHSV Agent-based modeling (ABM) may use either method. As these distinctions and the descriptions suggest, the process of creating and evaluating models in each system can differ VLJQL¿FDQWO\
Selecting a Dynamic Form for Broadband Adoption Modeling Each of the available modeling approaches has strengths and weaknesses. Differential equation models for example, although well established, become cumbersome if a large number of variables are required to explain the dynamic behavior. The
mathematical techniques required to formulate and solve them are also a barrier for the uninitiated. Stochastic process and DES models focus on random behaviors of the problem over time. Treating unknown motivations as random variables can be a useful way to investigate them. Stochastic process models can also be cumbersome for a large number of variables and may require the problem WREHVLPSOL¿HGLQRUGHUWREHWUDFWDEOH $GRPLDQ 1987). DES avoids these issues by using computer simulation. As an event-based approach however, HDFKLQGLYLGXDOWUDQVDFWLRQLQÀXHQFLQJEURDGEDQG adoption must be represented; this is likely to produce a high quantity of data for analysis. Statistical dynamic time series models, such as ARIMA, UHTXLUHDVLJQL¿FDQWQXPEHURIZHOOIRUPDWWHGGDWD observations to be used effectively. In the case of EURDGEDQGDGRSWLRQDFFHVVWRGDWDRIVXI¿FLHQW quantity and quality is not currently possible. ABM is a potentially effective way of representing the problem of broadband adoption. Subscribers could be represented individually, and the emergent effects of their behavior over time could be analyzed. However, emergent properties of the problem can only be assessed quantitatively using ABM. Due to its relative lack of maturity,
Table 1. Comparison of paradigms that may potentially be used for modeling broadband adoption Modeling Paradigm
Time
Basis
Agent Based
Dynamic
Simulation
Description Imitating the behavior of a problem using granular entities with encapsulated behavior
Differential Equation
Dynamic
Calculus
Using differential calculus and related techniques explicitly to formulate and analyze a problem
Discrete Event Simulation
Dynamic
Simulation
Imitating the behavior of a problem using processes and passive
Regression
Static
Statistical
Applying statistical techniques to test relationships between problem
entities, based on events
data variables Stochastic Process
Dynamic
Stochastic
Using theoretical properties of random distributions and processes to analyze a problem
System Dynamics
Dynamic
Simulation
Time Series
Dynamic
Statistical
,GHQWLI\LQJÀRZVVWRFNVDQGIHHGEDFNLQDSUREOHPLQRUGHUWR analyze systemic behavior Applying statistical techniques to problem data collected over successive time intervals
324
Developing a Dynamic View of Broadband Adoption
ABM and its tools can often be complicated to implement. System Dynamics is the chosen approach for developing a dynamic model of broadband adopWLRQ IRU WKUHH PDLQ UHDVRQV WKH ÀH[LEOH GDWD requirements of SD, support for feedback, and qualitative analysis. The implications of these properties are as follows: •
•
•
Problem structure rather than data alone is used to generate results from the model, and this may compensate for the current lack of appropriate data System Dynamics supports the interaction of variables over time including feedback loops, the effects of previous values and multicollinearity, and the effect of variables on each other Techniques such as causal loop analysis allow the problem to be analyzed qualitatively. The model therefore provides a guide to informed speculation on the development of the problem.
Contrasting Statistical Regression and System Dynamics Approaches Meadows (1985) compares the process and concepts of a regression modeling approach with that of System Dynamics. Regression modeling typically begins by representing the conceptual model as a hypothesis about relationships between variables. Those hypothesized to exert an effect are the independent, or predictor, variables, and those affected are the dependant variables. Data is collected for each variable, and the relationship between them is analyzed statistically for data quality and then using regression techniques. The results indicate ZKHWKHUWKHSUHGLFWRUYDULDEOHVDUHVLJQL¿FDQW,I so, the EYDOXHFRHI¿FLHQWLQGLFDWHVKRZLQÀXHQWLDO each predictor is on the dependant variable. The greater the number of observations with the appropriate qualities, included in the analysis, the more reliable the model is considered to be. System Dynamics by contrast is a systemsoriented approach (Forrester, 1961) that typically begins by mapping a conceptual model of
WKHIDFWRUVLQÀXHQFLQJWKHSUREOHP$Q\FDXVDO links between factors are included, and the map LVH[WHQGHGWRLQFOXGHDQ\VLJQL¿FDQWIDFWRUVWKDW affect the subject behavior. Two distinct classes of YDULDEOHVFDQEHLGHQWL¿HG exogenous variables, determined beyond the system boundary, and endogenous variables, with values determined by the system behavior. Time-delay factors are also included to indicate a delay between a change in one variable and its effect on others. The modeler attempts to replicate the feedback structure of the problem including the perceptions of actors in the system. Simulation is used to generate output for the PRGHORYHUD¿[HGWLPHSHULRGZLWKDJLYHQVHW of starting conditions. Data from the subject is compared with the output of the model. The result RIWKHFRPSDULVRQLVXVHGWRUH¿QHDQGGHYHORSWKH model. When the model is considered adequate, different policies are applied to the simulation to investigate their effect on the behavior of the system. Richardson (1991), Roberts (1994), and Coyle (1977) are recommended as an introductory texts.
Multicollinearity: An Illustrative Difference in Approach From their descriptions, clearly the qualities of models and criteria for testing them differ. Statistical models are based on positivist ideas of observation and inference, while SD models are based on LGHDVIURPV\VWHPVWKHRU\RIVWUXFWXUHGH¿QLQJ EHKDYLRULPSOHPHQWHGXVLQJ¿UVWRUGHUGLIIHUHQWLDO equations. Meadows (1985) even suggests the two paradigms may be irreconcilable. The issue of multicollinearity provides a practical example of the differing methods at work in the two approaches. In regression modeling, observations of the predictor variables are ideally independent. For accurate measures of the relationship to the dependant variable, predictor variables should affect the dependant variable, but not each other (Greene, 2003). Multicollinearity refers to instances where correlation occurs between predicWRUYDULDEOHVRIWHQEHFDXVHWKH\DUHLQÀXHQFHGE\ each other or through a common source. There are
325
Developing a Dynamic View of Broadband Adoption
many tests to establish whether multicollinearity is present in a data set, and some strategies exist that attempt to resolve it. For example variables are chosen and data collected in such a way as to minimize multicollinearity. By contrast, in a system dynamics model, creating patterns of interdependence between variables is seen as essential to establishing the structure of the problem. All endogenous variables are dependant on other variables in the system. 9DOXHVPD\DOVREHDXWRUHJUHVVLYHLQÀXHQFHGE\ their own previous values through feedback loops. These properties in SD models are the fundamenWDOVRXUFHRIWKHPRGHOV¶EHKDYLRUDQGFRQVLGHUHG key to replicating the observed behavior of the problem. In regression models however, both auto regression and multicollinearity are considered problematic.
The ability of SD to represent diverse and F\FOLFDO UHODWLRQVKLSV PD\ SURYH VLJQL¿FDQW LQ developing a dynamic view of broadband adoption, as such relationships clearly exist. Factors which DSSHDULQGHSHQGHQWPD\LQÀXHQFHHDFKRWKHUDQG their own future values. For example, Aron and Burnstein (2003) observe that a fall in the price of broadband subscription may increase the number of subscribers, an increase in subscribers may improve the economies of scale, which may in turn lead to lower prices.
BROADBAND ADOPTION: TESTING A CONCEPTUAL MODEL Before presenting the dynamic model, a previous study of broadband adoption which applies a rig-
7DEOH0%$FRQVWUXFWVDQGGH¿QLWLRQV DGDSWHGIURP'ZLYHGL 'H¿QLWLRQVRI&RQVWUXFWVDQG6RXUFHV Construct Behavioral Intention (BI)
$FRQVXPHU¶VLQWHQWLRQWRVXEVFULEH RULQWHQWLRQWRFRQWLQXHWKHFXUUHQWVXEVFULSWLRQ DQGPDNHXVHRI broadband Internet in the future
Relative Advantage (RA)
Degree to which broadband Internet is perceived as being better than its predecessor, narrowband Internet
Utilitarian Outcomes (UO)
Extent to which broadband Internet usage enhances the effectiveness of household activities such as XQGHUWDNLQJRI¿FHZRUNDWKRPHFKLOGUHQ¶VKRPHZRUNLQIRUPDWLRQRUSURGXFWVHDUFKDQGSXUFKDVHDQG home business
Hedonic Outcomes (HO)
Pleasure derived from the consumption or use of broadband Internet, for example, the entertainment potential of the Internet via offerings such as online radio, streaming audio and video, electronic greetings, online games, and online casino
Service Quality (SQ)
The perceived quality of service a consumer obtained or is obtaining from the current Internet service providers; service quality is measured in terms of speed of connection and security problem with Internet connections, virus, and pop-up problems, with connection and customer support obtained from the ISP providers
3ULPDU\,QÀXHQFHV
7KHSHUFHLYHGLQÀXHQFHVIURPIULHQGVDQGIDPLO\WRVXEVFULEHWRDQGXVH RUQRWWRVXEVFULEHWRDQGXVH
(PI)
broadband Internet services
6HFRQGDU\,QÀXHQFHV
7KHSHUFHLYHGLQÀXHQFHRILQIRUPDWLRQIURPVRXUFHVVXFKDVDGVDQGQHZVRQ79RULQQHZVSDSHUVWR
(SI)
subscribe to and use (or not to subscribe to and use) broadband Internet services
Knowledge (K)
7KHSHUFHLYHGOHYHORINQRZOHGJHDERXWEURDGEDQG,QWHUQHWDQGLWVULVNVDQGEHQH¿WV
6HOI(I¿FDF\ 6(
The perceived ability or skill to operate computers and the Internet (narrowband or broadband) without the assistance of others
326
Developing a Dynamic View of Broadband Adoption
orous and comprehensive statistical methodology to create a static model of broadband adoption is reviewed. An overview of this study is provided to introduce some key concepts in broadband adoption. Findings from the static model are adapted to inform the dynamic model presented in the next section. The study included in Dwivedi (2005) provides a detailed analysis of broadband adoption behavior in the UK and focuses on three areas: factors affecting the take-up of domestic broadband, factors affecting customers changing service provider, and the patterns of Internet usage.
Forming the MBA Model Dwivedi (2005) develops and tests a new conceptual model, the model of broadband adoption (MBA). Based on existing concepts of technology adoption and diffusion, MBA is formed to examine three key factors: behavioral intention to adopt broadband (BI), broadband adoption behavior (BAB), and behavioral intention to change service provider (BISP). BI is used as both a predictor and a dependant variable in the tests. The model proposes that these key factors are affected by three categories
of construct: attitudinal, normative, and control. A GH¿QLWLRQRIHDFKFRQVWUXFWLVJLYHQLQ7DEOH The constructs are categorized as attitudinal (RA, UO, HO, SQ), normative (PI, SI), and control (K, SE, FCR) constructs. Fourteen hypotheses are formed to express the relationships proposed by the model; each expresses a reasoned assertion that a particular construct or variable is expected to have a positive effect on BAB or BISP. For example, hypothesis H1 states that “Overall atWLWXGLQDOIDFWRUVZLOOKDYHDSRVLWLYHLQÀXHQFHRQ the behavioral intention to adopt broadband” (Dwivedi, 2005). BAB is treated as a dependant variable. BI is treated as a dependant variable in relation to the constructs and a predictor variable in relation to BAB.
Data Collection and Testing Three relevant rounds of the regression analysis were performed, regressions I-III. A number of statistical measures were applied to the collected data, including least squares regression (R 2). Regression I used the individual factors of each construct as predictor variables and BI as the dependant variable. The K construct was found
Figure 1. Diagram of MBA hypotheses (taken from Dwivedi, 2005) ATTITUDINAL CONSTRUCTS
Behavioural Intention to change service provider (BISP)
H5
Relative Advantage
H2
Utilitarian Outcomes
H3
H8
Hedonic Outcomes H4
Service Quality H1 NORMATIVE CONSTRUCTS
H7
Primary Influence
Behavioural Intention to Adopt Broadband (BI)
H6 Secondary Influences
H13
Broadband Adoption Behaviour (BAB)
H9 CONTROL CONSTRUCTS Knowledge Self-efficacy
H11 H12
H10b
H10a
Facilitating Conditions Resources
327
Developing a Dynamic View of Broadband Adoption
7DEOH0%$FRQVWUXFWFRHI¿FLHQWVIURP5HJUHVsion II (adapted from Dwivedi, 2005) E
&RHI¿FLHQWV
RA
0.255
PI
0.195
FCR
0.180
SE
0.165
UO
0.113
HO
0.100
QRWWREHVLJQL¿FDQWLQGHWHUPLQLQJ%,7KH82 factor was also found to be an unreliable indicator in Regression I by a small margin, but not omitted from Regression II which produced the E values in Table 3 for RA, PI, FCR, SE, UO, HO on BI. Regression III consisted of two stages of testing: combined constructs against BI and testing BI against BAB using logistic regression. The constructs were combined by taking the mean of attitudinal, normative, and control constructs in each case and applied in regression on BI. BI and FCR were applied in regression on BAB using a logistic regression approach.
Conclusions from the MBA Study Dwivedi (2005) concludes by making a number of suggestions to the broadband industry and to policymakers more broadly. The study found that NQRZOHGJHGLGQRWVLJQL¿FDQWO\DIIHFWLQWHQWLRQWR adopt broadband and that, of the nine constructs, UHODWLYHDGYDQWDJHKDVWKHPRVWLQÀXHQFHRQLQWHQtion. It is suggested that differentiating broadband and narrowband packages more distinctly could exploit this market feature and encourage take-up. The importance of facilitating conditions leads to the suggestion that low-cost packages could increase take-up more broadly. Dwivedi (2005) also suggests that increasing skills to improve VHOIHI¿FDF\FRXOGDOVRGULYHDFFHSWDQFH%DODQFHG content packages to support both hedonic and utilitarian outcomes were also advised.
328
UK BROADBAND ADOPTION: BUILDING A DYNAMIC VIEW Using the MBA study as a basis, a dynamic view of UK broadband adoption can be developed using the System Dynamics approach. The dynamic model excludes demographic factors covered by MBA. The limits were chosen to keep the dynamic model simple and to focus on the main issue of the broadband adoption process. The static regression model implies a relationship between those intending to adopt broadband (BI) and broadband customers (BAB). The dynamic model attempts to present the process by which this transition occurs. Formal concepts from MBA are retained, including each of the nine constructs, however the dynamic model also includes informal ideas from the study not included in the regression PRGHO)RUH[DPSOHWKHLQÀXHQFHRIJRYHUQPHQW and the effect of broadband subscription price are included. Some concepts from the MBA model and study are included directly and others indirectly. Table 4 maps the relationship between the constructs and variables in the dynamic model. The dynamic model also integrates and connects the variables, to show their logical interdependence; the number of broadband customers impacts on the EURDGEDQGVHUYLFHSURYLGHU %63 SUR¿WVZKLFK in turn affects levels of investment. Concepts such DVSULPDU\LQÀXHQFHDQGNQRZOHGJHVHSDUDWHGLQ the regression model, have a direct relationship in the dynamic view. Time-delay features are added for the dynamic model that are unnecessary in the static view. For example there is a notable delay before investment in innovation, and infrastructure affects service quality and the price of broadband. The completed model is presented in Figure 2.
REVIEWING THE NEW MODEL: CAUSAL LOOP ANALYSIS The implications of the problem structure can be explored through causal loop analysis. This is the analysis of the interactions between variables and
Developing a Dynamic View of Broadband Adoption
Figure 2. Broadband adoption causal loop diagram +
Training in PC and internet skills
Access to broadband in an alternative place Support from government and private companies Relative advantage
3 +
+ Service Quality +
Competition among BSPs
1 +
Investment in innovation and infrastructure +
2
-
Broadband Customers + BSP profit
+
Available BSPs
Investment
+ Time spent on internet + + Word of mouth
+
Volatily of market
4
7
Awareness of broadband benefits + +
-
+
Use of broadband for utilitarian outcomes +
+ Customers perception of broadband need
+
+ Potential Broadband customers
+
5
-
-
+
-
+
6
Potential customers + perception of broadband need
+
Price of broadband access -
Use of broadband for hedonic outcomes
Time spent on other daily activities +
+ Investment in advertising +
Customer Satisfaction +
+ <Service Quality>
Table 4. Dynamic model variable descriptions including references to MBA model MBA Construct
Related Variable in Dynamic Model
Behavioral Intention to Change Service
Customer satisfaction
Provider Broadband Adoption Behavior
Broadband customers
Behavioral Intention
3RWHQWLDOFXVWRPHUV¶SHUFHSWLRQRIEURDGEDQGQHHG Potential broadband customers
Relative Advantage
Relative advantage
Utilitarian Outcomes
Use of broadband utilitarian outcomes
Hedonic Outcomes
Use of broadband hedonic outcomes
Service Quality
Service quality
3ULPDU\LQÀXHQFHV
Word of mouth
6HFRQGDU\,QÀXHQFHV
Investment in advertising
Knowledge
$ZDUHQHVVRIEURDGEDQGEHQH¿WV
6HOI(I¿FDF\
Training in PC and Internet skills
Facilitating Conditions Resources
Investment in innovation and infrastructure
Additional factors included in the dynamic
$FFHVVWREURDGEDQGLQDQDOWHUQDWLYHSODFH$YDLODEOH%63V%63SUR¿W&RPSHWLWLRQ
model
among BSPs, Customer satisfaction, Customers perception of broadband need, Investment, Price of broadband access, Support from government and private companies (House of Commons European Scrutiny Committee, 2006), Time spent on Internet, Time spent on other daily activities, Volatility of market
329
Developing a Dynamic View of Broadband Adoption
)LJXUH,QÀXHQFHH[DPSOHV
Income
+ +
Shortage of Doctors -
Outgoings
-
Bank Account
Bank account +
Interest
+ Spending
Disposable income +
+ Medical Students
delays in a qualitative manner. Flows between variables are indicated by marked arrows. A posiWLYHPDUNPHDQVDQLQFUHDVHLQWKHµIURPYDULDEOH¶ FDXVHVDQLQFUHDVHLQWKHµWRYDULDEOH¶$QHJDWLYH PDUN PHDQV DQ LQFUHDVH LQ WKH µIURP YDULDEOH¶ FDXVHVDGHFUHDVHLQWKHµWRYDULDEOH¶)LJXUHD provides an example. A path may lead a variable back to itself causing a feedback loop. Figure 3b illustrates the two kinds of loop possible: reinforcing (+) loops and balancing (-) loops. Roberts (1994) provides a comprehensive introduction to these concepts. The loops in the examples are obvious and direct; in real models, loops often pass through several other variables and may not be obvious.
Doctors
(c)
(b)
(a)
+
Delays are another feature of the problem that affects how the model behaves. The notation for a delay is demonstrated in Figure 3c: there is a VLJQL¿FDQW GHOD\ EHWZHHQ WKH PHGLFDO VWXGHQWV beginning their studies and their qualifying as doctors. When there is a shortage of doctors, more students apply to be trained. By the time they all TXDOLI\WKHVKRUWIDOOPD\DOUHDG\EH¿OOHGE\D previous cohort of students.
Dynamics of the Broadband Market 6HYHQ IHHGEDFN ORRSV KDYH EHHQ LGHQWL¿HG WKDW DIIHFW WKH RYHUDOO EHKDYLRU LQ D VLJQL¿FDQW ZD\ 7KH¿UVWORRSLQ)LJXUHVKRZVDVLPSOH³YLUWXRXV
Figure 4. Fundamental features of the broadband market (left) including market competition and government support (right) Support from government and private companies Price of broadband access -
- Potential Broadband customers
Price of broadband access
-
+ Broadband Customers
1 Investment in innovation and infrastructure
+ Investment in innovation and infrastructure +
2
+ Broadband Customers
Volatily of market + + Customer Available BSPs Satisfaction
+
-
+ Investment
330
-
Competition among BSPs +
1
-
- Potential Broadband customers
+
BSP profit
Investment
+
BSP profit +
Developing a Dynamic View of Broadband Adoption
circle” view of the broadband market suggested by Aron and Burnstein (2003); an increase in FXVWRPHUVLQFUHDVHVWKHSUR¿WIRU%63VDQGSUR¿W is translated into investment which eventually lowers prices. In the United States, where broadband adoption has been relatively slow, solutions ranging from state subsidy to increased market competition have been suggested (Stanton, 2004). In the UK both mechanisms are at work, as shown in Figure 4, yet adoption has none the less been slow (Dwivedi, 2005). The second causal loop suggests why this may be: as the price for falls and up-take increases, the pool of potential customers, inhibited only by price, is steadily drained. This suggests to policymakers that, whether based on subsidy or competition, solutions to poor take-up based on price alone have only a limited scope for success.
Attracting New Adopters Figure 5 highlights the basic mechanism for attracting new adopters, based around the perception of need for broadband. Several variables stimulate the
perception of need, but clearly in order to become a potential customer, one must believe that broadband access would be useful for some reason. Aron and Burnstein (2003) discuss stimulating the perception of need through a broadband µNLOOHUDSS¶$µNLOOHUDSS¶LQPDUNHWLQJLVDQDSplication so useful in itself that customers adopt the base technology just to use that application. Killer apps may drive adoption initially, but once the base technology is established, other applications may prosper. For example, electric lighting ZDV DUJXDEO\ WKH µNLOOHU DSS¶ GULYLQJ WKH LQLWLDO adoption of domestic electric power. Once power was established however, other household appliances, such as electric ovens, irons, and washing machines, were able to prosper. It is as still unclear ZKHWKHUDµNLOOHUDSS¶IRUEURDGEDQGFDQEHIRXQG EXWORZFRVWÀDWUDWHWHOHSKRQ\DSSOLFDWLRQVVXFK as SKYPETMPD\SURYHLQÀXHQWLDOLQLQFUHDVLQJ the number of broadband subscribers. The growing popularity of high-bandwidth multimedia services, VXFKDV
Figure 5. Mechanisms for attracting potential customers to adopt broadband Relative advantage +
Support from Government and PC
3
Service Quality +
Price of broadband access +
1 +
Investment in I&I +
+ Potential customers perception of broadband need + + + Potential Broadband customers -
+ Broadband Customers
Volatily of market +
+ Customer Satisfaction
Available BSPs
-
2
BSPs Competition +
-
Awareness of broadband benefits
+
BSP profit Investment
+
331
Developing a Dynamic View of Broadband Adoption
VXEVFULEHUVPXVWEHPDGHDZDUHRIWKHEHQH¿WV $OVR WKH DSSOLFDWLRQ RU LWV NH\ EHQH¿WV ZRXOG need to apply broadband in particular rather than narrowband to increase the relative advantage. An alternative approach suggested by Dwivedi WR LQÀXHQFH WKH SHUFHSWLRQ RI QHHG LV WR improve the relative advantage of broadband by developing service packages that clearly differentiate broadband from other Internet access methods. The role of relative advantage in the PRGHOLVYHU\VLJQL¿FDQW$WRQHH[WUHPHLIWKH advantage is small, it does not affect take-up very much. At the other, if broadband is the most effective or only access method, the effect on take-up will be constant. Control over relative advantage is not only in the hands of the BSPs; competing access methods, such as narrowband, may resist losing subscribers by reducing costs and improving services. This is unlikely to be the case with narrowband suppliers in the UK market, as most of them also have
DVWDNHLQEURDGEDQGDQGZRXOGEHQH¿WIURPLWV wider acceptance. Clearly using relative advantage as the key adoption driver would require sensitive management to be effective in the long term. The other path for increasing the perception of need in potential subscribers is by increasing the DZDUHQHVVRIWKHEHQH¿WVRIEURDGEDQGEHQH¿WV The next sections expand on this and explain why the approach may prove especially potent as the market develops.
Limitations to Growth Figure 6 adds paths which develop the awareness RI EURDGEDQG EHQH¿WV 7KLV LV VWLPXODWHG LQ D number of ways: through the advertisements of the BSPs, through government support for skills WUDLQLQJ DQG WKURXJK WKH LQÀXHQFH RI H[LVWLQJ broadband users. All of these contribute indirectly to the perception of need. However, the need for domestic adoption is affected by having access
Figure 6. Mechanisms potentially balancing or driving broadband adoption +
Training in PC and internet skills Internet Access to broadband in an alternative place
Support from Government and PC
customers -Potential customers¶ + perception of broadband need
Relative advantage +
3 Service Quality +
1 + Investment in I&I +
Price of broadband access -
Awareness of broadband benefits + Potential Broadband + + customers -
+ Broadband Customers
Volatily of market + + Available BSPs
-
Investment
4 Customer Satisfaction +
+
BSP profit +
Word of mouth +
-
2
BSPs Competition +
Customer Satisfaction
332
+
+
-
+
+
+
Investment in advertising <Service Quality>
Developing a Dynamic View of Broadband Adoption
to broadband through some other meansfor example, convenient broadband access through work, a place of study, or local Internet cafes. The development of alternative access sources is EH\RQGWKHVFRSHRIWKHVWXG\EXWLWVLQÀXHQFHRQ adoption cannot be ignored. Observing the reinforcing loops, in particular loop 4, through investment in advertising as well DV WKH LQÀXHQFH RI WUDLQLQJ DQG ZRUG RI PRXWK recommendations, there is certainly the potential to overcome the effect of alternative access sources. This may, perhaps, develop in the same way, as domestic washing machines have proved to be more convenient than alternative access to laundry facilities. Of the reinforcement mechanisms for adopWLRQSHUKDSVWKHPRVWVLJQL¿FDQWDQGLQWHUHVWLQJ LVWKURXJKµZRUGRIPRXWK¶DVLWLVDIIHFWHGE\WKH usage patterns of existing broadband adopters.
use for hedonic outcomes reducing time in daily activities and increasing the perception of the need for broadband. The sixth loop shows the effect of utilitarian outcomes of increasing time spent on other activities. The seventh loop balances the effect of the sixth. Service quality permitting, either hedonic or utilitarian use leads to potential EHQH¿WVIRUWKHXVHU7KHLUVDWLVIDFWLRQLVSDVVHG on to others through word of mouth, a normative SULPDU\ LQÀXHQFH LQ WKH FRQFHSWXDO PRGHO $V FXVWRPHUVEHQH¿WIURPEURDGEDQGVHUYLFHVQHZV RIWKHEHQH¿WVVSUHDGWRRWKHUV&XVWRPHUSHUFHStion of the need for broadband feeds into to the perception of need in potential adopters through word of mouth. The more people using broadband regularly encourages others to do so. Notably however, consistent or improving service quality is required for this normative effect to become an effective driver of adoption. The model also shows that it is possible that utilitarian Internet use creates opportunities to spend time in other ways. It is equally possible to spend any time saved by utilitarian uses on hedonic outcomes if desired. In both cases, the resultant effect is that the more use customers make of broadband services, the more they perceive that they need them and to some extent pass this on to
Patterns of Usage and Their Effects Figure 7 shows the effect of usage patterns. Patterns of usage are formed, as time spent in hedonic and utilitarian outcomes impact on time spent in daily activities. The model represents this as the LQWHUDFWLRQ RI WKUHH ORRSV 7KH ¿IWK ORRS VKRZV
Figure 7. Mechanisms affecting the patterns Internet usage and ‘word of mouth’ Use of broadband for hedonic outcomes
-
5
Time spent on other daily activities +
Use of broadband for utilitarian outcomes +
6 Awareness of broadband benefits +
+
7
-
+ Customers perception of broadband need + + Time spent on internet + Internet
Word of mouth + +
Customer Satisfaction
<Service Quality>
333
Developing a Dynamic View of Broadband Adoption
RWKHUV7KH¿UVWSDUWRIWKLVHIIHFWWKHUHIRUHJRHV VRPHZD\WRH[SODLQWKHSKHQRPHQDRIµ,QWHUQHW DGGLFWLRQ¶ GHVFULEHG E\ .LP 5\X DQG &KRQ (2006), and suggests that excessive Internet use may be a trigger of the phenomenon rather than just the main symptom.
CONCLUSION Researchers interested in the adoption of broadband have many research tools available to them including quantitative models. This chapter proposed that dynamic models may offer a compromise for the UHVHDUFKHUDQGVXJJHVWHGKRZH[LVWLQJ¿QGLQJVRI static studies of the current state of the broadband market could be deconstructed and reused in the development of a dynamic view. A number of modeling approaches with different features may be suitable to model the broadband market, and this study used System Dynamics due to the close match between the features of the approach and those inherent in the broadband market such as the endogenous variables described by Aron and Burnstein (2003). The statistical model MBA (Dwivedi, 2005) was used as the basis for a less formal dynamic model using system dynamics notation. The implications of the dynamic view were analyzed using the qualitative technique causal loop analysis and FRPSDUHGWRWKH¿QGLQJVRIDQXPEHURIH[LVWLQJ studies. We based this model on the study of the UK broadband market, but the same principles could be applied to any similar study. Researchers attempting to develop a dynamic model for another national market are advised to base their model on a thorough and well-grounded initial study. Most of the variables from the UK study are generic enough to be found in most broadband markets worldwide, however their LQGLYLGXDOVLJQL¿FDQFHDQGWKHUHODWLRQVKLSVEHtween them may vary, which will naturally affect the analysis. The UK model suggests several potential mechanisms for developing and stabilizing the broadband market. Although largely consistent, the static and dynamic models occasionally came to
334
differing conclusions. For example the conclusion RI'ZLYHGL LGHQWL¿HVUHODWLYHDGYDQWDJHDV the dominant factor in broadband adoption, and suggests that industry and policy exploit this to stimulate adoption. The dynamic view suggests that while this factor may help establish a base of subscribers, over the long term normative facWRUVFRXOGFRPHWRKDYHWKHJUHDWHVWLQÀXHQFHRQ the adoption process as long as service quality is maintained. A notable limitation of the dynamic model is that, unlike the empirical models of Aron and Burnstein (2003), Choudrie and Dwivedi (2005), and Stanton (2004), it did not address demographic factors directly. This is partly because qualitative model analysis is of limited use in this area, and aggregation in general is a feature of system dynamics modeling. An attempt to represent the complexity of demographic factors using an agent-based modeling approach may prove fruitful however. The development of a system dynamics VWRFNÀRZVLPXODWLRQPRGHOZDVQRWDWWHPSWHGEXW may have been a useful tool to evaluate the effects of different policies on a more detailed quantitative version of the dynamic model. Although the construction process for a system dynamics model is not as rigorous or as compatible with hypothesis testing as a statistical model, it may SURYLGHVRPHFRPSOHPHQWDU\EHQH¿WVDQGLQVLJKWV for researchers. The ability to include informal concepts, for example, allows the researcher to LQFOXGHUHODWLRQVKLSVREVHUYHG¿UVWKDQGGLUHFWO\ into the model. The emphasis placed on the interaction of components may be particularly useful in problems where interactions play a key role in the overall behavior. In conclusion, dynamic models provide a structure to guide prediction and speculation, and highlight possible areas of research and experimentation, or conditions under which the observed behavior may change.
REFERENCES Adomian, G. (1987). Modelling and solving physical problems. Mathematical Modelling, 8, 57-60.
Developing a Dynamic View of Broadband Adoption
Aron, D., & Burnstein, D. (2003). Broadband adoption in the United States: An empirical analysis. Retrieved November 1, 2006, from http://ssrn. com/abstract=386100 BSG & ANALYSYS. (2006). BSG green paper: Predicting UK future residential bandwidth requirements. Broadband Stakeholder group (www. broadbanduk.org) Choudrie, J., & Dwivedi, Y. (2005). The demographics of broadband residential consumers of a British local community: The London Borough of Hillingdon. Journal of Computer Information Systems, 45(4), 93-101. COM. (2006). Bridging the broadband gap. COM 129, The Commission to the Council, The European Parliament, The European Economic and Social Committee, The Committee of the Regions. Coyle, R. (1977). Management system dynamics. London/New York: John Wiley & Sons. Dwivedi, Y. (2005). Investigating consumer adoption, usage and impact of broadband: UK households. Thesis, Brunel University, UK. Forrester, J. (1961). Industrial dynamics. Cambridge MA: MIT Press. Greene, W. (Ed.). (2003). Econometric analysis. Englewood Cliffs, NJ: Prentice Hall. +HHMLQ/2¶.HHIH5 .\RXQJOLP< The growth of broadband and electronic commerce in South Korea: Contributing factors. In the Information Society, 38(1), 81-93. House of Commons European Scrutiny Committee. (2006). Twenty-sixth report of session 2005–06. /RQGRQ 6WDWLRQHU\2I¿FH/LPLWHG
Richardson, G.P. (1991). Feedback thought in social science and systems theory. Philadelphia: University of Pennsylvania Press. Roberts, N. (1994). Introduction to computer simulation: A system dynamics modeling approach. Portland, OR: Productivity Press. 6WDQWRQ/ -DQXDU\ )DFWRUVLQÀXHQFLQJ the adoption of residential broadband connections to the Internet. Proceedings of the 37th Annual International Conference on System Sciences (p. 10), Hawaii.
KEY TERMS Balancing Loop: A feature of a problem that PD\EHREVHUYHGXVLQJLQÀXHQFHGLDJUDPV/RRSV exist where a chain of cause and effect leads back to the same variable. In a balancing loop an even number of negative relationships between each of the variables in the chain causes quantities to decrease throughout the loop. In a model these relationships may represent potential goal-seeking behavior in the problem. Causal Loop Analysis: A qualitative method IRUHYDOXDWLQJLQÀXHQFHGLDJUDPVZKLFKDUHDOVR know as causal loop diagrams. Re-enforcing and EDODQFLQJORRSVDUHLGHQWL¿HGLQWKHGLDJUDPDQG causal loop analysis is the process of analyzing their interactions and the possible effects of loops and delays on the dynamic behavior of the problem. Roberts (1994) provides extensive coverage on this subject.
Kim, K., Ryu, E., & Chon, M. (2006). Internet addiction in Korean adolescents and its relation to depression and suicidal ideation: A questionnaire survey. International Journal of Nursing Studies, 43(2), 185-192.
Dynamic Modeling: An approach to modeling of a problem that includes concept of change in the problem over time. There are many different methods available and the principles they based on varies. Their suitability to model a particular problem may be determined by the match between the features they support and the features of the problem.
Meadows, D. (1985). The electronic oracle: Computer models and social decisions. Chichester: John Wiley & Sons.
,QÀXHQFH Diagrams: A method for conceptualizing a problem by representing the main YDULDEOHVWKDWFDXVHRULQÀXHQFHWKHPWRLQFUHDVH
335
Developing a Dynamic View of Broadband Adoption
or decrease. Relationships between variables are marked with either positive or negative polarity. Delays are also included, showing time lags in the effect of change on other variables. Coyle (1977) provides extensive coverage. Also known as causal loop diagrams. Re-Enforcing Loop: A feature of a problem WKDWPD\EHREVHUYHGXVLQJLQÀXHQFHGLDJUDPV Loops exist where a chain of cause and effect leads back to the same variable. In a re-enforcing loop, a positive relationship between each of the variables in the chain, or an even number of negative relationships, cause quantities to increase throughout the loop. In a model these relationships may represent potential exponential growth, or possibly decline, in the problems behavior.
336
Static Modeling: An approach to modeling RIDSUREOHPEDVHGRQWKHVWDWHDWD¿[HGSRLQWLQ time. A variety of mathematical approaches exist with different properties. For example, empirical or statistical models use collected data to create a view of the problem, whereas stochastic models reason about uncertainty in the problem. System Dynamics: A systems-oriented G\QDPLF PRGHOLQJ DSSURDFK ¿UVW SURSRVHG LQ Forrester (1961). Models are based on the causal structure of the problem including the perceptions of the actors. Two levels of modeling are possible: TXDOLWDWLYHPRGHOLQJXVLQJLQÀXHQFHGLDJUDPVDQG causal loop analysis or quantitative modeling using Stock-Flow diagrams and computer simulation.
337
Chapter XXI
Employing the Content Validity Approach for Improving the Content of the Broadband Adoption Survey Instrument Yogesh K. Dwivedi Swansea University, UK Banita Lal Nottingham Trent University, UK Khalil Khoumbati University of Sindh, Pakistan
ABSTRACT Studies on broadband adoption are just beginning to emerge and are exploratory in nature. Progress has been made to develop conceptual models to understand the adoption of broadband from the consumer perspective. However, in order to test the conceptual model of broadband adoption, a reliable survey instrument is yet to be developed and validated. Therefore, the overall aim of this research is to validate content of the broadband adoption survey instrument from the household consumer’s perspective. The REMHFWLYHVRIWKLVFKDSWHUDUH ¿UVWWRHQVXUHWKDWWKHLGHQWL¿HGFRQVWUXFWVDQGWKHLUUHVSHFWLYHLWHPV adequately cover relevant dimensions of factors that affect consumers in the domain of broadband DGRSWLRQ VHFRQGWRGHWHUPLQHZKHWKHUWKHLGHQWL¿HGFRQVWUXFWVDQGWKHLUUHVSHFWLYHLWHPVDGHTXDWHO\ cover relevant dimensions; and third, to conduct a pre- and pilot test on the resulting survey instruPHQWVLQRUGHUWRREWDLQIHHGEDFNOHDGLQJWRLPSURYHPHQWVLQWKH¿QDOTXHVWLRQQDLUH7KHLQLWLDOLWHPV IRUHDFKFRQVWUXFWZHUHLGHQWL¿HGIURPERWKWKHWHFKQRORJ\DGRSWLRQOLWHUDWXUHDQGH[SORUDWRU\VWXGLHV RQEURDGEDQGDGRSWLRQ9DOLGDWLRQRIWKHLGHQWL¿HGLWHPVZDVWKHQSHUIRUPHGHPSOR\LQJDYDULDWLRQ RIDTXDQWLWDWLYHDSSURDFKWRFRQWHQWYDOLGLW\7KH¿QGLQJVREWDLQHGIURPWKHFRQWHQWYDOLGDWLRQDUH then presented and subsequently discussed. Finally, emphasizing the limitations of content validation concludes the chapter.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Employing the Content Validity Approach
INTRODUCTION The widespread availability and diffusion of broadband are considered to be measures of international competitiveness and national economic growth (Oh, Ahn, & Kim, 2003; Sawyer, Allen, & Heejin, 2003). Furthermore broadband is also regarded as an important means for accelerating the growth and deployment of emerging electronic services including e-commerce, e-government, and e-health 2I¿FHRIWKHH(QYR\ ,QRUGHUWRDSSUHFLDWH WKH VRFLRHFRQRPLF EHQH¿WV WKDW EURDGEDQG offers, governments of many countries including the United Kingdom (UK) have established ambitious targets for the deployment and diffusion of broadband services to the consumers and end users (BAG, 2003; Computer Science and Telecommunications Board, 2002; National Broadband Task )RUFH 2I¿FHRI7HFKQRORJ\3ROLF\ 2I¿FHRIWKH((QYR\ The nationwide efforts from the UK government and competition among the Internet service providers (ISPs) have made broadband access widely available at affordable prices (Choudrie & Lee, 2004). However, consumer demand has not increased in line with expectations (Crabtree, 2003; OECD, 2001). This suggests that the current JURZWKDQGGLIIXVLRQRIEURDGEDQGDUHµGHPDQG FRQVWUDLQHG¶DQGQRWµVXSSO\FRQVWUDLQHG¶ &UDEtree, 2003; Haring, Rohlfs, & Shooshan, 2002). The issue of demand constraints has encouraged researchers with a motivation to investigate the following questions: • • • •
Why are consumers slow to adopt broadband? What is the usage of broadband for the household users? What is the impact of broadband usage on household users? How can broadband demand be accelerated?
Therefore, currently, research towards investigating the diffusion of broadband including the adoption, usage, and impacts from the perspective of the household consumer is imperative.
338
Studies on adoption and diffusion of broadband are just beginning to emerge (Oh et al., 2003; Stanton, 2004) and are exploratory in nature (Choudrie & Dwivedi, 2004ab). Progress has been made in developing conceptual models to understand FRQVXPHUV¶DGRSWLRQ 'ZLYHGL &KRXGULH and diffusion (Choudrie & Dwivedi, 2004c) of broadband. However, in order to test the conceptual model of broadband diffusion, a reliable survey instrument has yet to be developed and validated. Validating the data collection instrument is a critical step before testing the conceptual model. This LVEHFDXVHWKHULJRURI¿QGLQJVDQGLQWHUSUHWDWLRQV of positivist, quantitative research is based on solid validation of the instruments that are used to gather the data (Boudreau et al., 2001; Straub, Boudreau, & Gefen, 2004). Therefore, the overall aim of this research is to validate content of the broadband adoption VXUYH\LQVWUXPHQWIURPWKHKRXVHKROGFRQVXPHU¶V perspective. This will be achieved utilizing the approach of content validity and pre- and pilot test on a survey instrument for broadband diffusion UHVHDUFK6SHFL¿FDOO\WKLVFKDSWHUDLPVWRDFKLHYH WKHIROORZLQJWZRREMHFWLYHV ¿UVWWRGHWHUPLQH ZKHWKHUWKHLGHQWL¿HGFRQVWUXFWVDQGWKHLUUHVSHFtive items adequately cover relevant dimensions of the factors affecting consumers in the domain of broadband adoption; and second, to conduct a pre-test and pilot test on the resulting survey instruments in order to determine if the questions are understandable to the participants and to obtain feedback for future improvements. By achieving the set objectives, the contributions of this research are to provide a reliable measure to the academic and practitioner communities who hold a particular interest in the study and management of broadband diffusion from the household consumer perspective. The survey instrument developed in this chapter is expected to provide assistance to practitioners from the telecommunications industry who are interested in determining how to improve their current strategies for increasing consumer base. This can also help policymakers in minimizing the digital divide by understanding the reasons of non-adoption and accelerating the diffusion
Employing the Content Validity Approach
process. This chapter also contributes to theory by FRQ¿UPLQJWKHDSSOLFDWLRQRIWKHFRQWHQWYDOLGLW\ approach in a novel context. Having introduced the topic of interest, this FKDSWHUQRZSURFHHGVWREULHÀ\GHVFULEHWKHEURDGband adoption constructs. The next section offers a brief review of the content validation procedure. We then offer a brief discussion of the research method followed to conduct content validation and DSUHWHVWVWXG\1H[WWKH¿QGLQJVRIWKHFRQWHQW validation and pre-test study are presented and GLVFXVVHGDQG¿QDOO\WKHOLPLWDWLRQVDQGIXUWKHU improvement of this research are offered in the concluding section.
BROADBAND ADOPTION CONSTRUCTS The constructs included in this study were adopted from the model of adoption of technology in households (utilitarian outcomes, hedonic outcomes, knowledge, and skills) (Venkatesh & Browns, 2001), diffusion of innovations (relative advantage) (Rogers, 1995), and the theory of planned behavior EHKDYLRUDOLQWHQWLRQVRFLDOLQÀXHQFHIDFLOLWDWLQJ conditions) (Ajzen, 1991). The proposed model asVXPHGWKDWWKHGHSHQGHQWYDULDEOHµEHKDYLRUDOLQWHQWLRQ¶WRZDUGVEURDGEDQGDGRSWLRQLVLQÀXHQFHG by several independent variables, including attitudinal (relative advantage, utilitarian outcomes, hedonic outcomes, and service quality), normative SULPDU\LQÀXHQFHDQGVHFRQGDU\LQÀXHQFH DQG control factors (knowledge, skills, and facilitating conditions). A detailed discussion on each construct is not possible within the scope of this chapter; a list of constructs is provided in Table 3.
VALIDATION IN I.S. RESEARCH: CONTENT VALIDITY &RQWHQWYDOLGLW\LVGH¿QHGDVWKH³GHJUHHWRZKLFK LWHPVLQDQLQVWUXPHQWUHÀHFWWKHFRQWHQWXQLYHUVH to which the instrument will be generalized” (Straub et al., 2004). In general, content validity
involves evaluation of a new survey instrument in order to ensure that it includes all the items that are essential and eliminates undesirable items to a particular construct domain (Boudreau et al., 2001; .LWFKHQKDP 3ÀHHJHU /HZLV6Q\GHU Rainer, 1995; Straub et al., 2004). Although content validity is a highly desirable and recommended practice that ensures rigor in any empirical research (Straub et al., 2004), its application is limited in IS research (Boudreau et al., 2001; Straub, 1989; Straub et al., 2004). Examples of a few studies that have utilized the content validity approach are summarized in Table 1. However, content validity is not employed in any of the previous studies which focus on broadband diffusion and adoption-related issues. Although only two approachesnamely judgmental and statisticalare available to determine content validity, their application is unique to each study (Emory & Cooper, 1991; Torkzadeh & Dhillon, 2002). The application of content validity differs in terms of when it is utilized, how it is conducted, and how many experts evaluated the content. The judgmental approach to establish content validity involves literature reviews and then follow-ups, with the evaluation aspects being undertaken by expert judges or panels. The validation of the items is based on a high degree of consensus among expert panels who judge the items in question; therefore, it is judgmental in nature (Boudreau HWDO 'DYLV .LWFKHQKDP 3ÀHHJHU 2002; Moore & Benbasat, 1991; Smith, Milberg, & Burke, 1996; Storey, Straub, Stewart, & Welke, 2000; Torkzadeh & Dhillon, 2002; Straub et al., 2004). An empirical or quantitative approach was ¿UVWLQWURGXFHGE\/DZVKH ZKLFKLQYROYHV estimating the statistical validity ratio (see also Lewis et al., 1995). In order to perform content validity for broadband diffusion research, a quantitative approach was considered to be more suitable in comparison to a judgmental approach (Lawshe, 1975; Lewis et al., 1995). In research terms, broadband diffusion studies are still emerging, academic experts are few in numbers and located in different places, therefore a quantitative approach was considered
339
Employing the Content Validity Approach
Table 1. Studies within information systems that utilized the content validity approach
Author & Year
Stage Prior/
Subject of the Study
Approach
Who Were the Experts
Technology acceptance
Judgmental
Computer users: 5 secretaries, 5
During Davis (1989)
Prior
by users Moore & Benbasat I
Prior
(1991)
IT innovation
students, 5 professional staffs Judgmental
1RWVSHFL¿HG
acceptance
Kappelman (1995)
Adopted
User involvement
Judgmental
1RWVSHFL¿HG
Lewis et al. (1995)
Prior
Information resource
Quantitative using
13 MIS professionals from both
management
CVR
industry and academia 3 expert and 15 doctoral students
Smith et al. (1996)
Prior
Information privacy
Judgmental
Storey et al. (2000)
Prior
E-commerce
Judgmental
40 academics studying eFRPPHUFH QRWVSHFL¿HGZKRWKH\ are?
Torkzadeh & Dhillon
Prior
(2002) Kim, Lee, Han, &
Success factors of
Judgmental
Internet commerce
2 experts and 12 Internet online shopping consumers
Prior
Internet business
Interviews
45 experts
Prior
IT acceptance
Judgmental
2 IT professionals
Prior
Broadband diffusion
Quantitative using
First stage: 12 experts
CVR
Second stage: 20 experts and
Lee (2002) Gefen & Ridings (2003) Current Study
consumers
to be most appropriate. The next section offers a description of the pursued research method.
RESEARCH METHOD Content validity of the broadband diffusion instruments was performed employing a quantitative approach (Lawshe, 1975). In terms of IS research, such an approach has been successfully applied when validating information resource management instruments (Lewis et al., 1995). In order to validate the content of constructs, the quantitative approach was undertaken in the following manner. First, relevant items from existing literature were LGHQWL¿HG7KLVOHGWRWKHFRQVWUXFWLRQRIWKHTXHVtions and content validity questionnaire. Second, a content evaluation panel that consisted of experts
340
from academia and/or industry who were related to the desired research area was selected. Third, each member of the panel was then provided with the questionnaire formed in step 1. The panel members were requested to respond independently to each item in relation to a particular construct on a three-point scale where: 1 = not necessary, 2 = useful but not essential, and 3 = essential. Fourth, responses from all the panelists were then pooled. This step also included counting responses that LQGLFDWHGµHVVHQWLDO¶IRUHDFKLWHP)LIWKWKH content validity ratio (CVR) for each item was estimated utilizing the formula: CVR=(n-N/2)/(N/2) (Lawshe, 1975); N is the total number of respondents, and n is the frequency count of the number of panelists rating the item as 3 = essential. Finally, the CVR values obtained for each item were examined for WKHLU VLJQL¿FDQFH HPSOR\LQJ WKH VWDQGDUG WDEOH
Employing the Content Validity Approach
provided by Lawshe (1975). If the estimated CVR value was equal to or above the standard value, then the item was accepted, otherwise it was eliminated. 7KH VLJQL¿FDQFH OHYHO VWDQGDUG YDOXH GHSHQGV upon the number of experts rating the item. The minimum number of experts required to rate each LWHPVKRXOGEH¿YH7KHYDOXHRI&95UDQJHGIURP 0 to 1 (Lawshe, 1975; Lewis et al., 1995). The aforementioned steps were followed to evaluate the content of broadband adoption survey instruments. A sample of items for each construct ZDVLGHQWL¿HGHPSOR\LQJDQH[KDXVWLYHUHYLHZRI literature on technology adoption in general, and EURDGEDQGDGRSWLRQDQGGLIIXVLRQVSHFL¿FDOO\7KH OLWHUDWXUHUHYLHZLGHQWL¿HGDWRWDORI LWHPVIRU 15 adoption-related constructs. These 95 items included those obtained from the exploratory survey conducted in stage 1. A content validity questionnaire was then generated that comprised GH¿QLWLRQVRIWKHFRQVWUXFWVDQGDVVRFLDWHGLWHPV on a 1-3 scale. A total of 12 academic experts ZHUH LGHQWL¿HG RQ WKH EDVLV RI SXEOLFDWLRQV LQ peer-reviewed journals and leading telecommunication and broadband conferences (10 experts) or their engagement (2 experts) with the research area related to broadband adoption and diffusion. 7KHLGHQWL¿HGH[SHUWVZHUHWKHQDSSURDFKHGDQG subsequently the content validity questionnaire was sent to them using e-mail. The purpose of the study and instructions to complete the questionnaire were detailed in a covering e-mail. The experts were requested to rate each item in relation to different constructs of broadband diffusion on a three-point scale: 1 = not necessary, 2 = useful but not essential, and 3 = essential. They were also requested to provide comments if the items were not understandable, if the items needed to be reworded, or if new items needed to be added. Then, responses from all the experts were collated by counting the number of ratings that indicated “essential” for each item. Following that, the CVR was estimated and evaluated for a VWDWLVWLFDOVLJQL¿FDQFHOHYHORIE\HPSOR\LQJ /DZVKH¶V PHWKRGPHQWLRQHGDERYH7KLV process was undertaken for each item. For items ZKHUH VLJQL¿FDQFH OHYHO DW OHYHO ZDV QRW achieved, elimination occurred. The list of items
along with their associated CVR values is presented and discussed in Tables 2 and 3.
Pre-Test A pre-test of the resulting instrument was conducted with participants from the broadband industry (3), an IT manager of a county council (1), academics and researchers (10), and household consumers (6). The respondents were asked to judge whether they could understand the wording of the questions and to suggest future improvements.
Pilot Test 3ULRUWRGLVVHPLQDWLRQRIWKH¿QDOTXHVWLRQQDLUH a pilot study was conducted to determine the response rate and learn of any discrepancies within the questions, which included determining whether the format of the questionnaire and questions were suitable. Additionally, the duration that completion of the questionnaire would require was also established. The questions of the pilot questionnaire were divided into three categories: (1) multiplechoice questions examining the demographics of the respondents, Internet connection types, and frequency and duration of Internet access on a daily basis; (2) yes/no questions that determined the location of the Internet at home or elsewhere; and (3) Likert-scale questions to assess the perception of the adopters and non-adopters of broadband. Four other questions were also asked in order to GHWHUPLQH D UHVSRQGHQW¶V XQGHUVWDQGLQJ RI WKH questionnaire. The four questions were: (1) Is the length of the questionnaire appropriate? (2) Are the questions understandable? (3) Is the layout of the questionnaire acceptable? (4) How long did it take the respondent to complete the questionnaire? The responses of the respondents on these four questions are summarized in Table 4. The pilot questionnaire was delivered via the postal service to a total of 200 randomly selected participants from “UK-Info Disk V11” in December 2004. A total of 40 usable replies were obtained from the UHVSRQGHQWVZLWKLQWKHVSHFL¿HGGXUDWLRQ
341
Employing the Content Validity Approach
FINDINGS AND DISCUSSIONS Content Validity A brief summary of the CVR is provided in Tables DQG7KH¿QGLQJVSUHVHQWHGLQWKHWDEOHVLOlustrate that of the 95 items from the adoption domain, the majority of respondents considered LWHPV HVVHQWLDO WR LQFOXGH LQ WKH ¿QDO TXHVtionnaire. This was due to the CVR value being VLJQL¿FDQWDWWKHOHYHO7DEOHLOOXVWUDWHVWKH overall items, average CVR, and average mean for each construct. The average CVR value for all 10 constructs fell between the maximum value of 0.83 and minimum value of 0.56 at the 0.05 level RIVWDWLVWLFDOVLJQL¿FDQFH 7DEOH 7KLVLOOXVWUDWHV that the constructs possess a high level of content validity, which means that the items are representative of construct universe (Table 3). The experts also provided a number of suggestions regarding rewording of the items and addition of three new items for utilitarian outcomes and one new item for hedonic outcomes. These suggestions were incorporated during the pre-test
questionnaire design stage. However due to the word or page limitations, the discussion on the improvements is not provided in this chapter. The H[SHUWVDOVRDJUHHGWKDWIRUWKH¿QDOTXHVWLRQQDLUH the seven-point scale would be more suitable in FRPSDULVRQWRWKH¿YHSRLQWVFDOH7KLVLVEHFDXVH the seven-point scale values are widely spread in FRPSDULVRQ WR ¿YHSRLQW VFDOH DQG UHVSRQGHQWV would have more choices to select. This prevents a UHVSRQGHQW¶VELDVE\MXVWVHOHFWLQJDQHXWUDOYDOXH Therefore, seven-point is considered to be the most VXLWDEOH/LNHUWVFDOHIRUWKH¿QDOVWXG\ The experts who evaluated the content of the instrument belonged to several countries, namely, the UK, Denmark, the United States, Australia, and Canada, making the content of the questionnaire not only valid to the UK, but to the aforementioned countries as well. Therefore, in order to conduct a comparative study, it may be possible to pre-test and further validate the questionnaire in the context of a number of countries such as the United States, Australia, Canada, and EU member states. 7KH¿QGLQJVDOVRVXJJHVWWKDWWKHFRQWHQWYDOLGity experts rated essential mainly those items that
Table 2. Summary of content validity ratio CVR
Total Items
0.90-0.99
5
0.80-0.89
10
0.70-0.79
0
0.60-0.69
10
0.50-0.59
13
0.40-0.49*
0
0.30-0.39*
19
0.20-0.29*
0
0.10-0.19*
12
0-0.09*
13
Total
82
RLH
13
Grand Total
95
QRWVLJQL¿FDQW5/+ LWHPVUDWHGHVVHQWLDOE\OHVVWKDQKDOIRIWKHSDUWLFLSDQWV
342
Employing the Content Validity Approach
7DEOH6XPPDU\RIFRQVWUXFWVWRWDOLWHPVVLJQL¿FDQWLWHPV$&95DQG$0 Constructs
TI
SI
ACVR
AM
Behavioral Intention
3
2
0.83
2.83
Relative Advantage
9
4
0.61
2.79
Utilitarian Outcomes
14
7
0.69
2.88
Hedonic Outcomes
4
3
0.78
2.89
Service Quality
13
5
0.73
2.80
3ULPDU\,QÀXHQFHV
4
3
0.57
2.78
6HFRQGDU\,QÀXHQFHV
4
2
0.75
2.88
Requisite Knowledge
6
3
0.61
2.69
Skills
7
3
0.56
2.61
Facilitating Conditions
9
5
0.70
2.78
Total
95
38
---
---
7, WRWDOQXPEHURILWHPV6, QXPEHURIVLJQL¿FDQWLWHPV$&95 DYHUDJHFRQWHQWYDOLGLW\UDWLR AM = average mean
were adopted from the previous exploratory studies on broadband adoption (Choudrie & Dwivedi, 2004; Oh et al., 2004). Contrastingly, the items adopted from the general technology adoption studies (Davis, 1989) were mostly rated but not considered HVVHQWLDO7KHUHIRUHWKLVVWXG\FRQ¿UPVWKDWWKH items investigated in the exploratory studies are LPSRUWDQWWRXQGHUVWDQGWKHFRQVXPHUV¶EURDGEDQG adoption and usage behavior.
2.
Pre-Test
1.
The pre-test participants suggested that all the items for one construct should be measured in the same direction. The content validity experts also made this comment. Therefore, this particular issue was carefully considered and such changes were made wherever required. The following is an example of such a change:
2.
The items for the facilitating condition construct ‘before pre-test’ 1.
I cannot subscribe to broadband at home because it is too costly to purchase a new computer or to upgrade my old computer
3. 4.
My annual household income level is enough to afford subscribing to broadband It is too costly for me to subscribe to broadband at its current subscription fee I would be able to subscribe to broadband if I wanted to
The items for facilitating condition construct ‘after pre-test’
3. 4.
It is not too costly to purchase a new computer or to upgrade my old computer My annual household income level is enough to afford subscribing to broadband It is not too costly for me to subscribe to broadband at its current subscription fee I would be able to subscribe to broadband if I wanted to
Another issue that the respondents commented on was the length of the questionnaire. The questionnaire sent to the pre-test participants was 11 pages long. The respondents expressed concern about the length of the questionnaire. They suggested that in the current form, the questionnaire length was extensive and may lead to a low response
343
Employing the Content Validity Approach
rate. Therefore the length of the questionnaire should be reduced without losing the measurement content. Bearing the suggestion of the pre-test participants, the total length of the questionnaire ZDVUHGXFHGWR¿YHSDJHV7KLVZDVDFKLHYHGE\ making the following change in the structure and format of the questions. In the pre-test questionnaire, the options for Likert-scale questions were arranged in two rows, but after undertaking the change, for each item the Likert-scale options were managed within one row. This helped to save a lot of space. For example:
and pre-test participants, the length of the resultant (pilot) questionnaire was reduced to six pages. More GHWDLOVDERXWWKHFRQWHQWVRIWKH¿QDOTXHVWLRQQDLUH are provided in the next sub-section.
Pilot Test The majority of the respondents reported that the questionnaire was easily understandable and required 10 to 15 minutes for completion (Table 4). Additionally, the majority of the respondents
Arrangement of the Likert Scale items in the pre-test questionnaire Broadband has an advantage over dial-up/narrowband because it offers faster access to Internet
7 = Extremely agree
5 = Slightly agree
3 = Slightly disagree
6 = Quite agree
4 = Neutral
2 = Quite disagree 1 =Extremely disagree
Arrangement of the Likert Scale items in the ¿QDOTXHVWLRQQDLUH Broadband has an advantage over dial-up/narrowband because it offers faster access to Internet Extremely disagree
1
2
3
4
5
6
7
Extremely agree
Apart from the aforementioned change and a few spelling and typographical errors, the respondents from the pre-test studies supported the content of the questionnaire. After incorporating all the suggested changes by the content validity experts
344
validated the content of the questionnaires, alWKRXJKPLQRUFKDQJHVWRWKH¿QDOGHVLJQRIWKH questionnaire were undertaken based on the UHFHLYHGIHHGEDFNDQGD¿QDOTXHVWLRQQDLUHZDV developed. 7KHSXUSRVHRIWKHSLORWWHVWZDVDOVRWRFRQ¿UP WKHUHOLDELOLW\RIWKHLWHPV7KH¿QGLQJVREWDLQHG
Employing the Content Validity Approach
Table 4. Respondent perception of survey instrument (N = 40) Questions
Frequency
Percent
Yes
No
Yes
No
1. Is the length of the questionnaire appropriate?
35
5
87.5
12.5
2. Are the questions understandable?
39
1
97.5
2.5
3. Is the layout of the questionnaire OK?
37
3
92.5
7.5
Time required to complete questionnaire 4. How long did it take to complete the questionnaire?
10 Min.
15 Min.
20 Min.
25 Min
Percent
37.5
30.0
30.0
2.5
Table 5. Summary of statistics obtained from pilot test (N = 40) Constructs
Number of Items
Scale
Scale
Mean
SD
RELIABILITY Cronbach’s D
Behavioral Intention
2
5.5
1.7
0.95
Relative Advantage
4
6.3
0.7
0.75
Utilitarian Outcomes
10
5.7
1.0
0.91
Hedonic Outcomes
4
3.7
1.7
0.88
Service Quality
4
4.2
1.3
0.78
3ULPDU\,QÀXHQFH
3
4.5
1.6
0.88
6HFRQGDU\,QÀXHQFH
2
3.6
1.9
0.95
Facilitating Conditions Resources
4
5.0
1.2
0.72
Knowledge
3
5.7
1.3
0.85
6HOI(I¿FDF\
3
6.2
1.2
0.94
from the pilot test demonstrated an acceptable level of reliability for all the constructs (see Table 5).
CONCLUSION 7KLVUHVHDUFKLVDQLQLWLDOVWHSWRZDUGVDFRQ¿UPDtory study that examines broadband diffusion in the household context. Further, this study validated the contents of a survey instrument utilizing a quantitative approach. It also pre- and pilot tested the survey questionnaire which illustrated that instrument measures were internally consistent. $OWKRXJKWKLVLVDQLQLWLDOVWHSWRZDUGVWKHFRQ¿Umatory study, the chapter contributes to both theory and practice. A theoretical contribution is offered E\FRQ¿UPLQJWKHDSSOLFDWLRQRIWKHFRQWHQWYDOLG-
ity approach in the broadband adoption context. The survey instrument developed and validated in this chapter will contribute to practice by assisting professionals from the telecommunications LQGXVWU\7KLVZLOOEHIXO¿OOHGE\WKLVUHVHDUFKE\ demonstrating how an improvement to current services and the consumer base can be obtained. This can also provide assistance to policymakers by minimizing the digital divide. This will be achieved by understanding the reasons of nonadoption and adoption.
Limitations The following three limitations were encountered while conducting content validity for the broadband adoption survey instruments. These are:
345
Employing the Content Validity Approach
¿UVW ORFDWLQJ WKH H[SHUWV UHODWHG WR WKH VSHFL¿F research area; second, conducting content validity with experts located at different places; and third, the length of content validity instruments. Since research in broadband adoption and diffusion from the demand perspective is novel, researchers involved in examining this issue are minimal, especially in one location or country. This formed the initial problem for locating the experts. This problem was overcome by considering researchers located in other countries such as the United States, Australia, Canada, and Denmark. The second limitation was due to the distance and lack of face-to-face interaction with the majority of the researchers who perceived initially that the TXHVWLRQQDLUHVHQWWRWKHPLVWKH¿QDOLQVWUXPHQW This created confusion while evaluating the content of the questionnaire. To overcome this problem, several e-mails were sent to each expert to answer their queries and clarify the context of content validity. Also, since the content validity questionnaire FRPSULVHGDGH¿QLWLRQRIHDFKFRQVWUXFWDQGUHODWHG items, it increased the length of the instrument. For example, the content validity instruments in the pre-test study were 18 pages long. That discouraged many experts from participating in the content evaluation in the initial instance. However, after repeated contacts and requests, the researchers agreed to assist this research.
REFERENCES
Proceedings of the 12th European Conference of Information Systems (ECIS), Turku, Finland. Choudrie, J., & Dwivedi, Y.K. (2004b). Investigating the socio-economic characteristics of residential consumers of broadband in the UK. Proceedings of the American Conference on Information Systems (AMCIS), New York. Choudrie, J., & Dwivedi, Y.K. (2004c). Towards a conceptual model of broadband diffusion. Journal of Computing and Information Technology, 12(4), 323-338. Choudrie, J., & Lee, H.J. (2004). Broadband development in South Korea: Institutional and cultural factors. European Journal of Information Systems, 13(2), 103-114. Computer Science and Telecommunications Board. (2002). Broadband: Bringing home the bits. Washington, DC: National Academy Press. Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. Dwivedi, Y.K., & Choudrie, J. (2004). Developing a model of broadband adoption in the household. Proceedings of the ETHICOMP Conference, Syros, Greece. Gefen, D., & Ridings, C.M. (2003). IT acceptance: Managing user–IT group boundaries. The DATA BASE for Advances in Information Systems, 34(3), 25-40.
Ajzen, I. (1991). The theory of planned behaviour. Organisational Behaviour and Human Decision Processes, 50(2), 179-211.
Haring, J., Rohlfs, J., & Shooshan, H. (2002). Propelling the broadband bandwagon. MD: Strategic Policy Research.
B.A.G. (2003). Australia’s broadband connectivity. Retrieved September 20, 2004, from http://www. noie.gov.au/publications/NOIE/BAG/report/index. htm
Horrigan, J.B., & Rainie, L. (2001). Online communities: Networks that nurture long-distance relationship and local ties. Retrieved March 11, 2004, from http://www.pewinternet.org/pdfs/ PIP_Communities_Report.pdf
Crabtree, J. (2003). Fat Pipes, connected peoplerethinking broadband Britain. iSOCIETY Report, London. Choudrie, J., & Dwivedi, Y.K. (2004a). Analysing the factor of broadband adoption in the household.
346
Kappelman, L.A. (1995). Measuring user involvement: a diffusion of innovation perspective. The DATA BASE for Advances in Information Systems, 26(2&3), 65-86.
Employing the Content Validity Approach
Kim, J., Lee, J., Han, K., & Lee, M. (2002). Businesses as buildings: Metrics for the architectural quality of Internet businesses. Information Systems Research, 13(3), 239-254.
Smith, H.J., Milberg, S.J., & Burke, S.J. (1996). ,QIRUPDWLRQSULYDF\ 0HDVXULQJLQGLYLGXDOV¶FRQcern about organizational practices. MIS Quarterly, 20(2), 167-197.
.LWFKHQKDP% 3ÀHHJHU6/ 3ULQFLSOHV of survey research part 4: Questionnaire evaluation. Software Engineering Notes, 27(3), 20-23.
Shih, C.F., & Venkatesh, A. (2004). Beyond adoption: Development and application of a use-diffusion model. Journal of Marketing, 68, 59-72.
Lawshe, C.H. (1975). A quantitative approach to content validity. Personnel Psychology, 28, 563-575.
Straub, D., Boudreau, M.-C., & Gefen, D. (2001). Validation guidelines for IS positivist research. MIS Quarterly, 25(1), 1-23.
Lewis, B.R., Snyder, C.A., & Rainer, K.R. Jr. (1995). An empirical assessment of the information resources management construct. Journal of Management Information Systems, 12(1), 199-223.
Straub, D., Boudreau, M.-C., & Gefen, D. (2004). Validation guidelines for IS positivist research. Communications of the Association for Information Systems, 13, 380-427.
Moore, G.C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, 2(3), 192-212.
Straub, D. (1989). Validating instruments in MIS research. MIS Quarterly, 13(2), 147-169.
National Broadband Task Force. (2001). The new national dream: Networking the nation for broadband access. Canada: Ottawa Industry. OECD. (2001). Working party on telecommunication and information services policies: The development of broadband access in OECD countries. Paris. 2I¿FH RI WKH ((QYR\ UK online: The broadband future. 2I¿FHRI7HFKQRORJ\3ROLF\ Understanding broadband demand: A review of critical issues. Washington, DC: U.S. Department of Commerce. Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitude. Journal of Information Technology, 18, 267-280. Rogers, E.M. (1995). Diffusion of innovations. New York: Free Press. Sawyer, S., Allen, J.P., & Heejin, L., (2003). Broadband and mobile opportunities: A socio-technical perspective. Journal of Information Technology, 18(2), 121-136.
6WDQWRQ/- )DFWRUVLQÀXHQFLQJWKHDGRStion of residential broadband connections to Internet. Proceedings of the 37th Hawaii International Conference on System Sciences, Hawaii. Storey, V.C., Straub, D.W., Stewart, K.A., & Welke, R.J. (2000). A conceptual investigation of the ecommerce industry. Communications of the ACM, 43(7), 117-123. Torkzadeh, G., & Dhillon, G. (2002). MeasurLQJIDFWRUVWKDWLQÀXHQFHWKHVXFFHVVRI,QWHUQHW commerce. Information Systems Research, 13(2), 187-204. Venkatesh, V., & Brown, S. (2001). A longitudinal investigation of personal computers in homes: Adoption determinants and emerging challenges. MIS Quarterly, 25(1), 71-102.
KEY TERMS Behavioral Intention: $FRQVXPHU¶VLQWHQWLRQ to subscribe (or intention to continue the current subscription) and make use of broadband Internet in the future (Ajzen, 1991). Content Validity: The “degree to which items LQ DQ LQVWUXPHQW UHÀHFW WKH FRQWHQW XQLYHUVH WR
347
Employing the Content Validity Approach
which the instrument will be generalized” (Straub et al., 2004). Facilitating Conditions Resources: The perceived level of resources when subscribing to broadband (Ajzen, 1991). Hedonic Outcomes: The pleasure derived from the consumption or use of broadband Internetfor example, the entertainment potential of the Internet via offerings such as online radio, streaming audio and video, electronic greetings, online games, and online casino (Venkatesh & Brown, 2001). Knowledge: The perceived level of knowledge DERXW EURDGEDQG ,QWHUQHW LWV ULVNV DQG EHQH¿WV (Venkatesh & Brown, 2001). 3ULPDU\,QÀXHQFH 7KHSHUFHLYHGLQÀXHQFH from friends and family to subscribe to and use (or not to subscribe and use) broadband Internet services (Venkatesh & Brown, 2001).
348
Relative Advantage: The degree to which broadband Internet is perceived as being better than its predecessor, narrowband Internet (Rogers, 1995). 6HFRQGDU\,QÀXHQFH 7KHSHUFHLYHGLQÀXHQFH of information from secondary sources such as ads and news on TV, as well as newspapers, to subscribe to and use (or not to subscribe to and use) broadband Internet services (Venkatesh & Brown, 2001). Utilitarian Outcomes: The extent to which broadband Internet usage enhances the effectiveness of household activities such as undertaking RI¿FHZRUNDWKRPHFKLOGUHQ¶VKRPHZRUNLQIRUmation or product search and purchase, and home business (Venkatesh & Brown, 2001).
349
Chapter XXII
Inside the Microcosm: A Case Study of a Wireless Internet Hotspot Pierre Vialle GET-INT, France Olivier Epinette GET-INT, France Olivier Segard GET-INT, France
ABSTRACT The objective of this chapter is to highlight critical elements affecting the diffusion of broadband wireless Internet at a hotspot location, through a case study. The research deals with a wireless Internet services project in the main Paris airports, and comprises two components. First, we analyze the emergence of DYDOXHFKDLQIRUDQHZVHUYLFHEDVHGRQWKHFRRSHUDWLRQRIVHYHUDO¿UPV,QSDUWLFXODUZHVKRZKRZ different actors can or cannot position themselves on this value chain, according to their resources and capabilities. Second, we explore the perceptions and attitudes of business passengers in order to better understand the potential adoption and use of hotspot services, and provide a preliminary framework of analysis. The research is drawn from a qualitative survey, made up of in-depth interviews of potential suppliers (airline companies, service providers, airport managers) and business passengers.
INTRODUCTION In this chapter we explore micro-level factors affecting the adoption and diffusion of broadband data transmission at wireless Internet hotspots. The number of commercial hotspots enjoys a worldwide popularity and was expected to reach a total number 143,000 by the end of 2006, showing a growth of
47% compared to 2005. To our knowledge, little academic-oriented research addresses the topic of public hotspots; it tends to focus on market, implementation, and business model considerations (i.e., Chang, Yu, & Tsai, 2006; Vandepeutte, 2005). Valuable insights for our project have been provided by Thorngren (2002), who has analyzed various business models of WLAN hotspots in
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Inside the Microcosm
airports. The key issue in the adoption and usage RIZLUHOHVV,QWHUQHWRXWVLGHKRPHVDQGRI¿FHVLV the attractiveness of each hotspot. This is a more complex situation than in the case of broadband access in the home, where demand can be seen as responding at an aggregate level to factors such as availability of infrastructure and service, price level, or penetration rate of home PCs factors which in turn may depend on market characteristics and regulatory policy. Diffusion here depends on a mix of “general decisions” to adopt and use wireless Internet in a situation of mobility, and of micro decisions to adopt it and/or use it in a VSHFL¿F KRWVSRW GHSHQGLQJ RQ VSHFL¿F FRQGLtions. Moreover, some hotspots, as in the case of DLUSRUWVPD\GHOLYHUVSHFL¿FFRQWHQWDQGVHUYLFHV in addition to the plain Internet and e-mail access, and decide to bill for service provision. Therefore, all the issues dealt with in the case of large-scale service provision are also encountered at a lower VFDOHLQYHU\VSHFL¿FFRQVXPSWLRQVLWXDWLRQV,WLV essentially a question of designing and implementing networks and services, of identifying potential users, and ensuring that services provision meets the expectations and usage constraints of target customers. This chapter is based on a research project carried out from January to December 2003, and ¿QDQFHGE\WKH/RXLV/H3ULQFH5LQJXHW)RXQdation. It highlights critical aspects in providing broadband access with what was a new technology at the time, through the analysis of a case study. The objective of this project was to draw lessons from the launch of a new telecommunications service. The partnership between ADP Télécom and the research team was based on mutual interest. The researchers had the opportunity to gather information on several aspects of the project during its successive phases, while ADP Télécom EHQH¿WHGIURPWKHVXSSRUWRIWKHUHVHDUFKWHDP in particular concerning the marketing component of the project. The research deals with a wireless Internet services project in the main Paris airports, the provision of which began in the year 2003, and comprises two components. First, we analyze the emergence of a value chain for a new service and show how different actors can or cannot position
350
themselves on this value chain, according to their resources and capabilities. Second, we explore the perceptions and attitudes of business passengers in order to better understand the potential adoption and use of hotspot services, and provide a preliminary framework of analysis. The research is drawn from a qualitative survey, made up of in-depth interviews of potential suppliers (airline companies, service providers, airport managers) and business passengers.
BACKGROUND INFORMATION Wireless Internet Hotspots $KRWVSRWLVGH¿QHGDVDJHRJUDSKLFDUHDWKDWLVFRYered by a wireless network. Inside this area, Internet access is made available to any device equipped with a wireless card. Hotspots are located in public venues such as coffee shops, hotels, railway stations, or airports. Most industry experts forecast that airports are promising hotspots for wireless Internet because of the time the passengers spend checking in and waiting before boarding. Our study relates to a hotspot service in the main airports of Paris, using wireless local area network (WLAN) technology. WLAN is a short-range, broadband data communication system using radio waves. Various standards coexist, but the 802.11b VWDQGDUGGH¿QHGE\WKHInstitute of Electrical and Electronics Engineers (IEEE), tends to be domiQDQW,WDOORZVDWKHRUHWLFDOÀRZRI0EVLQWKH frequency band of the 2.4 GHz, on a distance of up to several hundred meters. The IEEE 802.11b standard is better known under the name of WiFi. The deployment of a wireless Internet network is relatively easy, and it offers a much higher level RIÀH[LELOLW\WKDQ¿[HGQHWZRUNV7KHQXPEHURI accesses can be increased without needing to redimension the network, as would be the case with ¿[HGQHWZRUNV7KLVV\VWHPDOORZVVRPHOLPLWHG mobility inside one cell, but contrary to cellular networks such as GSM networks, does not allow roaming between cells without service interruption. Such networks can be accessed by different types of terminals equipped with wireless Internet chips,
Inside the Microcosm
such as mobile phones, personal digital assistants (PDAs), or laptop computers. There are three possible uses of wireless InWHUQHW7KH¿UVWLVSULYDWHXVHHLWKHUIRUDKRPH network or for a company local area network (LAN). The second is community networks, usually in rural areas. The third is hotspots in public venues such as coffee shops, hotels, railway stations, or airports. A wireless Internet network can give access to the following services: •
•
•
Internet DFFHVVLQFOXGLQJVXU¿QJPHVVDJLQJ and virtual private network services allowing access to an intranet Localized information services, that is, reception of “push” information, such as infrastructure usage information and schedules (transport hotspots), or commercial information (for example, from nearby shops) Multimedia services, such as movies or musicas the available bandwidth is shared between current users, a large number of users can result in unacceptable download time
As users of a particular hotspot may be customers of another hotspot, or of a mobile operator, the establishment of roaming agreements is necessary. A roaming agreement essentially ensures that a customer of network A is able to use the service provided by network B, while being billed by network A, with some mechanism of revenue sharing between the two networks. Therefore, users can pay for service usage in three different ways. They can be billed by the hotspot manager, either directly through payment by credit card, or indirectly with SUHSDLGFDUGVWKDWFDQEHERXJKWLQWKHDLUSRUW¶V newspaper shops. They can also be billed on their mobile phone account, as most mobile operators offer subscription schemes bundling 2.5G (i.e., GPRS) cellular services or 3G (i.e., UMTS) services with wireless Internet services.
Case Study Context An airport covers different activities and geographic zones. Several activities can be carried
out on an airport platform, such as transport, control, cleaning, maintenance, or passenger care. Geographically, these activities broadly relate to three zones on the airport area: tarmac (services concerning airplanes), air terminals (services for passengers on site), and airport periphery (activities EHQH¿WLQJIURPFORVHORFDWLRQWRWKHDLUSRUWVXFK DV KRWHOV RU RI¿FHV )LQDOO\ WKUHH FDWHJRULHV RI people can be found on an airport site: employees of businesses and administrations located on the airport platform, passengers, and employees from RXWVLGH¿UPVRQVSHFL¿FPLVVLRQVDWWKHDLUSRUW (for example, for repair activities). In this study, we IRFXVRQEXVLQHVVWUDYHOHUVWKDWKDYHEHHQLGHQWL¿HG as the main target market by ADP, the organization managing the various airports around Paris. Several airport actors are concerned by this project with different expectations according to their characteristics, such as the owner of the venue, the wireless Internet service provider (WISP), the businesses in the airport, and the consumers. The ADP (Aéroports de Paris) group is a public utility whose mission is the management of civil airports within a 50-kilometer radius of Paris. It is responsible for the management of the facilities enabling the arrival and the departure of passengers, goods, and airmail. It manages service activities and assists airline companies. It grants concessions to its customers (airline companies, commercial companies, or administrations). ADP manages one of the most important airports in continental Europe with approximately 70 million travelers each year. Its revenues amounted to 1.356 million ¼LQ2QHRIWKHPDLQFKDOOHQJHVRIDLUSRUW management is to ensure maximum security while giving a sense of freedom to passengers. The reasons evoked for the wireless Internet project were to contribute to the image of ADP (among the general public, the airport-related businesses, and political instances), to provide a supplementary service to support the core business of airport management, and additionally to provide some extra revenues on a longer term. Telecommunications in the airport zone are managed by a subsidiary company created in 2001, ADP Télécom. Its mission is to design, operate, and market telecommunications networks and services,
351
Inside the Microcosm
mainly on the airport sites of Roissy-Charles De Gaulle and Orly. ADP Télécom was chosen as the WISP by ADP. For ADP Télécom, this project was an occasion to extend its current capabilities to a new technology, to reach scope economies with other projects, and to gain expertise that could be used in other markets. 2WKHUDLUSRUWDFWRUVFDQEHFODVVL¿HGLQWRWZR types according to their expectations concerning wireless Internet services. On the one hand, the WLAN service only constitutes an enhancement of the basic local services offered by companies such as the airport authorities, shops, and restaurants. These companies are more interested in enhancing the service and thus they try to offer suitable content. On the other hand, companies such as airline companies, car rentals, and hotels are focused on enhancing customer relationships. Therefore, offering a continuous range of valueadded services through several media is their most important expectation. The difference between the motivations of the actors and the services suggested at the Paris DLUSRUWVKRWVSRWVLVVLJQL¿FDQW VHH7DEOH DQG can be a hindering factor for the development of new services.
On-site companies do not constitute a privileged WDUJHWIRUD:/$1RIIHU7KH\KDYHVSHFL¿FWHOHcommunications requirements (e.g., communications with a plane), and are extremely demanding in terms of quality and reliability. As their needs DUHDOUHDG\VDWLV¿HGE\FXUUHQWVHUYLFHVWKH\DUH not likely to change for WLAN services. The main target would be transit passengers and passengers GHSDUWLQJIRUORQJÀLJKWVLQSDUWLFXODUVHQLRUH[HFXWLYHVDQG³IUHTXHQWÀ\HUV´RIWHQHTXLSSHGZLWK computer equipment. They broadly account for one-third of the passengers in the target airports.
THE EMERGENCE OF A NEW VALUE CHAIN In this section we analyze the emergence of a network of actors in order to design and implement a new value chain necessary to provide wireless Internet services in airports. In particular, we identify the critical resources that are necessary IRU¿UPVWRSRVLWLRQWKHPVHOYHVRQWKHYDOXHFKDLQ of activities.
Table 1. Main actors’ motivations for wireless Internet services ADP
ADP Télécom
Companies Willing to Reach
Companies Willing to Provide
Airport Manager and Venue
Subsidiary of ADP
Customers by Providing Local
Value-Added Services (i.e.,
Owner
Content (i.e., shops, restaurants) hotel chains, car rental companies)
•
Support on-site mobility
•
of employees and passengers •
Complement source of income
•
Stand out from competitors
•
352
Improve public image
M a ke u s e of c u r r e nt •
Stimulate consumption by
•
Answer customer requests
capabilities
promoting products and
•
Increase relational
•
Gain economies of scale
•
Gain expertise in order to be able to penetrate new markets
services • •
Provide transaction
marketing with customers •
Differentiate from
support
competitors with value-
Improve public image
added services
Inside the Microcosm
Conceptual Background In order to analyze this case, we have mobilized two complementary streams of research: the Resource-Based View (RBV) and the Industrial Network Approach (INA). The latter proposes a conceptualization of industrial systems as networks of relationships between interdependent ¿UPVZKLOHWKHIRUPHUSXWVWKHHPSKDVLVRQWKH resources and capabilities contributing to competitive advantage. The Resource-Based View, initialized by Barney (1991), Grant (1991), and Wernerfelt (1984), puts the emphasis on resources and capabilities as main sources of competitive advantage. Resources and capabilities are bundles of tangible and intangible assets, and can include management skills, organizational processes and routines, and information and knowledge. They are distributed KHWHURJHQHRXVO\DPRQJ¿UPVDQGFDQQRWEHWUDQVIHUUHGIURP¿UPWR¿UPZLWKRXWFRVW0DUNHWLQJ scholars (e.g., Day, 1994; Srivastava, Fahey, & Christensen, 2001) have stressed the importance of market-based assets, among which relational assets play a key role in the current competitive environment. The basic argument of RBV is that when resources are rare and valuable (in terms of ¿UPHI¿FLHQF\RUHIIHFWLYHQHVV WKH\FDQEHWKH source of competitive advantage. If these resources DUH DOVR QRW LPLWDEOH GLI¿FXOW WR UHSOLFDWH QRW substitutable (by other resources), and not transferable (they cannot be purchased), the competitive advantage produced can be sustainable. While our focus is not on sustainability of competitive advantage, the RBV provides useful insights to RXUFDVHLQRUGHUWRH[SODLQKRZDSDUWLFXODU¿UP may get a competitive advantage. The Industrial Network Approach (INA) (e.g., Axelsson & Easton, 1992; Hakansson & Snehota, 1995; Möller & Wilson, 1995) allows us to shift WKHXQLWRIDQDO\VLVIURPDVLQJOH¿UPWRWKDWRI D QHW RI ¿UPV ,W YLHZV LQGXVWULDO QHWZRUNV DV QHWZRUNV RI LQWHU¿UP UHODWLRQVKLSV LQ ZKLFK one actor is connected directly or indirectly to other actors through exchange relationships. The INA provides an analytical framework, the actorresource-activity model, in which companies are
closely interrelated by actors bonds, resources ties, and activity links. A basic assumption in the netZRUNPRGHOLVWKDWLQGLYLGXDO¿UPVDUHGHSHQGDQW RQUHVRXUFHVFRQWUROOHGE\RWKHU¿UPV7KH¿UP gains access to these external resources through its network position, which is built over time and can be considered as a market asset. Largely based on a cumulative process of interaction and mutual adaptation, the evolution of networks may not be intentional (Hakansson & Ford, 2002). However, Möller, Rajala, and Swann (2005) argue that it is possible, to a limited extent, to manage networks, depending on the type of network involved. The INA complements the RBV by shedding light on how resources and capabilities can be valuable in a network context. In this perspective, the emergence of a new value chain is an opportunity to capture two esVHQWLDODVSHFWVRILQGXVWULDOQHWZRUNV7KH¿UVWLV KRZD¿UPLQDSRVLWLRQRIOHDGHUVKLSLVDEOHWR mobilize the necessary resources and capabilities IURPRWKHU¿UPVLQRUGHUWRUHDFKLWVREMHFWLYHV 7KHVHFRQGLVKRZ¿UPVFDQSRVLWLRQWKHPVHOYHVLQ a value chain, and which resources and capabilities play a critical role. In situations such as this case VWXG\¿UPVDUHSOD\LQJDFRPSOHPHQWDU\UROHLQ a value chain, but may also be in competition for some roles, in a climate of “co-opetition” (Brandenburger & Nalebuff, 1996). Being able to reach a favorable position is a kind of acid test indicating if the resource or capability concerned is valuable and contributes to competitiveness.
Positioning of Actors on the Value Chain We shall now analyze the emergence of the value chain that is necessary to provide wireless Internet services in the airport. As in the case of m-commerce (see Vialle & Epinette, 2006), cooperation EHWZHHQ ¿UPV FRQWUROOLQJ GLIIHUHQW UHVRXUFHV LV necessary in order to be able to provide and market services. In Table 2, we draw inspiration from the actor-resource-activity model of industrial networks (Axelsson & Easton, 1992), where each actor can be positioned in a network according to
353
Inside the Microcosm
the resources that are mobilized to produce some value-creating activities. This presentation also complements current value chain and business models which distinguish between activity (or operations) models and role models, by adding a resource dimension.
hotspot strategy is clearly related to the management of airport activities, rather than that of a standalone activity. At the end of 2002, ADP decided to launch a wireless Internet service aimed at the general public in its main airports. The provision of Internet access and other services to the public was part of a more general project aiming also at providing services to the various businesses and administrations located in the airport premises. $'3DOVRGH¿QHGWKHUHTXLUHPHQWVWREHPHWE\ suppliers. Apart from traditional airport require-
Hotspot Leadership The hotspot leadership is usually assumed by the venue owner or manager, in our case ADP. The
Table 2. The wireless Internet value chain in the Paris airports case Activity
Functions
Critical Resources
Actors Involved
Hotspot
'H¿QHKRWVSRWVWUDWHJ\DQG
•
Venue ownership
•
ADP
negotiate with suppliers and
•
Relationships with partners
Design, install, and maintain
•
Technical capabilities
•
ADP Télécom
infrastructure
•
Knowledge of airport
(with Siemens)
Leadership
partners Network Provisioning
• •
Negotiate with subcontractors and partners
•
activities •
Ensure access provision and
Relationship with venue owner
interoperability with roaming agents Authentication and
•
Ensure authentication of users
•
Network ownership
•
Security
•
Ensure security of
•
Technical capabilities
(with Netinary)
ADP Télécom
Provide technical and billing
•
Network ownership
•
interconnection with other
•
Technical capabilities
hotspot and mobile operators
•
communications •
Provide integration with other networks
Billing and Roaming
•
•
Manage data for billing
ADP Télécom
Ability to negotiate roaming agreements
purposes Content Provision
•
Provision of content services
•
Content ownership
•
Aggregation of content from
•
Ability to relate content and
•
Airlines, shops… (provision)
various sources
customer expectations
•
ADP (provision and aggregation)
Marketing and
'H¿QLWLRQRIRIIHUZLWKSDUWQHUV •
Sales network
•
ADP
Customer Service
•
Promotion and sales of access
•
Customer ownership
•
ADP Télécom
and services
•
Brand image
•
Shops
•
Other operators
•
Management of customer relationships
354
Inside the Microcosm
ments such as security or non interference with airplane operations, ADP stressed the importance of offering a wide range of advanced services, beyond the plain Internet access service. ADP also insisted that the network should be open and subject to multiple roaming agreements, so that customers of other hotspots could use the ADP facilities and vice versa. The relationships established with partners such as airlines and the various businesses located in the venue are critical in order to be able to motivate them to participate in this project.
that a customer of network A is able to use the service provided by network B, while being billed by network A, with some mechanism of revenue sharing between the two networks. In our case, as international travelers may be concerned, international roaming is critical, which requires the collaboration of international brokers such as Boingo. Concerning roaming with GSM operators, the platform allows the hotspot to inter-operate in a similar way to a foreign GSM network.
Content Provision Network Provisioning ADP naturally contracted with its subsidiary ADP Télécom, because of its know-how in telecomPXQLFDWLRQV DQG LWV H[SHULHQFH RI WKH VSHFL¿F constraints of operating in an airport. ADP Télécom sub-contracted the provision of equipment to the Siemens manufacturer, and the server platform to Netinary, a partner of Siemens. The QHWZRUNZDVLQVWDOOHGZLWKLQWKHVSHFL¿FDLUSRUW areas prescribed by ADP, as well as the platform. Alpha testing (by ADP Télécom employees) was conducted, and a plain Internet access service was quickly provided.
Authentication and Security Intrusion (for example, by hackers) is easier to DFKLHYH IRU UDGLR OLQNV WKDQ IRU ¿[HG OLQHV $Q intruder could pretend to be another user for the server, or alternatively appear as the server to a user. Therefore the authentication of users and the security of communications are critical functions. The platform provided by Netinary is based on a system similar to the one implemented on GSM networks, in order to achieve a high level of security. These functions are also critical for billing purposes.
Billing and Roaming As users of a particular hotspot may be customers of another hotspot, or of a mobile operator, the establishment of roaming agreements is necessary. Basically, a roaming agreement ensures
Content provision can be split into two functions: content provision itself and content aggregation. Content provision is naturally linked to content ownership. Critical resources are exclusive and “indispensable” content: in our context the timetable, gate numbers, and other operational information owned by ADP are very critical. Other airportrelated content provided by airlines, car rental FRPSDQLHVKRWHOVRUVKRSVQDWXUDOO\¿WZLWKWKH needs of customers. Content aggregation is assembling content from various sources to answer the needs of a target population. ADP, because of its position as venue manager and because of the very critical content it owns, is the content aggregator. In fact most of the information was already available on the ADP Internet portal and could be provided by wireless access with minor changes. To provide “advanced” services contributing to the image of the airport, it was decided to establish a partnership with Virgin Mega and the Hachette media group, in order to provide entertainment services. Such services include provision of movies, video clips, music, TV programs, and games. Movie provision may not be offered because of downloading time exceeding customer acceptance. Other advanced services include services targeted at business travelers such as access to e-mail or DFFHVVWRWKHLUFRPSDQ\¶VLQWUDQHW/RFDWLRQEDVHG services, selecting content according to the geographical location of the user, could be offered in the future.
355
Inside the Microcosm
Marketing and Customer Service ADP has a general responsibility concerning the marketing policy, whereas ADP Télécom is in charge of customer acquisition, support, and retention. Acquisition is subject to “network externalities,” where the use of one service is dependant on the use of other goods and services. In order to be able to use this particular service, the user must have a laptop computer or PDA on hand, HTXLSSHGZLWKDVSHFL¿FZLUHOHVV,QWHUQHWFKLS$ strong hypothesis is also that the use of wireless Internet at this particular hotspot may be dependant on its use in other hotspots or cellular networks. Initially, ADP Télécom intended to rely only on its own sales, either directly through payment with a credit card, or indirectly with prepaid cards that FRXOGEHERXJKWLQWKHDLUSRUW¶VQHZVSDSHUVKRSV However, it was immediately obvious that the lack of existing direct relationships with end-customers was a strong barrier, and that it would be more ef¿FLHQWWRFRQWUDFWZLWKFRPSDQLHVDOUHDG\KDYLQJ customers. Therefore, agreements were made with mobile operators, so that the customers could be billed on their mobile phone account, thus making the usage simpler. Most mobile operators offer subscription schemes bundling GPRS cellular services with wireless Internet services.
Discussion The limited scope of the project, and hence of the business net concerned, allowed avoiding problems usually encountered by researchers analyzing ODUJHUQHWZRUNVZKHUHWKHERUGHUVDUHGLI¿FXOWWR delimitate. This business net was basically an extension of the current business net of actors involved in core airport activities. New relationships were only indirect and mediated by ADP Télécom: they concerned network providers such as Siemens and Netinary, and roaming partners such as Boingo and Orange. This (case) highlights the importance RISRVVHVVLQJWKHULJKWUHVRXUFHVLQRUGHUWR¿QGD viable position in a new value chain of activities. In particular, relational assets were critical in two LQVWDQFHV)LUVWDVKRWVSRWOHDGHU$'3EHQH¿WHG from preexisting relationships which allowed the
356
company to mobilize a network of partners on this project. Second, ADP Télécom realized that the lack of relational assets with consumers would QRWHQDEOHLWWRSOD\DVLJQL¿FDQWUROHLQPDUNHWLQJ and billing services to the public. It is interesting to note the learning effect from this experience for ADP Télécom, called Hub Télécom. First, the supplier network for this project and its experience in coordination allowed the company to expand its business beyond the airport, by providing wireless Internet and other types of broadband access technologies to other customers. Hub Télécom, a telecom service provider for complex sites, extended its business to seven other airports, to hotels (Alliance Hospitality chain, Concorde Lafayette Paris), and one convention center (Palais des Congrés, Paris). It also carried out trials on the :L0$;WHFKQRORJ\RQWKHDLUSRUWSUHPLVHVDQG participated in the Bolloré Télécom consortium, bringing together Hub Télécom, Bolloré, and Antalis. This consortium was awarded 13 regional :L0$;OLFHQVHVLQ)UDQFHLQ-XO\ The success of such a project is conditioned by the ability to attract customers and stimulate them to adopt and use the services supplied. Due to the GLI¿FXOW\RIIRUHFDVWLQJWKHOHYHODQGWKHQDWXUHRI demand, an interactive process usually takes place where: (1) pre-conceptions of demand are factored into the project and then subjected to some form of market research, usually leading to some more RUOHVVVLJQL¿FDQWUHRULHQWDWLRQRIWKHRIIHULQJ the offering is then compared with experimental market responses which are factored back into the project; and (3) market monitoring is implemented after an extended service launch. The second SDUWRIRXUUHVHDUFKWDNHVSODFHLQWKH¿UVWSKDVH where actors had some broad conceptions about user expectations and of how potential customers would react to the intended offer.
EXPLORATION OF THE INTENT TO ADOPT AND USE SERVICES &DVH6SHFL¿FLW\ Understanding and anticipating potential consumer behavior is critical for suppliers of new high-tech-
Inside the Microcosm
nology services. However, this is a particularly GLI¿FXOW H[HUFLVH EHFDXVH RI IDFWRUV VXFK DV WKH complexity of the products involved, the presence of interdependencies, the low or non-existing consumer experience, or the high uncertainty of the impact of innovations. 0RUHVSHFL¿FDOO\FRQVXPHUEHKDYLRUFRQFHUQing wireless Internet services in public venues is hard to understand for the following reasons. First, these services can be used by a large variety of traveling professionals, even if senior executives currently constitute the main target of hotspots. Second, adoption and usage are dependent both on the decision of the individual users and on the company employing them. Third, from the indiYLGXDOXVHU¶VSHUVSHFWLYHWKHUHLVQRFOHDUERUGHU between professional and more personal expectations. Family, entertainment, and business topics must be equally considered. Fourth, a multitude of companies can coexist at a hotspot, as in the case of an airport. Therefore the spectrum of localized mobile Internet services which can be proposed is broad. Fifth, the adoption and consumption of such services are subject to network externalities: consumption in one hotspot is dependent on consumption in other geographical areas. Finally, in addition to the services to be provided, the physical FKDUDFWHULVWLFVRIWKHKRWVSRWFRXOGKDYHDQLQÀXence on the consumption experience. Various theoretical approaches have been used for analyzing information and communication technology (ICT) adoption and diffusion, such as the theory of reasoned action (Fishbein & Ajzen, 1975), innovation diffusion theory (Rogers, 1983), social cognitive theory (Bandura, 1986), the technology acceptance model (TAM) (Davis, 1989), the theory of planned behavior (Ajzen, 1991), and TAM 2 (Venkatesh & Davis, 2000). Some authors also attempted to synthesize several traditions by SURSRVLQJDQLQWHJUDWHGPRGHOVXFKDVWKHXQL¿HG theory of acceptance and use of technology (Venkatesh, Morris, Davis, & Davis, 2003). Competing theories and empirical studies are, for example, reviewed in Venkatesh et al. (2003) and Jeyaraj, Rottman, and Lacity (2006). Several articles adopting the TAM model perspective are related to mobile and wireless Internet (Bruner & Kumar,
)DQJ &KDQ %U]H]LQVNL ;X /X Yao, & Yu, 2005; Wu & Wang, 2005; Yoon & Kim, 2006). These general approaches provide valuable insight in understanding the multiple facets of innovation adoption and diffusion. However our concern was different and led us to choose a different approach. Firstly, our aim was not to design, test, or improve a general model, but to capture WKHFKDUDFWHULVWLFVRIDVSHFL¿FFDVH,QSDUWLFXODU the intertwining between potential adoption and use of a new service, travel situations, personal LQÀXHQFHVSURIHVVLRQDOLQÀXHQFHVDQGWKHVSHFL¿F characteristics of the hotspot consideredthese all render this case particularly complex. Secondly, with regard to our partnership with ADP Télécom, our objective was dominantly to provide a broad insight into understanding potential adoption and use of wireless Internet services in an airport, and to suggest guidelines for managerial action. On the one hand, a modeling exercise may have had a limiting effect on the scope of factors taken into account. On the other hand, some factors which are critical in modeling adoption and use, such as imitation effects, may not have been relevant for PDQDJHULDODFWLRQLQWKLVVSHFL¿FFDVH
Research Approach To explore this complex subject, we made a qualitative study. Because of the novel character of wireless Internet in France at that time, our concern was WRLGHQWLI\IDFWRUVLQÀXHQFLQJERWKDGRSWLRQDQG use of such services in airports. The qualitative study is based on 24 interviews from a sample of middle and senior management executives likely to be equipped with a related communications device in situations of mobility. From our preliminary results, we were able to conceive a simple framework aiming at explaining the intent of adoption and use of wireless Internet services. 7KLVIUDPHZRUNVSHFL¿FDOO\FRQFHUQVDQDLUSRUW environment, but can easily be adapted to other types of public hotspots. The potential sample was made up of several OLVWV RI EXVLQHVV VFKRROV¶ DOXPQL 3HRSOH ZHUH selected in order to ensure the widest variety of viewpoints on our study subject. The objective of
357
Inside the Microcosm
the qualitative study being to explore the adoption and use of this new service, the composition of the sample did not aim to be statistically representative of the reference population. Therefore, the results presented here cannot be generalized. This selection took into account two discriminating variables: the function in the company and the degree of mobility. The objective was to get a variety of levels of involvement in IT for the function, as well as a variety of travel frequencies. After having contacted these 200 people, we were able to carry out 27 interviews. Three were unusable because of bad audio recording quality. To collect information, individual in-depth interviews were carried out. These interviews occurred at in the working place of the interviewees, GXULQJWKH¿UVWVHPHVWHURI2XULQLWLDOLQWHQW was to conduct interviews in the passenger zone of the airport premises. From a methodological perspective, it would have offered the advantage of getting respondents in real travel conditions. However, because of tightened security measures during this period, we were not able to do so. Therefore, ZHKDYHWULHGWRFUHDWHD¿FWLRQDOPRELOLW\VLWXDWLRQ by asking respondents to imagine a travel story, from the organization of the journey to the arrival DWWKH¿QDOGHVWLQDWLRQ7KHLQWHUYLHZHHZDVDEOHWR express himself freely within the framework of the WRSLFVVXJJHVWHGE\DWKUHHSDUWJXLGHOLQH7KH¿UVW part made it possible for the person interviewed to express stereotypes concerning the Internet. The second part centered the discussion on travel situations and the different usage opportunities. The last part was concerned with imagining a possible range of mobile Internet services.
Results In order to understand potential adoption and use of Wireless data services in a situation of mobility, we have focused our analysis on identifying causal relations between variables, and several propositions have emerged from our analysis. A QH[WVWHSZRXOGEHWRUH¿QHDQGWHVWWKHSURSRVHG relations (existence, direction, intensity) with a representative sample of the target population, in a quantitative study.
358
Four categories of factors emerge from our qualitative analysis, namely individual characteristics and experience of business travelers, the policy of their organization concerning information and communication technology, situational factors relative to the hotspot, and service availability inside and outside the hotspot (see Table 1).
Individual Characteristics and Experience 9DULRXVIDFWRUVVHHPWRKDYHDQLQÀXHQFHRQWKH intent to use wireless Internet at a hotspot: hierarchical position, function, age, sex, personality, experience, and user status. The hierarchical position plays an ambiguous role. On the one hand, people with a higher position are more likely to be equipped with terminals by their organization, and particularly with PDAs. On the other hand, their comments suggest that travel time is often perceived as an opportunity to read, think, and relax, rather than dealing with a computing device. The function in the organization also has an LQÀXHQFHRQLQWHQGHGXVDJH)LUVWIRUVRPHIXQFtions such as sales or technical support, communicating in situations of mobility can be critical. People carrying out such functions are therefore more likely to be equipped, to be experienced in communicating while being mobile, and to have urgent needs at a hotspot. In general, executives with technical functions or in the ICT sector are more experienced in using computing devices and therefore more likely to use wireless Internet services. $JHFDQKDYHDPRGHUDWHLQÀXHQFHRQXVHDV intent to use seems to decrease with age. Men exhibit a higher propensity to use wireless Internet than women. One explanation could be that laptop computers and their connection do not appear practical. PDAs are considered as an interesting alternative, but are still not perceived as user friendly enough by women. 6RPH SHUVRQDOLW\ IHDWXUHV VHHP WR LQÀXHQFH use, and particularly one characteristic that we call “need to obtain reassurance.” Even when the trip is organized in advance, some respondents
Inside the Microcosm
Table 3. Factors explaining the intent to use wireless Internet in an airport Factors
&DXVDO5HODWLRQV,GHQWL¿HGLQWKH6DPSOH
Individual Characteristics and Experience Position
Managers with higher positions are more likely to be equipped with the technology, but less likely to use it
Function
Mobile communications are more crucial for some functions (e.g., sales, technical support)
Age
Older people are less likely to use it
Gender
Men more likely to use it than women
Personality
Desire to keep in touch, need to anticipate developments and obtain reassurance
Experience
Experience in using the Internet and mobile devices (learning effect, familiarity), related to personal interest, technical functions, sector of industry
User Status
Effect of positive/negative experience on general attitude towards computing and wireless Internet; wireless Internet user in other contexts
Organization Policy Attitude towards ICT
Strategic resource or cost to minimize
Equipment and Access Policy
,QÀXHQFHHTXLSPHQW
Service Policy
Policy to spur or restrict use
Security Policy
May lead to restrict availability of remote access to IS
Situational Factors Perceived Available Time Travel Characteristics
The more perceived available time, the more likely to use The type of travel, in terms of duration, destination, transit, determine the available time; a low level of organization induces need for information and transaction
Physical Environment
Privacy, comfort, noise level, availability of sockets
Service Availability At the Hotspot
$WWUDFWLYHQHVVDQGUDQJHRIVHUYLFHVDYDLODEOHLQÀXHQFHXVH
At Other Hotspots
The more hotspots, the more adoption and use (network externalities)
At Company Premises
Availability of company WLAN encourages adoption and use
Roaming Availability
Reduce transaction costs, more spontaneous use
expressed the need to connect to the Internet in order to check information or to be able to deal with unexpected events, concerning, for example, a hotel or car reservation. The role of “orientation towards communicating with others” can be ambiguous. On the one hand, such people can be expected to communicate electronically; on the other, they may prefer to communicate directly, for example, with their neighbor. The latter case seems to be more likely.
Experience plays a role through two distinct but UHODWHGDVSHFWV7KH¿UVWFRQFHUQVWKHDFFXPXODWHG experience, and hence familiarity and skills with the use of computing and telecommunications products and services. The more experienced the user, the higher the propensity to use other products and services such as wireless Internet. This is apparent, as explained before, through the type of function (e.g., technical) or the sector of industry. The second concerns the positive or negative attitude generated by former experience. Interestingly enough,
359
Inside the Microcosm
some respondents having tested wireless Internet in another context (WLAN) had a negative attitude due to the problems they encountered. As for user status, being already a wireless Internet user in other hotspots naturally leads to an increase in the propensity to use it in an airport.
Organization Policy Concerning Information and Communication Technology For business travelers the policy of their company or organization concerning ICT can have a decisive impact on the availability of equipment and of usDJH,Q9LDOOHDQG(SLQHWWH ZHLGHQWL¿HG that the adoption behavior for new ICT products and services was strongly dependent on the way these investments and expenses were considered, and that information technology (IT) managers IRXQG LW GLI¿FXOW WR DVVHVV WKH UHWXUQ RQ LQYHVWment for this type of project. The attitude towards ICT investment and expenses broadly oscillates between two extreme positions. One possible attitude is to consider ICT as a strategic investment (at least for a given number of projects), whose EHQH¿WVDUHORFDOL]HGRXWVLGHWKH,7IXQFWLRQVXFK as improving internal and external coordination, reducing time to market, or increasing customer satisfaction. Another attitude is to treat these resources as a cost to be minimized, according to the available budget. In this case, investments will EHFRQVLGHUHGLIEHQH¿WVDUHORFDOL]HGZLWKLQWKH IT function, such as a lower cost per bit or lower maintenance cost. The attitude towards ICT will impact on the equipment and access policy (deciding who will be equipped with a terminal equipment and allowed to access a service), as well as the service policy, such as the range of services available. Another critical point is the security policy of the IT department. A tight security policy may lead IT departments to restrict remote access to their information system, in order to avoid malevolent intrusions (for example, by hackers). IT departments may be even more cautious in the case of wireless Internet access, as intrusion is easier to achieve IRU UDGLR OLQNV WKDQ IRU ¿[HG OLQHV $Q LQWUXGHU
360
could pretend to be another user for the server, or alternatively appear as the server to a user.
Situational Factors 7KH PDLQ LQÀXHQFLQJ IDFWRUV DUH WKH SHUFHLYHG available time, the travel characteristics, and the characteristics of the physical environment. The perceived available time relates to two aspects: the user personality (e.g., tolerance of inactivity), and the competing entertainment or other activities at hand. Travel characteristics, in terms of duration, GHVWLQDWLRQDQGWUDQVLWWLPHKDYHDQLQÀXHQFHRQ both the time available and need for communication and information. For example, a passenger WUDQVLWLQJ EHWZHHQ WZR LQWHUFRQWLQHQWDO ÀLJKWV who needs to keep informed about his family or his business also has enough time available for it. A low level of pre-organization of travel should also lead to stronger needs for communicating, for H[DPSOHLQRUGHUWRFRQ¿UPDUHVHUYDWLRQRUWRJHW additional tourist information. Consumption of wireless Internet services should also depend on the physical environment. 7KLVUHIHUVWRDPELHQFHIDFWRUVVXFKDVEHQH¿WLQJ from a comfortable enough environment, in terms of furniture, silence, and isolation from other people. It also refers to more critical issues such as the availability of sockets for terminal equipment to be plugged in to. These factors which appear WREHFULWLFDODUHDOVRGLI¿FXOWWRSURYLGHDVWKH organization of space in an airport is strongly determined by considerations of security, passenger management, and scarcity of available space.
Service Availability The expected consumption will be dependent on the availability of attractive services both inside and outside the hotspot considered. Service expectations are rather standard, and mainly concern e-mail and Internet access. For PDA users, the provision of information in a suitable format is required. In terms of content, respondents seem to focus more on travel-related information. “Push” information, for example to keep informed about accurate board-
Inside the Microcosm
ing and departure time, also raised interest. We did not identify much attraction towards multimedia services, such as downloading movies.1 Expected consumption is also related to service availability outside the hotspot considered, through the number of other hotspots and/or the existence of WLANs in the organization employing travelers. This is a typical example of network externalities where the utility of a service is dependent on the availability of similar or complementary products and services (Katz & Shapiro, 1985). In fact it is unlikely that potential users would DGRSWZLUHOHVV,QWHUQHWVSHFL¿FDOO\LQWKHFDVHRI the Paris airports. We would expect a two-stage GHFLVLRQSURFHVV ¿UVWDGHFLVLRQWRDGRSWZLUHOHVV Internet at a general level or in a company WLAN, and then to use it in the airport under a given number of conditions. Multi-usage means the existence of roaming agreements between the various hotspot owners and mobile operators, including the posVLELOLW\WREHELOOHGE\RQH¶VXVXDOSURYLGHU
Discussion 2XU¿QGLQJVDUHLQOLQHZLWKWKHFXUUHQW¿QGLQJV on innovation adoption and use. In their review of the literature, Jeyaraj et al. (2006) cite gender, personal innovativeness, experience, anxiety, attitudes, age, education, and motivation as individual characteristics playing the role of independent variables in various papers. Venkatesh et al. (2003), in WKHLUDWWHPSWWRSURSRVHDXQL¿HGPRGHOLGHQWLI\ performance expectancy, effort expectancy, social LQÀXHQFH and facilitating conditions as main direct determinants, while gender, age, experience, and voluntariness of use play a role as key moderators. Similarly, the literature on organizational adoption RI,7KDVLGHQWL¿HGWKHRUJDQL]DWLRQDOSROLF\IDFtors outlined in our study. +RZHYHU ZH DOVR RXWOLQH VSHFL¿F DVSHFWV RI adoption and use of wireless Internet services at KRWVSRWVLQDVLWXDWLRQZKLFKVLJQL¿FDQWO\GLIIHUV from that prevailing at home or at the workplace. 7KLV VLWXDWLRQ FOHDUO\ DIIHFWV WKH LQÀXHQFH RI some of the factors cited above and also raises the issue of situational factors. Our analysis suggests the importance of perceived available time,
travel characteristics, and physical environment as situational variables. The use of such situational variables may also be relevant in the case of other hotspot in hotels, railway stations, and other public venues. They may also be used as inputs to the design of marketing policies and programs aimed at increasing consumption of wireless Internet services in hotspots. However, this must be done in accordance with the marketing policy for the core business of hotspot owners. For example, a railway company may not consider it desirable to increase the perceived available time in order to increase wireless Internet services consumption. Similarly, an organization of the physical environment aiming at improving the condition of use of such services may not be in accordance with the UHTXLUHPHQWV RI PDQDJLQJ SDVVHQJHU ÀRZV LQ D limited geographical zone.
CONCLUSION 7KLV ¿QHJUDLQ DQDO\VLV RI D VSHFL¿F SURMHFW sheds light on micro-level factors affecting the diffusion of broadband data transmission at hotspots. The adoption and use of wireless Internet in public venues depends on the attractiveness of each hotspot. Diffusion here depends on a mix of “general decisions” to adopt and use wireless Internet in situations of mobility, and of micro GHFLVLRQV WR DGRSW LW DQGRU XVH LW DW D VSHFL¿F KRWVSRW GHSHQGLQJ RQ VSHFL¿F FRQGLWLRQV 2XU UHVHDUFKVXJJHVWVVRPHVSHFL¿FDVZHOODVPRUH general conclusions pertaining to supply-side and demand-side elements. 2QWKHVXSSO\VLGHLW¿UVWVKRZVWKDWWKHDELOLW\ to mobilize a network of partners and suppliers is paramount to the success of projects that imply a combination of resources from different types of actors. ADP, as venue manager, was de facto in a position to mobilize various resources. Second, it stresses the importance of preexisting relationships with other actors and customers. The relative inability of ADP Télécom to commercialize by itself wireless Internet services to consumers, because of the lack of a preexisting clientele and customer relationships, is a striking example. Third, it high-
361
Inside the Microcosm
lights the importance of the interaction between the venue environment and the hotspot: on the one hand, the hotspot service must take into account VSHFL¿FJHRJUDSKLFDODQGVHFXULW\LPSHUDWLYHV2Q the other hand, it is not enough to install a wireless network in a public venue; the venue must also be PRGL¿HGWRDOORZFRPIRUWDEOHDQGHI¿FLHQWXVHRI service. This interaction may give rise to a “virtuous” or “vicious” cycle: low consumption because of the inadequacy of the environment may result in a lack of incentive for investing to improve the venue environment. 2QWKHGHPDQGVLGHD¿UVWNH\UHVXOWRIRXU research is the importance of situational factors, such as the availability of electrical connectors, a comfortable environment, or the perception of the serviceable time. The provision of localized services and of a related physical environment seems to be able to trigger the consumption of wireless Internet services at hotspots. A second NH\UHVXOWLVWKDWIRFXVLQJRQO\RQWKHLQGLYLGXDO¶V needs and expectations, a traditional characteristic RIFRQVXPHUPDUNHWLQJLVQRWVXI¿FLHQWLQRUGHU WRXQGHUVWDQGWKHSURGXFW¶VSRWHQWLDOIRUDGRSWLRQ The individual adoption behavior will be obviously LQÀXHQFHGE\WKHFRPSDQ\¶V,7SROLF\VXFKDVWKH terminal equipment policy, the security policy, or the services provided by the organization. A third NH\UHVXOWFRQVLVWVRIWKHFRQ¿UPDWLRQRIWKHLPportance of network effects. The consumption of VHUYLFHVDWDVSHFL¿FKRWVSRWLVGHSHQGHQWRQWKH consumption in other places, related to geographical range of availability of similar services, as well as the provision of adequate billing solutions.
ACKNOWLEDGMENT This research was supported by the “Louis Le Prince Ringuet” Foundation.
REFERENCES Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision Process, 50(2), 179-211.
362
Axelsson, B., & Easton, G. (1992). Industrial networks—a new view of reality. London: Routledge. Barney, J.B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99-120. Brandenburger, A.M., & Nalebuff, B.J. (1996). Co-opetition. New York: Doubleday. Bruner, G.C., & Kumar, A. (2005). Explaining consumer acceptance of handheld Internet devices. Journal of Business Research, 58, 553-558. Chang, S.C., Yu, H.C., & Tsai, J. (2006). How to operate public WLAN business: The case of Taiwan. Journal of American Academy of Business, Cambridge, 8(1), 253-259. Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. Dholakia, N., Rask, M., & Dholakia R.R. (Eds.). (2006). M-commerce—global experiences and perspectives. Hershey PA: Idea Group. Dineen, R., & Anderton, S. (2002). Public wireless LAN: The business opportunity. London: Ovum. )DQJ;&KDQ6%U]H]LQVNL- ;X6 Moderating effects of task type on wireless technology acceptance. Journal of Management Information Systems, 22(3), 123-157. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. Reading, MA: AddisonWesley. Grant, R.M. (1991). The Resource-Based Theory of Competitive Advantage: Implications for strategy. California Management Review, 33(Spring), 114-135. Hakansson, H., & Snehota, Y. (1995). Developing relationships in business networks. London: Routledge. Hakasson, H., & Ford, D. (2002). How should companies interact in business networks? Journal of Business Research, 55, 133-139.
Inside the Microcosm
Jeyaraj, A., Rottman, J.V., & Lacity, M.C. (2006). A review of the predictors, linkages, and biases in IT innovation adoption research. Journal of Information Technology, 21, 1-23.
Thorngren, B. (2002). Public WLAN: The interaction between venues and WISPs.0DVWHU¶V7KHVLV Department of Business Administration, Lund School of Economics and Management, Sweden.
Katz, M.L., & Shapiro, C. (1985). Network externalities, competition and compatibility. American Economic Review, 75(3), 424-440.
Vandepeutte, P. (2005). The expansion of hotspots in Belgium. Technology in Society, 27, 243-250.
Kim, H., Kim, J., Lee, Y., Chae, M., & Choi, Y. (2002). An empirical study of the use contexts and usability problems in mobile Internet. Proceedings of the 35th Hawaii International Conference on System Sciences. King, W.R., & He, J. (2006). A meta-analysis of the Technology Acceptance Model. Information & Management, 43, 740-755. Lu, J., Yao, J.E., & Yu, C.S. (2005). Personal inQRYDWLYHQHVV VRFLDO LQÀXHQFHV DQG DGRSWLRQ RI wireless Internet services via mobile technology. Journal of Strategic Information Systems, 14, 245-268. Möller, K., & Wilson, D. (Eds.). (1995). Business marketing: An interaction and network perspective. Boston: Kluwer Academic. Möller, K., Rajala, A., & Svahn, S. (2005). Strategic business netstheir type and management. Journal of Business Research, 58, 1274-1284. Pagani, M. (2006). Determinants of adoption of high speed data services in the business market: Evidence for a combined Technology Acceptance Model with Task Technology Fit Model. Information & Management, 43, 847-860. Rogers, E.M. (1983). Diffusion of innovations. New York: The Free Press. Sabat, H.K. (2002).The evolving mobile wireless value chain and market structure. Telecommunications Policy, 26, 503-535. Srivastava, R.K., Fahey, L., & Christensen, H.K. (2001). The resource-based view and marketing: The role of market-based assets in gaining competitive advantage. Journal of Management, 27(November-December), 777-802.
Venkatesh, V., & Davis, F.D. (2000). A theoretical extension of the Technology Acceptance Model: )RXUORQJLWXGLQDO¿HOGVWXGLHVManagement Science, 46(2), 186-204. Venkatesh, V., Morris, M.G., Davis, G.B., & Davis, F.D. (2003). User acceptance of information WHFKQRORJ\ 7RZDUGDXQL¿HGYLHZMIS Quarterly, 27(3), 425-478. Vialle P. (1998). Stratégies des opérateurs de télécoms. Paris: Hermès. Vialle, P., & Epinette, O. (2001, July). The attitudes DQGSHUFHSWLRQVRI,&7PDQDJHUVRIODUJH¿UPV towards m-business Internet adoption: Beyond the magic…. Proceedings of the Conference on Telecommunications and Information Markets (COTIM), Karlsruhe, Germany. Vialle, P., & Epinette, O. (2006). France: Mobile communications and emerging m-commerce. In N. Dholakia, M. Rask, & R.R. Dholakia (Eds.), M-commerce—global experiences and perspectives. Hershey, PA: Idea Group. Wernerfelt, B. (1984). A resource-based view of WKH ¿UP Strategic Management Journal, 5(2), 171-180. Wu, J.H., & Wang, S.C. (2005). What drives mobile commerce? An empirical evaluation of the revised Technology Acceptance Model. Information & Management, 42, 719-729. Yoon, C., & Kim, S. (2006). Convenience and TAM in a ubiquitous computing environment: The case of wireless LAN. Electronic Commerce Research and Applications, 1-11.
363
Inside the Microcosm
KEY TERMS Hotspot: A geographic area that is covered by a wireless network. Inside this area, Internet access is made available to any device equipped with a wireless card. Local Area Network (LAN): A network covering a limited geographic distance, such as an RI¿FHEXLOGLQJRUDJURXSRIEXLOGLQJVLQFORVH proximity to each other. Roaming: A subscriber of one service provider can use the network of another service provider in another location, thanks to an agreement between service providers, called a roaming agreement. In this case the customer is still billed by his regular service provider. Wireless Fidelity (WiFi): A trade name promulgated by the WiFi Alliance, formerly called WECA (Wireless Ethernet Compatibility Alliance). ,WFRUUHVSRQGVWRWKHEVWDQGDUGGH¿QHGE\ the IEEE (Institute of Electrical and Electronics (QJLQHHUV ,WDOORZVDWKHRUHWLFDOÀRZRI0EV in the frequency band of the 2.4 GHz, on a distance of up to several hundred meters.
364
Wireless Internet Service Provider (WISP): An Internet service provider using radio wave systems to provide services. Wireless LAN (WLAN): A short-range, broadband data communication system using radio waves, serving as a LAN. Worldwide Interoperability for Microwave Access (WiMAX): Corresponds to IEEE 802.16 technology. With a range of up to 30 miles, it provides a viable alternative to ADSL for subscriber connection, in particular in low population density areas such as the countryside.
ENDNOTE 1
However, this may result from a possible bias in the interviews, as we felt that respondents were rather reluctant to mention entertainment while discussing professional travel.
Inside the Microcosm
APPENDIX Sample Description (24 Respondents) Gender
Men: 17 Women: 7
Hierarchical Level
Middle management: 14 Senior management: 10
Direct IT Involvement
Rather low direct IT involvement: 6
(function and use)
Rather high direct IT involvement: 18
Level of Mobility
Rather low mobility: 9
(frequency, distance)
Rather high mobility: 15
365
366
Chapter XXIII
7KH8VHVDQG*UDWLۋFDWLRQV of Broadband Internet Karianne Vermaas Utrecht University / Dialogic innovatie & interactie, The Netherlands Lidwien van de Wijngaert Utrecht University, The Netherlands
ABSTRACT The objective of the study in this chapter is to identify a small number of relatively homogeneous groups of Dutch Internet users (both broadband and narrowband), based on their usage pattern. Using individual and behavioral characteristics, we further investigate the nature of the different groups. The Uses and *UDWL¿FDWLRQVDSSURDFKLVHPSOR\HGDVDVWDUWLQJSRLQWIRUDQRQOLQHTXHVWLRQQDLUH 1 WKDWZDV conducted. Cluster-analysis and logistic regression were used for data analysis. We were able to identify ¿YHFOXVWHUVRIGLIIHUHQW,QWHUQHWXVHUVEDVHGRQSDWWHUQVRIEHKDYLRU5HVXOWVVKRZWKDWWKH,QWHUQHWXVers in our sample consist of a large group that more or less conducts the same online activities. Results also show that narrowband and wideband users differ in their Internet behavior.
INTRODUCTION Someone might use the Internet to look up information like train departure times, telephone numbers, and so forth. Someone else may not see that as the main function of the Internet at all. He might like to listen to online music and look at pictures of his grandchildren. Yet another person may go online WRVHQGHPDLOVDQG¿QGLQIRUPDWLRQZKLOHRWKHUV use the Internet more professionally. According to Anderson, Gale, Jones, and McWilliam (2002), broadband users make more frequent use of a wider range of applications. In a
recent study, Dwivedi, Choudrie, and Brinkman (2006) also found that broadband consumers differ from narrowband users in terms of the variety of Internet use. In the research conducted by Wales, Sacks, and Firth (2003), respondents universally said that they were not driven to broadband by any single application. Rather, they found that broadband enabled them to use standard Internet applications (e-mail, chat, browsing) more ef¿FLHQWO\ Different from Wales et al. (2003), there are researchers that do mention killer applications (an application that is so useful or desirable that
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
LWMXVWL¿HVWKHXQGHUO\LQJWHFKQRORJ\ &KRXGULH and Dwivedi in a recent study (2006) found the adoption of broadband is driven by a combinaWLRQRIIDFWRUVWKH¿UVWEHLQJUHODWLYHDGYDQWDJHV (e.g., faster access), the second utility outcomes (work purposes), and the third hedonic outcomes (entertainment). While narrowband Internet is adequate for many current residential applications, some applications will be hard to use and cause annoyances by the users. In order to keep those annoyances to a minimum, many users that use peer-to-peer applications (including games, sending and receiving of large ¿OHV DQG SRUQRJUDSK\ ZLOO ¿QG WKH\ UHTXLUH D broadband connection. These applications are said to be the killer applications for broadband to the home (Firth & Kelly, 2001; Thierer, 2002; Wales et al., 2003, Anderson et al., 2002). Clearly, many different kinds of people use the Internet for a variety of things. The question however is which kinds of people use which kinds of Internet applications to gratify which needs? Can groups of Internet users be recognized that have similar needs and therefore are for example typical µJDPHUV¶ RU µVHULRXV LQIRUPDWLRQ VHHNHUV¶" $QG what kind of people are they? Do broadband users IRUPDVSHFL¿FFOXVWHUEDVHGRQWKHLU,QWHUQHWXVDJH that is very different from narrowband users? The objective of this study is to identify a small number of relatively homogeneous groups of Internet users based on their usage patterns. Insights into these patterns make it possible to better understand and predict (broadband) Internet usage. Adding demographics and the kind of Internet connection (broadband vs. narrowband) to these patterns of uses make this information even more valuable. Among others, Vermaas and van de Wijngaert (2007) and Choudrie and Dwivedi (2005) have suggested that demographics, such as gender and education, might be of importance with regard to the adoption of broadband. With this information, Internet service and content providers can offer their target groups DSSOLFDWLRQV WKDW EHWWHU ¿W WKH QHHGV RI HDFK RI WKRVH JURXSV PRUH VSHFL¿FDOO\ ZLWK UHJDUG WR broadband service development. Moreover, we aim at contributing to the further development of
approaches or theories explaining the motivation for media usage in regard of Internet usage. The questions that are to be answered are: 1. 2.
Which individuals can be grouped based on their Internet usage patterns? How can the clusters be characterized (demographics, Internet experience, and type of connection)?
Using the answers to these questions, we intend to establish insight into differences between different groups of Internet users. This insight into, for example, more basic and more advanced users may help us to understand how broadband will evolve in the course of time.
THEORETICAL BACKGROUND 8VHVDQG*UDWL¿FDWLRQV An extensively used approach that address the motivation for media usage is the 8VHVDQG*UDWL¿FDtions (U&G) approach. This theoretical framework has been applied to various mass media such as television, radio, and books (Katz, Blumler, & Gurevitch, 1974; Katz, Gurevitch, & Haas, 1973). U&G is not as much a theory as it is an approach with a number of basic assumptions that can be used as a starting point for various types of research. The basic idea of the U&G approach is that people use media in order to gratify needs. U&G research generally generates lists of needs that precede certain media usage. There are many examples of needs that motivate the use of media. Katz et al. (1973) give the following categorization RIQHHGJUDWL¿FDWLRQVZKLFKDUHEDVHGRQVRFLDO and psychological functions of the media: cognitive needs (acquiring information, knowledge, and understanding), affective needs (emotional, pleasurable, or aesthetic experience), personal integrative needs (strengthening credibility, FRQ¿GHQFH VWDELOLW\ DQG VWDWXV VRFLDO LQWHUDFtive needs (straightening contacts with family, friends, and so forth), and tension release needs (escape and diversion). McQuail (1987) suggests
367
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
another typology of common reasons for media use: information, personal identity, integration and social interaction, and entertainment. Rubin (1994) suggests information acquisition, escape, emotional release, companionship, reality exploration, and value reinforcement as needs that are to EHJUDWL¿HGE\PHGLD
Although the media landscape changed dramatically as a consequence of many new technologies in recent years, the U&G approach applies quite well to the usage of Internet. One of the basic assumptions of the U&G theory is that users are actively involved in media usage and interact highly with the media (Katz et al., 1974). For this reason the U&G approach seems even more applicable to the Internet than it is to other mass media like televiVLRQ YDQGH:LMQJDHUW 2QOLQHVXU¿QJIRU example requires great user cognitive involvement because of the constant demands to point, click, comprehend, and select the hypertext links. In this light the U&G applies to the Internet even better than to mass media for which it was meant. A second basic assumption of U&G is that media compete with other sources of need satisfaction (Katz et al., 1974). This is also the case with the Internet. The Internet competes with face-to-face, phone, DQGSDSHUPHGLDDVVRXUFHVRIQHHGJUDWL¿FDWLRQV However, the Internet offers a very wide variety of services.
derstanding on the use and diffusion of broadband Internet and other future innovations. Another problem lies in the fact that it is hard WR FOHDUO\ GH¿QH ZKLFK EHKDYLRUV DUH UHODWHG WR which underlying needs and motivations. Motivations for using the Internet differ from person to person (Flanagin & Metzger, 2001). Whereas some users go online because of various needs, other users go online for one single need. Because of the complexity of the networks of needs, we KDYHGHFLGHG¿UVWWRH[DPLQHDOLPLWHGDPRXQWRI needs. Based on studies by Choi, Watt, Dekkers, and Park (2004), Kaye and Johnson (2001), and 3DSDFKDULVVLDQG5XELQ ZHKDYHLGHQWL¿HG three important motivations for using the Internet: information, social interaction (communication), and entertainment. To this we have added a fourth motivation to go online: shopping (or transactions). In this study we assume that using certain types of Internet services is strongly related with a certain type of motivation. For example, people who have high entertainment motivation are more likely to expose themselves to gaming, online TV, and so forth than those who have low entertainment motivation (Alwitt & Prabhaker, 1992). Likewise, people with high information needs will visit informative Web sites more frequently than people without a strong need for information. Figure 1 outlines the scope of our theoretical framework. The variables and relations that were actually measured are depicted in black, whereas the underlying variables and relations are drawn in grey.
Patterns of Usage as a Starting Point
The Need for Information
A problem of U&G research lies in the assumption RI8 *WKDWSHRSOHDUHVXI¿FLHQWO\VHOIDZDUHWR report their needs (Katz et al., 1974). People often cannot indicate what their needs are (Buntz, 2001). It is much easier for them to state what they do on the Internet than it is to state what need underlies particular behavior. In order to solve these problems, this research focuses on Internet usage patterns (what people say they do online) as the source of information and after that attempt to deduct patterns of needs and motivations. Understanding these needs and motivation will increase our un-
Virtually all U&G researchers have mentioned in one way or the other the need for information. In terms of Katz et al. (1974), this need is a cognitive need. But also McQuail (1987), Rubin (1994), and Papacharissi and Rubin (2000) mention this need. It is often proved that information gathering is an important reason to go online (Maltha et al., 2002, 2003). Information is made more accessible by the Internet, and an abundant amount of information can be found online.
U&G and the Internet
368
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
)LJXUH$8VHVDQG*UDWL¿FDWLRQVDSSURDFKWRZDUGV,QWHUQHWXVH Demographics - Gender - Education - Income
Underlying motivations
Internet behaviour - Information - Communication - Entertainment - Transaction
Internet behaviour - Internet experience - Frequency of use - Time online - Connection
Gratified needs
Other effects
Other behaviour
Legend
This project
Uses and Gratifications
The Need for Communication
transactions online. The need to complete transactions is not easily comparable with needs stated by U&G research. This is understandable because with traditional media, it was often not possible to shop (although TV now offers programs in which products are sold).
Social interactive needs (Katz et al., 1974), social interaction (McQuail, 1987), social companionship (Rubin, 1994), interpersonal utility (Papacharissi & Rubin, 2000)it all comes down to the need people feel to be in contact with other people. Also for this need, the Internet has brought tremendous changes.
RESEARCH METHOD
The Need for Entertainment
Data Collection
Besides the need for information and communication, also the need for entertainment is an imporWDQWQHHGIRUPDQ\SHRSOHWREHJUDWL¿HG2WKHU U&G researchers also pay attention to this need: entertainment (McQuail, 1987), escape (Rubin, 1994), affective and tension release needs (Katz et al., 1974), and to pass time and for entertainment (Papacharissi & Rubin, 2000).
Since it is the purpose of this study to recognize groups of Internet users, the data was collected through an online questionnaire. This research is based on a convenience sample. In this case this means that we have asked people that had already responded to one of our studies and left their email addresses. We provided links to our online questionnaires on several Web sites in order to get a diverse group of respondents. The following Web sites were involved in our research: Web sites of ISPs (such as the Dutch telecom company KPN and FDEOHFRPSDQLHV RI¿FLDO:HEVLWHVRI'XWFKFLWLHV (Amsterdam, Almere); and Web sites focused on the elderly, women, and immigrants. Data gathering took place from January to March 2003. We used
The Need for Transactions Completing online transactions is an increasing important driver to go online. For many people this need may be more like an obligation, but apparently people feel the need to shop and complete
369
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
a multi-method approach, but the most important part was an online questionnaire, which also forms the basis for this chapter. The objective of this online questionnaire is to obtain insight into current Internet behavior. Questions regard type of Internet access, activities on the Internet (information seeking, communication, entertainment, and transactions), skills and experiences, wishes, and expectations. The response consisted of 2,404 completed and usable questionnaires; 75% of the respondents had a broadband connection. Mainly men reacted to this questionnaire (81% male, 19% female). Also people with different kinds of connections answered the questions. The mean age is 40 years with a standard deviation of almost 15 years. Education level is 4% primary school, 20% lower secondary, 39% higher secondary, and 37% higher education. This is of course not a representative sample of the Dutch population. However, the proportions RIWKHVHFKDUDFWHULVWLFVDUHLQOLQHZLWK¿JXUHVRQ the overall image of Internet users in The NethHUODQGV$OWKRXJKUHFHQW¿JXUHVDUHQRWDYDLODEOH ¿JXUHVIURPWKH'XWFK6WDWLVWLFDO,QVWLWXWLRQVKRZ a similar picture (De Haan & Steyaert, 2003). In sum, we will generalize the results of this research to the Dutch Internet population, rather than to the population as a whole.
Method The questionnaire was clustered around the four needs that we established in the previous section: information, communication, entertainment, and transaction. This is also an apprehensive way to group the large variety of the functional uses of the Internet: •
•
370
Information: Search engines, portals, reference works, downloads, info audio/images, forms, info newsletters, known sites, info from discussion groups, own Web site Communication: Chat via ICQ/IM, chat via special Web site, IP telephony, video phone, reading Web log, writing Web log, discussion groups, SMS/MMS, e-mail
•
•
Entertainment:*DPHVGRZQORDGLQJ¿OPV XSORDGLQJ ¿OPV FUHDWLQJ DQG PDQDJLQJ D community, participating in communities, downloading television programs, downloading short clips, uploading short clips, downloading music, uploading music, downloading photos/drawings, uploading photos/ GUDZLQJVIXQPDLOIXQVXU¿QJ Transactions: Buying products and services, online-making reservations marketplaces for individuals, auction sites, swapping, telebanking
For each of these 38 services, respondents were asked whether or not they have used it.
Cluster Analysis Many U&G researchers use factor analysis in order to cluster needs (Leung & Wei, 1998; Lin, 2002; Leung, 2003). The data of this research, however, is not suitable for a factor analysis, because of the binary values of the variables: people were asked to state whether or not they use certain Internet services, rather than to what degree. Therefore we used cluster analysis to organize the data into meaningful structures. &OXVWHUDQDO\VLVVXJJHVWVDFODVVL¿FDWLRQVFKHPH of grouping cases into a certain amount of classes (Everitt, 1977). Here cluster analysis is used as a pattern recognition technique to summarize relatively homogeneous Internet usage patterns. In the hierarchical clustering, simple matching was used as a measure of similarity. This means that all matches are countedthat is, two people are similar when they both do use a portal, for example, but people are also similar when they both do not download movies. Furthermore, the cluster method used is furthest neighbor or complete linkage. This method tends to produce relatively tight clusters of similar cases. In this research that means that Internet users are not considered similar because they do the same things, but also because they do not do the same things. For example within a group of people that make extensive use of e-mail, the few people that do not e-mail at all are also connected with each other.
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
After examination we concluded that the data FRXOG EH EHVW GLYLGHG LQWR ¿YH FOXVWHUV )HZHU clusters would leave out information, while more clusters did not add more information (clusters that are already small are broken up in even smaller sub clusters).
•
Logistic Regression • After the cluster analysis of the usage patterns, we took a closer look at which kinds of people are in the clusters. This was done by logistic regression. In our data analysis we focus on the question how clusters can be characterized. However, the dependent variDEOHPHPEHUVKLSRIDFOXVWHULVGH¿QHGDV PHPber) or 0 (non-member). Because the dependent variable is measured as a binary variable instead of interval or ratio level, ordinary least square regression cannot be used. Instead, logistic regression will be used. Logistic regression basically uses the same type of model as ordinary least square regression: a linear function of the explanatory variables predicts the dependent variable. However, the outcome is not simply the score of the dependent variable. Instead, the logarithm of the odds (probability divided by one minus the probability) is used.
RESULTS ,QWKLVVHFWLRQZHZLOO¿UVWSUHVHQWWKHUHVXOWVIURP the hierarchical clustering. After that we will investigate how the clusters can be characterized in terms of individual and behavioral characteristics.
Clustering Internet Use Patterns Five groups emerged from the cluster analysis. We can describe the activities of the individuals in the ¿YHFOXVWHUVDVIROORZV •
Cluster 1 (N=1178): Large middle bracket group, uses the Internet for a variety of things, EXWGRHVQRWXVHLWH[WHQVLYHO\)XQVXU¿QJ search engines, and e-mail are keywords.
•
•
Cluster 2 (N=173): Individuals in this cluster use the Internet for applications that require a fair amount of bandwidth, like downloads, music, photos/drawings/fun mail, and video clips. They do not often buy products and services from e-shops, but rather exchange products and use Web sites for individuals to buy and sell products. Cluster 3 (N=143): Contains individuals that like to participate in discussion groups and communities, who either need opinions of other people or want to share their own opinions Cluster 4 (N=729): Also a large group of individuals that shows progressive Internet usage patterns. They use downloads for information, chat, to download music and ¿OPVSOD\RQOLQHJDPHVDQGVRIRUWK Cluster 5 (N=211): These individuals are mainly information seekers (search engines, portals, reference guides), but also make reservations, buy via the Internet, download music, and send fun mail.
Although cluster analysis was conducted on all reported activities at once, the differences in Internet usage between the clusters are presented in four spider charts (one for each of the basic needs of information, communication, entertainment, and transactions). Also the activities that were hardly conducted by any of the clusters (less than 10%) were left out. This is done because, otherwise, too much information is presented at once, making it WRRGLI¿FXOWWRVHHWKHDFWXDOGLIIHUHQFHV Figure 2 shows that the clusters have distinct patterns of Internet usage concerning information. The individuals in cluster 1 appear to be the middle bracket, using mainly search engines and portals for their information need. In clusters 2 and 4, individuals are grouped that mainly use downloads and search engines. What makes cluster 5 a distinct group of Internet users is the fact that they use mostly portals and online reference works. The Internet users in cluster 3 distinguish themselves by looking for information in discussion groups.
371
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
Figure 2. Patterns of information use Search engine 90% Own website
Downloads 40%
Newsletter
-10%
C luster C luster C luster C luster C luster
1 2 3 4 5
(N=1178) (N=173) (N=143) (N=729) (N=211)
Discussion groups
Type in URL
Portals
Reference guides
Figure 3. Patterns of communication use C hat IC Q 100%
50% E-mail
C hat public
C luster C luster C luster C luster C luster
1 2 3 4 5
(N=1178) (N=173) (N=143) (N=729) (N=211)
0%
SMS/MMS
Discussion group
When we look at communication activities (see Figure 3), we see a more uniform image. The cluster that differs most from the other clusters is cluster 3, again by using discussion groups. Cluster 1 shows to be the middle group once more. Remarkably, most online ways to communicate are not used, but e-mail and chatting are used often to satisfy communication needs. The Internet users in cluster 4 state that they chat more often than people in the other clusters. Figure 4 shows a variety of entertainment activity patterns. Individuals in cluster 4 reveal themselves as true progressive Internet users by conducting real broadband activities like gaming, DQGGRZQORDGLQJ¿OPVDQGPXVLFPRUHWKDQWKH
372
other clusters. Also cluster 2 stands out by some bandwidth-consuming activities. Uploading of material is not practiced a lot. Transactions are conducted less frequently that the other activities. Also the patterns of the different clusters are quite similar. Tele-banking is most popular for most clusters, except for cluster 5, where buying of products and making reservations is more often done. Although the results presented so far show that WKHUHDUHGLIIHUHQFHVEHWZHHQWKH¿YHFOXVWHUVLWLV hard to explain these differences in a few keywords. The differences appear to be only slight differences and do not allow us to characterize the clusters
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
Figure 4. Patterns of entertainment use Games 90% Download TV program
Download movie 40%
Download film (short)
C luster C luster C luster C luster C luster
1 2 3 4 5
(N=1178) (N=173) (N=143) (N=729) (N=211)
-10%
Fun surfing
Funmail
Download music Download pictures
Figure 5. Patterns of transaction use Buy products and services 90%
40%
Make a reservation
Buy from private person -10%
Telebanking
E\DSSO\LQJQDPHWDJVVXFKDVW\SLFDOµJDPHUV¶RU µVHULRXVLQIRUPDWLRQVHHNHUV¶%HFDXVHWKHXVDJH SDWWHUQVGRQRWSURYLGHDGH¿QLWHDQVZHUZHZLOO now examine demographical characteristics and Internet experience in order to further characterize the clusters.
Individual and Behavior Characteristics The second research question is related to how the clusters that were distinguished in the previous section can be characterized in terms of sociodemographics and Internet experience, as these
Market
C luster C luster C luster C luster C luster
1 2 3 4 5
(N=1178) (N=173) (N=143) (N=729) (N=211)
characteristics might explain more about the different individuals in the different clusters. Also the question is whether broadband users are assigned to different clusters than narrowband users. In the large middle bracket group (cluster 1) that uses the Internet for a variety of things, to gratify a variety of needs, the analogue modem is used more often to connect to the Internet than in other clusters. That may explain why these individuals do not carry out activities that require a lot of bandwidth. Forty percent of the people in this cluster have an education on the university/HBO level. Most people in the cluster are between 35 and 49 years old (39.8%). The proportion male/female
373
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
is almost equal (49/51%). Furthermore there are mainly average (28.4%) to experienced (30%) users in this cluster that go online several times a day (61.3%) for about one to two hours per visit. Cluster 2 includes individuals that use the Internet for applications that require a fair amount of bandwidth, like downloads for information, downloading music, photos and drawings, video FOLSV DQG XVH IXQ PDLO ,Q WKLV FOXVWHU ZH ¿QG more broadband connections than in cluster 1 (35.3% cable and 38.2% ADSL). But 17% have a traditional telephone modem. Again, in this cluster the proportions of men and women are almost the same (51% and 49%) and their ages are between 35-49 years (36%). Education is slightly lower: mainly HAVO/VWO (46%) and university (35.8%). Forty-six percent are experienced users, and 25.4% are average users. More than in any other cluster, they go online once a day (23.7%). Most of them however go online more than once a day (65.9%) for two to four hours per session (35%). Cluster 3 contains individuals that like to participate in discussion groups and communities: they either need opinions of other people or want to share their opinions with others (personal integrative needs). This group consists mainly of very experienced users (40%) and also has the highest percentage of professional Internet users. There are no beginners in this cluster, which may explain why almost 80% have broadband connections (cable 32.9% and ADSL 46.9%). Furthermore individuals in this cluster tend to have a slightly higher education than other clusters (university/ HBO42.7%). They use the Internet mostly more than one time a day (79%) for two to four hours per session (38.5%). For 14% one session takes longer than eight hours. That is more often and longer than in other clusters. The ages in cluster 3 are between 18-34 (35.5%) or between 35-49 (36.9%). Cluster 4 also includes a large group of individuals that show quite progressive Internet usage patterns. They use downloads for information, chat, GRZQORDG PXVLF DQG ¿OPV SOD\ RQOLQH JDPHV and so forth. The Internet users in this cluster say they are mainly experienced (24.2%) and very experienced users (30%). They use the Internet more than once a day (74.1%) and the sessions take
374
two to four hours (33.1%). In this cluster the main education level is HAVO, VWO, MBO (41.2%). This may be due to their rather young ages, mainly between 18-34 (41.6%). Not surprisingly 84% of these progressive users have broadband connections (cable 39.4% and ADSL 44.6%). The information seekers in cluster 5 reckon themselves to be experienced (39.3%): 60.7% use the Internet more than once a day. That is less than in the other clusters. They also do not stay on the Internet very long: one to two hours is mostly enough to carry out their activities (35.1%), and sessions of eight hours or longer are less common than in other clusters. A total of 74.4% have a broadband connection, and education is high (HAVO, VWO, MBO is 39.8% and university is 37%). Their ages are between 35-49 (33.8%) and 75% are married. This group consists for the larger part of women (60%).
Logistic Regression In order to answer this question, we obtained a more detailed view on the clusters by comparing them on the following variables: age, gender, education, experience, frequency of use, duration of use, income, and type of connection. The results of this logistic regression analysis are presented in Table 1. Logistic regression shows that education, experience, time per session, and type of connection KDYHVLJQL¿FDQWLQÀXHQFHRQDVVLJQPHQWWRFOXVWHU $OOKDYHQHJDWLYHLQÀXHQFHH[FHSWIRUHGXFDWLRQ People with high education, narrowband connection, who do not spend too much time online, and are not very experienced usersthey are likely to be part of cluster 1. Table 2 provides a comprehenVLYHRYHUYLHZRIWKHVLJQL¿FDQWUHODWLRQV The results from the logistic regression on demographical characteristics, experience, and connection do not fully explain the variance. Especially gender and income do not explain much of the variance. Education and time online seem to explain more, but not for every cluster. When we look at the clusters, we see that especially cluster LVGLI¿FXOWWRH[SODLQEDVHGRQWKHGHPRJUDSKLF characteristics. The type of connection is relevant
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
7DEOH/RJLVWLFUHJUHVVLRQIRUHDFKRIWKH¿YHFOXVWHUV Cluster 1
B
S.E.
Wald
df
Sig.
Exp(B)
Gender
-.107
.110
.962
1
.327
.898
Education
.135
.053
6.603
1
.010
1.145
Experience
-.190
.044
18.312
1
.000
.827
Frequency
-.064
.057
1.273
1
.259
.938
Time Online
-.251
.041
37.324
1
.000
.778
Connection
-.157
.042
13.854
1
.000
.855
Income
.011
.016
.428
1
.513
1.011
Constant
2.175
.358
36.838
1
.000
8.799
Cluster 2
B
S.E.
Wald
df
Sig.
Exp(B)
Gender
-.037
.208
.031
1
.860
.964
Education
.151
.102
2.209
1
.137
1.163
Experience
-.132
.084
2.461
1
.117
.876
Frequency
.203
.113
3.194
1
.074
1.225
Time Online
-.009
.077
.014
1
.907
.991
Connection
-.060
.079
.575
1
.448
.942
Income
.003
.031
.010
1
.921
1.003
Constant
-3.101
.693
20.020
1
.000
.045
Cluster 3
B
S.E.
Wald
df
Sig.
Exp(B)
Gender
.224
.231
.943
1
.332
1.252
Education
.206
.116
3.182
1
.074
1.229
Experience
.312
.093
11.281
1
.001
1.367
Frequency
.208
.144
2.081
1
.149
1.231
Time Online
.267
.081
10.816
1
.001
1.306
Connection
-.025
.092
.071
1
.790
.976
Income
-.012
.033
.120
1
.729
.989
Constant
-6.845
.884
59.970
1
.000
.001
Cluster 4
B
S.E.
Wald
df
Sig.
Exp(B)
Gender
-.107
.124
.745
1
.388
.899
Education
-.251
.057
19.590
1
.000
.778
Experience
.187
.049
14.731
1
.000
1.205
Frequency
.079
.066
1.410
1
.235
1.082
Time Online
.262
.044
35.790
1
.000
1.299
Connection
.163
.048
11.466
1
.001
1.177
Income
-.023
.017
1.753
1
.185
.977
Constant
-2.590
.408
40.366
1
.000
.075
Cluster 5
B
S.E.
Wald
df
Sig.
Exp(B)
Gender
.430
.172
6.268
1
.012
1.538
Education
-.030
.090
.114
1
.736
.970
Experience
-.001
.077
.000
1
.992
.999
Frequency
-.164
.092
3.198
1
.074
.849
Time Online
-.149
.072
4.272
1
.039
.862
Connection
.197
.073
7.282
1
.007
1.218
Income
.031
.028
1.248
1
.264
1.031
Constant
-2.518
.590
18.228
1
.000
.081
375
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
IRUWKUHHRXWRI¿YHFOXVWHUV FOXVWHUWKHODUJHVW group of not very advanced users is a group of narrowband users; clusters 4 and 5 are broadband users. These groups show more elaborate types of Internet use.
DISCUSSION AND CONCLUSION The results of this research show that it is possible WRGLVWLQJXLVK¿YHJURXSVRI,QWHUQHWXVHUVEDVHGRQ their patterns of Internet use. Also, it was possible to characterize some of these groups by a number of demographical characteristics, Internet behavior, and type of Internet connection. The most important difference that we have found is between clusters 1 and 4, the two clusters containing the largest groups of Internet users. The largest group of inexperienced Internet users FOXVWHU 1 IXO¿OOV D OLPLWHG QXPEHU RI basic functions using a narrowband connection. These Internet users do not use their connection extensively and do not really feel the need for a faster broadband connection. In contrast, the second largest group (cluster 4, N=729) is a group of people that does use the Internet for a variety of functions. They are experienced but low educated broadband users. Other interesting clusters are 3 (N=143) and 5 (N=211). These are interesting because cluster 3 differs from cluster 5 with regard to education. The Internet users in cluster 3 are higher educated and
show more interest in (online) discussions. Clusters 5 and 1 are alike in the sense that these clusters both contain a relatively large amount of women, and they use the Internet quite functionally. The difference, however, is that cluster 5 has already made the step to broadband. Although different groups of Internet users FDQ EH LGHQWL¿HG ZLWK GLIIHUHQW XVDJH SDWWHUQV and demographical characteristics, type of connection, and experience, this does not mean that VLPSO\ NQRZLQJ RQH¶V GHPRJUDSKLF FKDUDFWHULVWLFVSURYLGHVDGH¿QLWHDQVZHUDERXWWKHFOXVWHU RQH¿WVLQWR Future research could also focus on the question of whether some people consider using the Internet their hobby and others mainly see it as a functional aid. This is probably related to their motivations, which play such an important role in WKH8VHVDQG*UDWL¿FDWLRQVDSSURDFK The main problem with the U&G approach, however, is that the approach has to deal with the following paradox. On the one hand, two individuals with the same need, for example the need for entertainment, can choose very different ways of gratifying these needs. While one individual chooses to do so by downloading his favorite music, the other enters a chat environment. On the other hand, the usage of the same service by two individuals can originate from completely different needs. E-mailing with friends can be entertainment for one individual, while for the other it is a matter of the need for social inclusion. Also one activity can gratify dif-
7DEOH6LJQL¿FDQWUHODWLRQV Cluster 1
Cluster 2
Cluster 3
Cluster 4
High
High
Low
Lay
Experienced
Experienced
Time Online
Short
Long
Connection
Narrow
Gender Education
Cluster 5 Women
Income Experience Frequency
376
Long
Short
Broadband
Broadband
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
ferent needs of one individual. E-mailing a joke can JUDWLI\DSHUVRQ¶VQHHGIRUHQWHUWDLQPHQWSHUVRQDO integration, and social interaction. Based on the results of our research, it appears that there is a relationship between the switch to broadband and the differences in usage patterns. The question however is what causality there is. Do certain people (wish to) use the Internet in such an extensive way that they feel the need for a broadband connection, resulting in switching to broadband, or does using a broadband connection result in changes in the way people use the Internet, EHFDXVHRIPRUHSRVVLELOLWLHV"7R¿QGWKHDQVZHU to this question, we will, in future (longitudinal) research, compare the data of several years. By doing so we hope to gain more insight into the development of Internet usage.
Choudrie, J., & Dwivedi, Y.K. (2006). InvestigatLQJIDFWRUVLQÀXHQFLQJDGRSWLRQRIEURDGEDQGLQ the household. Journal of Computer Information Systems, 46(4), 25-34. De Haan, J., & Steyaert, J. (2003). Jaarboek ICT en samenleving 2003. De sociale dimensie van technologie. Amsterdam: Boom. Dwivedi, Y.K., Choudrie, J., & Brinkman, W.P. (2006). Consumer usage of broadband in British households. International Journal of Services and Standards, 2(4), 400-416. Eighmey, J., & McCord, L. (1998). Adding value LQWKHLQIRUPDWLRQDJH 8VHVDQG*UDWL¿FDWLRQVRI sites on the World Wide Web. Journal of Business Research, 41(3), 187-194. Everitt, B. (1977). Cluster analysis. London: Heinemann.
REFERENCES Alwitt, L.F., & Prabhaker, P.R. (1992). Functional and belief dimensions of attitudes to television. Journal of Advertising Research, 32(5), 30-42.
Firth, L., & Kelly, T. (2001). The economic and regulatory implications of broadband. Geneva, ITU. Retrieved August 5, 2005, from http://www. LWXLQWRVJVSXQLEURDGEDQGZRUNVKRSEULH¿QJSDSHU¿QDOGRF
Anderson, B., Gale, C., Jones, M.L.R., & McWilliam, A. (2002). Domesticating broadbandwhat FRQVXPHUVUHDOO\GRZLWKÀDWUDWHDOZD\VRQDQG fast Internet access. BT Technology Journal, 20(1), 103-114.
Flanagin, A.J., & Metzger, M.J. (2001). Internet use in the contemporary media environment. Human Communication Research, 27(1), 153-181.
$QJOHPDQ6 8VHVDQG*UDWL¿FDWLRQVDQG ,QWHUQHWSUR¿OHV $IDFWRUDQDO\VLV,V,QWHUQHWXVH and travel to cyberspace reinforced by unrealized JUDWL¿FDWLRQV"Proceedings of the Western Science Social Association 2001 Conference, Reno, NV. %XQ] 8. 8VDELOLW\ DQG JUDWL¿FDWLRQV Towards a Web site analysis model. Proceedings of the National Communication Association Convention, Atlanta, GA. Choudrie, J., & Dwivedi, Y.K. (2005). The demographics of broadband residential consumers of a British local community: The London Borough of Hillingdon. Journal of Computer Information Systems, 45(4), 93-101.
Katz, Gurevitch, & Haas. (1973). On the use of mass media for important things. American Sociological Review, 38(2), 164-181. Katz, Blumler, & Gurevitch. (1974). Uses and *UDWL¿FDWLRQVUHVHDUFKPublic Opinion Quarterly, 37(4). /LQ&$ 3HUFHLYHGJUDWL¿FDWLRQVRIRQOLQH media service use among potential users. Telematics and Informatics, 19(1), 3-19. Leung, L. (2003). Impacts of Net-generation attributes, seductive properties of the Internet, and JUDWL¿FDWLRQVREWDLQHGRQ,QWHUQHWXVHTelematics and Informatics, 20, 107-129. /HXQJ/ :HL5 7KHJUDWL¿FDWLRQVRI pager use: Sociability, information-seeking, enter-
377
7KH8VHVDQG*UDWL¿FDWLRQVRI%URDGEDQG,QWHUQHW
tainment, utility and fashion and status. Telematics and Informatics, 15, 152-264. Maltha S., Schuurman, K., Vermaas, K., Vandeberg, R., Bongers, F., Bekkers, R., & van de Wijngaert, L. (2002). Breedband en de gebruiker. Utrecht: Dialogic. Maltha, S., Bongers, F., Schuurman, K., Vandeberg, R., Vermaas, K., & van de Wijngaert, L. (2003). Breedband en de gebruiker 2003. Utrecht: Dialogic. McKinsey & Company & Digital Age Design. (2002). Retrieved from http://www.dad.be/press_ corner/pdf/persbericht_econsument_NL.pdf McQuail, D. (1987). Mass communication theory: An introduction (2nd ed.). London: Sage. Multiscope. (2002). Retrieved May 2, 2003, from http://www.multiscope.nl Papacharissi, Z., & Rubin, A.M. (2000). Predictors of Internet use. Journal of Broadcasting and Electronic Media, 44(2), 175-220. Rubin, A.M. (1994). Media uses and effects: A 8VHVDQG*UDWL¿FDWLRQVSHUVSHFWLYH,Q-%U\DQW & D. Zillmann (Eds.), Media effects advances in theory and research. Hillsdale, NJ: Lawrence Erlbaum. Thierer, A. (2002). Solving the broadband paradox. Issues in Science and Technology, 18(3), 57-62. Vermaas, K., & van de Wijngaert, L. (2007). A longitudinal study to investigate consumer/user adoption and use of broadband technology in The Netherlands. In Y. Dwivedi et al. (Eds.), Consumer adoption and usage of broadband. Hershey, PA: Idea Group. van de Wijngaert (1999, January 18-23). Matching media. Enschede: Telematica Instituut.
378
Wales, C., Sacks, G., & Firth, L. (2003). Killer applications versus killer attributes in broadband demand. Proceedings of the PTC Conference, Hawaii.
KEY TERMS Broadband:$FRQQHFWLRQWKDWLVµDOZD\VRQ¶ SDLGIRUDWDµÀDWUDWH¶DQGVXLWDEOHIRUJRRGTXDOLW\ video and audio applications, and for exchangLQJ H[WHQVLYH GDWD ¿OHV $'6/ FDEOH DQG ¿EHU optics. Cluster Analysis: With cluster analysis cases, for example, people are clustered into groups so that the relationship is strong between members of the same cluster and weak between members of different clusters. Internet: Worldwide accessible network of interconnected computer networks that transmit data by using the standard Internet protocol (IP). Users: All people that use broadband or narrowband Internet. 8VHVDQG*UDWL¿FDWLRQV 8 * The basic idea of the U&G approach is that people use media in order to gratify needs.
ENDNOTE 1
Note that the Dutch education system applies: MAVO (pre-vocational secondary education), HAVO (senior general secondary education), VWO (pre-university education), MBO (professional training/middle management training), HBO (universities of professional education (UPEs), and university (Source: The Dutch Ministry of Education, Culture, and Science).
Division III
Middle East
380
Chapter XXIV
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia Vishanth Weerakkody Brunel University, UK
ABSTRACT Utilizing a survey approach, this research set out to explore the reasons for the slow progress in broadband adoption and investigates the factors that may be affecting the adoption of broadband by KSA consumers. Particular emphasis was placed on individual-level factors such as social and cultural LQÀXHQFHV7KHNH\¿QGLQJVZHUHWKDWWKHIDFWRUVZLWKWKHPDLQLQÀXHQFHRQDWWLWXGHWRZDUGVDGRSWLRQ of broadband were: (1) usefulness, (2) service quality, (3) age, (4) usage, (5) type of connection, and (6) type of accommodation. Contrary to prediction, although socio-cultural factors such as regulation WKURXJK¿OWUDWLRQRIEURDGEDQGZHUHIRXQGWRKDYHQRVLJQL¿FDQWLQÀXHQFHRQWKHDGRSWLRQRIEURDGEDQG consumers were aware and largely did not like the regulation. The chapter also provides a discussion on research implications, limitations, and future directions.
INTRODUCTION Broadband is commonly used to refer to highspeed Internet access via cable modems or digital subscriber line (DSL), which is faster than dial-up (or narrowband). Broadband is considered vital for the growth and diffusion of emerging e-services including e-learning and e-government services (Bose, 2004; Choudrie & Ghinea, 2005; Choudrie & Dwivedi, 2005a, 2006a). This is because broadband users are more likely to be aware of new e-government services (Dwivedi, Choudrie, & Brinkman, 2006b). However, the Internet in general
has not been adapted globally at the same time or the same rate. Some countries are the leaders and others simply follow them. The adoption and diffusion literature suggests many factors are likely WRLQÀXHQFHWKHSURFHVVRIDGRSWLQJWKH,QWHUQHW Such factors include cultural, social, economical, skills, service quality, resources, and technological factors (Choudrie & Dwivedi, 2005a, 2006a, 2006b; Oh, Ahn, & Kim, 2003). This chapter is concerned with the adoption of broadband in the Kingdom of Saudi Arabia (KSA), therefore there is a need to have some background information about the KSA. The following offers a
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
brief description and discussion on the development of information and communication technology ,&7 LQJHQHUDOLQWKH.6$ZLWKVSHFL¿FIRFXVRQ broadband technology. The Internet is a relatively new technology in the region. King Abdulaziz city for Science & Technology is responsible for starting the Internet service in 1997. It has put in place the policies and procedures for using the Internet in arrangement with other related government and private sector organizations. The city also trained its staff and put the layout and design of the new network that become the main vessel that transfers information all over the kingdom. Currently, there are three main types of Internet access are available in the KSA: (1) dial-up, which is the most common one; (2) DSL broadband, which is the focus of this research; and (3) satellite, which is comparatively expensive but not as popular. The Internet Services Unit (ISU) in King Abdulaziz city for Science & Technology has many roles to play, such as connecting the service provider with the Internet so to provide it to the customer, connecting all the Saudi colleges and universities to the Internet, and the most important role is the ¿OWUDWLRQ RI :HE VLWHV ,Q WKH Council of Ministers Resolution prohibited users within the Kingdom of Saudi Arabia from publishing or accessing certain content on the Internet. The ISU operates the high-speed data links that connect the country to the international Internet. While Saudi Internet users may subscribe to any of a number of ORFDO,QWHUQHWVHUYLFHSURYLGHUVDOO:HEWUDI¿FLV forwarded through a central array of proxy servers at the ISU, which implements Internet content ¿OWHULQJURXJKO\LQOLQHZLWKSDUWVRIWKHUHVROXWLRQ The ISU blocks any Web site that contains sexually explicit content and sites that are related to drugs, bombs, alcohol, gambling, and pages insulting the Islamic religion or Saudi laws and regulations. In the KSA the Internet has taken a while to diffuse and is therefore seen as a relatively new WHFKQRORJ\ 7KH .6$ ¿UVW VWDUWHG ZLWK GLDOXS connections and then moved on to adopt broadband and satellite connections to provide better data communication services to its citizens. However, even with broadband technology, the number of Internet connections is considered to be relatively
poor in comparison to other developed countries such as the UK, as well as newly industrialized leading broadband users such as South Korea (Oh et al., 2003). This poor connectivity is often claimed WREHFDXVHGE\:HEVLWH¿OWUDWLRQLQWKHUHJLRQ Consequently, broadband adoption has been slower than expected in the region. Furthermore, research that examines broadband adoption from the individual perspective has not yet been undertaken in the KSA and other developing countries in the region. Therefore, this research aims to explore the reasons for the slow adoption by examining the individual-level factors affecting the adoption of broadband in the KSA. The research will thereby seek to identify individual-level factors, DQGDWWHPSWWRH[DPLQHZK\DQGKRZWKHLGHQWL¿HG IDFWRUV DIIHFW FRQVXPHUV¶ DWWLWXGHV WRZDUGV WKH adoption of broadband in the region. The chapter then provides some implications and recommendations to the government and ISPs for encouraging broadband adoption within the KSA. Having introduced the topic of interest, this chapter now provides a brief discussion on the background literature in the next section. That is followed by a brief discussion on the theoretical basis for this research and a brief discussion of the XWLOL]HG UHVHDUFK PHWKRGV 7KH ¿QGLQJV DUH WKH SUHVHQWHGIROORZHGE\GLVFXVVLRQDQG¿QDOO\D conclusion including the contributions and limitations to the research are provided.
BACKGROUND LITERATURE The primary aim of this study was to investigate the factors that affect the adoption of broadband by consumers of the KSA. There are relatively few studies that have examined broadband adoption from the micro perspective (i.e., individual level) in general, and as per our understanding, no such undertaking has occurred in the KSA. This section EULHÀ\GLVFXVVHVWKHDYDLODEOHOLWHUDWXUHLQWKHDUHD and outlines the theoretical basis of this study. The adoption literature discussed henceforth provides discussions of both macro and micro factors that drive the success or slow uptake of broadband deployment in the context of leading
381
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
countries in terms of broadband penetration. As stated earlier, research on the topic of broadband adoption at the micro level in general (Crabtree, 2003; Oh et al., 2003; Stanton, 2004) and particularly in developing countries such as the KSA is minimal. In an initial study of broadband deployPHQWLQ6RXWK.RUHD/HH2¶.HHIHDQG<XQ LGHQWL¿HGWKUHHPDMRUIDFWRUVFRPSULVLQJSXEOLF sector actions, private sector actions, and the sociocultural environment factors that explained the high rate of broadband adoption in South Korea. Further research suggested that six success factors are responsible for driving the high penetration rate of broadband within South Korean residential consumers (Choudrie & Lee, 2004). These six key IDFWRUVFRQVLVWRIWKHJRYHUQPHQW¶VYLVLRQVWUDWHJ\ and commitment, facilities-based competition, pricing, the PC Baang phenomenon, culture and geography, and demographics (Choudrie & Lee, 2004). Among initial studies to understand the adoption of broadband, the work by Oh et al. (2003) is of high relevance. This study examined the individual-level factors affecting the adoption of broadband access in South Korea by combining factors taken from 5RJHUV¶GLIIXVLRQWKHRU\DQG the technology acceptance model (Oh et al., 2003). 7KH¿QGLQJVRIWKLVVWXG\VXJJHVWWKDWFRQJUXHQW experiences and opportunities in adopting a new technology affect user attitudes through the three extended technology acceptance model constructs, namely, perceived usefulness, perceived skills, and perceived resources (Oh et al., 2003). In a recent study Stanton (2004) emphasized the need to conduct an inquiry that addresses the access question and the demographics that would be useful to observe any changes in the likelihood of socioeconomic groups when adopting broadband. Following on from these studies, demographics of the UK broadband consumers were examined at both the local level (Choudrie & Dwivedi, 2005a) and the national level (Choudrie & Dwivedi, 2006a). Both the aforementioned studies highlighted the role of demographic variables such as age, income, education, and occupation in broadband adoption (Choudrie & Dwivedi, 2005a, 2006a). Further exploratory studies by the same authors
382
examined attitudinal (i.e., relative advantage, utilitarian outcomes, and hedonic outcomes) and control (skills, knowledge, facilitating conditions resources) variables to understand behavior of broadband adopters and nonadopters. These studies suggested that the aforementioned variables are important for understanding broadband adoption behavior (Choudrie & Dwivedi, 2006b). Building XSRQSUHYLRXVVWXGLHVDFRQ¿UPDWRU\VWXG\ZDV also conducted to develop and validate a reliable survey instrument for measuring the perception of consumers regarding adoption of broadband in the UK (Dwivedi et al., 2006a). While the abovementioned discussion suggests that although both macro- and microlevel studies were conducted to understand deployment of broadband in the developed world and leading broadband users (countries) such as South Korea, no studies have focused upon developing countries (Dwivedi, Khan, & Papazafeiropoulou, 2006c). The reason could be slow infrastructure development in these countries and the very low rate of adoption as mentioned before. This has provided the motivation for the researchers to undertake this exploratory study in order to understand the perception of consumers regarding broadband adoption and its usefulness in the KSA. Its worthwhile contribution would be to understand reasons for consumer adoption and nonadoption of subscription-based technologies such as broadband in the developing world. The QH[WVHFWLRQEULHÀ\GLVFXVVHVWKHWKHRUHWLFDOEDVLV for examining the factors of consumer adoption and nonadoption of broadband.
THEORETICAL BASIS The theoretical constructs included in this study are based on the discussion presented above. This study considered that the constructs such as usefulness, resources, skills, service quality, compatibility, relative advantage of technology, usage social cultural factors (for example, language DQGUHJXODWLRQWKURXJK¿OWUDWLRQ DQGFRQWUROYDULables such as age, gender, education, occupation, and income (Choudrie & Dwivedi, 2005a, 2006a; Dwivedi, 2005; Dwivedi et al., 2006b; Dwivedi,
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
Papazafeiropoulou, & Brinkman, 2006d; Oh et al., 2003; Rogers, 1995), type of accommodation, and whether one works from home or not (Choudrie /HH PD\LQÀXHQFHFRQVXPHUV¶FLWL]HQV¶ attitudes towards broadband adoption in the context of the KSA. All aforementioned metric constructs were expected to have a positive impact on attitude towards broadband adoption, with the exception of sociocultural factors that are expected to have a negative impact. However, the impact of each of the constructs (factors) was likely to vary; that means some factors have more impact than others. Detailed discussion on each factor is not possible within the scope of this chapter due to limited space. However, interested readers may refer to original sources that have successfully utilized the aforementioned constructs in existing studies.
RESEARCH METHODOLOGY To examine broadband adoption in the KSA, the researchers considered a survey as a suitable research method (Cornford & Smithson, 1996; Choudrie & Dwivedi, 2005b). A number of techniques are available to capture the data; however, a self-administered questionnaire was considered to be the primary survey instrument for data collection in this investigation. This is because it addresses the issue of reliability of information by reducing and eliminating differences in the way that the questions are asked and how they are presented (Cornford & Smithson, 1996). Furthermore, questionnaires facilitate the collection of data within a short period of time from the majority of respondents, and this was a critical issue for this research (Fowler, 2002). Therefore mainly multiple and closed questions were included in the questionnaire. The literature review provided an initial understanding of the broadband adoption and usage, and the basis for WKHGHYHORSPHQWRIDGUDIW7KH¿QDOTXHVWLRQQDLUH consisted of a total of 18 questions that included close-ended, multiple, and Likert-scale-type quesWLRQV 7KH /LNHUWVFDOHW\SH ¿YHSRLQW VFDOH questions were adapted from various sources mentioned in the theoretical basis subsection,
and demographic categories were adapted from Choudrie and Dwivedi (2005a, 2006a). Due to the uncertainty regarding personnel using the broadband facility, the researchers adopted the snowball or chain sampling (Fridah, 2002) method when selecting the respondents IRU WKH VXUYH\ ,Q RUGHU WR LGHQWLI\ WKH ¿UVW IHZ respondents with Internet connection, researchers approached friends and colleagues who possess the broadband connections at home using e-mail in order to complete the questionnaire. Aforementioned respondents were also requested to recommend their friends and family contacts that had Internet connections at home. This strategy led to the questionnaire being administered to a total of 150 broadband users via e-mail attachment during July and August 2005. The majority of respondents who replied were located in two of the large cities of the KSA which are Riyadh, the capital, and Jeddah. Of the 150 questionnaires administered, 138 respondents returned the completed questionnaire, which helped to obtain a response rate of 92%.
Data Analysis The initial stage of data analysis involved checking WKHUHVSRQVHVDQGSURYLGLQJDXQLTXHLGHQWL¿FDWLRQ number to each response. Using SPSS (version 11.5) the researchers generated the descriptive statistics (i.e., percentage and tables) and conducted regression analysis to analyze and present the research data obtained from the questionnaire in order to determine the extent to which the factors, proposed HDUOLHU LQÀXHQFH LQGLYLGXDOV¶ DWWLWXGHV WRZDUGV broadband adoption in the KSA. Multivariate linear regression was selected as the most appropriate technique because it allowed examining the impact of the factors on broadband adoption. In order to PHDVXUHWKHLQÀXHQFHRIDGRSWLRQIDFWRUVRQDWtitude towards broadband adoption, the data had to be transformed in the following manner. For the factors with Likert-scale questions (i.e., usefulness, skills, resources, compatibility, technology, service quality, and social/cultural factors), the average for each factor was computed and then subsequently used. For example, usefulness had four key statements or items, which were averaged to compute
383
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
a mean score and a new IV was created, which was called usefulness. The control variables with multiple-type questions were transformed to binary format. In the case of male/female, the control variables were allocated 1 (male) and 2 (female) and treated as categorical data format. The similar categorization was done IRUDFFRPPRGDWLRQW\SH LHKRXVHÀDW 9DULDEOHV with multiple categories were transformed into two categories (binary format) as follows. For age, those below 25 years were grouped as 1 (young) and those above 25 (middle to old) were grouped as 2. For education, those up to secondary level were put in 1 (less educated) while those above secondary (undergraduate and postgraduate) were put as 2 (highly educated). For usage, those who XVHXSWRWZRKRXUVZHUHFODVVL¿HGDV ORZXVers), while those who use more than two hours per GD\ZHUHFODVVL¿HGDV KLJKXVHUV )RUWKHXVHRI Internet, broadband and satellite were combined DV KLJKVSHHG ZKLOHGLDOXSZDVFODVVL¿HGDV 2 (slow speed).
RESEARCH FINDINGS This section begins by presenting demographic SUR¿OHRIUHVSRQGHQWV7KLVLVWKHQIROORZHGE\D detailed statistical analysis, namely linear regresVLRQ /LQHDU UHJUHVVLRQ LGHQWL¿HG WKH YDULDEOHV WKDW KDG VLJQL¿FDQW LPSDFW RQ WKH LQGLYLGXDO¶V attitude.
5HVSRQGHQWV3UR¿OH Of the 138 completed questionnaires that were returned, two questionnaires were discarded because they were not completed in full. This meant that D¿QDOVDPSOHRITXHVWLRQQDLUHVZDVXVHGIRU all subsequent analysis. Of the 136 respondents, 92 (68%) used broadband, while 32% did not use broadband, although all respondents had knowledge about broadband technology. This meant that about 32% of Internet users still used dial-up as opposed to broadband. The respondents were also fairly well represented, with women being 44% of the respondents. The majority of the respondents
384
were aged between 18 years and 50 years. With regard to education, only 5% were educated to the secondary level, while 57% were graduates and 38% had postgraduate levels of education. The usage level (number of hours on the Internet per day) was normally distributed, and the majority of respondents spend on average three hours a day on the Internet. Of these, 72% also work from home, while the remaining 28% work away from KRPH LHLQDQRI¿FHHQYLURQPHQW $ERXW of them use the Internet for personal use, while the remaining 77% use it for both business and personal use. Of the respondents, 67% live in houses, ZKLOHOLYHLQÀDWVDSDUWPHQWV:LWKUHJDUGWR the connection type, 55% use broadband, 12% use satellite, while 32% still used dial-up. This means that 68% use high-speed connections.
Adoption of Broadband: Descriptive Statistics Before conducting any tests to determine the impact of factors affecting broadband adoption on WKHUHVSRQGHQW¶VDWWLWXGHWRDGRSWEURDGEDQGLWLV important to do a descriptive analysis of the data. Descriptive data analysis is important because it provides the reader with an appreciation of the actual numbers and values, and hence the scale the researcher is dealing with. This assists in the conceptualization of the results of advanced statistical tests such linear regression analysis realistically. This assists the user in selecting a strategy that is QRW MXVW EDVHG RQ VWDWLVWLFDO VLJQL¿FDQFH EXW RQ SUDFWLFDOVLJQL¿FDQFHDQGLWVLPSOLFDWLRQV+HQFH this subsection presents the basic descriptive analysis results. As shown in Table 1, the average scores for UHVSRQGHQWV¶DWWLWXGHWRZDUGVEURDGEDQGUDQJHG from 4.43 to 4.72 (where 1 = minimum and 5 = maximum). Descriptive statistics show that these scores are quite high, and that respondents tend to have a positive attitude towards broadband. $VVKRZQLQ7DEOHWKHVFRUHVIRUµ8VHIXOQHVV RI%URDGEDQG¶UDQJHGIURP WRLQGLFDWing that the scale is quite uniform. These scores DUHDOVRTXLWHKLJK)RUµ6NLOOV¶WKHVFRUHUDQJHG from 3.99 to 4.10, which are average scores. This
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
PHDQVWKDWUHVSRQGHQWV¿QGLWUHODWLYHO\HDV\WR use broadband, and that they can also learn by WKHPVHOYHV:LWKUHJDUGWRµ5HVRXUFHV¶WKHUHVXOWV show that the scores ranged from 3.63 to 4.04. The lowest score of 3.63 indicates that quite a few of the respondents have some limitations on resources, LQGLFDWLQJWKDWWKH\PD\¿QGWKHSULFHRIEURDGEDQG to be too high for them. Other limitations could EHDFFHVVDQGVRPHXQVSHFL¿HGEDUULHUVRIVRPH sorts. However, on balance, the majority of the UHVSRQGHQWVGRQRW¿QGWKHVHOLPLWDWLRQVWREHD hindrance to broadband adoption. The scores for µ&RPSDWLELOLW\¶ZHUHTXLWHKLJKZLWKDQDYHUDJH RI7KLVLQGLFDWHVWKDWPRVWSHRSOH¿QGWKHLU situations (both home and work) to be compatible with the use of broadband. :LWKUHJDUGWRµ5HODWLYH$GYDQWDJHRI7HFKQRORJ\¶UHVSRQGHQWVVWURQJO\DJUHHWKDWEURDGEDQG is faster than dial-up, with an average score of a very high 4.67. However, a sizeable proportion of respondents do not think that broadband is a very secure technology, with a neutral score of 3.36, which is quite low compared to the rest of the scores described thus far. For most respondents, EURDGEDQGLVUHDGLO\DYDLODEOH7KHµ6HUYLFH QualLW\¶VFRUHVZHUHQHXWUDOUDQJLQJIURP WR 7KLVLQGLFDWHVWKDWDVLJQL¿FDQWSURSRUWLRQRIWKH respondents do not think that the service quality of broadband is not satisfactory. This may not be surprising, given the fact that in the KSA, there is only one broadband provider. Both reliability and the level of technical support had low scores, indicating that there are important issues relating to WHFKQLFDOHI¿FLHQF\DQGUHOLDELOLW\WKDWWKHSURYLGHU should address. A few respondents also do not have broadband in their areas (work or home). µ6RFLDODQG&XOWXUDO¶VFRUHVDUHWKHLPSRUWDQW tenet of this study. Table 1 presents the mean scores for each of the key statements (i.e., items). As shown in Table 1, respondents tended to score the social and cultural factors quite low, with scores ranging from 2.11 to 2.75, with only one factor scoring above 3.46. Respondents do not think that UHOLJLRQKDVDQLPSRUWDQWLQÀXHQFHRQWKHVORZUDWH at which broadband is being adopted in the KSA (mean score = 2.11), despite that the KSA is one of the most religious countries in the world, with
UHOLJLRQKDYLQJVWURQJLQÀXHQFHRQWKHGD\WRGD\ lives of people. The respondents also do not feel that Saudi traditions (mean score = 2.59) have inÀXHQFHRQWKHUDWHDWZKLFKEURDGEDQGLVDGRSWHG by individuals. A similarly low score (mean = 2.58) was recorded for conservatism, in which the respondents do feel that conservatism has a negaWLYHLQÀXHQFHRQWKHVORZDGRSWLRQRIEURDGEDQG Resistance to change also has a low score (mean = 2.53), which means that although KSA has always been associated with resistance to change, partly H[SODLQHGE\WKHFRXQWU\¶VFXOWXUHUHVSRQGHQWVGR QRWIHHOWKDWWKLVUHVLVWDQFHWRFKDQJHLQÀXHQFHV the adoption of broadband. The same applies to WKH QDWXUH RI WKH VRFLHW\ LQÀXHQFLQJ EURDGEDQG awareness (mean score = 2.75). However, there is one sociocultural factor that respondents feel affects the adoption of broadband. This factor is the UHJXODWLRQRIEURDGEDQGWKURXJK¿OWUDWLRQZKLFK UHVSRQGHQWVEHOLHYHKDVVWURQJLQÀXHQFHRQWKHVORZ uptake of broadband. Preliminary evidence seems to indicate that regulation appears to have negative LQÀXHQFHRQEURDGEDQGDGRSWLRQ+RZHYHUJLYHQ that this score is 3.46, just above the neutral score, its impact can only be measured using advanced statistical techniques, which is the subject of the next section. Table 1 also illustrates the average means of the above discussed constructs, rated according to their average score. Average scores of all items for each construct suggest that the respondents considered usefulness of broadband most important (average mean = 4.3). Two constructsskills and compatibilitywere considered equally important with average mean of 4.1. The relative advantage of technology was considered third most important (average mean = 4.0) factor among seven that LQÀXHQFHVDWWLWXGHWRZDUGVEURDGEDQG5HVRXUFHV construct was considered fourth most important (average mean = 3.8) followed by service quality (average mean = 3.4). The least important rated construct was sociocultural factor, with average mean of 2.7. To conclude this section, descriptive statistics have provided the indications of the levels of agreement of respondents on the various broadband DGRSWLRQIDFWRUV7KHQH[WVHFWLRQSUHVHQWV¿QGLQJV
385
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
Table 1. The descriptive statistics for the attitude and factors affecting attitude towards broadband adoption Factors
Mean
AM
SD
4.6
0.73
ATTITUDE TOWARDS BROADBAND Using broadband is a good idea.
4.72
8VLQJEURDGEDQGLVEHQH¿FLDO
4.65
0.66
Using broadband is pleasant.
4.43
0.95
Using broadband generally increases my productivity.
4.30
0.75
Using broadband improves my work performance.
4.26
FACTORS A FFECTING ATTITUDE TOWARDS BROADBAND A DOPTION Usefulness of Broadband
0.87 4.3 (1)
Using broadband makes my work easier.
4.26
0.79
Using broadband is useful for my work.
4.19
0.89
Skills Learning to operate broadband is easy for me.
4.10
,¿QGEURDGEDQGHDV\WRXVH
4.10
0.96
I clearly understand how to use broadband.
3.99
1.24
I would be able to use broadband if I wanted to.
4.04
1.14
I have access to resources that I need for broadband.
3.72
There are no barriers to me using the broadband.
3.66
1.27
I have the resources, opportunities, and knowledge for using broadband.
3.63
1.16
4.1 (2)
1.02
Resources
1.24 3.8 (4)
Compatibility Using broadband is compatible with my work and life.
4.19
8VLQJEURDGEDQG¿WVZHOOZLWKWKHZD\,OLNHWRZRUN
4.10
0.75 4.1 (2) 0.95
Relative Advantage of Technology Broadband is very fast compared to dial-up.
4.64
Broadband is a very reliable and robust technology.
3.90
Broadband is a very secure technology.
3.36
0.69 4.0 (3)
0.82 0.79
Service Quality Broadband is readily available in the area that I live/work.
3.54
7KHEURDGEDQGWHFKQLFDOVXSSRUWLVYHU\HI¿FLHQW
3.41
1.41
The broadband service providers are very reliable.
3.29
1.09
5HJXODWLRQRIEURDGEDQGWKURXJK¿OWUDWLRQKDVVWURQJLQÀXHQFHRQWKHVORZXSWDNHRIEURDGEDQG
3.46
1.12
7KHQDWXUHRIWKH6DXGLVRFLHW\KDVVWURQJLQÀXHQFHRQWKHORZOHYHOVRIDZDUHQHVVDERXWEURDGEDQG
2.75
1.31
6DXGLWUDGLWLRQVKDYHDVWURQJLQÀXHQFHRQVORZDGRSWLRQRIEURDGEDQG
2.59
1.30
7KHFRQVHUYDWLVPRI6DXGLVKDVLQÀXHQFHRQWKHVORZDGRSWLRQRIEURDGEDQG
2.58
5HVLVWDQFHWRFKDQJHLQWKH6DXGLVRFLHW\KDVVWURQJLQÀXHQFHRQWKHVORZDGRSWLRQRIEURDGEDQG
2.53
1.27
5HOLJLRQKDVDVWURQJLQÀXHQFHRQWKHVORZDGRSWLRQRIEURDGEDQG
2.11
1.28
3.4 (5)
0.99
Social and Cultural Factors
2.7 (6) 1.32
SD = standard deviation; scores range from 1 to 5, where 1 = strongly disagree and 5 = strongly agree
386
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
from regression analysis, which was conducted to assess the overall impact of the adoption factors on attitude towards broadband adoption.
Regression Analysis: Factors ,QÀXHQFLQJWKH$WWLWXGH7RZDUGV Adoption of Broadband A regression analysis was conducted with attitude towards broadband adoption as the dependent variable and usefulness, skills, resources, compatibility, technology, service quality, social/cultural, and control factors as predictor variables. A total of 136 cases were analyzed. From the analysis, a VLJQL¿FDQWPRGHOHPHUJHG F (16, 136) = 4.544, p < .001) (see Table 2). The second important statistic that was obtained from the analysis is the R2, which ranges from 0 to 1, with 1 being a perfect ¿WPRGHO7KHR2 for this analysis was found to be 0.396. This means that the factors (IVs) explain 40% RIWKHFKDQJHVLQWKHDWWLWXGH2WKHUXQLGHQWL¿HG factors account for the remaining 60%. The R2 of 0.4 (40%) is considered as a good value for a crosssectional data involving many predictor variables. The overall regression analysis results suggest that RXWRIWKHYDULDEOHVLGHQWL¿HGLQWKHEURDGOLWHUDWXUH as having an impact on attitude towards broadband DGRSWLRQLQWKHFRQWH[WRIWKH.6$WKHVLJQL¿FDQW YDULDEOHVZHUH WKHUHVSRQGHQWV¶SHUFHLYHGXVHIXOness of broadband, the service quality, the age of the respondent, the type of connection, and the type of accommodation. The remaining variables ZHUHIRXQGWRKDYHLQVLJQL¿FDQWRUQRLPSDFWRQ dependent variable. Table 2 shows that of all the IDFWRUVWKHIROORZLQJKDYHVLJQL¿FDQWLPSDFWRQ attitude towards adoption of broadband: usefulness (Eҏ p = .015), service quality (E ҏ p = .036), age (E= -.384, p < .001), usage (E ҏ p = .009), connection type (E ҏ p = .010), and type of house (Eҏ ҏ p = .013). Also, as shown in Table 2, the sociocultural factor was found to EHLQVLJQL¿FDQW7KLVLQGLFDWHVWKDWZKLOH6DXGLV feel that there is some level of control of broadband by the authorities (the state), they do not feel that this control affects their attitudes adopting the broadband technology.
DISCUSSION This research established that the key factors that affect attitude towards broadband in the KSA are usefulness, service quality, age, usage, type of connection, and the accommodation type. The rest RIWKHYDULDEOHVZHUHIRXQGWRKDYHLQVLJQL¿FDQW LPSDFW7KH¿QGLQJVGLIIHUDVZHOODVFRPSOHPHQW previous published work by other authors. That is, VRPHRIWKH¿QGLQJVIURPWKLVUHVHDUFKVXSSRUW VRPHSUHYLRXVUHVHDUFKZKLOHWKHRWKHU¿QGLQJV do not support previous empirical research by other DXWKRUV7KHOLWHUDWXUHKDVLGHQWL¿HGDOOWKHDERYH variables to affect the adoption of a technology such as broadband (Choudrie & Dwivedi, 2005a, 2006b; Choudrie & Lee, 2004; Oh et al., 2003). However, in this study, only a few of these many IDFWRUVZHUHIRXQGWRVLJQL¿FDQWO\LQÀXHQFHWKH .6$FRQVXPHUV¶DWWLWXGHVWRZDUGVWKHDGRSWLRQ of broadband. The following paragraphs attempt WRSURYLGHH[SODQDWLRQVIRUWKH¿QGLQJV The usefulness construct of a technology is LPSRUWDQW LQ WKH GHFLVLRQ SURFHVV DQG ¿QGLQJV clearly indicate that the KSA consumers ranked this factor strongly. It may be an indication that the service provider is not creating enough awareness about the products and services that broadband can offer (Dwivedi, 2005; Oh et al., 2003). Despite the fact that the service quality was previously LGHQWL¿HGDVEHLQJFULWLFDOWRWKHDGRSWLRQRIDQHZ innovation including broadband (Dwivedi et al., 2006d) on a larger scale, there is strong evidence that the KSA broadband service provider is not satisfying customers with the service quality they provide. A possible explanation for this poor service is the lack of competition, since the KSA has only a single broadband provider. Early studies carried out in the UK on the broadband market suggest that possibly lack of competition caused by incumbent monopolist of the market was leading to poor quality service, and also that consumers were not getting value for their money (Choudrie & Lee, 2004). On the other hand, studies carried out in South Korea where broadband penetration was most successful in the world (Oh et al., 2003) showed that the quality of service was excellent. This indicates that high quality of service has a
387
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
Table 2. Regression analysis results ANOVA Sum of Squares
df
F
Sig.
19.50
16
4.54
.000a
Model
Ra
R Square
Adjusted R
Std. Error of the
Square
Estimate
1
.629
.396
.309
.52
Standardized
T statistic
6LJQL¿FDQFH
2.293
.024*
Regression Regression Model
Factors Affecting Broadband Adoption Broadband Adoption Factor
Beta
(Constant)
2.201
Usefulness
.249
.205
2.467
.015*
Skills
.111
.136
1.347
.181
Resources
.038
.049
.512
.610
Beta
Compatibility
.088
.109
1.210
.229
Technology
.151
.119
1.403
.164
Service Quality
.152
.200
2.123
.036*
Socio-Cultural Factor
-.005
-.006
-.072
.943
Age
-.502
-.384
-3.612
.000*
Education
.017
.005
.063
.950
Usage
.294
.213
2.676
.009*
Purpose
-.049
-.033
-.433
.666
Connection Type
.292
.220
2.622
.010*
Gender
-.051
-.041
-.493
.623
Language
-.299
-.102
-1.299
.197
Home Working
-.003
-.002
-.027
.979
House Type
-.273
-.204
-2.537
.013*
a
Predictors: (Constant), House Type, Language, Home Working, Resources, Purpose, Usage, Usefulness, Gender,
Education, Connection, Socio-Cultural, Technology, Service Quality, Compatibility, Skills, and Age; Dependent Variable: Attitudes; * 9DULDEOHLVVLJQL¿FDQW (p <= 0.05)
YHU\LQÀXHQWLDOLPSDFWRQWKHPDVVDGRSWLRQRI broadband technology. Therefore, it can be argued that the situation in the KSA could be mirroring that in the UK in comparison to South Korea, where services were seen to be of high quality. In South Korea the high level of broadband penetration was partly attributed to the level of competition and also the general level of education that broadband users possessed.
388
In line with the arguments presented earlier, age ZDVIRXQGWREHVLJQL¿FDQWZLWK\RXQJHUSHRSOH being more likely to adopt broadband compared to older people. It was also found that demographic YDULDEOH µW\SH RI DFFRPPRGDWLRQ¶ VLJQL¿FDQWO\ LQÀXHQFHGWKHDWWLWXGHWRZDUGVEURDGEDQGDGRSWLRQ7KRVHZKROLYHLQÀDWVDSDUWPHQWVDUHPRUH likely to adopt broadband compared to those who live in houses. Although it is not entirely clear why
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
this is the case, a possible explanation can be that often the young and more educated are likely to be city workers and therefore more likely to live LQÀDWVDSDUWPHQWV+HQFHWKH\DUHPRUHOLNHO\WR adopt a new innovation than, for instance, citizens who live in houses in more rural areas. Another important demographic variable that has an impact on the attitude to adopt broadband is the usage factor. High users (more than two hours on the Internet per day) are more likely to adopt broadEDQGFRPSDUHGWRORZOHYHOXVHUV7KLV¿QGLQJLV consistent with literature predictions (Choudrie & Lee, 2004; Oh et al, 2003). The important fact KHUHLVWKHLGHQWL¿FDWLRQRIWKHFXWRIISRLQWRIWZR hours, which providers can use to identify low users from high users. However, contrary to expectations, many of the demographic variables did not have an impact on broadband adoption attitudes as claimed by previous studies outlined earlier. For example, in this VWXG\ HGXFDWLRQ ZDV QRW VLJQL¿FDQW $ SRVVLEOH explanation for this is that only 7 of the 136 respondents had secondary education or lower. Their LQÀXHQFHLQWKHUHJUHVVLRQPD\KDYHEHHQVPDOO because of their small representation. Another possible explanation is that Internet technology is now a standard technology; hence the impact of education may have declined because of the blurring of the gap between the educated and noneducated in their knowledge of the Internet. Therefore, the overall conclusion is that in the KSA, education has no VLJQL¿FDQWLPSDFWRQDWWLWXGHWRZDUGVEURDGEDQG adoption. Other demographic control variables WKDWZHUHQRWVLJQL¿FDQWLQFOXGHJHQGHUSXUSRVH language, and home working. With regards to gender, this research has provided indications that the KSA women do get involved in innovative products at the same rate as men. While society, particularly in developing countries, expects men to know more about technological innovation, it may be that since most people who use broadband are educated, and that women too are well educated in the KSA, the gender differences when it comes to innovative products are not as important. There is an increasing number of women in IT jobs in the KSA, which may explain why gender is not as important for explaining the attitude towards adoption of broadband. As for language, most edu-
cated Saudis can communicate in English as well as Arabic. Home working is not an issue because the respondents use Internet both at home and at work. Again, this may be because the Internet is becoming a standard service at home, similar to television for example. :LWKUHJDUGWRWKHLQVLJQL¿FDQWFRQVWUXFWVVXFK as skills, resources, compatibility, technology, and socioculture, there could be some possible explanations for these. For skills, it may be possible that broadband is perceived as an extension for the Internet, and consumers are very familiar ZLWK WKH WHFKQRORJ\ DQG DUH FRQ¿GHQW WKDW WKH\ can learn this as they adopt the technology. This is the view that the researcher observed during some interviews conducted during the data collecWLRQSURFHVV6DXGLVDUHFRQ¿GHQWRIEHLQJDEOHWR use broadband, to the extent that it does not affect their attitudes towards adoption of broadband. This is an indication that broadband technology may not be as innovative as we perceived, but just an extension of a technology that consumers are well familiar with. :LWKUHJDUGWRUHVRXUFHVRXU¿QGLQJVVKRZWKDW these have not adversely affected the attitude to DGRSWEURDGEDQG:KLOHSULFHZDVLGHQWL¿HGDVDSRtential deterrent, in the KSA this does not seem to be the case. This is not surprising, given the comments by Crabtree (2003) that the price of broadband had fallen substantially since its introduction, and this should provide some stimulus for take-up. Crabtree (2003) argues that there are many micro barriers that were preventing consumers from migrating to broadband. However, the assumption here is that there should be a natural shift from dial-up to broadband. However, this assumption needs to be WHVWHGDQGUHVHDUFKGRQHWR¿QGRXWZK\GLDOXS users may not migrate to broadband. Indeed, the KSA still has 32% of those on dial-up who have not migrated to broadband although they can, and resources are not the limiting factor in this case. This means that these consumers may need to be FRQYLQFHGRIWKHEHQH¿WVRIEURDGEDQGIRUWKHPWR migrate to broadband. Furthermore, compatibility is not an issue because consumers have access to broadband both at home and at work, and the FRQVXPHUV¿QGWKDWWKHLUOLIHVW\OHVDUHQRWDIIHFWHG
389
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
by broadband, hence there is no or minimal impact on attitude to adopt broadband. 7KH¿QGLQJVRQVRFLRFXOWXUDOIDFWRUVZHUHXQanticipated, because the prediction was that these IDFWRUV KDYH D QHJDWLYH LQÀXHQFH RQ EURDGEDQG DGRSWLRQ7KHSUHYLRXVOLWHUDWXUHDOVRLGHQWL¿HGWKDW the South Korean culture in general was a critical factor in the success of broadband adoption (Oh et al., 2003). Given the strong traditional and conservative nature of Saudi culture, and the general information that was gathered during brief interviews with some of the survey respondents, there was strong and persuasive argument to predict that these factors have a negative impact on the broadband adoption attitude. A possible explanation for the lack of impact by sociocultural factors could be that most Saudis may have in fact learned to live with the problems imposed by their sociocultural background in the region. While they are aware of regulations, most Saudis are determined not to let this affect their attitudes towards the adoption of broadbandthey will use the permissible aspects of a technology to their best ability. In this context, it is possible to suggest that most consumers are FRQYLQFHGWKDWWKHEHQH¿WVRIEURDGEDQGIDURXWweigh the limitations caused by regulations such as ¿OWUDWLRQRI,QWHUQHWFRQWHQW$QRWKHUH[SODQDWLRQ FRXOGEHWKDWVRPHFRQVXPHUVVHH¿OWUDWLRQDVD necessity to protect their sociocultural values and their children in particular from harmful Internet content. However, the extent to which this is exercised raises important issues for government to EHPRUHXSIURQWRQZKDWH[DFWO\LV¿OWHUHG7KLV coupled with an effective publicity campaign by the government, will no doubt encourage further take-up of broadband in the KSA.
CONCLUSION Broadband technology has been heralded as a natural successor to dial-up Internet technology. However, migration to broadband has not been as fast as predicted. This research set out to investiJDWHWKHIDFWRUVWKDWDIIHFWLQGLYLGXDOV¶DWWLWXGHV towards the adoption of broadband in the KSA. Using survey data from the KSA Internet users
390
and employing multivariate regression analysis, WKHVWXG\IRXQGWKDWIDFWRUVWKDWKDYHDVLJQL¿FDQW LQÀXHQFHRQWKHDGRSWLRQRIEURDGEDQGDUHXVHIXOness, service quality, age of consumer, connection type, and the accommodation type. The research established that the service quality provided by the monopolist provider in the KSA was having a negative impact on the adoption of broadband. The research also established that contrary to prediction, the sociocultural factors, although being important for consumers, did not negatively affect the adoption of broadband in the KSA. This research has important implications for academic research, the government, and the broadband service provider in the KSA. The important implications are discussed below.
Implications for the Government It is quite clear in this study that the government has an important role to help both the single broadband provider and the consumers. The implication for the government in terms of the broadband provider is to ensure that the company provides a better service quality. In a monopoly situation like this, it is often very tempting for the monopoly supplier WREHLQHI¿FLHQW7KHJRYHUQPHQWVKRXOGWKHUHIRUH consider increasing the number of service providers and help create a more competitive market for broadband provision. The UK has done this in two ZD\VLQHIIRUWVWREUHDNGRZQ%7¶VPRQRSRO\RI EURDGEDQG,QWKH8.WKH¿UVWDFWLRQKDVEHHQWR allow cable providers such as NTL and TeleWest to provide broadband services. This has worked well, as quite a substantial proportion of consumers, including public services such as schools and hospitals, use NTL/TeleWest broadband. Although BT still dominates the market, there has been a VLJQL¿FDQWPLJUDWLRQWR17/E\ERWKEXVLQHVVHV and domestic consumers. However, in the KSA, broadband is provided by only one supplier due to government regulation that seems not to encourage competition in this sector. The second action in WKH8.KDVEHHQWRµIRUFH¶%7WRVHOOEURDGEDQGDW wholesale price to many local providers who use the BT system to run their own broadband. This has proved very popular, with the emergence of
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
companies such as Wanadoo and VirginNet supplying broadband at cheaper prices and tailoring packages to individuals depending on their need levels. For example, there are very cheap packages for light users (limit on download megabytes per month) to expensive high-speed services for those who download large volumes of content such as music and multimedia. These tailored packages are proving to be very popular with consumers. This action has been encouraging as these companies are also offering other telephony services, and this industry is growing rapidly resulting in a higher uptake rate of broadband. In contrast in the KSA, there is no wholesale pricing of broadband by a sole supplier. At the individual level, the government has to conduct a publicity campaign to make consumHUVPRUHDZDUHRIWKHEHQH¿WVRIEURDGEDQG2I particular importance is the need for the governPHQWWRDGGUHVVWKHLVVXHVUHODWLQJWR¿OWUDWLRQ,W was quite clear in this study that while regulation WKURXJK ¿OWUDWLRQ KDV QR VLJQL¿FDQW LPSDFW RQ DGRSWLRQRIEURDGEDQGLQWKH.6$TXLWHDVLJQL¿cant proportion are not happy with this regulation. The government must therefore conduct a strong publicity campaign that will explain and educate FRQVXPHUVDVWRZK\¿OWUDWLRQLVQHFHVVDU\DQG KLJKOLJKW WKH ORQJWHUP EHQH¿WV RI WKLV DFWLRQ to the KSA society. There is ample evidence in literature (Lee et al., 2003) in which parents fear for Internet security, especially when children can access and download adult material with less effort using broadband. The government should be more forthright when handling what exactly is ¿OWHUHGVRWKDWLWFDQJDLQSXEOLFFRQ¿GHQFHZKLFK will ultimately lead to the increased adoption of broadband. Finally, government responsibility to increase awareness with regard to the usefulness of broadband among citizens is also important, particularly in relation to addressing the critical issue of the digital divide (Yu & Wang, 2004; ,¿QHGR 'DYLGUDMXK
Implication for the KSA Broadband Provider
broadband provider needs to convince potential XVHUVRIWKHEHQH¿WVWKDWWKH\FDQJDLQIURPXVLQJ broadband. To do this, the company must continually produce and disseminate information about WKH EHQH¿WV RI EURDGEDQG DQG WKH SURGXFWV DQG services that can be accessed using broadband. It is important for the company to realize that content cannot be separated from infrastructure matters, and hence both need to continually develop. Also, the speed of broadband connections needs to continually evolve like in the case of South Korea, where broadband speeds are up to 10 times faster than those in the UK (Oh et al., 2003). Regrettably, the broadband speed is quite limited in the KSA, and the KSA provider therefore has much to learn from the South Korean experience. $OVRWKHVHUYLFHTXDOLW\ZDVLGHQWL¿HGDVEHLQJ DVLJQL¿FDQWIDFWRULQLQÀXHQFLQJWKHEURDGEDQG adoption decision. The service quality had low scores, indicating that the provider must improve VLJQL¿FDQWO\ $OWKRXJK D PRQRSRO\ FDQ OHDG WR LQHI¿FLHQF\DQDGYDQWDJHRIDPRQRSRO\LVWKDW the single provider can divert resources aimed at ¿JKWLQJRIIFRPSHWLWLRQWRLPSURYLQJWKHVHUYLFH quality. In this context, it may be necessary for the provider to seek the services of or partner with external organizations to provide high-quality, value-added services to consumers, like in the case of BT in the UK. Finally, the KSA broadband provider also needs WRWDUJHWVSHFL¿FJURXSVRISHRSOH)LUVWWKHUHLVHYLdence that younger people are more likely to adopt broadband compared to older people. Therefore, the provider needs to devise marketing strategies to target older people in order to increase migration to broadband. Secondly, the provider needs to target people who live in houses compared to those ZKR OLYHLQ ÀDWVDSDUWPHQWV VLQFHWKLV UHVHDUFK LQGLFDWHVWKDWWKRVHZKROLYHLQÀDWVDSDUWPHQWV are more likely to adopt broadband compared to those who live in houses. However, the provider PD\ ¿UVW QHHG WR FRQGXFW VXUYH\V WR GHWHUPLQH why this is the case. Finally, the provider needs to target low users (less than two hours on the Internet), as high users are more likely to adopt broadband anyway.
The usefulness of broadband was found to have DQLPSRUWDQWLQÀXHQFHLQWKH.6$$VVXFKWKH 391
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
Research Limitations and Suggestions for Future Research Like most research, a study can hardly be perfect. As such, this study has two main limitations. However, these limitations also present opportuniWLHVIRUIXWXUHUHVHDUFK7KH¿UVWOLPLWDWLRQLVWKDW WKH PDMRULW\ RI WKH IDFWRUV ZHUH LGHQWL¿HG IURP literature developed in the West, as there is hardly any academically eminent literature on broadband adoption that has focused on the KSA. As such, LWPD\EHWKDWVRPHYDULDEOHVZHUHQRWVLJQL¿FDQW simply because they are less relevant to the KSA, and that some important variable may have been omitted, which may partly explain why the explanatory power of the model (R2) is not 100%. Therefore an opportunity exists in which the variables that are relevant to the KSA are further explored (in addition to socioeconomic issues) so as to come XSZLWKDPRUHUH¿QHGVHWRIYDULDEOHVIDFWRUVWKDW affect the KSA consumers. This needs a large-scale and expensive survey, which this study did not have the resources for, but that which creates an opportunity for others in the future. Furthermore, due to resource constraints, this research did not investigate the individual products that would make people migrate to broadband. A new survey can be FRQGXFWHGZKHUHWKHLPSDFWRIVSHFL¿FSURGXFWV on the adoption of broadband can be measured. This is important because research in Korea (Oh et al., 2003) showed that the adoption of broadband ZDVVWURQJO\LQÀXHQFHGE\WKHSURGXFWVWKDWZHUH provided by the broadband providers.
ACKNOWLEDGMENT The authors wish to thank Mrs. Shatha Makki for administering the survey questionnaire in the Kingdom of Saudi Arabia and acknowledge the overall contribution made to this research.
REFERENCES Bose, R. (2004). E-government infrastructure and technologies for education and training. Electronic
392
Government, An International Journal, 1(4), 349361. Choudrie, J., & Dwivedi, Y.K. (2005a). The demographics of broadband residential consumers of a British local community: The London Borough of Hillingdon. Journal of Computer Information Systems, 45(4), 93-101. Choudrie, J., & Dwivedi, Y.K. (2005b). Investigating the research approaches for examining technology adoption issues. Journal of Research Practice, 1(1), Article D1. Retrieved June 20, 2005, from http://jrp.icaap.org/content/v1.1/choudrie.pdf Choudrie, J., & Dwivedi, Y.K. (2006a). A comparative study to examine the socio-economic characteristics of broadband adopters and non-adopters. Electronic Government, An International Journal, 3(3), 272-288. Choudrie, J., & Dwivedi, Y.K. (2006b). InvestigatLQJIDFWRUVLQÀXHQFLQJDGRSWLRQRIEURDGEDQGLQ the household. Journal of Computer Information Systems, 46(4), 25-34. Choudrie, J., & Ghinea, G. (2005). Integrated views of e-government Web site usability: Perspectives from users and Web diagnostic tools. Electronic Government, An International Journal, 2(3), 318-333. Cornford, T., & Smithson, S. (1997). Project research in information systems: A student’s guide. London: Macmillan. Crabtree, J. (2003). Fat pipes, connected peoplerethinking broadband Britain. iSOCIETY Report, London. Retrieved March 30, 2004, from http:// www.theworkfoundation.com/pdf/1843730146. pdf Dwivedi, Y.K. (2005). Investigating consumer adoption, usage and impact of broadband: UK households. Unpublished Doctoral Dissertation, Department of Information Systems and Computing, Brunel University, UK. Dwivedi, Y.K., Choudrie, J., & Brinkman, W.P. (2006a). Development of a survey instrument to examine consumer adoption of broadband. In-
Factors Affecting Attitudes Toward Broadband Adoption in the Kingdom of Saudi Arabia
dustrial Management and Data Systems, 106(5), 700-718. Dwivedi, Y.K., Choudrie, J., & Brinkman, W.P. (2006b). Consumer usage of broadband in British households. International Journal of Services and Standards, 2(4), 400-416. Dwivedi, Y.K., Khan, N., & Papazafeiropoulou, A. (2006c). Examining the socio-economic determinants of adoption of an e-government initiative µJRYHUQPHQWJDWHZD\¶Electronic Government— An International Journal, 3(4), 404-419. Dwivedi, Y.K., Papazafeiropoulou, A., & Brinkman, W.-P. (2006d, August 4-6). Examining the second wave of broadband use: Service quality DQGVHFRQGDU\LQÀXHQFHRQFRQWLQXHGEURDGEDQG subscription. Proceedings of the 12th Americas Conference on Information Systems, Acapulco, Mexico. Fridah, M.W. (2002). Sampling in research. Retrieved April 29, 2004, from http://www.socialresearchmethods.net/tutorial/Mugo/tutorial.htm Fowler, F.J. (2002). Survey research methods. London: Sage. ,¿QHGR3 'DYLGUDMXK5 'LJLWDOGLYLGH in Europe: Assessing and comparing the e-readiness of a developed and an emerging economy in the Nordic region. Electronic Government, An International Journal, 2(2), 111-133.
Conference on System Sciences, Hawaii. Yu, C.C., & Wang, H.I. (2004). Digital divide in Taiwan: Evidence, comparisons and strategies. Electronic Government, An International Journal, 1(2), 179-197.
KEY TERMS Broadband: Always-on access, at work, at home, or on the move, provided by a range of ¿[HGOLQHZLUHOHVVDQGVDWHOOLWHWHFKQRORJLHVWR progressively higher bandwidths capable of supporting genuinely new and innovative interactive content, applications, and services, and the delivery of enhanced public services (Dwivedi, 2005). Narrowband: Internet access through a standard modem (dial-up speeds can vary between 28.8 kbps and 56.6 kbps). Relative Advantage: The degree to which broadband Internet is perceived as being better than its predecessor, narrowband Internet (Rogers, 1995). Resources: The perceived level of resources when subscribing to broadband (Ajzen, 1991).
/HH+2¶.HHIH% <XQ. 7KHJURZWK of broadband and electronic commerce in South Korea: Contributing factors. The Information Society, 19, 81-93.
Service Quality: The perceived quality of service a consumer obtained or is obtaining from the current Internet service providers. Service quality is measured in terms of speed of connection, security problems with Internet connections, virus and popup problems with connection, and customer support obtained from the ISP providers (Dwivedi, 2005).
Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitude. Journal of Information Technology, 18, 267-280.
Skill: The perceived ability to operate computers and the Internet (narrowband or broadband) without the assistance of others (Dwivedi, 2005).
Rogers, E.M. (1995). Diffusion of innovations. New York: The Free Press.
Usefulness: The extent to which broadband Internet usage enhances the effectiveness of KRXVHKROG DFWLYLWLHV VXFK DV XQGHUWDNLQJ RI¿FH ZRUNDWKRPHFKLOGUHQ¶VKRPHZRUNLQIRUPDWLRQ or product search, and purchase and home business (Dwivedi, 2005).
6WDQWRQ/- )DFWRUVLQÀXHQFLQJWKHDGRStion of residential broadband connections to Internet. Proceedings of the 37th Hawaii International
393
Division IV
North America
395
Chapter XXV
Characteristics of Farm and Rural Internet Use in the United States Peter L. Stenberg1 Economic Research Service, USDA, USA Mitchell Morehart Economic Research Service, USDA, USA
ABSTRACT The Internet became enmeshed in U.S. businesses management practices over the last decade. During this period access and use of the Internet increased for all regions of the United States, most types of work places, and all income groups. In this study we examine Internet use by farm and rural workers and proprietors using descriptive statistics and market demand analysis. In our market demand analysis approach, the primary methodology we use is categorical dependent variable analysis. The results indicate income is a critical element, though other factors such as age of proprietor and rural-urban ORFDWLRQDUHDOVRVLJQL¿FDQWLQPDUNHWGHPDQGGHWHUPLQDWLRQ
INTRODUCTION The Internet has become a major management tool in the United States in the years since the Internet was spun out into the private sector. For instance, today more than 13% of U.S. wholesale trade is conducted over the Internet. The increase in Internet activity has been especially rapid over the last decade. The prevalence of Internet activity goes across all regions, income groups, and most types of work places. Internet access and use, however, is not unvarying across the economic-geographic spectrum. For instance, while wholesale trade over
the Internet in the United States is at 13%, wholesale trade in the farm sector is only around 3%. In this study we examine factors that underlie U.S. farm and nonfarm business Internet use, and may be causal for the observed spatial and economic sector patterns. Our study addresses these underlying factors in Internet use by examining market demand for information technologies, that is, the Internet. We explicitly test the factors in logistic regression models of Internet use in the farm and nonfarm sectors.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Farm and Rural Internet Use in the United States
BACKGROUND
and not simply going from nonuse to use (Greenstein & Prince, 2006; Stenberg, 2006).
Diffusion in the Use of the Internet Private Sector Internet Use Theories of diffusion offer an effective guide to the observed spatial and socioeconomic patterns in Internet usage. According to the theories, adopWLRQ UDWHV DUH MRLQWO\ GHWHUPLQHG E\ FRQVXPHUV¶ willingness to pay for the new product and service SURYLGHUV¶SUR¿WDELOLW\IURPHQWHULQJQHZPDUNHWV Often, observed behavior is explained by the concept of the S-Curve. The S-Curve is the path that the introduction of new technology takes as it is adopted across the economy over time (Davies, 1979; Mahler & Rogers, 1999; Rogers, 1995). The S-Curve is the result of three well-known stages in the adoption of new technology when the technology becomes broadly spread in the economy. 7KH¿UVWVWDJHUHSUHVHQWVHDUO\DGRSWHUV,QWKH case of Internet use when the Internet entered the public domain, it was the more entrepreneurial businesses and wealthier households that were ¿UVWWRXVHLW(DUO\DGRSWHUVZHUHDPRQJWKHPRVW educated and had more knowledge about the new technology (Leamer & Storper, 2001; Greenstein, 1999; MacKie-Mason & Varian, 1997). In the second stage, the middle of the S-curve, the rate of adoption picked up. The technology PRYHGEH\RQGHDUO\DGRSWHUV7KH,QWHUQHW¶VYDOXH and risk became better known. The slope of this middle portion of the S-Curve is thus the steepest of the three stages. All consumer and business groups had rapid increases in penetration rates prior to 2002 (Brown, 2000; Forman, Goldfarb, & Greenstein, 2002; Greenstein & Prince, 2006). Eventually the new adoption rate slows down. No penetration rate, after all, will ever exceed 7KLVLVWKHWKLUGDQG¿QDOVWDJHRIWKH6 curve and the one that the United States appears to be in now. The higher income groups already show signs of leveling off (NTIA, 2004; Malecki, 2002; Stenberg, 2006). The results for Internet service are consistent with the normalization model of technology diffusion. While the Internet continues to spread across businesses and consumer groups, most changes that are occurring today involve the improvement in Internet service and types of use,
396
While some studies have described the changing user demographics of Internet use, such as the series of studies by the National Telecommunications and Information Administration and Economic and Statistics Administration (NTIA) that started in 1994, more recent studies describe the current and more static situation or examine the adoption of newer broadband technologies. Household studies by Choudrie and Dwivedi (2005, 2006a), Stanton DQGWKH86*HQHUDO$FFRXQWLQJ2I¿FH (2001) tested socioeconomic factors distinguishing adopters and nonadopters of broadband Internet use. Choudrie and Dwivedi (2005) found age, gender, and social grade were important when distinguishing between adopters and nonadopters. Their 2006 study found that characteristics such as income and education were just as important factors. Stanton (2004) tests for a digital divide and ¿QGVLWWKHZLGHVWIRUFRPSXWHURZQHUVKLSDQGWKH narrowest for broadband Internet connections. Most studies on Internet adoption have focused on the household. The series of studies by NTIA and the PEW Internet & American Life Project, for example, have described differences across many demographic and geographic groupings, not only for households, but also Internet activity in the workplace. The NTIA and PEW studies have described the increasing universality of workplace Internet use. All workplace studies hypothesize that busiQHVVHV DFFUXH EHQH¿WV IURP WKH DGRSWLRQ RI WKH Internet into their management practices. In the case of this study, it can be further posited that IDUPDQGUXUDOEXVLQHVVHVPD\DFFUXHPRUHEHQH¿WV from Internet use than urban businesses because of their relative isolation. The Internet compensates for their distance from major markets. Internet use increases market choices, information sources, and FRQWLQXLQJHGXFDWLRQRSSRUWXQLWLHV7KHVHEHQH¿WV however, will not necessarily readily translate into higher demand. Demand depends on many factors such as gross income, educational attainment, and age of the managers.
Farm and Rural Internet Use in the United States
Little systematic research in the small business environment, the common business type for farm and rural businesses, has been completed on the causal demand factors for Internet use. Much of the literature, as Forman (2005) has pointed out, lies in large corporate organizations. Bresnahan, Brynjolfsson, and Hitt (2002), for example, emphasized how prior information technology investments and new organizational structures affect new investments in information technology. Our study examines the hypothesized determinants of Internet access for farm and rural businesses.
DEFINITIONS AND DATA ,QWKH8QLWHG6WDWHVWKHFODVVL¿FDWLRQRIPHWURpolitan and nonmetropolitan areas is made by the 2I¿FHRI0DQDJHPHQWDQG%XGJHW 20% :HXVH WKH 20%GH¿QLWLRQKHUH0HWURSROLWDQDUHDV are counties with one or more urbanized areas or adjacent counties that are economically tied to the core counties as measured by work commuting. An urbanized area, as determined by the Bureau of the Census, is any county where there is an ur-
ban population nucleus of 50,000 or more people. Adjacent counties are included if 25% of workers living in the county commute to the central counties or if 25% of the employment in the county originate in the central counties. The data analyzed in our study come from two sources. Nonfarm data comes from the Current Population Survey (CPS) conducted by the Bureau of the Census (Department of Commerce). The most recent CPS data were released in November 2004 with a sample size of roughly 55,000. The farm business data come from the annual U.S. 'HSDUWPHQWRI$JULFXOWXUH¶V$JULFXOWXUH5HVRXUFH Management Survey (ARMS). The data here are from the 2004 ARMS. Approximately 18,000 farm operators were asked questions about Internet use as part of the survey. ,QWHUQHW FRQQHFWLRQ HI¿FLHQF\ LV QRW GLUHFWO\ addressed in this study. Access to the Internet may be dial-up or high-speed, access may be hard-wired or wireless systems. The study scope is restricted to what socioeconomic factors may be deterministic LQWKHMXGJPHQWWRXVHWKH,QWHUQHW(I¿FLHQF\ZLOO only be inferred by measures of geographic location introduced into the equations.
Figure 1. Availability of Internet at work by household income and metropolitan status, 2003 80
Percent of Households
70 60 50 Total
40
Urban Rural
30 20 10 0 10,000 to 12,499
15,000 to 19,999
25,000 to 29,999
Source: authors using Bureau of the Census data.
35,000 to 39,999
50,000 to 59,999
75,000 to 99,999
Household Income
397
Farm and Rural Internet Use in the United States
FACTORS IN INTERNET USE A number of socioeconomic characteristics have UHSHDWHGO\EHHQLGHQWL¿HGDQGSRVWXODWHGDVGHWHUminates of Internet use in the literature (e.g., the NTIA studies and Choudrie & Dwivedi, 2006b). The most common postulated socioeconomic determinants are income, education, and age. Earlier gender, ethnicity, and other factors had also been postulated as important factors. Observed differences in these latter factors, however, have largely been, it has been argued, the result of variances in income and education levels and other socioeconomic factors (Greenstein, Prince, 2006; Malecki, 2003; Stenberg, 2006). The most common characteristic in the literaWXUHLVLQFRPH+HUHUXUDOZRUNHUV¶VDODULHGKRXVHhold income is used. As salaried income increases, the likelihood of workplace Internet use increases (see Figure 1). The greater the income, the greater the likelihood the work is more highly skilled. The more highly skilled the work, the more likely the Internet is in the work environment irrespective of workplace location. Rural and urban work places, however, are not identical. The Census survey data shows a consistent, though not great, difference
within income groups between rural and urban, with rural workers less likely. This may indicate a variance in types of work within income groups across regions. Conceptually, farm income is the same as nonIDUPZRUNHUV¶VDODULHGLQFRPHEXWLWLVQRWPHDsured the same. Here we use a measure of gross farm income, economic class based on sales. The same positive relationship between income and Internet use appears. As the economic class of the farm increases, so does the likelihood of farm Internet use increase (see Figure 2). Large operations are both more likely to use the Internet to make farm purchases as well as make purchases for household consumption than small farm operations. Education attainment is positively related to, and causality inferred to, income (Becker, 1964). Educational attainment would therefore be expected to interact positively with Internet adoption. In addition the prevalence of Internet and computer technologies in educational institutions provides additional exposure and experience in Internet use as years of formal education increase. In Table 1 the higher the educational attainment, the more likely a worker would use the Internet. The relationship holds within the metropolitan and
Figure 2. Farm Interent access, 2004. 90 80
Percent of farms
70 60 50 40 30 20 10 0 $4,999 or less
$10,000-$19,999
$50,000-$99,999
$250,000-$499,999
Economic class of farm Source: Authors using 2004 USDA Agricultural Resource Management Survey
398
$1,000,000 or more
Farm and Rural Internet Use in the United States
Table 1. Internet use by educational attainment, age, and metropolitan status, 2003 Total (percent)
Metropolitan (percent)
Nonmetropolitan (percent)
Less than a high school diploma High school graduates, no college Some college or associate degree Bachelor's degree or higher
38.5 49.3 68.1 80.9
40.0 50.9 69.2 81.6
32.8 43.7 63.1 75.8
19 to 25 26 to 35 36 to 45 46 to 55 56 to 65 over 65 Total
58.3 62.0 67.9 67.1 57.3 33.4 60
60.2 63.0 69.6 69.0 59.6 35.8 62
48.1 56.3 59.7 58.9 48.6 25.6 52
Education
Age
Source: authors based on Current Po ulation Surve , Bureau of the Census
nonmetropolitan regions. The difference within education attainment levels between metropolitan and nonmetropolitan regions may be explained by the lower incomes in nonmetropolitan areas vis-à-vis metropolitan areas. As can be observed in Table 2, the higher the IDUP RSHUDWRU¶V OHYHO RI HGXFDWLRQDO DWWDLQPHQW the greater the likelihood of Internet use in farm management. Variance in online activity proclivity is also associated with education. The higher his or her educational attainment, the more likely a farm operator will use the Internet to make household SXUFKDVHVZLWKDVLJQL¿FDQWGLIIHUHQFHLQEHKDYLRU between college grads and noncollege grads. A more mixed picture appears for purchasing farm inputs over the Internet with lower incidence of using the Internet for farm inputs as compared to household purchase Internet activity. One causal factor may be household item/farm input price differentials. Farm operators who have obtained a college degree are more likely than those without a college degree in using the Internet to make purchases for the farm business. The difference in Internet proclivity between college grads and noncollege grads, however, was less for farm input purchases than household purchases. Oden and Strover (2002), Grant and Meadows (2002), Choudrie and Dwivedi (2006a), and many
others have cited age as a factor in determining the likelihood of Internet use. The literature suggests that older individuals are more reticent about using the Internet than the young. As may be observed in Table 1, age may be factor but it does not follow the same linear rules as education or income. In Table 1, a bell-shaped curve can be observed. The bell-shaped curve may be due to the fact that younger workers may less likely have jobs that require the use of the Internet, while workers at the other end of the age spectrum may indeed be more reticent or perhaps feel less need to adopt the Internet into their work environment. The observed behavior holds for both metropolitan and nonmetropolitan workers, with lower Internet use across all age groups for rural workers vis-à-vis urban workers. The age factor occurs for farm operators as it did for nonfarm workers (see Table 2), and the same bell-shaped curve for Internet access can be observed. The youngest and oldest farm operators were the least likely to use the Internet, but more than 70% of farm operators between the ages of 35 and 54 had Internet access. Observed Internet purchase behavior, however, does not exhibit a bell-shaped curve (see Table 2). Irrespective of what age group a farm operator belongs to, once he or she is on the Internet, an
399
Farm and Rural Internet Use in the United States
Table 2. Farm Internet use by education and Age, 2004.
Number of farms
Percent of farms
Percent with Internet access
Used for farm purchases
Used for household purchasses
Some high school or less Completed high school Some college Completed college (BA, BS) Graduate school
236,577 800,933 499,929 343,736 141,363
11.7 39.6 24.7 17.0 7.0
24.5 46.4 66.8 76.6 79.7
14.1 18.7 18.0 24.0 20.7
22.6 21.1 25.7 39.8 41.4
Less than 35 years 35 to 44 45 to 54 55 to 64 years 65 years or older
81,112 286,713 531,834 553,202 569,676
4.0 14.2 26.3 27.4 28.2
60.5 71.9 70.2 57.4 34.0
16.8 21.1 18.4 20.8 19.5
36.7 29.5 30.3 28.3 24.3
19.7
28.9
Education
Age
48-State total 2,022,538 100.0 56.4 Source: Authors using 2004 USDA Agricultural Resource Management Survey.
operator of one age group is nearly as likely to make farm purchases over the Internet as an operator of another age group. With respect to making household purchases, however, young operators are more likely to make household purchases over the Internet than older operators. It has been hypothesized that young Internet users are less risk averse with respect to Internet security than older Internet users (PEW).
MODELING THE DETERMINANTS OF INTERNET USE Income, educational attainment, age, metropolitan/nonmetropolitan place of residence, and some other factors we hypothesize are determinants in Internet use. Our null hypothesis is that Internet use is a random event with no determinants. As is often the case when the dependent variable is categorical, we employ the logit model to examine IDFWRUVWKDWLQÀXHQFH,QWHUQHWXVHE\IDUPRSHUDWRUVDQGUXUDOZRUNHUV7KHORJLVWLFVSHFL¿FDWLRQ is well suited to this type of application and has been used in similar studies. See for example Gloy and Akridge (2000).
400
It should be noted that estimates of goodQHVVRI¿W DUH JLYHQ LQ PRGHO HVWLPDWLRQV KHUH R-squared estimates are traditionally given for logistic regressions, but they are not the same as in non-categorical-dependent variable regression models, such as in OLS. A number of different methods have been used to proxy the R-square of noncategorical regression models. Nevertheless R-squared estimates used in logistic regressions are highly controversial, with no broad acceptance of any one estimation methodology over another, and as many statisticians argue, may be misleading and should only very carefully be used to compare models. We estimate two models: nonfarm rural workers and farm businesses. The farm business model is analyzed more extensively due the richness of the ARMS data.
NONFARM RURAL WORKER The independent variables of our nonfarm rural worker model are income, age, educational attainment, and metropolitan/nonmetropolitan place of residence. The independent variables are also
Farm and Rural Internet Use in the United States
categorical. The inclusion of place of residence formally tests the hypothesis that where one lives makes a difference in the observed likelihood of Internet use.
Results The full model (with all independent variables included) predicts correctly the likelihood of Internet use 69% of the time. The model has an adjusted R-square value of 0.224. The Wald test indicated no reason to a priori exclude any of the variables. They all have some predictive power in the observed likelihood of Internet use. As mentioned in the literature, such as articles by Becker (1964), education and age are highly deterministic for income. As a consequence we also test a simple model with only income as the independent variable for the observed likelihood of Internet use. The simple model predicts correctly the likelihood of Internet use 73% of the time. The adjusted R-square value is 0.281. The results are nearly identical for the full and simple models. Exclusion of metropolitan/nonmetropolitan place of residence has no affect on the predictability of either model. We can reject the null hypothesis that Internet use is a random event.
FARM BUSINESSES Most farm operators have Internet access. Only a minority of farm operators, however, make farm and household purchases online. We separated farm operators into three groups based on their Internet activity: those without Internet access, Internet users that did not conduct e-commerce, and those with Internet access that made sales or purchases. :H¿WWHGDPXOWLQRPLDOORJLVWLFUHJUHVVLRQPRGHO with the independent variables: • • • • •
Government payments for farm operations Spouse works off-farm Number of persons in the household Combined years of education of farm operator and spouse Gross farm income
• •
Age of farm operator Miles from farm for shopping
8QGHU WKH 86'$¶V HFRPPHUFH LQLWLDWLYH transactions by farms participating in government programs can be conducted more quickly DQG HI¿FLHQWO\ RQOLQH :H K\SRWKHVL]H WKDW WKLV may encourage adoption by those who receive government payments. A spouse who works off the farm may be LQGLFDWLYH RI ¿QDQFLDO VWUHVV DQG OHVV ¿QDQFLDO ZKHUHZLWKDOWRLQYHVWLQIDUPVSHFL¿HG,QWHUQHW usage. On the other hand, off-farm employment may provide more exposure to Internet technologies and spur home or farm adoption. The number of persons in the household greater than two would be indicative of children or young adults in the farm operations. It is commonly hypothesized that the presence of children or young adults increases the rate of Internet usage among their elders in any household. Larger households may also indicate multigenerational farms that may be more receptive to Internet use. The combined farm operator and spouse years of education variable are one of the many proxies for educational attainment. The combined factor allows both operator and spouse, who are frequently involved in day-to-day farm operations, to be taken into account. It is hypothesized that the higher the educational attainment, the greater the likelihood of farm Internet use. It is hypothesized that the greater the gross farm income, the larger the business increasing the likelihood the Internet is used in the farm operation. 0LOHVIURPIDUPIRUVKRSSLQJUHÀHFWVKRZFORVH the farm is to the core of the telephone infrastructure and the ease of obtaining Internet service. It is hypothesized that the further away from shopping, WKHPRUHGLI¿FXOWLWLVWRREWDLQ,QWHUQHWVHUYLFHDQG the less the likelihood of Internet use. If the farm has the Internet, it is hypothesized that the further away the farm is from town, the more likely the farm will make purchases over the Internet.
Results 7KH0F)DGGHQ¶V5VTXDUHIRUWKHPRGHOLV The gross farm income, age of operator, and the 401
Farm and Rural Internet Use in the United States
Table 3. Farm model results, 2004 Variable (Intercept):1 (Intercept):2 Government Payments:1 Government Payments:2 Spouse works off-farm:1 Spouse works off-farm:2 Household size: 1 Household size: 2 Combined education: 1 Combined education: 2 Gross farm income: 1 Gross farm income: 2 Age of operator: 1 Age of operator: 2 Miles to shopping: 1 Miles to shopping: 2 Source: Authors usin 2004 USDA A
Estimate Std. Error 4.2946 1.0673 2.4473 1.3665 -0.1437 1.3998 0.0909 0.9554 -0.3029 1.9054 0.1492 0.8205 -0.0393 0.3277 -0.0585 0.3108 -0.1895 0.0254 -0.0703 0.0410 0.0000 0.0000 0.0000 0.0000 0.0340 0.0105 0.0032 0.0134 -0.0138 0.0810 -0.0050 0.1313 ricultural Resource Mana ement Surve .
T value 4.0237 1.7909 -0.1027 0.0951 -0.1590 0.1818 -0.1199 -0.1883 -7.4731 -1.7148 -1.6884 -0.1669 3.2298 0.2417 -0.1701 -0.0378
Marg. Effects 0.4638 0.0495 -0.0359 0.0376 -0.0707 0.0697 0.0000 -0.0087 -0.0249 0.0064 0.0000 0.0000 0.0056 -0.0033 -0.0018 0.0005
Figure 3. Predicted probability by gross farm income, 2004. 0.6
Predicted probability
0.5
Internet w ith purchase
0.4
No computer Computer w ith Internet
0.3 0.2 0.1 0 100,000
200,000
300,000
400,000
500,000
600,000
Gross farm incom e (U.S. dollars)
Source: Authors using 2004 USDA Agricultural Resource Management Survey
Predicted probability
Figure 4. Predicted probability by educational attainment, 2004 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
Internet w ith purchase No computer Computer w ith Internet
8
12
16
20
24
28
Combined education (years) Source: Authors using 2004 USDA Agricultural Resource Management Survey
402
32
36
40
Farm and Rural Internet Use in the United States
combined years of education of the operator and VSRXVHZHUHIRXQGWREHVLJQL¿FDQWDQGWKHVLJQV supported the hypotheses (see Table 3). The results of miles to shopping variable do not support the hypothesis that distance is an inhibiting factor in Internet use. Predicted probabilities are the most conclusive and clearest results that may be obtained from the multinomial logistic regression, and here we get some of our most substantive and interesting results. They show clearly the interaction of one variable, constricted within the logistics model, on farm information technology behavior. As gross farm income increases, so does the likelihood of farm operator Internet use (see Figure 3note that graphs sum vertically to 1 over entire range). As farm income increases, the likelihood of not having a personal computer or Internet access steadily decreases. Also as gross farm income increases, so does the likelihood of using the Internet for farm and household purchases. The increases are uniform over the entire range UHÀHFWLQJ WKH ORJ WUDQVIRUPDWLRQ RI WKH LQFRPH data. Across all income levels a majority of farm operators use the Internet. Level of educational attainment has a more complex relationship than has gross income with information technologies (see Figure 4). As educational attainment increases, the likelihood
that the farm operation does not have a personal computer decreases sharply. Type of Internet use, however, is more dynamic. As educational attainment increases, the probability of having a personal computer with Internet access increases whether the Internet is used in making farm and household purchases or not. Initially the probability of having Internet access while making no online purchases increases in unison with making online purchases. After a certain point of educational attainment, however, the probability of making no online purchases begins to decline, while the probability of making online purchases sharply increases. The likelihood of making online purchases increases exponentially with increasing level of educational attainment and exceeds not making online purchases for those with a post-graduate level of HGXFDWLRQDO DWWDLQPHQW (GXFDWLRQ VLJQL¿FDQWO\ determines online behavior. The probability diagram for age of farm operator shows a more uniform relationship than what appeared in the simple descriptive statistics (see Figure 5). The older the farm operator is, the greater the likelihood he or she will not use a personal computer on the farm. The decline in likelihood of Internet use is gradual but increases more rapidly with age. Only by the age of 85, however, would a majority of farm operators be expected not to have a computer or Internet access.
Figure 5. Predicted probability by age of farm operator, 2004.
Predicted probability
Internet w ith purchase 0.7
No computer
0.6
Computer w ith Internet
0.5 0.4 0.3 0.2 0.1 0 25
35
45
55
65
75
85
95
Age of farm operator
Source: Authors using 2004 USDA Agricultural Resource Management Survey
403
Farm and Rural Internet Use in the United States
For any given age, a farm operator with the Internet is more likely not to make online purchases than will make online purchases (see Figure 6). Interestingly, as age increases, the decline in probability is less for making purchases online than for not making purchases. Online purchase behavior is determined less by age than is getting Internet DFFHVVLQWKH¿UVWSODFH Access to shopping is a proxy for how far out of town the farm is situated and where the core of telecommunication services would be located. Distance to shopping appears to have no effect on the likelihood of Internet access. Farms far out of town were more likely to have Internet access than farms close to town. For policymakers this means Internet access is not the issue, though quality of access may still be an important issue. Distance to shopping does affect the probability of online purchase activity: the greater the distance to shopping, the greater the likelihood of making online purchases.
he or she would have used the Internet. Differences in demographic characteristics across rural-urban space account for the observed gap in the aggregate rural-urban worker Internet rates. In 2004, 56% of farm operators reported that they had a computer with Internet access. A sigQL¿FDQWSURSRUWLRQRIWKRVHZLWK,QWHUQHWDFFHVV used the Internet for farm and household purchases. Difference in Internet use across farm size (as measured in farm sales) was striking. The share of farms with both farm and household Internet purchases was the highest for the largest farms, the lowest for the smallest farms. Results from the logistics regression model suggest that income is a critical determinate, though other factors such as age and education DUHVLJQL¿FDQWLQPDUNHWGHPDQGGHWHUPLQDWLRQ Distance from urban centers was not a factor in Internet access. Distance, however, is a factor in the proclivity of making online purchases for the farm and household. Differences in socioeconomic characteristics across metropolitan/nonmetropolitan space may largely explain the observed gap in aggregate Internet access rates. The results suggest policies targeting income and education likely would have some positive effect on take-up rate for Internet use. Continuing or extension education may reduce some of the barriers of entry to Internet access for some of the smaller farm and rural businesses. It is not clear that policies targeting assistance to more remote
CONCLUSION AND POLICY IMPLICATIONS In 2003, 42% of all employees in the United States had accessed and used the Internet through work over the previous year. Rural workers, however, were less likely than urban workers to use the InterQHW7KHKLJKHUDZRUNHU¶VLQFRPHWKHPRUHOLNHO\
Figure 6. Predicted probability by distance to shopping, 2004. Internet w ith purchase No computer
0.6
Predicted probability
Computer with Internet 0.5 0.4 0.3 0.2 0.1 0 15
25
35
45
55
Distance to shopping (m iles)
Source: Authors using 2004 USDA Agricultural Resource Management Survey
404
65
75
Farm and Rural Internet Use in the United States
IDUPDQGUXUDOEXVLQHVVHVZRXOGKDYHVLJQL¿FDQW effect. The research suggests more explicit analysis is needed on the effect better technologies have on the proclivity to use the Internet as part of business management practices for farms and rural small enterprises. The answer has been clear in large enterprises in the urban business environment, but not in the farm and rural business environment, and it is the subject of ongoing research.
REFERENCES Becker, G. (1964). Human capital. New York: Columbia University Press. Bresnahan, T.F., Brynjolfsson, E., & Hitt, L.M. (1999). Information technology, workplace organization, and the demand for skilled labor: Firm-level evidence. Working Paper 7136, National Bureau of Economic Research, USA. Brown, D. (2000). Communication technology timeline. In A.E. Grant & J.H. Meadows (Eds.), Communication technology update (8th ed., pp.745). Woburn, MA: Elsevier Science. Choudrie, J., & Dwivedi, Y.K. (2006a). InvestigatLQJIDFWRUVLQÀXHQFLQJDGRSWLRQRIEURDGEDQGLQ the household. Journal of Computer Information Systems, 46(4), 25-34. Choudrie, J., & Dwivedi, Y.K. (2006b). A comparative study to examine the socioeconomic characteristics of broadband adopters and non-adopters. Electronic Government, 3(3), 272-288. Choudrie, J., & Dwivedi, Y.K. (2005). The demographics of broadband residential consumers in a British local community: The London Borough of Hillingdon. Journal of Computer Information Systems, 45(4), 93-101. Davies, S. (1979). The diffusion of process innovations. Cambridge: Cambridge University Press. Forman, C. (2005). The corporate digital divide: Determinants of Internet adoption. Management Science, 51(4), 641-654.
Forman, C., Goldfarb, A., & Greenstein, S. (2002). Digital dispersion: An industrial and geographic census of commercial Internet use. Proceedings of the Meeting of the Telecommunications Policy Research Conference, Arlington, VA. *HQHUDO$FFRXQWLQJ2I¿FH )HEUXDU\ Telecommunications: Characteristics and choices of Internet users. Washington, DC: (GAO-01-345). Gloy, B.A., & Akridge, J.T. (2000). Computer and Internet adoption on large U.S. farms. International Food and Agribusiness Management Review, 3(3), 323-338. Grant, A.E., & Meadows, J.H. (Eds.). (2002). Communication technology update (8th ed.). Woburn, MA: Elsevier Science. Greenstein, S. (1999). On the Net: The recent commercialization of access infrastructure. Information Impacts, (December). Greenstein, S., & Prince, J. (2006). The diffusion of the Internet and the geography of the digital divide in the United States. Working Paper 123182, National Bureau of Economic Research, USA. Leamer, E.E., & Storper, M. (2001). The economic geography of the Internet age. Working Paper 8450, National Bureau of Economic Research, USA. Mahler, A., & Rogers, E.M. (1999). The diffusion of interactive communications innovations and the critical mass: The adoption of telecommunications services in German banks. Telecommunications Policy, 23, 719-740. Malecki, E.J. (2003). Digital development in rural areas: Potentials and pitfalls. Journal of Rural Studies, 19, 201-214. MacKie-Mason & Varian, H. (1997). Economic FAQs about the Internet. In L.W. McKnight & J.P. Bailey (Eds.), Internet economics (pp.27-62). Cambridge/London: MIT Press. NTIA (National Telecommunications and Information Administration & Economics and Statistics Administration). (2000). Falling through the Net: Toward digital inclusion. Washington, DC: U.S. Department of Commerce.
405
Farm and Rural Internet Use in the United States
NTIA. (2002). A nation online: How Americans are expanding their use of the Internet. Washington, DC: U.S. Department of Commerce. NTIA. (2004). A nation online: Entering the broadband age. Washington, DC: U.S. Department of Commerce.
nology or practices through the economy into businesses, households, and governments. E-Commerce: Economic transactions taking place through the Internet. Farm Businesses: Farms in the business of producing and selling agricultural products.
Oden, M., & Strover, S. (2002). Links to the future: The role of information and telecommunications technology in Appalachian economic development. Washington, DC: Appalachian Regional Commission.
Logistic Regression: A regression model where the dependent variable takes on a limited number of discrete values, often two values representing yes and no.
PEW Internet & American Life Project. (n.d.). Retrieved from http://www.pewinternet.org/index. asp
Metro and Nonmetro: Terms indicating degrees of urbanization. Metro areas are considered densely populated; nonmetro areas are not densely populated.
Rogers, E.M. (1995). Diffusion of innovations. New York: The Free Press. 6WDQWRQ/- )DFWRUVLQÀXHQFLQJWKHDGRStion of residential broadband connections to the Internet. Proceedings of the 37th Hawaii International Conference on System Sciences. Stenberg, P.L. (2006). Investment and household adoption of communication and information services across the United States. In B. Johansson, C. Karlsson, & R. Stough (Eds.), The emerging digital economy: Entrepreneurship, clusters and policy (pp. 263-276). New York: Springer. 86*HQHUDO$FFRXQWLQJ2I¿FH Characteristics and choices of Internet users. Washington, DC (GAO-01-345).
KEY TERMS Diffusion: The adoption process of new tech-
406
Online Purchases: Goods and services that are bought over the Internet. Predictive Probability: In the logistic regression model, a predictive probability is the odds the GHSHQGHQWYDULDEOHWDNHVDVSHFL¿FYDOXHZKHQRQH independent variable value changes while all other LQGHSHQGHQWYDULDEOHVDUHKHOG¿[HG
ENDNOTE 1
The views expressed are those of the authors, DQGGRQRWQHFHVVDULO\UHÀHFWWKHYLHZVRI the Economic Research Service of the U.S. Department of Agriculture.
Division V
South America
408
Chapter XXVI
Broadband User Behavior Characterization Humberto T. Marques Neto Federal University of Minas Gerais (UFMG) Leonardo C.D. Rocha Federal University of Minas Gerais (UFMG) Pedro H.C. Guerra Federal University of Minas Gerais (UFMG) Jussara M. Almeida Federal University of Minas Gerais (UFMG) Wagner Meira Jr. Federal University of Minas Gerais (UFMG) Virgilio A.F. Almeida Federal University of Minas Gerais (UFMG)
ABSTRACT This chapter presents a broadband user behavior characterization from an Internet service provider VWDQGSRLQW8QGHUVWDQGLQJWKHVHXVHUEHKDYLRUSDWWHUQVLVLPSRUWDQWWRWKHGHYHORSPHQWRIPRUHHI¿FLHQW applications for broadband users. Our characterization divides the users into two categories: residential and VPDOORI¿FHKRPHRI¿FH 62+2 ,WHPSOR\VIRXUFKDUDFWHUL]DWLRQFULWHULD VHVVLRQDUULYDOSURFHVV session duration, number of bytes transferred within a session, and user request patterns. Our results show that both residential and SOHO session interarrival times are exponentially distributed, and point out that a typical SOHO user session is longer and transfers a larger volume of data. Our analysis also uncovers two main groups of session request patterns within each user category: (i) sessions that comprise traditional Internet services, such as WWW services, e-mail, and instant messenger, and (ii) VHVVLRQVWKDWFRPSULVHSHHUWRSHHU¿OHVKDULQJDSSOLFDWLRQVEDVLFDOO\:HDOVRDQDO\]HGDQGFODVVL¿HG WKHHEXVLQHVVVHUYLFHVPRVWFRPPRQO\DFFHVVHGE\XVHUVZKLFKGLGQRWYDU\VLJQL¿FDQWO\DFURVVWKH user categories.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband User Behavior Characterization
INTRODUCTION The use of household Internet through broadband connections has grown over past recent years (Fukuda, Cho, & Esaki, 2005; OECD, 2006). This growth improves the use of typical applications, such as e-mail, WWW, and instant messengers, and, moreover, creates a good environment to foster others. Videoconferencing, interactive video and television, collaborative gaming, peer-to-peer (P2P) applications, grid-oriented computing, network-based backups, data-capable wireless network with portable networked gadgets that will used them, and voice over Internet protocol (VoIP) are some potential services which have been used in residential broadband network (MIT, 2005). Peer-to-peer applications deserve special attention. Some studies analyze the growth of peerto-peer usage in broadband networks (Gummadi, Dunn, Saroiu, Gribble, Levy, & Zahorjan, 2004; Hamada, Chujo, Chujo, & Yang, 2004; Lakshminarayanan & Padmanabhan, 2003). An environment with a good availability of Internet resources and with a good quality of service, compared with dial-up connections, encourages some users to VHDUFK VKDUH GRZQORDG DQG XSORDG ODUJH ¿OHV through the Internet. With the increasing demand on resources, the Internet service providers (ISPs) should create mechanisms to avoid the congestion of their backbones, because the growth in resiGHQWLDOXVHUWRXVHUWUDI¿FDQGLQSRSXODUL]DWLRQ RI33V\VWHPVKDVWDNHQXSDVLJQL¿FDQWSRUWLRQ of backbone networks (Cho, Fukuda, Esaki, & Kato, 2006). Understanding the nature and characteristics of broadband user behavior is a crucial step to improve the quality of service offered to users in broadband environments. Broadband user behavior characterization can lead to a better understanding of the interaction between users and service providers. It can also help the design of systems with better QoS metrics, such as performance, availability, security, and cost. Studies of broadband user behavior are scarce LQOLWHUDWXUHPDLQO\EHFDXVHRIWKHGLI¿FXOW\LQ obtaining actual logs from Internet service providers. Most of the service providers on the Internet
consider logs as very sensitive data. Existing studies, such as the one performed by Pew Internet & American Life (Pew, 2004), concentrate on qualitative analysis based on surveys. The Pew UHSRUW VKRZV KRZ $PHULFDQV¶ RQOLQH EHKDYLRU changes with high speed connections at home. The study also shows that broadband services allow users to distinguish themselves from dial-up counterparts in the following ways: (i) broadband users engage in multiple Internet activities on a daily basis, (ii) high speed users become creators and managers of different types of online content, and (iii) broadband users perform a large variety of queries for information. Yet, in another report of Pew Internet & American Life (Horrigan, 2005), the author points out growth in broadband adoption; however, he also shows that the growth rate is decreasing if compared with early years. In spite RI3HZ¶VUHSRUWVTXDQWLWDWLYHVWXGLHVRIEURDGEDQG user behavior are still lacking. 7KLV FKDSWHU LQWHQGV WR ¿OO WKLV JDS UHYLHZing the work presented in Marques-Neto, Rocha, Guerra, Almeida, Meira, & Almeida (2004). To understand the broadband user behavior, we present a characterization from a broadband ISP (a TV cable company that provides broadband VHUYLFHVWRLWVXVHUV ZKLFKFODVVL¿HVWKHLUXVHUV into two major categories: residential and smallRI¿FHKRPHRI¿FH 62+2 )RU HDFK FDWHJRU\ ZH LGHQWLI\ XVHU VHVVLRQV ZKLFK DUH GH¿QHG DV the period during which a user is connected to the broadband network. Basically, the behavior RIXVHUVLVGH¿QHGDVDIXQFWLRQRIWKHZD\XVHUV arrive at the ISP, how long they remain online, the number of bytes they transfer and what they do while connected, that is, the request pattern within a session. Thus, the characterization process is performed along four criteria: (i) session arrival process, (ii) session duration, (iii) number of bytes transferred within a session, and (iv) user request pattern. The broadband user behavior characterization is based on logs collected on an authentication server and by 1HWÀRZ 1HWÀRZ UXQQLQJLQ a border router. The data collecting architecture implemented in the ISP allows us to identify the services used by each user category.
409
Broadband User Behavior Characterization
In order to analyze the service request patterns, we use a state transition graph called customer behavior model graph (CBMG) (Menascé & Almeida, 2000), which describes the behavior of groups of customers who exhibit similar navigational patterns. We then applied clustering algorithms to user session data (both residential and SOHO) to determine groups of users that exhibit similar behavior graphs. Finally, we look further into the HTTPbased Web services most frequently accessed by the residential and SOHO broadband users in our workload and characterize the popularity of different categories of e-business services. 7KH PDLQ ¿QGLQJV RI WKH FKDUDFWHUL]DWLRQ study are: •
•
•
•
•
410
Both residential and SOHO session interarrival times are exponentially distributed during periods of stable arrival rates. Residential session arrival rates remain relatively high during the day, whereas SOHO session arrival UDWHVYDU\VLJQL¿FDQWO\WKURXJKWKHGD\ Residential user session durations are accurately approximated by a lognormal distribution. On the other hand, the duration of SOHO user sessions is better modeled with a combination of a lognormal distribution for the body and a pareto for the tail. For both residential and SOHO sessions, the numbers of incoming and outgoing bytes are modeled by lognormal distributions. In addition to that, the typical ratio of the average number of incoming bytes to the average number of outgoing bytes per session falls LQWKHWKUHHWR¿YHUDQJH The use of a state transition graph (CBMG) XQFRYHUHG VL[ FODVVHV RI VLJQL¿FDQWO\ GLIferent patterns in the user behavior of both categories. One class is dominated by HTTP requests. A second class has a lot of HTTP requests but include some requests to other services, such as P2P and instant messengers. Another class is dominated by P2P requests, which lasts much longer than HTTP sessions. The vast majority of the e-business services requested by both residential and SOHO
users in our workloads (79 out of 164 classi¿HG:HEVLWHV SURYLGHFRQWHQWDQGVHUYLFHV either on a subscription basis or mixed with advertisements. The rest of the chapter is organized as follows. In the next section we discuss related work. In the third section we describe the data collection process and the characterization methodology. The fourth section presents the results from our characterization. Concluding remarks are offered in the last section.
BACKGROUND Several workload and user behavior characterizations are available in literature. Traditional Web workloads, consisting mainly of HTTP requests to GRFXPHQWVDQGLPDJH¿OHVDUHDQDO\]HGLQVHYHUDO previous studies, focused on either server-side (Arlitt, 2000; Arlitt & Jin, 2000) or client-side workloads (Barford, Bestavros, Bradley, & Crovella, 1999; Cunha, Bestravos, & Crovella, 1995). More recent studies characterize the workloads of other types of applications like on-demand and live distribution of streaming media (Costa et al., 2004; Veloso, Almeida, Meira, Bestravos, & Jin, 2006) and peer-to-peer (P2P) services (Gummadi et al., 2004; Leibowitz, Ripeanu, & Wierzbicki, 2003; Saroiu, Gummadi, Dunn, Gribble, & Levy, 2002; Sen & Wang, 2004), which are both becoming increasingly popular, possibly due to the availability of broadband “last mile” connections (Lakshminarayanan & Padmanabhan, 2003). Previous streaming media workload characterizations propose hierarchical models to capture the most UHOHYDQWDVSHFWVRIXVHUEHKDYLRUIRUWKHVSHFL¿F type of workload studied (live [Veloso et al., 2006] and on-demand [Costa et al., 2004]) and produce extensive characterizations of each component of the proposed models. Previous peer-to-peer workORDGVWXGLHVDQDO\]HVHYHUDODVSHFWVRIWKHWUDI¿F generated by these applications such as object popularity, object size, bandwidth utilization, and session durations (Gummadi et al., 2004; Hamada et al., 2004; Leibowitz et al., 2003; Saroiu et al., 2002; Sen & Wang, 2004).
Broadband User Behavior Characterization
These previous workload analyses focus on a VSHFL¿FW\SHRIDSSOLFDWLRQ,QFRQWUDVWRXUZRUN looks into a client-side broadband workload including requests to a multitude of different applications. In that sense, the extensive characterization of a broadband ISP Web proxy (Arlitt, Friedrich, & Jin, 1999) is possibly the previous work that is most closely related to ours. However, that work IRFXVHVRQWKHWUDI¿FJHQHUDWHGE\WKHDQDO\]HG EURDGEDQGFRPPXQLW\FKDUDFWHUL]LQJ¿OHW\SHV sizes, popularity, and frequency of requests to different services (i.e., HTTP, FTP, etc). ,QFRQWUDVWRXUIRFXVLVQRWRQO\RQWKHWUDI¿F generated by broadband users but, especially, on the patterns of user requests to different services that most accurately represent the typical behavior within a broadband user session. In other words, we characterize not only traditionally analyzed aspects such as user session arrival process, duraWLRQDQGWUDI¿FYROXPHEXWDOVRWKHPRVWFRPPRQO\ observed patterns of user requests to different services within the same session. Furthermore, ZH DOVR FRQWUDVW RXU ¿QGLQJV IRU WZR GLIIHUHQW categories of broadband users: residential and SOHO users.
CHARACTERIZATION METHODOLOGY This section presents our characterization methodology and describes how it is applied to the ISP environment. The goal is to analyze the user activity while connected to the Internet, quantifying and qualifying the workload they generate. Our characterization methodology is based on four criteria: session arrival process, session GXUDWLRQWUDI¿FYROXPHDQGXVHUUHTXHVWSDWWHUQ The session arrival process and session duration provide temporal information about the workload generated by users, since we may estimate how frequently and for how long a user is connected. The WUDI¿FYROXPHSURYLGHVOHYHUDJHRQKRZXVHUVDUH using their connection regarding a critical resource for any ISP: bandwidth. Finally, the user request SDWWHUQTXDOL¿HVWKHQDWXUHRIWKHVHUYLFHVEHLQJ
requested by users and how they are distributed across the connection time. We employ three sources of data in the proposed characterization: user authentication log, user daWDEDVHDQGWUDI¿FORJ7KHXVHUDXWKHQWLFDWLRQORJ is compatible with the RADIUS protocol (Rigney, 2000; Rigney, Willens, Rubens, & Simpson, 2000). It has an entry for each user session containing the following information: start date and time, duration, number of bytes transferred, and the dynamic IP assigned to the user. The user database is a table that informs the user category, namely residential RU VPDOORI¿FHKRPHRI¿FH 62+2 LQ WKH ,63 DQDO\]HG7KHWKLUGORJLVFROOHFWHGXVLQJ1HWÀRZ 1HWÀRZ 7KHWUDI¿FLVGLYLGHGLQWRÀRZV DQG HDFK ÀRZ LV FKDUDFWHUL]HG E\ D WLPHVWDPS WKDWLQGLFDWHVZKHQWKHÀRZZDVUHFRUGHGVRXUFH and destination IPs, protocols and ports, and the volume of bytes transferred. Figure 1 shows our characterization methodology strategy process. Access data, data about user FODVVL¿FDWLRQVHUYLFHGDWDDUHFOHDQHGKDQGOHG transformed, and joined to analyze the four characterization criteria. The broadband user behavior is established through these analyses. 7KHWUDI¿FORJIURPZKLFKWKHUHVXOWVSUHVHQWHG in the next section are generated, was collected at one of the three backbone routers of the ISP and FRUUHVSRQGVWRDERXWRIWKHRYHUDOOWUDI¿F Since the user population is equally spread across the three routers, we believe that the data gathering at a single router does not affect the statistical meaning of the results. In other words, the user population analyzed is representative of the ISP user community. Before characterizing the workload along each criterion, we divide the data into two sets according WRWKHXVHUFDWHJRULHVGH¿QLQJWKXVWZRVHSDUDWH broadband workloads. Our residential workload consists of all sessions initiated by users categorized as such by the service provider. Similarly, the SOHO workload consists of all SOHO user sessions. We then characterize the four user behavior criteria for each workload. The session arrival, session duration, and traf¿FYROXPHFULWHULDDUHFKDUDFWHUL]HGE\XVLQJWKH authentication log. For the sake of characterization,
411
Broadband User Behavior Characterization
Figure 1. Characterization methodology
we take into consideration only sessions that start DQG¿QLVKZLWKLQWKHFROOHFWLRQWLPHLQWHUYDO:H FKDUDFWHUL]HWUDI¿FYROXPHVHSDUDWHO\GHSHQGLQJ on its direction: inbound and outbound. For each analyzed criterion, we determine the statistical distribution that best approximates the measured GDWDXVLQJOHDVWVTXDUH¿WPHWKRG 7ULYHGL and visual inspection. The service request pattern is characterized IURPWUDI¿FORJVLQWHUPVRIWKHVHUYLFHVWKDWDUH requested by the users within each of their sessions. A service is a request to an application or application class, such as HTTP, e-mail, and P2P, and is usually LGHQWL¿HGE\RQHRUPRUHSRUWQXPEHUVZKHUHLWV server answers requests to the service. We use and extend the IANA taxonomy1 to match ports to services. The extension is necessary EHFDXVHWKHLGHQWL¿FDWLRQRIVRPHSURWRFROVSRUWV W\SLFDOO\XVHGLQ33WUDI¿FLVQRWHDV\7KHWUDI¿FJHQHUDWHGE\WKHVHNLQGVRIV\VWHPVPXVWEH FDUHIXOO\LGHQWL¿HGLQGDWDVRXUFHVXVHGWRFKDUacterize user behavior. Peer-to-peer applications GRQRWIROORZDGH¿QHGSDWWHUQOLNHXVLQJNQRZQ 7&3RU8'3SRUWWR¿QGSHHUVRUWRWUDQVIHU¿OHV ,GHQWLI\LQJ33WUDI¿FFRXOGEHGRQH L DQDO\]LQJ7&38'3SD\ORDGWR¿QGNQRZQVLJQDWXUHV
412
of peer-to-peer applications, like that proposed in Sen, Spatscheck, and Wang (2004), or (ii) applying heuristics to understand and to model SDFNHWÀRZG\QDPLF 'DQJ3HUpQ\L*HIIHUWK Molnar, 2006; Karagiannis, Broido, Faloutsos, & Claffy, 2004; Sen & Wang, 2004). For the sake of our characterization work, we use one heuristic, similar to the one presented by Dang et al. (2006), WRLGHQWLI\WKH33WUDI¿F We analyze user sessions, and, if the following FKDUDFWHULVWLFVDUHPHWWKHVHVVLRQÀRZVDUHFRQVLGHUHG33WUDI¿F H[FHSWWKHZHOONQRZQRQHV
• •
7KHUHDUHDVLJQL¿FDQWQXPEHURI8'3ÀRZV in the session, which is explained by the frequent use of such ports by P2P protocols There are accesses to the default ports of P2P applications (i.e., 1214 for Kazaa) There are accesses to several apparently unrelated ports in the same session
Although this heuristic is not optimal (i.e., 33WKURXJKSRUWUHPDLQVGLI¿FXOWWRGHWHFW our preliminary results are satisfactory. By using the resulting port-to-service translation table with RXUKHXULVWLFWRLGHQWLI\33WUDI¿FZHZHUHDEOH
Broadband User Behavior Characterization
to transform each user session into a sequence of services. )RU HDFK LGHQWL¿HG XVHU VHVVLRQ ZH EXLOG D customer behavior model graph (CBMG) (Menascé & Almeida, 2000). The CBMG is a state transition graph that has one node for each possible service and transitions between these services. A probability is assigned to a transition between two services representing the frequency at which the user requested the services consecutively in the session. The CBMG is a condensed and semantically rich representation of the user behavior since different types of users may be characterized by different CBMGs in terms of the transition probabilities. 5HSUHVHQWDWLYH VHVVLRQ SUR¿OHV DUH LGHQWL¿HG E\ clustering the session CBMGs. We employ the kmeans algorithm and choose the number of clusters based on the ȕCV metric, as described in Menascé and Almeida (2000). We further characterize user request patterns in terms of e-business services accessed within HDFKXVHUVHVVLRQ,QRWKHUZRUGVWKHWUDI¿FORJV are further processed considering only requests to HTTP or HTTPS services, that is, ports 80 and 443, respectively, typically used by e-business applications. Each distinct IP address requested within a broadband user session (destination IP) is translated into an URL using the host Unix command. )LQDOO\85/VVKDULQJWKH¿UVWFODVVGRPDLQ LH the Web site) are grouped together. For instance, chat.msn.com, intl.msn.com, and help.msn.com are grouped together into the msn.com Web site. We then characterize a number of the most popular sites (i.e., most frequently requested domain names), using the e-business model categorization SURSRVHGLQ5DSSD ,QWKLVSURSRVDOZH¿QG nine models: brokerage, advertising, information intermediary, merchant, manufacturer direct, af¿OLDWHFRPPXQLW\VXEVFULSWLRQDQGXWLOLW\7KH\ are described below: Brokerage model: Includes Web sites that intermediate business-to-business (B2B), business-to-consumer (B2C), and/or consumer-to-consumer (C2C) markets and that charge or receive commission for this service. This category, which includes Web sites
such as ebay.com and priceline.com, may be further broken into eight subgroups, namely PDUNHWSODFHH[FKDQJHEX\VHOOIXO¿OOPHQW demand collect system, auction broker, transaction broker, distributor, search agent, and virtual marketplace. Advertising model: Includes Web sites that provide content and services in conjunction with advertising messages in the form of EDQQHUDGV3RUWDOVFODVVL¿HGVXVHUUHJLVWUDtion, query-based paid placement, contextual advertising, and content-targeted advertising are subcategories that fall within this general model. Yahoo, Hotmail, and MSN are Web sites that operate in this model. Information intermediary model: Supports Web sites whose main goal is to help improve the relationship between buyers and sellers by providing information about both business parts. This e-business Web site model includes advertising networks, audience measure services, and incentive marketing services. The Web sites akamai.com, nielsennetratings.com, and doubleclick.com are examples of this model. Merchant model: Includes Web sites such as amazon.com and barnesandnoble.com that provide a means for wholesalers and retailers to sell their goods and services. Subcategories of this e-business model include virtual merchant, catalog merchant, click and mortar, and bit vendor. Manufacturer direct model: Includes Web sites of product makers or services that sell GLUHFWO\ WR WKH ¿QDO FRQVXPHU 3XUFKDVH lease, license, and brand integrated content are categories of Web sites which work with this model, such as dell.com and microsoft. com. $I¿OLDWH PRGHO Includes Web sites that provide direct links to merchant (partner) :HE VLWHV 7KH DI¿OLDWH :HE VLWH RIIHUV D percentage of their revenue to their partner sites. Subcategories of this model include banner exchange, pay-per-click, and revenue sharing.
413
Broadband User Behavior Characterization
Community model: Includes Web sites that are based on user loyalty and that generate revenue with the sale of secondary products and services or voluntary contributions. This model includes Web sites involved with open source, public broadcasting, and knowledge networks. A typical example is the open source computing Web site (i.e., redhat. com). Subscription model: Includes content providers, person-to-person networking services, trust services, and Internet service providers. Web sites, such as aol.com, falling into this major category typically charge the user subscription for a period (day, week, month, or year) to provide content and various services. Utility model: Includes Web sites that provide services and charge based on the amount of use.
RESULT ANALYSIS This section shows the most relevant results of our EURDGEDQG ZRUNORDG FKDUDFWHUL]DWLRQ 7KH ¿UVW subsection provides an overview of the residential and SOHO workloads. User session interarrival times and durations as well as number of incoming and outgoing bytes transferred within each session LQERWKZRUNORDGVDUHFKDUDFWHUL]HGLQDVSHFL¿F
section. The last subsection analyzes service request patterns and e-business activities.
Workload Overview An overview of our residential and SOHO workloads is provided in Table 1. Our logs cover a period of 28 days (12/23/2003 to 01/21/2004), during which a total of over 310 thousand user sessions were completed. More than 80% of them are from residential users. Similarly, over 73% of all incoming bytes and over 78% of all outgoing bytes are from residential users. A user initiates a session by authenticating KLPVHOIKHUVHOI DW WKH ,63 $ VHVVLRQ LV ¿QLVKHG either explicitly by the user or by timeout after a period of inactivity, which is 4 hours in the case of the ISP that provided the data. In Figure 2, (a) and (b) show the number of simultaneous active (open) sessions during one week for residential and SOHO users, respectively. In Figure 2, (c) and (d) show the same metric over a single day (a Wednesday). Note that most SOHO sessions are active during the day and on weekdays. On the other hand, the fraction of residential sessions active over night and over the weekend is much higher. On average, a residential user completes 1.62 sessions per day and a SOHO user completes only 1.28 sessions per day. This indicates that residential users close their sessions (or are interrupted by timeout) more frequently, during a typical day. On average, a residential user session lasts
7DEOH6XPPDU\RIWKHZRUNORDGV &9 FRHI¿FLHQWRIYDULDWLRQ
Period Total # user sessions completed Total # incoming bytes (GB) Total # outgoing bytes (GB)
SOHO
12/23/03 -01/21/04
12/23/03 -01/21/04
256,239
61,112
11,422
4,135
4,135
1,128
45 (0.76)
36 (0.74)
9.80 (5.00)
13.41 (4.29)
Mean (CV) # incoming bytes per session (MB)
46 (5.02)
70 (3.79)
Mean (CV) # outgoing bytes per session (MB)
20 (8.47)
18 (7.89)
Mean (CV) # sessions completed per user Mean (CV) # session duration (hours)
414
Residential
Broadband User Behavior Characterization
Figure 2. Number of simultaneous active sessions
(a) Typical Week (Residential)
(b) Typical Week (SOHO)
(c) Weekday (Residential)
(d) Weekday (SOHO)
Figure 2: Number of Simultaneous Active Sessions.
approximately 9.8 hours, during which 46 MB of data are received and 19 MB of data are sent out. In contrast, typical SOHO user sessions last longer and receive much more data. On average, a SOHO user remains connected for approximately 13 hours, receives 70 MB of data and sends 18 MB of data. Finally, it is also interesting to note the KLJKYDULDELOLW\ LHKLJKFRHI¿FLHQWRIYDULDWLRQ in the number of sessions as well as in the number of bytes transferred within residential and SOHO user sessions. This implies that there might be some heterogeneity among different user sessions within the same category (residential or SOHO).
Session Characteristics 7KLVVHFWLRQDQDO\]HVWKH¿UVWWKUHHFULWHULDXVHGLQ our characterization of broadband user behavior:
(i) session arrival process, (ii) session duration, DQG LLL WKHLQERXQGDQGRXWERXQGWUDI¿FZLWKLQ a user session.
Session Arrival Process This subsection characterizes the user session arrival process during periods of roughly stable session arrival rate in order to avoid spurious effects due to data aggregation. We carefully selected a large number of stable periods covering different times of the day and different days of the week, including weekends. We found that user session interarrival times are exponentially distributed2 for both residential and SOHO users, as illustrated in Figure 3, (a) and (b), for typical periods of stable arrival rate in each workload. For residential workload, the 415
Broadband User Behavior Characterization
Figure 3. Distribution of session inter-arrival times (seconds)
(a) Residential
mean interarrival time was from 4.81 to 10.20 VHFRQGVZLWKFRHI¿FLHQWRIYDULDWLRQ &9 IURP WRUHVSHFWLYHO\DQGȜSDUDPHWHU3 of the EHVW¿WWHGH[SRQHQWLDOGLVWULEXWLRQYDULHGIURP to 0.21. On the other hand, for SOHO workload the mean interarrival time was from 4.63 to 42.19 VHFRQGVZLWK&9YDULHGIURP WR DQGȜ varied from 0.02 to 0.22. This result is consistent with those presented in Floyd and Paxson (2001) and in Veloso et al. (2006) for session interarrival times. 7KH WLJKWHU UDQJH RI Ȝ YDOXHV REVHUYHG IRU residential users indicates that residential sessions are initiated at relatively high rates (one each 4 to 10 seconds, on average), throughout the day. In contrast, SOHO users usually initiate their sessions during working hours (as discussed in the previous section). In other words, the traditional daily access pattern with peaks in the middle of the day and on weekdays, pointed out by Floyd and Paxson (2001), is more pronounced among SOHO users.
Session duration The durations of residential and SOHO user sessions are characterized separately for different
416
( b) SOHO
days to avoid data aggregation. For each of the two workloads, we separately characterize the distribution of the durations of all sessions that are initiated on a given day, for a large number of days. We found that the durations of residential user sessions can be accurately approximated, both at the body and at the tail of the measured data, by a lognormal distribution4, as illustrated (a) of Figure 4 for a typical day. This is consistent with results presented by Floyd and Paxson (2001) and Veloso et al. (2006). In contrast, the duration of SOHO user sessions are better modeled with a combination of a lognormal distribution, for the body, and a pareto distribution5 for the tail. As illustrated in (b) of Figure 4, the breaking point is around KRXUV:HVSHFXODWHWKLVEHKDYLRUUHÀHFWVWZR different classes of SOHO users: (1) users who remain connected mostly while at work, and (2) users who either work longer journeys or remain connected even after leaving the workplace. For residential workload, the mean session duration was from 4.71 to 13.09 hours, with CV from 1.75 WR7KHSDUDPHWHUVıDQGȝRIWKHEHVW¿WWHG lognormal distribution varied from 1.18 to 1.52 and from 0.48 to 1.86, respectively. On the other hand, for SOHO workload the mean session duration was from 6.95 to 19.21 hours, with CV varied from WR7KHSDUDPHWHUVıDQGȝYDULHGIURP
Broadband User Behavior Characterization
Figure 4. Distribution of session duration on a typical day (a and b) and histogram of session duration (c and d)
(a) Distribution of Session Duration (Residential)
(b) Distribution of Session Duration (SOHO)
(c) Histogram of Session Duration (Residential)
(d) Histogram of Session Duration (SOHO)
Figure 4: Distribution of Session Duration on a Typical Day (a and b) and Histogram of Session Duration (c and d).
0.92 to 1.45 and from 1.04 to 2.30, respectively. 7KHSDUDPHWHUVNDQGĮRIWKHEHVW¿WWHGSDUHWR distribution varied from 1.82 to 7.18 and from 1.28 to 1.95. In Figure 4, (c) and (d) show a clear difference between SOHO and residential users. SOHO users tend to establish sessions that last between 10 and 12 hours, which is approximately the number of typical working hours in companies and organizations (and computers remain turned on during lunch time). This difference regarding SOHO sessions justify the need for a more complex model, mixing lognormal and pareto.
,QERXQGDQG2XWERXQG7UDI¿F This section characterizes the total numbers of incoming and outgoing bytes, transferred within each user session. As in the previous section, the analysis is performed for different days. We found that, for both residential or SOHO sessions, the number of incoming bytes and the number of outgoing bytes can be each well modeled with lognormal distributions, as illustrated in )LJXUH7KHVHUHVXOWVFRQ¿UPWKRVHSUHVHQWHGLQ Arlitt et al. (1999), Barford et al. (1999), and Floyd and Paxson (2001).
417
Broadband User Behavior Characterization
Figure 5. Distribution of the number of incoming (a and b) and outgoing (c and d) bytes per user session (MB)
(a) Distribution of Incoming Bytes (Residential)
(b) Distribution of Incoming Bytes (SOHO)
(c) Distribution of Outgoing Bytes (Residential)
(d) Distribution of Outgoing Bytes (SOHO)
Figure 5: Distribution of the Number of Incoming (a and b) and Outgoing (c and d) Bytes per User Session (MB).
Table 2 presents a summary of our results. Compared to residential users, SOHO users typically receive and send larger amounts of data within each session, possibly due to the longer average session duration. Moreover, for each workload, the ratio of the average number of incoming bytes to the average number of outgoing bytes per session is QRWYHU\KLJKIDOOLQJW\SLFDOO\LQWKHWKUHHWR¿YH range. This may be due to the use of services which transfers large amounts of data in both directions, such as peer-to-peer applications.
418
In conclusion, compared to SOHO users, residential users usually initiate a larger number of sessions throughout the day. Moreover, residential sessions are typically shorter and transfer fewer bytes, both downstream and upstream.
User Request Patterns We now turn to the analysis of the most commonly observed user request patterns within a session in our two broadband workloads. Our analysis focuses
Broadband User Behavior Characterization
Table 2. Summary of the distributions of the numbers of incoming and outgoing bytes per session Transferred Bytes Workload
Lognormal Parameters
Metric Mean(MB)
CV
ı
ȝ
Residential
Incoming
28 -44
3.95 -4.63
1.62 -1.83
1.76 -2.46
SOHO
Incoming
47 -80
3.31 -3.40
1.47 -1.70
2.39 -3.27
Residential
Outgoing
10 -16
6.82 -8.27
1.84 -2.09
0.31 -1.09
SOHO
Outgoing
9 -23
2.98 -6.82
1.51 -2.09
0.41 -1.31
Lognormal PDF:
p X ( x)
on the classes of services (i.e., HTTP, POP3, P2P, etc.) most commonly requested by the users, and looks further into the HTTP requests in search of the e-business services most frequently accessed by the residential and SOHO broadband users in our workloads. :H¿UVWORRNLQWRWKHSRSXODULW\RIGLIIHUHQW services in the SOHO and residential workloads. The popularity of a service is assessed in terms of the percentage of sessions that include at least one request addressed to a port number that idenWL¿HVWKHVHUYLFH HJSRUWQXPEHUIRU+773 service). Although HTTP appears in over 95% of all sessions in both workloads, e-mail (i.e., POP3, SMTP) as well as interactive applications such as instant messenger and ICQ are also popular among both residential and SOHO users. We also SRLQWRXWWKHVLJQL¿FDQWIUDFWLRQRIVHVVLRQVWKDW include requests to P2P services, such as Kazaa. In particular, around 23% of the residential sessions and 12% of the SOHO sessions contain requests addressed to P2P services, illustrating the growth of the popularity of these applications among broadband users, previously discussed in Hamada et al. (2004) and Lakshminarayanan and Padmanabhan (2003). Interestingly, applications that have higher bandwidth requirements such as streaming media are very modestly used by the users we observed. The following subsections characterize the most common request patterns within a user ses-
ln x
1 x
2
e
2
2
2
VLRQLQHDFKZRUNORDG7KH¿UVWRQHFKDUDFWHUL]HV the patterns of Internet services requested within a session. The second subsection focuses on the e-business activities within each user session, characterizing the most frequently requested HTTP-based e-business services.
Service Request Pattern The service request pattern is characterized in terms of the frequency of requests to each service and the frequency at which a user switches between different services, within the same session. To do so, we represent the sequence of service requests within each session with a CBMG, as described in the section which describe our characterization methodology, and use standard clustering techQLTXHVWR¿QGWKHPRVWUHSUHVHQWDWLYHSHUVHVVLRQ service request patterns. 2XUDQDO\VLVXQFRYHUHGVL[FODVVHVRIVLJQL¿cantly different request patterns in each workload, summarized in Tables 3 and 4 for residential and the SOHO workloads. Due to space constraints, we focus only on the classes that accounted for at least 3% of the sessions, omitting two unpopular residential classes and one unpopular SOHO class. (DFKFODVVLVGH¿QHGE\WKHIUHTXHQF\RIUHTXHVWV WRHDFKVHUYLFHZLWKLQDVHVVLRQ ¿UVWURZ Within both residential and SOHO workloads, the session classes can be further grouped into two major super-classes. One super-class represents those sessions that are dominated by HTTP
419
Broadband User Behavior Characterization
Table 3. Summary of the main classes of residential user request patterns Class 1
Class 2
HTTP(41%) Services requested within session
Class 4
HTTP(73%)
P2P(28%)
HTTP(97%)
MSN(18%)
P2P(76%)
MS-DS(18%)
HTTPS(3%)
ICQ(6%)
HTTP(24%)
EPMAP(13%) Total # sessions (%)
Class 3
POP3(3%)
3,249 (4%)
51,606 (66%)
9,797 (13%)
10,640 (14%)
Total # incoming bytes (GB) (%)
214 (7%)
1,436 (47%)
361 (12%)
806 (26%)
Total # outgoing bytes (GB) (%)
75 (6.5%)
451.6 (39%)
111 (10%)
405 (35%)
Mean (CV) duration (hours)
9.01 (2.16)
6.06 (2.61)
8.13 (2.48)
10.43 (2.24)
Mean (CV) incoming bytes (MB)
65.88 (3.76)
27.82 (5.09)
36.80 (4.37)
75.79 (3.69)
Mean (CV) outgoing bytes (MB)
23.14 (4.89)
8.75 (10.33)
11.29 (8.18)
38.06 (4.99)
Table 4. Summary of the main classes of SOHO user request patterns Class 1 HTTP(75%) Services requested within sessions
POP3(11%) P2P(10%) SMTP(4%)
Class 2
Class 3
HTTP(96%)
HTTP(60%)
HTTPS(3%)
MSN(34%)
POP3(1%)
POP3(4%)
Class 5
HTTP(56%) DNS(22%)
P2P(70%)
HTTPS(18%)
HTTP(30%)
POP3(4%)
Total # sessions (%)
1,644 (7%)
16,010 (67%)
2,458 (10%)
1,603 (7%)
1,572 (7%)
Total # incoming bytes (GB) (%)
162.8 (12%)
573.6 (44%)
199.3 (15%)
122.7 (9%)
235.7 (18%)
Total # outgoing bytes (GB) (%)
36.9 (11%)
134.2 (38%)
35.3 (10%)
39.1 (11%)
101.4 (29%)
Mean(CV) duration (hours)
12.23 (2.07)
7.89 (2.53)
11.99(1.50)
11.40 (3.51)
14.82 (1.98)
Mean(CV) incoming bytes (MB)
99.03 (2.60)
35.83 (3.60)
81.08 (2.70)
76.56 (3.26)
149.92 (2.58)
Mean(CV) outgoing bytes (MB)
22.44 (4.81)
8.38 (11.35)
14.37 (3.02)
24.37 (6.85)
64.5 (3.91)
requests but also may include some requests to other services such as e-mail, instant messenger, ICQ and, generally speaking, P2P services. This category consists of classes 1, 2, and 3 in the residential workload, and classes 1, 2, 3, and 4 in the SOHO workload. Compared to sessions consisting mostly of HTTP requests (class 2 in both workloads), the use of e-mail and interactive chatting applications (class 3 in both workloads) increase VLJQL¿FDQWO\WKHDYHUDJHVHVVLRQGXUDWLRQDQGWKH DYHUDJHYROXPHRIWUDI¿FUHFHLYHGDQGVHQWRXW In other words, users remain connected for longer periods communicating with other people. Sessions that include some requests to P2P services (class 1 in both workloads) are even longer and transfer more data, as one might expect. 420
Class 4
The second user session super-class is dominated by P2P requests (classes 4 in the residential work-load and 5 in the SOHO workload). They last, on average, much longer than the HTTPEDVHG VHVVLRQV DQG WUDQVIHU VLJQL¿FDQWO\ ODUJHU volume of data. :HQRWHWKHVLJQL¿FDQWO\ORZHUFRHI¿FLHQWVRI variation for the session duration and numbers of incoming and outgoing bytes, for each session class, compared to the variations observed in the aggregated SOHO and residential workloads, shown in Table 1. This means that categorizing users based on their request patterns is probably a more effective strategy for QoS analysis and capacity planning than simply relying on its status in the user database.
Broadband User Behavior Characterization
Figures 6 and 7 show the CBMGs of the request SDWWHUQFODVVHVLGHQWL¿HGIRUUHVLGHQWLDODQG62+2 sessions, respectively. As an example, (c) in Figure 6 shows that a class-3 residential user requesting a HTTP service will switch to a POP3 service with probability 0.03, will switch to ICQ with probability 0.05, will start using instant messenger (MSN) with SUREDELOLW\DQG¿QDOO\ZLWKSUREDELOLW\ he / she will remain requesting HTTP services. Note that, in each class, for both residential and SOHO users, the self-loop transitions have typically high probabilities. In other words, a user tends to request the same kind of service repeatedly.
E-Business Activities We have shown that the vast majority of user sessions in both residential and SOHO workloads
are dominated by HTTP requests. This section further analyzes user requests to HTTP services, IRFXVLQJVSHFL¿FDOO\RQUHTXHVWVWR+773EDVHG e-business services. As discussed in the characterization methodology section, we translate each distinct IP address requested through the HTTP protocol into an URL, group URLs belonging to the same e-business site together, and characterized the popularity of different sites according to the categorization proposed in Rappa (2004). We found approximately 250 thousand distinct IPs in the two workloads. We were able to translate 78% of them into the corresponding host names using the Unix host command. URLs sharing the same third level domain6 were grouped into e-business sites. Perhaps surprisingly, we found that the most popular e-business sites among broadband users were mostly the same across all classes of
Figure 6. Main classes of residential user request patterns.
(a) Class 1
(b) Class 2
(c) Class 3
(d) Class 4
Figure 6: Main Classes of Residential User Request Patterns.
421
Broadband User Behavior Characterization
XVHUUHTXHVWSDWWHUQ DVGH¿QHGLQFKDUDFWHUL]Dtion methodology), in both residential and SOHO ZRUNORDGV7KHUHIRUHZH¿UVWGLVFXVVHEXVLQHVV activities considering all user sessions in both ZRUNORDGV/DWHUZHSUHVHQWVRPHVSHFL¿F¿QGLQJV that are a function of the user class. We were able to classify 164 out of the 300 most popular e-business sites using the categorization proposed in Rappa (2004) (36 of them are universities or .gov sites and 100 are not reachable from
D:HEEURZVHU 7KHVHFODVVL¿HGHEXVLQHVVVLWHV account for approximately 60% of all user HTTP requests. Furthermore, the popularity of different e-business sites is highly skewed. For instance, WKH ¿YH PRVW SRSXODU FODVVL¿HG HEXVLQHVV VLWHV account for 30% of all HTTP requests. 7DEOH VKRZV WKH SHUFHQWDJH RI FODVVL¿HG e-business sites that fall into each e-business category, as well as the percentage of HTTP requests they account for. Note that no e-business
Figure 7. Main classes of SOHO user request patterns.
(a) Class 1 (
b) Class 2
(c) Class 3 (
d) Class 4
(e) Class 5 Figure 7: Main Classes of SOHO User Request Patterns.
422
Broadband User Behavior Characterization
Table 5. Probability of e-business site categories for all users RIFODVVL¿HG
DOO+773
e-business sites
requests
Subscription
22
26
Advertising
27
23
Information Intermediary
11
7
Manufacturer Direct
16
2
Merchant
23
1
Brokerage
1
1
E-business Category
VLWHLVFDWHJRUL]HGDVHLWKHUDI¿OLDWHFRPPXQLW\ or utility. In contrast, e-business sites categorized as either subscription or advertising account for RIDOOFODVVL¿HGHEXVLQHVVVLWHVDQG RI all user requests to HTTP and HTTPS services, DQGIRXURIWKH¿YHPRVWSRSXODUFODVVL¿HGHEXVLness sites follow the subscription model, providing text, audio, and video content to subscribed users. 0RUHVSHFL¿FDOO\ZHQRWLFHGWKDWWKHPRVWSRSXODU VLWHVDUH%UD]LOLDQ¶VFRQWHQWVHUYLFHVDQGSRUWDOV Yahoo, Hotmail, and Google. We also analyzed how the site categories vary across user behavior classes introduced in our methodology. We observed that the probabilities of the HEXVLQHVVVLWHFDWHJRULHVGRQRWYDU\VLJQL¿FDQWO\ among user classes. The most popular e-business sites considering the whole user population are also popular in all user classes with slightly variations. The same observation applies to interactive services, such as MSN and ICQ, which present similar popularities for the classes where these services are invoked by users. Finally, there are no differences between the category distribution when contrasting residential and SOHO users, as an indication that the preferences for e-business services are not affected by people being at the RI¿FHRUDWKRPH
CONCLUSION AND FUTURE WORK This chapter used a quantitative approach for characterizing the behavior of broadband users.
The characterization relies on data collected at YHU\VSHFL¿FORFDWLRQVLQDVHUYLFHSURYLGHU7KH sources of data are the authentication logs, the user GDWDEDVHDQGWUDI¿FORJV7KHFKDUDFWHUL]DWLRQZDV performed at the session level and at the request level and considered residential and SOHO users separately. .H\¿QGLQJVRIWKLVVWXG\DUH L ERWKUHVLGHQWLDO and SOHO session interarrival times are exponentially distributed, (ii) residential user session duration is well modeled with a lognormal distribution, while the durations of SOHO sessions are better approximated by the combination of a lognormal and a pareto distributions, (iii) for both residential and SOHO sessions, the number of incoming and outgoing bytes may be modeled with a lognormal distribution, (iv) the use of a state transition graph &%0* XQFRYHUHGVL[FODVVHVRIVLJQL¿FDQWO\GLIferent patterns in the user behavior, and (v) user e-business activities concentrate on subscriptionbased content and services providers, mostly on services that rely strongly on advertisement for generating revenue. 7KHUHVXOWVSUHVHQWHGLQWKLVFKDSWHUDUHD¿UVW attempt to characterize the behavior and e-business activities of broadband users. It can lead to a better understanding of the interaction between users and service providers and help the design of systems with better QoS measures, such as performance, availability, security, and cost. :HDUHLQWKHSURFHVVRIUH¿QLQJWKHFKDUDFterization of the CBMGs, to evaluate the behavior of other services such as games and operating
423
Broadband User Behavior Characterization
system-oriented services. We are also working to create CBMGs that group classes of users, instead of classes of sessions, and the characterization of broadband daily pattern use.
REFERENCES Arlitt, M. (2000). Characterizing Web user sessions. ACM SIGMETRICS Performance Evaluation Review, 28(2), 50-56. Arlitt, M., Friedrich, R., & Jin, T. (1999). Workload characterization of a Web proxy in a cable modem environment (Technical Report HPL-1999-48). Palo Alto, CA: Internet Systems and Applications Laboratory - HP Laboratories Palo Alto. Arlitt, M., & Jin, T. (2000). Workload characterization of the world cup Web site. IEEE Network, 14(3), 30-37. Barford, P., Bestravos, A., Bradley, A., & Crovella, M. (1999). Changes in Web client access patterns: Characteristics and caching implications. World Wide Web, Special Issue on Characterization and Performance Evaluation, 2(1-2), 15-28. Cho, K., Fukuda, K., Esaki, H., & Akira, K. (2006). The impact and implications of the growth in UHVLGHQWLDOXVHUWRXVHUWUDI¿F,QSIGCOMM ‘06: Proceedings of the 2006 Conference on Applications, Technologies, Architectures and Protocols for Computer Communications (pp. 207-218). Costa, C., Cunha, I., Borges, A., Ramos, C., Rocha, M., Almeida, J., et al. (2004). Analyzing client interactivity in streaming media. In Proceedings 13th World Wide Web Conference, New York (pp. 534-543). Cunha, C., Bestravos, A., & Crovella, M. (1995). Characteristics of WWW client-based traces (Technical Report TR-95-010). Boston: Department of Computer Science - Boston University.1 Dang, T., Perényi, M., Gefferth, A., & Molnar, S. 2QWKHLGHQWL¿FDWLRQDQGDQDO\VLVRI33 WUDI¿FDJJUHJDWLRQ,QNETWORKING 2006 (LNCS 3976) (pp. 606-617).
424
)OR\G 6 3D[VRQ 9 'LI¿FXOWLHV LQ simulating the Internet. IEEE/ACM Transactions on Networking, 9(4), 392-403. Fukuda, K., Cho, K., & Esaki, H. (2005). The LPSDFW RI UHVLGHQWLDO EURDGEDQG WUDI¿F RQ -DSDnese ISP backbones. ACM SIGCOMM Computer Communications Review, 35(1), 15-21. Gummadi, K., Dunn, R., Saroiu, S., Gribble, S., Levy, H., & Zahorjan, J. (2004). Measurement, PRGHOLQJDQGDQDO\VLVRIDSHHUWRSHHU¿OHVKDULQJ workload. In Proceedings of the 19th ACM Symposium on Operating Systems Principles (SOSP-19), Bolton Landing, NY (pp. 314-329). +DPDGD 7 &KXMR . &KXMR 7
Broadband User Behavior Characterization
MIT. The Broadband Incentive Problem. (2005). MIT Communications Futures Program (CFP) and Cambridge University Communications Research Network. 1HWÀRZ 5HWULHYHG -XO\ IURP ZZZFLVFRFRPZDUSSXEOLF7HFKQHWÀRZ OECD. (2006). OECD broadband statistics. Retrieved July 10, 2007, from http://www.oecd.org Pew & American Life. (2004). The broadband difference. Retrieved July 10, 2007, from www. pewinternet.org Rappa, M. (2004). The utility business model and the future of computing services. IBM Systems Journal, 43(1), 32-42. Rigney, C. (2000). Radius accounting. RFC 2866, IETF. Rigney, C., Willens, S., Rubens, A., & Simpson, W. (2000). Remote authentication dial user service (RADIUS). RFC 2865, IETF. Saroiu, S., Gummadi, K., Dunn, R., Gribble, S., & Levy, H. (2002). An analysis of Internet content delivery systems. In Proceedings of the Fifth Symposium on Operating Systems Design and Implementation (OSDI 2002) (pp. 315-327). Sen, S., Spatscheck, O., & Wang, D. (2004). AcFXUDWHVFDODEOHLQQHWZRUNLGHQWL¿FDWLRQRISS WUDI¿FXVLQJDSSOLFDWLRQVLJQDWXUHs. In Proceedings of the 13th International Conference on World Wide Web (pp. 512-521). Sen, S., & Wang, J. (2004). Analyzing peer-to-peer WUDI¿FDFURVVODUJHQHWZRUNVIEEE/ACM Transactions on Networking, 12(1), 219-232. Trivedi, K. (2002). Probability & statistics with reliability, queueing, and computer science applications (2nd ed.). New York: John Wiley & Sons. Veloso, E., Almeida, V., Meira, W., Bestravos, A., & Jin, S. (2006). A hierarquical characterization of a live streaming media workload. IEEE/ACM Transactions on Networking, 14(1), 133-146.
KEY TERMS Characterization: Process based on actual measures (i.e., access logs) that aims to understand the nature and characteristics of a given scenario, VXFKDVWKHH[SORUDWLRQRIXVHUV¶,QWHUQHWDFFHVV data to understand the interaction between users and Internet service providers. CBMG: Customer behavior model graph is a state transition graph that has one node for each possible service and transitions between these services; a probability is assigned to a transition between two services representing the frequency at which the user requested the services consecutively in the session. Service Request Pattern: Is characterized in terms of the frequency of requests to each service and the frequency at which a user switches between different services, within the same session; it is represented with a CBMG. Session Arrival Process: Provides temporal information about the workload generated by users and estimates how frequently users start a new session. Session Duration: Provides temporal information about the workload generated by users and estimates how long a user is connected. 7UDI¿F9ROXPH Provides leverage on how users are using their connection regarding a critical resource for any ISP: bandwidth. User Behavior:7KHEHKDYLRURIXVHUVLVGH¿QHG as a function of the way users arrive at the ISP, how long they remain online, the number of bytes they transfer, and what they do while connected, that is, the request pattern within a session.
ENDNOTES 1
2 3
Internet Assigned Numbers Authority (http:// www.iana.org/assignments/port-numbers). e x Exponential (PDF): p X ( x) Session arrival rate.
425
Broadband User Behavior Characterization
4
Lognormal (PDF):
p X ( x)
426
ln x
1 x
2
e
2
2
5
2
6
Pareto (PDF): p X ( x)
k x
1
, where x t k .
For the sake of analysis, we considered domains such as .com, .gov, and .org as second level domains.
427
Chapter XXVII
Precisions about the Broadband Divide in Chile Sergio Godoy Universidad Catolica de Chile, Chile Soledad Herrera Universidad Catolica de Chile, Chile
ABSTRACT This chapter aims to quantify more exactly the adoption of broadband at the household level in Chile by assessing its impact on three types of digital divide: between users and nonusers of the Internet, between usage at home and elsewhere, and between home broadband users and modem home users. This was done by a statistical analysis of WIP-Chile surveys of 2003 and 2006. At least in Chile, the main digital gap is still between users and non users of the Internet, both in terms of age and education level. Income mainly affects the probability of having broadband access at home. Since broadband has rapidly expanded among all socioeconomic segments, it is becoming less relevant as a predictor of access and Web usage. Other factors are also weak predictors of both residential use of the Internet and broadband connections at home. New data and surveys are required to clarify the point.
INTRODUCTION This chapter examines different aspects affecting three levels or types of digital divide in Chile: between users and nonusers of the Internet in the ¿UVWSODFH EHWZHHQKRPHXVHUVDQGXVHUVHOVHZKHUH in the second place, and between home users with broadband vis-à-vis those who have a modem connection. The evidence examined is taken from surveys made by the local team of the World Internet Project (WIP).1 WIP is a longitudinal,
internationally comparable study about Internet usage in everyday life.2 According to the Organization of Economic CoOperation and Development (OECD), the digital divide refers to the gap in terms of both access and usage of information technologies (IT) among SHUVRQVKRXVHKROGV¿UPVDQGJHRJUDSKLFDODUHDV of different socioeconomic levels (OECD, 2001). That inequality exists even among individuals who have Internet access. Indeed, countries attached to WIP, such as Chile, have found that the place of access and speed of connection are increasingly
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Precisions about the Broadband Divide in Chile
among the most important differentiators of Web usage (Center for the Digital Future, 2007; WIPChile, 2006). Household connections, for instance, and in contrast to workplace, school, or cybercafés, generally allow longer times of usage as well as a wider repertoire of online activities. Among Web users at home, speed of connection is especially important—thus broadband users perform more online activities (Di Maggio & Hargittai, 2001, p. 79; Dwivedi & Choudrie, 2003; Fox, 2005; Husing & Selhofer, 2004). It is important to point out these differentiating aspects of Web access, since they imply different social, participatory, and empowering consequences (Norris, 2001). This chapter aims to quantify more exactly the adoption of broadband at the household level in Chile, assuming its potentialities are not completely IXO¿OOHG ZKHQ DFFHVV LV HOVHZKHUH²DXWRQRP\ and time of usage are limited at school, work, or cibercafés, therefore restricting how individuals can take advantage of the technology. We will therefore go beyond average broadband penetration in the country. We will distinguish in detail between those who have dial-up (modem) connections at home and those who have broadband, and what that implies. Our objectives are the following: 1.
Determine the magnitude of the three types of digital divide mentioned before, that is: (i) between users and nonusers of the Internet, (ii) between usage at home and elsewhere, and (iii) between usage at home through broadband and dial-up connections
2.
Study the factors affecting these divides, GHWHUPLQHWKHLULQÀXHQFHDQGDVVHVVZKHWKHU these factors are common to all the gaps or not
3.
Delve into the impact of broadband on daily activities, that is, on interpersonal communications and “off-line” activities, notably the traditional mass media
428
METHODOLOGICAL BACKGROUND This chapter is based on logistic regressions and ratios calculated from our WIP-Chile database, taken from surveys conducted in 2003, 2004, and 2006 among Web users and nonusers.3 We foFXVHGRXUDQDO\VLVRQGDWDIURP6DQWLDJR&KLOH¶V capital, which concentrates 40% of the national population of almost 16 million.4 The WIP-Chile surveys consist of a panel of 1000 face-to-face, random-sampled individuals aged 12 to 60 years. The questionnaire is shared by all countries participating in the project, yet each team adds some extra questions if required. The WIP questionnaire tries to determine who is online, how, for how long, and why; and the technological, social, and political attitudes of both users and nonusers of the Web, as well as their patterns of media usage, e-commerce, and other subjects. Web users are asked about their online behavior both online and off-line, the place where they access the Internet, and the connection they enjoy at home. In this context, an Internet user is someone who has personally used the Internet at least once in the last three months. Usage is understood as sending e-mails and/or Web browsing, at least.
INTERNET AND BROADBAND IN CHILE According to the United Nations, Chile had 28% of Internet users in 2003 using as a base the populaWLRQDERYHVL[\HDUVRIDJH7KH¿JXUHZDVLQ 2000 (United Nations, 2003). According to WIP&KLOH¶VHVWLPDWHV:HEXVHUVJUHZIURPLQ 2004 to 40% of the total population in 2006. These percentages are the highest in Latin America (see Figure 1), although lower than those of developed nations (WIP-Chile, 2006).
Precisions about the Broadband Divide in Chile
Figure 1. Internet users per 100 population; selected countries 1994-2003 70
Argentina
60
Bolivia 50
Brazil Chile
40
France 30
Germany United Kingdom
20
United States 10
Uruguay
0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
Source: United Nations (2003).
nomic level (the Ministry of Education has long since fostered Web access in public schools via an initiative called Plan Enlaces). According to WIP-Chile (2006), the household is the place of access which grew most, from 21% of users in 2003 to 24% in 2004 and 28% in 2006. Yet the place where the time of usage grew most was the workplace among those users who work. The expansion of broadband has also been vigorous in Chile. Household users connected through this device rose from 55% in 2003 to 85% in 2006. The growth rate has been stronger among the lower income C3-D groups—it reached 144% between both years, whereas it was 29% among the wealthier ABC1-C2 groups (WIP-Chile, 2006).5 Broadband growth at work, school, and other places has been also quite fast. Yet we have not conducted a deeper analysis of broadband adoption in the average Chilean household, the most important scenario for more advanced uses of the Web. Besides, the school is roughly as important as the household as the most important place of DFFHVV LQ &KLOH FRQ¿UPLQJ WKH FRXQWU\¶V FRQdition as a middle-income, developing nation (WIP-Chile, 2006). Perspectives about broadband growth remain high, with a 60% estimate of total connections for 2006 in contrast to almost nil in ,QWKLVWHFKQRORJ\UHDFKHGD¿IWKRIWKH Chilean population, that is, three million people (CCS, 2006).
A rather relevant issue is what is taken for broadband, a widespread concept that often has different meanings throughout the world. Anderson and Raban (2005) acknowledged as such any dedicated connection of at least 256 kilobytes per second (Kbps). That rate is often dismissed as too low by some specialists—in Japan, for instance, the DFFHSWHG¿JXUHLVPHJDE\WVSHUVHFRQG 0ESV while the ITU had recommended at least 1.5 Mbps. $UHSRUWE\&KLOH¶V0LQLVWU\RI7HOHFRPPXQLFDWLRQVSXEOLVKHGLQODWHGH¿QHGEURDGEDQGDV any permanent connection with a transfer speed of 128 Kbps at least. Ricardo Baeza, an expert who publicly disagreed with the report, argued there was no broadband in Chile, since only 12.5% of connections exceeded 1.0 Mbps (La Tercera, 2005). Meanwhile a report by the Latin American branch of Cisco Systems published in January 2007 stated that 45.5% of Internet connections were of broadband, reaching a total of 1,034,000 lines (Cisco Systems, 2007; Mouse.cl, 2007). Even though bit-rates offered by providers have been increasing in Chile in recent years (as well as “broadband” connections), confusion persists. 7KHUHSRUWE\&LVFRLQ&KLOHNHSWGH¿QLQJ broadband as any “always-on” connection even if they were lower than 128 kbps for either downloading or uploading (Cisco Systems, 2007). Therefore, for the sake of simplicity, we will understand as broadband those connections advertised as such by established, legally accountable local providers, provided they are “always on” in contrast to dial-up, pay-per-minute modem links. Respondents surveyed by WIP-Chile in 2003, 2004, and 2006 would have been unable to distinguish beyond WKLVGH¿QLWLRQ:HWKHUHIRUHVKDUHWKHGH¿QLWLRQRI broadband used by the Ministry of Telecommunications, which distinguishes between “dedicated” (“always-on”, a.k.a. broadband) and “commuted” (dial-up, modem) connections.6 By December 2005, 80.7% of connections were dedicated; this ZDVWKHODWHVWRI¿FLDO¿JXUHLQ&KLOHDYDLODEOHDW the time of writing.7
429
Precisions about the Broadband Divide in Chile
RESULTS AND DISCUSSION Measuring the Divides Using the Odds/Ratio Method 7KLVVHFWLRQDWWHPSWVWRGH¿QHWKHGLJLWDOGLYLGH in Chile by using ratios of indicators drawn from :,3&KLOH¶VVXUYH\VRIDQG Operationalization of the digital divide can be done in different ways, yet that would yield noncomparable results (Herrera, 2006). One possibility consists of calculating differences in terms of percentage points concerning a certain indicator, such as access to the Internet between the most favored group and the least favored one. Another involves calculating a ratio instead. 7KH¿UVWDOWHUQDWLYHLQGLFDWHVWKHUHDOGLIIHUHQFH LQDFFHVVEXWGRHVQRWUHÀHFWWKHRYHUDOOOHYHORI development of this variable within the country. For example, take Country A in which the least favored group has 2% of Internet access and a most favored group with 22%. This 20% gap would be exactly the same as that of Country B in which the least favored and the most favored groups show a 70% and 90% of access respectively. Yet the meaning RIWKH¿JXUHLVFRPSOHWHO\GLIIHUHQW In contrast, calculating a ratio implies comparing one group against the other in relative terms. This has the advantage of taking into account the stage of technological development or access to information technologies (IT) of the country anaO\]HG)ROORZLQJWKHSUHYLRXVH[DPSOH&RXQWU\$¶V UDWLRLV DQG&RXQWU\%¶VLV So the conclusion is that the gap is considerably lower in B, despite the percentage point difference is the same for both. Ratio-based gaps are often wider in countries where the general levels of access are lower. Based on the concept of a ratio, one can calculate the difference in percentage points between a country and a referential one used as a yardstick. Alternatively, a group of countries can be contrasted. For instance, Chile reduced its technological gap with respect to the United States from 77% in 2001 to 70% in 2005. The Santiago Chamber of Commerce (CCS, 2006), through its ICATI Index,8 GHWHUPLQHGWKDW&KLOH¶VWHFKQRORJLFDOJDSYLVjYLV
430
the developed nations fell from 61% in 2000 to 55% in 2005 (CCS, 2006). A similar calculation can be done with respect to the OECD countries, which include the most advanced economies as well as middle-income ones comparable to Chile (Mexico, Turkey, Hungary, Poland, Slovakia, and the Czech Republic). This gap reduction has been more considerable in terms of Internet and mobile-phone access LQVWHDGRI¿[HGWHOHSKRQHOLQHVDQG3&RZQHUVKLS 318' ,Q&KLOH¶VSHUFDSLWDFRQnectivity to mobile phones and the Internet was 77% and 48% of the OECD average respectively. According to the Ministry of Telecommunications, ZKLOH¿[HGWHOHSKRQHVVWDJQDWHGDWDERXWRI the population by the late 1990s, mobiles grew IURP DOPRVW QRQH LQ WR LQ WKH ¿UVW quarter of 2006. Based in this comparison in relative terms, Di Maggio, Hargittai, Russell, and Robinson (2001) calculated the odds ratio (or “advantage ratio”) as [pj/(1-pj)]/ [pk /(1-pk)], where pj is the probability of the most favored group of having access to IT and pk is the same probability for the less favored group. An odds ratio of 6.3 means that the probability of access to IT in contrast to the probability of not having access is 6.3 higher for the most favored group compared to the least favored one (Herrera, 2006). Table 1, Table 2, and Table 3 summarize the odds ratio between the most and least favored groups of Chileans surveyed by WIP Chile in 2003 and 2006 in terms of education (Table 1), income (Table 2), and age (Table 3). Preliminary evidence from WIP countries suggest these are the most relevant variables concerning use and access to the Internet.9 In each one of these tables, data from 2003 and 2006 are divided in rows indicating each type of gap we want to analyze, that is, users vs. nonusers of the Internet, home users vs. users elsewhere, and broadband home users vs. all residential users. Additionally, each table shows the contrast between the least and most favored group in terms of percentage point differences as well as ratios. Let us start with Table 1, which compares the gap according to education level. The least favored
Precisions about the Broadband Divide in Chile
Table 1. Digital gap ratios in terms of education % least favored
% most favored
Percentage
group
group
point
Favored/not
difference
favored ratio
2006
2003
Secondary
Odds ratio1
education or
Higher
less
education
Internet users / sample
27.4
72.4
45.0
2.6
7.0
Home users / users
31.4
68.4
37.1
2.2
4.7
Broadband home users / home users
56.9
59.0
*
*
*
Internet users / sample
20.7
71.2
50.5
3.4
9.5
Home users / users
35.9
73.4
37.6
2.0
4.9
Broadband home users / home users
91.7
86.6
*
*
*
Source: WIP-Chile surveys 2003 and 2006
'LIIHUHQFHVDUHQRWVWDWLVWLFDOO\VLJQL¿FDQWDWYDOXHVS 1 Odds ratio: [pj /(1-pj )]/ [pk /(1-pk )], where pj is the probability of the most favored group of having access to IT and pk is the same probability for the less favored group.
group reached secondary education or less, whereas their counterparts went to higher education (technical, college, or university). Data did not show a difference between broadband home users and the remaining home users, but it is interesting to note that the odds ratio deteriorated from 2003 to 2006 when analyzing Internet users over overall population. Table 2 examines the gap according to income. As usual, there is an important difference between the rich and the poor. Yet the fast spread of broadband in Chile, which has been even faster among the lower-income groups, account for lower ratios in terms of home broadband connections between 2003 and 2006. Finally, Table 3 looks at the age gap. Despite Internet users in general tend to be younger than nonusers, there are interesting differences: whereas the older appear in disadvantage when both “Internet users above sample” and “broadband home users above home users” categories are analyzed, they become the most favored group in the home user category. Aged people are more likely than the young to be home users who go online either at school and/or at public places in a higher proportion (WIP-Chile, 2006). Of course, wealthier
children and adolescents are heavy broadband users at home as well, but they are a minority among all Web users. 7KH8QLWHG1DWLRQV¶'HYHORSPHQW3URJUDPLQ &KLOH 318' FRQ¿UPV WKDW WKH LQFRPH based gap is narrower among the young, since most of them enjoy Web access at schools regardless of their socioeconomic level. Nevertheless, the wealthier youngsters who attend to private, feepaying schools get online mostly at their homes, while those attending to public-sector schooling do so at school where usage is more restricted. These three types of gap measurement—percentage point difference, ratio, and odds ratio—are rather consistent among them. Income and educaWLRQDFFRXQWIRUWKHELJJHVWJDSFRQ¿UPLQJRWKHU studies showing the high dependence of IT access on economic factors (see WIP-Chile, 2006). The probability of being an Internet user as well as a broadband home user is higher in the upper income group. In 2003, the main gap in terms of income was related to broadband. Yet this changed—in 2006, 72% of the lower income group had this technology at home. This is probably related to lower prices of connections and aggressive marketing strategies of providers (mostly cable TV operators).
431
Precisions about the Broadband Divide in Chile
Table 2. Digital gap ratios according to income using the WIP Chile surveys of 2003 and 2006 % least favored
% most favored
Percentage
group
group
point
Favored/not
High-upper
difference
favored ratio
Odds ratio1
2006
2003
middle income Lower income
(ABC1-C2
(C3-D groups)
groups)
Internet users / sample
22.1
69.7
47.6
3.1
8.1
Home users / users
22.3
77.2
54.9
3.5
11.8
Broadband home users / home users
4.9
72.2
67.3
14.6
50.1
Internet users / sample
31.5
73.3
41.9
2.3
6.0
Home users / users
23.8
86.8
63.0
3.6
21.1
Broadband home users / home users
71.8
91.5
19.7
1.3
4.2
'LIIHUHQFHVDUHQRWVWDWLVWLFDOO\VLJQL¿FDQWDWYDOXHVS 1 Odds ratio: [pj /(1-pj )]/ [pk /(1-pk )], where pj is the probability of the most favored group of having access to IT and pk is the same probability for the less favored group.
% least favored
% most favored
group
group
point
Favored/not
Odds
Older:
Younger:
difference
favored ratio
ratio1
46 - 60 years
12 - 29 years
Internet users / sample
12.1
61.5
49.4
5.1
11.6
Broadband home users / home users
64.3
57.9
23.1
75.1
Broadband home users / home users
84.1
86.2
2003
Internet users / sample
Home users (vs. nonhome users)
2006
2006
2003
Table 3. Digital gap ratios according to age using the WIP Chile surveys of 2003 and 2006
Home users (vs. nonhome users)
Younger:
Older:
12 - 29 years
46 - 60 years
50.7
70.4
51.5
79.6
Percentage
*
*
*
52.0
3.3
10.0
*
*
*
19.8
1.4
2.3
28.0
1.5
3.7
'LIIHUHQFHVDUHQRWVWDWLVWLFDOO\VLJQL¿FDQWDWYDOXHVS 1 Odds ratio: [pj /(1-pj )]/ [pk /(1-pk )], where pj is the probability of the most favored group of having access to IT and pk is the same probability for the less favored group.
432
Precisions about the Broadband Divide in Chile
Factors Affecting the Digital Divides Depending on which digital divide we are referring to, we can hypothesize that “pro-inclusion” factors are different. For example, in the United States, the biggest age divide occurs in terms of the general level of access to the Internet, whereas the biggest income gap exists in terms of the usage of high-speed, broadband Internet (Di Maggio et al., 2001). In Chile, as said before, the main education and age gap occurs between Internet users and nonusers. A higher income generally ensures Internet access at home. Based on the most important factors pointed out by literature which explain the differences in access, a regression analysis was done to differentiate among the three divides mentioned above and thus assess their importance. The variables analyzed include education, income, working status, age, gender, household composition, online social networks, knowledge of the English language, psychological aspects (such as technological motivation), and usage of other communication media (Di Maggio & Hargittai, 2001; Godoy & Herrera, 2004; Herrera, 2005; Keniston, 2004; Lenhart et al., 2003; Norris, 2001; OECD, 2001). According to literature, two main dimensions explain the differentiated access to the Internet: individual resources and social opportunities (De Haan, 2004). We now explain each of them in more detail. $PRQJWKHLQGLYLGXDOUHVRXUFHVD¿UVWW\SH relates to material ones such as money and time. Income and IT costs affect the ability to purchase these products—the higher the income and the lower the cost, the higher the chance of acquiring them (De Haan, 2004). In general, in countries where Internet penetration is low, the online population is more elitist and thus the socioeconomic gap is wider. Even in technologically advanced countries such as Japan and the United States, this break is still highly relevant, yet its relative LPSRUWDQFHLVGHFOLQLQJEHFDXVHRIWKHLQÀXHQFH of other factors (Chen & Wellman, 2004). +DYLQJIUHHWLPHLVDOVRLQÀXHQWLDO'H+DDQ and Huysmans (2002) found that Web users in
the Netherlands had an average of three hours of extra spare time than nonusers after controlling background characteristics. This relates both to marital status and paternity—those who enjoy more free time are mostly single with no children. According to Raban (2004), single persons go online more frequently than married ones. On the other hand, other studies discovered that marital status is affected by gender since single women are less online than those who are married because the latter have higher incomes than the former. Subsequently, married females are more likely to access to the Internet at their homes (Robinson, Di Maggio, & Hargittai, 2003). Another type of individual resources consists on cognitive ones, also called human capital. A ¿UVWRQHLVWKHVRFDOOHG³LQIRUPDF\´UHVRXUFHV that is, the ability to handle information required to make sense of the data available through digital technologies. In general, education is a proxy for these variables, given their high correlation (De Haan, 2004; Robinson et al., 2003). Another resource is related to the knowledge of the languages in which most online content is produced, mainly English, yet in this analysis we did not consider this factor because we had not comparable questions along the years. An additional factor is the disposition towards technology, since mental barriers can hamper both access and adoption (De Haan, 2004). 2WKHU W\SHV RI LQÀXHQWLDO LQGLYLGXDO IDFWRUV are age and gender. In particular, the relationship between age and Internet access is particularly strong. In 2002 in the United States, some 80% of individuals aged 3 to 15 years were Internet users. The percentage fell to 65% and 60% among those aged 25 and 55 respectively. After that age, the rate of users diminishes abruptly (Di Maggio, Hargittai, Celeste, & Shafer, 2003). This age gap is high in most countries, and it has even increased since new users are mainly young (Chen & Wellman, 2004). Meanwhile the link between gender and Internet usage was stronger at the beginning of the expansion of IT. Initially men had more access to the Web than women, mostly because of their higher share of the (paid) workforce. But after the grow-
433
Precisions about the Broadband Divide in Chile
ing access at home and school, the gender gap has decreased (Di Maggio et al., 2003). Nevertheless, in most countries apart from the United States, this gap still remains and it is even increasing in countries such as Germany and Italy (Chen & Wellman, 2004). Additionally, there are social contexts to consider. The household, the school, and the workplace represent opportunities for IT access. Households not only are endowed with economic resources enabling persons to be online (or not), but also are structured in ways that some of its members can help the rest to adopt the technology (De Haan, 2004). For instance, youngsters who have Web access at their schools or colleges can stimulate the purchase of a computer at their home, and teach their elders how to get online. In the United States, families with children have more computers than those without them (Di Maggio et al., 2003). In Chile, a middle-income country often categorized within the so called Third World, Web access at schools is a very important factor explaining the access to the Internet among the young (Godoy & Herrera, 2004). Besides, schools also provide knowledge and skills required for taking advantage of IT. Being employed also stimulates the access to IT (De Haan, 2004). The workplace is increasingly becoming more technologically endowed as economies move from manufacturing to services and then to information—even in Chile, according to the BIT-Chile project (Godoy, Herrera, Sepulveda, Lever, & Myrick, 2006). Businesses increasingly use IT to link the different steps of the value chain, linking consumers, suppliers, partners, and/or centers of production. The production process is also increasingly more dependent on information, ZKLOHRUJDQL]DWLRQDOVWUXFWXUHVEHFRPHÀDWWHUDQG PRUHÀH[LEOH 'L0DJJLRHWDO *RGR\HW al., 2006). All these changes make workers more dependent on digital technologies for their work. In the United States, according to the 2001 Current Population Survey, 65% of employees are Internet users in contrast with 37% of those who do not work (Di Maggio et al., 2003). Concerning social resources, an important one refers to online social networking. According to
434
De Haan, people will be more likely to buy and use new IT if more persons in their social networks share them (De Haan, 2004, p. 79). In Chile we also found that having relatives or friends online is a strong predictor of Internet adoption (Herrera, 2005). Usage of other (complementary) communication technologies is also relevant, particularly mobile phones (Di Maggio et al., 2001), in contrast to television. In fact, Rice and Katz (2003) note that mobile phone subscribers are more likely to use the Internet than non subscribers. Furthermore, the longer a person is a mobile subscriber, the longer that person is a Web user as well. Yet despite levels of usage may be similar, the populations concerned are not completely equivalent—especially in developing countries like Chile, where cellular handsets are far more widespread than PCs (70% and 40%, respectively). Subsequently, factors predicting the usage of both technologies are not equal.
Determining the Factors Affecting the Three Digital Divides Based in these considerations, the following independent variables from the WIP-Chile surveys of 2003 and 2006 were analyzed in order to determine WKHLULQÀXHQFH HGXFDWLRQLQFRPHZRUNLQJVWDWXV age, gender, household composition (whether the interviewed lived with a son/daughter below 18 years of age), perceived importance of IT in general, perceived political importance of the Internet, perceived importance of the Internet as a source of information and entertainment, and usage of mobile phones. As there were changes in the WIP questionnaire along the years, these were the only questions that remained constant and were therefore used in the analysis. Subsequently we developed three main models of logistic regression, each one for each type of digital divide. Table 4 summarizes the percentages of total users, residential users, and broadband users separately according to each category of independent variables. As in previous tables, 2003 and 2006 are compared. Income relates with all variables in both years: the higher the income, the higher the probability of being a broadband home user.
Precisions about the Broadband Divide in Chile
Table 4. Bivaried relationships between independent and dependent variables 2003
2006 Broadband home
Internet user
Home
Broadband home user
Home user
user
Internet user
user
Below secondary
27.4
31.4
56.9
20.7
35.9
91.7
Secondary level
23.2
45.7
49.6
39.3
41.9
85.9
university)
72.4*
68.4*
59.0
71.2*
73.4*
86.6
ABC1C2 (high)
69.7*
77.2*
72.2*
73.3*
86.8*
91.5*
C3 (middle)
31.8
46.4
33.9
48.7
53.4
81.7
D (low)
22.1
22.3
4.9
31.5
23.8
71.8
Works or studies
12.3
61.7
53.2
20.8
71.1
86.7
studies
40.6*
62.8
56.5
62.5*
57.4
86.3
Male
40.2
56.1
56.4
50.8
62.6
84.9
Female
35.5
55.0
56.8
45.7
54.9
88.1
12-29
61.4*
50.8
57.9
75.1*
51.5
86.2
Higher (technical, Education
Income
Neither works nor Working Status
Gender
30-45
29.2
62.7
51.5
38.5
65.2
88.1
Age
46-60
12.1
70.4
64.3
23.1
79.6*
84.1
Lives with a son/
Yes
37.3
58.3
86.3
daughter aged less than 18
No
Perceived importance
Pessimistic
36.5
61.6
56.8
55.7*
59.0
86.7
45.5
58.6
88.6
of IT Optimistic
39.1
53.3
57.1
49.5
58.6
85.1
Disagree
49.2*
54.7
57.6
46.2
58.1
82.2
are at a disadvantage
Agree
36.2
56.5
56.7
50.8
58.6
88.1
By using the
Disagree
52.1*
59.7*
58.2
57.8*
59.3
86.8
21.1
39.8
45.9
36.9
58.7
85.8
46.9*
57.4
58.3
59.6*
59.7
88.6
3HRSOHZKRGRQ¶W access to the Internet
Internet people will understand politics better
Agree
By using the Internet
Disagree
people will have more say in government
Agree
34.4
52.4
49.6
39.4
56.5
81.2
Yes
46.7*
66.3*
59.7
55.6*
61.0*
87.5
Has a mobile phone
No
27.6
34.5
44.6
25.7
45.4
75.9
Importance of
Not important/a little
23.5
46.8
46.1
35.3
52.9
81.9
information
Important
44.7*
57.6
58.4
54.5*
60.3
87.2
Importance of
Not important/a little
34.6
57.0
50.9
44.8
57.5
86.7
Important
44.9*
53.7
64.5
57.7*
60.7
85.9
Internet as a source of
Internet as a source of entertainment
&KLVTXDUHGWHVWVWDWLVWLFDOO\VLJQL¿FDQWDWOHYHOS 7KLVVWDWLVWLFDOWHVWZDVSHUIRUPHGDWLQGLYLGXDOOHYHOIURPFRQWLQJHQF\ tables of each “independent variable” with the respective “dependent variables”. For example: education (below secondary, secondary level, higher) by Internet user (user, non user) above total sample; education by home user (residential user, non residential user) above users; education by broadband home user (broadband user, modem user) above home users.
435
Precisions about the Broadband Divide in Chile
7DEOH5HJUHVVLRQFRHI¿FLHQWVSUHGLFWLQJWKHJDSEHWZHHQXVHUVQRQXVHUVDQG Independent
YEAR
variables
Income
Schooling
2003
High and uper middle ABC1C21
2006
B
Exp(B)
Sig.
B
Exp(B)
Sig.
1.088
2.970
.001
1.763
5.828
.000
Middle C31
.347
1.415
.273
.454
1.574
.028
Complete or incomplete secondary2
1.236
3.443
.072
.013
1.013
.965
Higher
2
4.011
55.185
.000
1.108
3.028
.001
12-29 years3
1.885
6.584
.000
2.845
17.199
.000
30-45 years3
.436
1.547
.241
.908
2.479
.000
Working status
Works or studies4
.706
2.025
.025
1.143
3.137
.000
Mobile phone
Has5
.686
1.986
.011
.634
1.886
.004
Importance of
Important for information, not important for
1.028
2.794
.001
.783
2.189
.001
Internet
entertainment6 1.164
3.201
.001
.704
2.022
.002
Age
Important for information and entertainment6
Cox & Snell R square
.439
.381
Nagelkerke R square
.612
.509
1
Category of comparison: lower income D. Category of comparison: lower than secondary level. 3 Category of comparison: 46-60 years. 4 Category of comparison: does not work nor study. Category of comparison: does not have mobile phone. 6 Category of comparison: no or little importance of Internet as a source of information and entertainment. 2
Despite education being correlated to income, it GLGQRWVKRZDVWDWLVWLFDOO\VLJQL¿FDQWUHODWLRQVKLS with broadband usage at home. Table 4 also reveals that those who work and study are more likely to be online. This is somehow related to age: the young are more likely to be Internet users as they have access to it at their schools. On the other hand, neither gender nor the perceived importance of IT are related to Internet usage. Table 4 also reveals that the lack of children in the household increases the probability of being a Web user, even after controlling by age. This apparently contradicts some conclusions from other studies commented above (Di Maggio et al., 2003). Therefore, having children seems to reduce the free time available as well as the extra-family sociability. When we checked the relationship between marital
436
status and Internet usage controlled by age, we found that not having a partner was linked to more usage only among the young. This is consistent ZLWK5DEDQ¶V(/LYLQJ3URMHFW ZKLFKIRXQG that marital status was the most important factor explaining the difference between “heavy users” and “light users.” The former group was mostly composed by unmarried persons who were not absorbed by the highly demanding child-caring. Age has different effects according to which gap is analyzed: the lower the age, the higher the probability of being a user. Yet more aged users get connected in a higher proportion at their homes, since they have less chance to do so at work or at school. Contrary to what may appear logical, users give less importance to the political implications of the Internet as well as to the very fact of be-
Precisions about the Broadband Divide in Chile
7DEOH 5HJUHVVLRQ FRHI¿FLHQWV SUHGLFWLQJ WKH JDS EHWZHHQ UHVLGHQWLDO XVHUQRQ UHVLGHQWLDO user, 2003 and 2006 Independent
YEAR
variables
2003 B
Income
Schooling
1
Sig.
B
Exp(B)
Sig.
High and uper middle ABC1C2
1.466
4.331
.003
2.745
15.569
.000
Middle C31
.394
1.482
.408
1.175
3.237
.000
-.162
.850
.905
.069
1.071
.878
.426
1.530
.755
.675
1.965
.141
.119
1.126
.826
-.315
.730
.408
-.239
.788
.660
-.380
.684
.357
Complete or incomplete secondary
2
Higher2 Age
2006
Exp(B)
12-29 years
3
30-45 years3 4
Working status
Works or studies
-.417
.659
.412
-.796
.451
.035
Mobile phone
Has5
.736
2.088
.049
.004
1.004
.991
Importance of
Important for information, not important for
.325
1.385
.467
.173
1.188
.580
Internet
entertainment6 .292
1.339
.526
.276
1.318
.371
Important for information and entertainment6
Cox & Snell R square
.142
.272
Nagelkerke R square
.194
.366
1
Category of comparison: lower income D. Category of comparison: lower than secondary level 3 Category of comparison: 46-60 years. 4 Category of comparison: does not work nor study. Category of comparison: does not have mobile phone. 6 Category of comparison: no or little importance of Internet as a source of information and entertainment. 2
ing connected (they agree less with the statement “persons who have no access to the Internet are at DGLVDGYDQWDJH´
of a mobile phone, and perceived importance of the Web as a source of entertainment and information. Therefore the following tables show how the different variables discussed affect the probabilities of being an Internet user (Table 5), a household user (Table 6), and a broadband home user (Table 7). All these variables have some independent as ZHOODVVWDWLVWLFDOO\VLJQL¿FDQWHIIHFWRQWKHSUREability of being an Internet user, with a rather high explanatory power (Nagelkerke R Square= .612 for the year 2003 and .704 for 2006). In 2003, having higher education was the main explanatory variable for Internet usage, even more than income. This is because education, despite being related to earnings, also means having technical skills required online. Thus being wealthier is not
437
Precisions about the Broadband Divide in Chile
7DEOH5HJUHVVLRQFRHI¿FLHQWVSUHGLFWLQJWKHJDSEHWZHHQEURDGEDQGUHVLGHQWLDOXVHUQRWEURDGEDQG residential user, 2003 and 2006 Independent
YEAR
variables
2003
2006
B
Exp(B)
Sig.
B
Exp(B)
Sig.
High and uper middle ABC1C21
3.721
41.320
.003
1.569
4.801
.003
Middle C31
2.057
7.819
.108
.695
2.003
1.67
Complete or incomplete secondary2
-.922
.398
.670
-.729
.482
.520
Higher2
-.845
.430
.690
-1.122
.326
.317
12-29 years3
.374
1.453
.569
.405
1.500
.401
30-45 years3
-.394
.675
.545
.566
1.761
.302
Works or studies4
-.183
.833
.786
-.183
.833
.739
Mobile phone
Has5
.525
1.690
.358
.628
1.875
.229
Importance of
Important for information, not important for
.261
1.298
.676
.398
1.490
.430
Internet
entertainment6 .670
1.953
.304
.413
1.511
.413
Income
Schooling
Age
Working status
Important for information and entertainment6
Cox & Snell R square
.226
.050
Nagelkerke R square
.303
.091
1
Category of comparison: lower income D. Category of comparison: lower than secondary level 3 Category of comparison: 46-60 years. 4 Category of comparison: does not work nor study. Category of comparison: does not have mobile phone. 6 Category of comparison: no or little importance of Internet as a source of information and entertainment. 2
enough for access–knowledge is also required. Nevertheless, in 2003, age became rather important as well: when controlling by the rest of the variables, the young showed a higher probability of being a user than those aged 46-60 years. The cause was the rather high Internet coverage at Chilean schools. In 2006, the effect of working and studying increased. These results were consistent with other studies showing that education is the most powerful predictor of Internet access, even above income (Robinson et al., 2003). Having a mobile SKRQHLVDOVRLQÀXHQWLDOWKRXJKOHVVWKDQLQFRPH education, age, or working status. Its effect stayed quite the same from 2003 to 2006. The perceived importance of the Internet as a VRXUFHRILQIRUPDWLRQLVDOVRVLJQL¿FDQWWRSUHGLFW
438
its usage. Though the predictive power of this variable diminishes in 2006, it is nevertheless higher than that of the perceived value of the Web as a source of entertainment. This logistical-regression model fared worse when trying to predict usage of the Internet at home (Nagelkerke R square of .194). In 2003, the only relevant probability was related to the highest income brackets ABC1-C2. The explanatory power of the model improved for 2006 data (Nagelkerke R square of .366) since both income and working status became more relevant. This might have been caused by the fact that many home users in 2006 were previously online at either school or work, and subsequently acquired home connections.
Precisions about the Broadband Divide in Chile
Finally, the type of home connection—either broadband or dial-up—is still mostly related to income, because users pay for it. Yet its predictive power decreased: the Nagelkerke R square of .303 in 2003 dropped to .091 in 2006, consistent with the fast growth of broadband among the lower income groups in Chile.
Impact of Broadband in E-Commerce, E-Government, Interpersonal Communications, and “Traditional” (Off-Line) Mass Media :HDOUHDG\H[DPLQHGWKHIDFWRUVLQÀXHQFLQJWKH probability of being an Internet user. Now we will try to determine whether Web usage, places of use, DQGVSHHGRIFRQQHFWLRQLQÀXHQFHSHRSOHLQWKH ¿HOGVRIHFRPPHUFHHJRYHUQPHQWLQWHUSHUVRQDO communications, and traditional mass media. Apart from comparing data from 2003 and 2006, we will compare between broadband home users vis-à-vis all home users. It seems logical to expect the Internet making an effect on “free” time, as the Web can steal much of it away from communications and social activities–especially face-to-face interactions.10 Some studies based on time-diaries suggest that Internet use at home has a strong negative impact on time spent with friends and family as well as on other social activities (Nie y Hillygus, 2002). On the other hand, Internet can save time spent ZRUNLQJ DQGRU VWXG\LQJ EXW WKDW LV GLI¿FXOW WR measure (we have not done so within the context of the WIP-Chile project). Yet it is not clear that the Internet affects negatively the use of other communication technologies as many of them can be used simultaneously (multitasking). In fact, Web users are often more intensive users of “traditional” mass media except television (see Robinson, Kestnbaum, Neustadtl, & Alvarez, 2000). Evidence gathered from the countries participating in the WIP project support this claim (see Godoy & Herrera, 2004). Both the place of access and the quality of the connection also affects times and types of Web usage. Some studies show that having a broadband connection at home is related to longer Internet
XVH DVZHOODVLWVÀDWUDWHVXEVFULSWLRQDOORZLQJ an “always on” connection, as pointed out by the WIP team in the United States). For instance, the E-Living project, based on a two-year panel study, concluded that a faster Web connection was related to more time spent online in the UK, Italy, Germany, Norway, and Israel (Raban, 2004). Choudrie and Dwinedi (2007) got a similar result from a mail sample of 358 cases in the UK. Yet, Anderson and Raban (2005) argued that these differences could have been caused by a selection effect: those who started using broadband might have spent more time online than those who had a modem because they anyway used the Web for longer before. In other words, broadband was not the cause of longer times of usage by itself—a claim which contradicts the conclusions drawn by Choudrie and Dwivedi (2007). Table 8 is related to e-commerce. It shows that neither home usage nor a broadband connection LQÀXHQFHV WKH SUREDELOLW\ RI PDNLQJ SXUFKDVHV online in Chile. Although both home and broadband home users buy more, this relationship disappears when controlling by income. This is because home users are generally wealthier than the rest. Table 9 focuses on time spent both on different places of access and on a broad categorization of online activities, according to type of user. Both in 2003 and 2006, all home users were connected nearly 10 hours per week longer than non home users. Yet home broadband users spent little more time online than their dial-up counterparts in 2003, DQGWKHGLIIHUHQFHGURSVWRVWDWLVWLFDOLQVLJQL¿FDQFH in 2006. This apparently contradicts the results of the E-Living project, commented above. Differences on average times spent online are mostly explained by usage patterns at home, as shown in the lower half of Table 9 (only 2006 data are included, as they were not comparable to previous years). The household became the most important place of Web use in Chile in 2004, when the school as the main location lost its lead. The same occurred in the late 1990s in the more developed WIP countries. On the other hand, there are no major differences regarding Web usage at the workplace. Besides, not all users work. The same happens for usage at school, except when the
439
Precisions about the Broadband Divide in Chile
Table 8. Online purchases controlled by income (% people who did online purchases in the last 12 months), 2003 and 2006 Income level
High and
Year
Home user? (base: users)
Broadband user? (base: home users) Dial-up home
Broadband
Nonhome user
Home user
Total
user
home user
Total
28.1%
34.1%
32.7%
24.5%
39.4%
35.0%
2006
21.0%
37.8%
35.6%
37.5%
37.6%
37.6%
2003
18.0%
18.8%
18.4%
15.1%
26.0%
18.8%
2006
14.6%
23.4%
19.3%
11.8%
26.3%
23.7%
2003
10.2%
6.4%
8.8%
6.8%
.0%
6.4%
2006
6.7%
11.2%
7.8%
.0%
15.6%
11.2%
2003
upper middle ABC1-C2
Middle C3
Lower D
1RWH QRUHODWLRQVKLSLVVWDWLVWLFDOO\VLJQL¿FDQWLIS Table 9. Online usage times per activity according to user type, 2003 and 2006 YEAR 2003
2006 Type of home
Home user?
Home user?
connection (base:
(base: users)
(base: users)
home user)
Non
Type of home connection (base: home user)
Non
home
Home
Dial
user
user
up
Broadband
home
Home
Dial
user
user
up
Broadband
HOURS PER WEEK ON THE INTERNET IN DIFFERENT PLACES OF ACCESS Any place (total users)
9.34
19.12*
15.66
22.13*
13.59
24.14*
23.17
24.20
Work (total users)
2.57
3.16
2.71
3.63
3.22
4.09
4.28
4.03
School (total users)
3.09*
1.59
1.58
1.58
2.41*
1.46
1.82
1.38
.00
11.33*
8.63
13.56
.00
13.12*
10.33
13.57
2.25
1.60
1.59
1.65
4.62*
2.84
3.72
2.65
Home (total users) Other places (total users) HOURS PER WEEK SPENT ON DIFFERENT ONLINE ACTIVITIES E-mail
.
.
.
.
2.26
2.78
3.71
2.61
Chats, discussion groups
.
.
.
.
1.59
3.77*
2.40
3.93
Entertainment
.
.
.
.
1.62
2.78*
1.64
2.99
Information-seeking for work or studies
.
.
.
.
2.86
3.81
3.48
3.75
Information-seeking not related to work or studies
.
.
.
.
.74
1.78*
1.99
1.75
'LIIHUHQFHVDUHVWDWLVWLFDOO\VLJQL¿FDQWDWDOHYHOS
440
Precisions about the Broadband Divide in Chile
Table 10. “Off-line” (non-Internet) usage times per activity according to user type, 2003 and 2006 2003
HOURS
2006
PER WEEK SPENT ON OFF-LINE
Internet user?
Home user?
(base: total sample)
(base: users)
ACTIVITIES
radio Reading newspapers Watching TV Sports, excercise Sharing with family Sharing with friends
Internet user?
connection (base:
(base: total
home user)
sample)
Non Nonuser
Listening to
Type of home
User
Home user? (base: users)
Type of home connection (base: home user)
Non
home
Home
Dial
user
user
up
Broadband Nonuser
User
home
Home
user
user
Dial up
Broadband
12,05*
8,52
10.22*
7.15
9.44*
5.54
15,84*
10,53
13.54*
8.42
9.46
8.23
1,95
1,90
1.45
2.26*
2.18
2.34
2,21
2,12
1.65
2.45*
1.96
2.50
16,56*
11,72
13.02
10.68
11.35
9.99
17,31*
12,56
15.42*
10.57
8.87
10.90
1,51
2,93*
3.12
2.78
2.46
3.07
1,28
2,31*
2.30
2.31
1.75
2.38
33,73*
26,04
27.97
24.50
26.96
22.56
29,19*
21,12
21.77
20.67
22.05
20.41
6,21
11,07*
10.85
11.26
11.49
11.10
5,25
9,38*
10.93*
8.29
7.70
8.25
'LIIHUHQFHVDUHVWDWLVWLFDOO\VLJQL¿FDQWDWDOHYHOS
incumbent ceases to be a student (and, as in the previous case, not all users are students). Table 9 also shows that home users dedicate more time performing all types of online activiWLHV :H FRXOG QRW SURYH VWDWLVWLFDOO\ VLJQL¿FDQW differences according to type of connection and time allocated to online activities in 2006. Finally, Table 10 shows whether broadband was related to other socialization, communication, and recreation activities in Chile—with no success. Usage was more relevant: nonusers listen to more radio, watch more television, and spend more time with their families than users, even after controlling by income and age. Web users, on the other hand, spend more time doing sports and sharing with friends–as they are younger, they are more likely to do physical exercise and socializing with people other than relatives. Home users spend less time with radio, yet this relationship disappears when data are controlled by income–the most devoted radio listeners in Chile belong to the poorer groups, therefore they enjoy fewer Web connections. The opposite happens with newspapers; home users read more printed
media, perhaps unsurprisingly as they are wealthier and more educated. As said before, television is the only “traditional” mass medium that shows consistently lower times of usage among all home XVHUVDVFRQ¿UPHGE\RWKHU:,3FRXQWULHV *RGR\ & Herrera, 2004). Yet broadband per se neither affects this nor the other activities analyzed.
CONCLUSION At least in Chile, the main digital gap is the one between users and nonusers of the Internet, both in terms of age and education level. Income mainly affects the probability of having broadband at home. A Chilean is more likely to be a user if younger, more educated, and wealthier. Also more prone to go online are those who work or study, have a mobile phone, and value the Web highly as a means of information and entertainment. Yet broadband has been rapidly expanding among all socioeconomic segments between 2003 and 2003, especially among the lower income
441
Precisions about the Broadband Divide in Chile
groups; therefore, the broadband gap at home is becoming less relevant. Factors predicting residential usage and broadband connections are less clear and seem to depend on the availability of disposable income. Other factors may be relevant, but new data and surveys are required. Among factors that may be relevant are the need to use the Internet for working at home, the presence of schoolchildren at the household, or patterns of usage requiring high speed connecWLRQVIRUKHDY\,QWHUQHWWUDI¿F SDUWLFXODUO\YLGHR Yet the WIP questionnaire, a broad survey about Internet usage in general, does not deepen into WKRVH H[WUD IDFWRUV \HW 0RUH VSHFL¿F TXHVWLRQV should be included in the next WIP-Chile survey, planned for 2008. The authors of this chapter also tried to measure the impact of residential access and broadband both over time spent online in different locations (home, work, school, and other places) and over off-line, non-Internet activities. Again, the impact of this WHFKQRORJ\ZDVQRWVLJQL¿FDQW7KLVPD\EHUHODWHG WR WKH UHODWLYHO\ OD[ GH¿QLWLRQ RI EURDGEDQG LQ Chile, as well as the fast growth of these dedicated, “always-on” connections in recent years.
REFERENCES Anderson, B., & Raban, Y. (2005). The social impact of broadband household Internet access (Chimera Working Paper, Vol. 2005-06). Colchester: University of Essex. CCS. (2006). La economía digital en Chile. Santiago de Chile: Centro de Estudios de la Economía Digital, Cámara de Comercio de Santiago. Center for the Digital Future. (2007). Surveying the digital future, year six. Los Angeles, CA: USC Center for the Digital Future. Chen, W., & Wellman, B. (2004, Spring/Summer). The global digital divide - within and between countries. IT & Society, 1(7), 39-45. Choudrie, J., & Dwivedi, Y.K. (2007). Broadband impact on household consumers: Online habits
442
and time allocation patterns on daily life. International Journal of Mobile Communications,(2), 225-241. Cisco Sysrtems. (2007). Estudio de Banda Ancha en Chile 2002-2010. Latinoamérica, Santiago: IDC/Cysco Systems. Retrieved online July 10, 2007, from http://www.ciscoredaccionvirtual. com/redaccion/multimedia/formatos.asp?materi al=1042&valor=documentos De Haan, J. (2004, Spring/Summer). A multifaceted dynamic model of the digital divide. IT & Society, 1(7), 66-88. De Haan, J., & Huysmans, F. (2002). Differences in time use between Internet users and non-users in the Netherlands. IT & Society, 1(2 ), 67-85. Di Maggio, P., & Hargittai, E. (2001). From the “digital divide” to “digital inequality”: Studying Internet use as penetration increases (Working Paper Series No. 15, Center for Arts and Cultural Policy Studies, Woodrow Wilson School). Princeton Univesity, Princeton. Di Maggio, P., Hargittai, E., Celeste, C., & Shafer, S. (2003, Fall). From unequal access to differentiated use: A literature review and agenda for research on digital inequality (Working Paper #29, Princeton University Center for Arts and Cultural Policy Studies). Retrieved July 10, 2007, from http://www. princeton.edu/~artspol/workpap29.html Di Maggio, P., Hargittai, E., Russell, W., & Robinson, J. (2001). Social implications of the Internet. Annual Review of Sociology, 27, 307-336. Dwivedi, Y., & Choudrie, J. (2003, June 24). Investigating the impact of broadband upon the users online habits and the usage of Internet services. Paper presented at the ITS Asia-Australasia Regional Conference, Perth, Australia. Fox, S. (2005). Digital divisions. There are clear differences among those with broadband connections, and no connections at all to the Internet. Paper from the Pew Internet & American Life Project, Washington, D.C.
Precisions about the Broadband Divide in Chile
Godoy, S. (2001). Tiempo, medios de comunicación y Sociedad de la Información, Cuadernos de Información, Nº14, pp. 121-131
cl, Santiago de Chile. Retrieved January 18, 2007, from http://www.mouse.cl/detail.asp?story=2007/ 01/18/16/50/28.
Godoy, S. & Herrera, M. S. (2004). Qué ocurre cuando se usa (y no se usa) Internet: resultadosdel World Internet Project-Chile. Cuadernos de Información, Vol.1, Nº16-17, 71-84.
Nie, N., & Hillygus, D. (2002, Summer). The impact of internet use on sociability: Time-diary ¿QGLQJVIT and Society, 1(1), 1-20.
Godoy, S. (Ed.), Herrera, M.S., Sepulveda, M., Lever, G., Myrick, A. (2006): El impacto de las tecnologías de la información en las empresas chilenas respecto a España y Estados Unidos. Resultados de la primera encuesta BIT-Chile , Report for Fondecyt project No. 1050769. 6DQWLDJRGH&KLOH 3RQWL¿FLD8QLYHUVLGDG&DWyOLFD de Chile, Instituto de Estudios Mediales, Instituto de Sociología y Escuela de Ingenería; Centro de Estudios de la Economía Digital, Cámara de Comercio de Santiago. Herrera, M.S. (2005). Dropouts and new users in the Internet community. Paper presented at the WIP/BIT CONFERENCE 2005, Santiago de Chile.
Norris, P. (2001). Digital divide: Civil engagement, information poverty and the Internet in democratic societies. New York: Cambridge University Press. OECD. (2001). Understanding the digital divide. Paris: OECD Publications. PNUD. (2006). Desarrollo humano en Chile. Las nuevas tecnologías: ¿un salto al futuro? Santiago de Chile: Programa de Naciones Unidas para el Desarrollo. Raban, Y. (2004). ICT uptake and usage: Panel data analysis. Retrieved July 10, 2007, from http://www.eurescom.de/e-living/deliverables/eliv-D11.1-Final.pdf
Herrera, M.S. (2006). ¿Está Disminuyendo la Brecha Digital en Chile? Economía & Administración, 1R GLFLHPEUHHQHUR , 30-37.
Rice, R., & Katz, J. (2003). Comparing Internet and mobile phone usage: Digital divides of usage, adoption, and dropouts. Telecommunications Policy, 27, 597-623.
Husing, T., & Selhofer, H. (2004, Spring/SumPHU ',',; $GLJLWDOGLYLGHLQGH[IRUPHDVXUing inequality in IT diffusion. IT & Society, 1(7), 21-38.
Robinson, J., Di Maggio, P., & Hargittai, E. (2003, Summer). New social survey perspectives on the digital divide. IT & Society, 1(5), 1-22.
Keniston, K. (2004). Introduction. In K. Keniston & D. Kumar (Eds.), IT experiences in India. Bridging the digital divide. Delhi, India: Sabe Publishers.
Robinson, J., Kestnbaum, M., Neustadtl, A., & Alvarez, A. (2000). Mass media use and social life among Internet users. Social Science Computer Review, 18(4), 490-500.
La Tercera. (2005). Nuestra Angosta Banda Ancha, La Tercera online/Mouse.cl, Santiago. Available on http://www.mouse.cl/2005/guiaweb/01/26/index.asp
United Nations. (2003). Millennium indicators database. Retrieved July 10, 2007, from http://mdgs. un.org/unsd/mdg/Home.aspx
Lenhart, A., Horrigan, J., Rainie, L., Allen, K., Boyce, A., & Madden, M. et al. (2003). The evershifting Internet population. A new look at Internet access and the digital divide. Washington: The Pew Internet & American Life Project. Mouse.cl. (2007). Conexiones de banda ancha superaron el millón en Chile, La Tercera/Mouse.
WIP-Chile. (2006). Monitoreando el futuro digital: resultados encuesta WIP-Chile 2003, 2004 y 2006. Proyecto Fondecyt 1050769. Santiago de &KLOH 3RQWL¿FLD 8QLYHUVLGDG &DWyOLFD GH &KLOH Instituto de Estudios Mediales, Instituto de Sociología y Escuela de Ingenería; Centro de Estudios de la Economía Digital, Cámara de Comercio de Santiago, Santiago. http://www.wipchile.cl
443
Precisions about the Broadband Divide in Chile
KEY TERMS Digital Divide: The digital divide is the gap between those with regular, effective access to digital technologies and those without. The term GLJLWDOGLYLGHUHIHUVWRWKRVHZKRFDQEHQH¿WIURP LWDQGWKRVHZKRGRQ¶W 6RXUFH :LNLSHGLD WIP-Chile Project: The Chilean chapter of the World Internet Project (WIP), an internationally-comparable, longitudinal academic study led by the USC Center for the Digital Future in Los Angeles, California. By mid 2006, 24 countries participated in WIP. Broadband: Broadband is a relative term, understood according to its context. The wider the bandwidth, the more information can be carried (source: Wikipedia). According to the Ministry RI 7HOHFRPPXQLFDWLRQV LQ &KLOH LQ WKH ¿HOG RI data communications broadband is considered to be those dedicated, always-on connections to the Internet (in contrast to modem, dial-up connections) of at least 128 kilobits per second (kbps).
Home Usage: Usage of the Internet at home by individuals, in contrast to usage at work, school, and other places. This distinction arises from the fact that, by allowing greater privacy and a better HQWDQJOHPHQWWRSHRSOH¶VHYHU\GD\OLIHWKHKRXVHhold allows a more intensive and complete usage of the Internet to the individual.
ENDNOTES 1
2
3
Internet in Chile: Refers to the stage of development of the Internet in Chile in comparison with other countries, including the deployment of broadband. Among other aspects, it includes penetration, usage, and adoption among individuals, industrial sectors, and organizations, as well as the availability and cost of the main digital technologies. Internet/broadband usage: Refers to the patterns of usage of the Internet in general and of broadband connections in particular by individuals (in contrast to organizations or industrial sectors) in Chile and other countries. Usage refers to time dedicated to the Web, the repertoire of activities performed online, location and places of usage, and their relationships to other aspects of the everyday life. Broadband/Internet adoption: Refers to the process of integration of the Internet in general and broadband in particular by individuals (in contrast to organizations or industrial sectors) to their everyday lives in Chile and other countries.
444
4
5
6
“WIP/BIT Chile. Uso y Aprovechamiento de Internet y Tecnologías de Información en Personas, Empresas y la Economía Chilena respecto al Mundo,” Sergio Godoy (director), Soledad Herrera, Marcos Sepúlveda. Fondecyt project Nº1050769. The World Internet Project was started in 1999 by reputed academic centers in the U.S.A., Singapore, and Italy, and added more than 20 countries afterwards. For more details of WIP, see the USC Center for the Digital Future at www.digitalcenter.org. See http://www.wipchile.cl Although WIP-Chile surveyed more cities than Santiago in 2003 and 2004, in 2006 we choose not to cover these extra cities since WKHGLIIHUHQFHVZHUHQRWVLJQL¿FDQW Throughout this chapter, we will use the advertising industry terminology of income level. The higher and upper middle income brackets are thus ABC1-C2 (13.5% of population, average monthly income per household above U.S. $3,000) middle is C3 (28% of population, average income of U.S. $1,275), lower middle is D (45% of population, U.S. $600), and poor is E (13.5% of population, income of U.S.$160). Chilean GNP per capita at purchasing power parity was U.S. $10,874 in 2004. In the latest WIP Project meeting at the Chinese Academy of Social Sciences in Beijing LQ-XO\SDUWQHUVGLVFXVVHGWKHGLI¿FXOWLHVRIDJUHHLQJRQDFRQVLVWHQWGH¿QLWLRQRI EURDGEDQG ZRUOGZLGH $OEHLW QR ³RI¿FLDO´ GH¿QLWLRQZDVDJUHHGXSRQWKH:,3&KLOH team kept considering “broadband” the dedi-
Precisions about the Broadband Divide in Chile
7
8
cated, nonmodem connections. According to the USC Center for the Digital Future, bit-per-second rates of Internet connections ZHUHOHVVVLJQL¿FDQWWKDQWKHLU³DOZD\VRQ´ characteristics in terms of Web usage in the 86$)RUPRUHGHWDLOVDERXWWKH&HQWHU¶V ¿QGLQJVVHHhttp://www.digitalcenter.org. Source: Sergio Espejo, Minister of Telecommunications, “Politica de Comunicaciones,” presentation at the seminar organized by the Public Works, Transport, and Telecommunications Commission of the Chamber of Deputies, Santiago de Chile, September 4, 2006. ICATI is the acronym in Spanish for a counWU\¶V,7$EVRUELQJ&DSDFLW\,QGH[FDOFXODWHG each year by the CCS. It considers infrastructure and costs of telecommunications, PC density, Web host density, and Internet user density in each nation (CCS, 2006).
9
10
Although not published yet, a tentative statistical analysis of data from the United States, Chile, Japan, Spain, Hungary, the UK, and South Korea discussed at the 2004 annual meeting of WIP partners in Tokyo, Japan, suggested income was the most relevant variable explaining Web access. Income is related to education. More recent data also indicates the relevance of age. Yet other variables, such as gender, varied in importance from country to country and was not always relevant. A previous study conducted in Chile almost D GHFDGH DJR VHH *RGR\ GLG ¿QG interesting relationships between the uses of time and communications technologies, yet Internet penetration was too low to be VLJQL¿FDQWIRUWKLVGLVFXVVLRQ
445
Section III
Small and Medium-Sized Enterprises (SMEs)
Division I
Asia
448
Chapter XXVIII
A Survey on the Adoption and Usage of Broadband Internet Roya Gholami Aston Business School, UK John Lim National University of Singapore, Singapore Sang-Yong Tom Lee Hanyang University, Korea
ABSTRACT ,QVSLWHRIWKHLQFUHDVLQJVLJQL¿FDQFHRIEURDGEDQG,QWHUQHWWKHUHDUHQRWPDQ\UHVHDUFKSDSHUVH[SOLFitly addressing issues pertaining to its adoption and postadoption. Previous research on broadband has mainly focused on the supply side aspect at the national level, ignoring the importance of the demand side which may involve looking more deeply into the use, as well as factors impacting organizational DQGLQGLYLGXDOXSWDNH,QDQDWWHPSWWR¿OOWKLVJDSWKHFXUUHQWVWXG\HPSLULFDOO\YHUL¿HVDQLQWHJUDWHG theoretical model comprising the theory of planned behavior and the IS continuance model to examine IDFWRUVLQÀXHQFLQJEURDGEDQG,QWHUQHWDGRSWLRQDQGSRVWDGRSWLRQEHKDYLRURIVRPHRUJDQL]Dtions in Singapore. Overall, strong support for the integrated model has been manifested by our results, SURYLGLQJLQVLJKWLQWRLQÀXHQWLDOIDFWRUV$WWKHDGRSWLRQVWDJHSHUFHLYHGEHKDYLRUDOFRQWUROKDVWKH JUHDWHVWLPSDFWRQEHKDYLRUDOLQWHQWLRQ2XU¿QGLQJVDOVRVXJJHVWWKDWDVFRPSDUHGWRDWWLWXGHVXEMHFWLYHQRUPVDQGSHUFHLYHGEHKDYLRUDOFRQWUROPRUHVLJQL¿FDQWO\DIIHFWWKHEURDGEDQG,QWHUQHWDGRSWLRQ decision. At the postadoption stage, intention is no longer the only determinant of broadband Internet FRQWLQXDQFH UDWKHULQLWLDOXVDJHZDVIRXQGWRVLJQL¿FDQWO\DIIHFWEURDGEDQG,QWHUQHWFRQWLQXDQFH
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
A Survey on Broadband Internet
INTRODUCTION There has been a recent surge in interest in highspeed, always-on broadband Internet connectivity both in industry and academia (Choudrie & Lee, 2004; Choudrie, Papazafeiropoulou, & Lee, 2003; Dutton, Gillett, McKnight, & Peltu, 2004; Middleton, 1999; Oh, Ahn, & Kim, 2003; Sawyer, Allen, & Lee, 2003; 2002a; 2002b; 2003), as narrowband technologies cannot meet the high bandwidth requirements for emerging applications that combine voice, video, and data. It is widely recognized that the next phase in the evolution of the Internet is broadband, including wireless and mobile Internet which will enable applications and services in relation to economy, education, entertainment, health care, government, patterns of the computer use, social and work interactions, and consumer behaviors (Sawyer et al., 2003). Economists and policy makers around the world believe widespread broadband deployment can broaden the ICT revolution and improve national productivity resulting in robust economic growth. They consider broadband connectivity as an important means to increase the international competitiveness, development of e-commerce and information economy. Not surprisingly, many industrialized countries are implementing a wide range of national broadband strategies trying to accelerate the deployment of broadband networks (Sawyer et al., 2003). Singapore, South Korea, and Malaysia are among the highly motivated nations to deploy broadband networks to create competitive advantage in attracting foreign direct investments by multinational corporations. One of the key objectives of IS research is to measure the value of ICT investment and to understand the determinants of this value. Understanding ICT adoption and usage is important because the H[SHFWHG EHQH¿WV RI ,&7 HJ LPSURYHPHQWV LQ HI¿FLHQF\HIIHFWLYHQHVVRUSURGXFWLYLW\ FDQQRW be realized if individual users do not accept and XVHWKHVHV\VWHPVIRUWDVNSHUIRUPDQFHLQWKH¿UVW place. It is evident that true business value from any ICT can only be derived through appropriate use by its target user group (Agarwal, 2000). ,QIRUPDWLRQV\VWHPV ,6 DGRSWLRQLVMXVWWKH¿UVW step toward overall IS success. An IS implementa-
tion can truly be considered as “a success” when DVLJQL¿FDQWQXPEHURIXVHUVKDYHPRYHGEH\RQG initial adoption and used the IS on a continued basis (Bhattacherjee, 2001; Davis & Venkatesh, 2004; Limayem & Hirt, 2003; Lyytinen & Hirschheim, 1987). However, previous research on broadband has mainly focused on the supply side such as pricing, promoting infrastructure, building, and establishing right policies, ignoring the importance of the demand side issues such as looking more deeply into the use and factors impacting organizational and individual uptake (U.S. Department of Commerce, 2002). This chapter tries to identify the key drivers of broadband adoption and continuance (after initial adoption) in Singapore. The Singaporean government is strongly committed to making Singapore one of the most connected cities in the world and offers aggressive plans to boost broadband penetration. As a developed city, Singapore provides a suitable setting for the purpose of the FXUUHQWUHVHDUFK7KH¿QGLQJVRIWKHFXUUHQWUHsearch will have important policy implications for governments, ISPs, and other supply side institutions responsible for enhancing broadband and provide lessons that may be applicable to similar developments elsewhere. The pioneering experience of Singapore should also be useful in guiding comparable efforts.
THEORETICAL BACKGROUND There exists an interesting phenomenon that while ICT is increasingly used in organizations, unsuccessful adoptions of ICT are not uncommon. The importance in understanding the factors determining success or failure of ICT adoption cannot be overemphasized. The research stream examining the adoption and usage of new ICT has evolved into RQHRIWKHULFKHVWDQGPRVWPDWXUHLQWKH,6¿HOG for over two decades (Venkatesh, Morris, Davis, & Davis, 2003). In order to study determinants of ICT adoption and usage, intention-based theories such as the theory of reasoned action (TRA), the theory of planned behavior (TPB), and the technology acceptance model (TAM) seem to be
449
A Survey on Broadband Internet
Figure 1. Theory of reasoned action (TRA)
appropriate tools which have been widely used in the literature.
(Fishbein & Azjen, 1975). Any other factors that LQÀXHQFHEHKDYLRULQWHQWLRQGRVRRQO\LQGLUHFWO\ E\LQÀXHQFLQJ$61RUWKHLUUHODWLYHZHLJKWV
Theory of Reasoned Action (TRA) Technology Acceptance Model (TAM) (PEHGGHGLQWKH¿HOGRIVRFLDOSV\FKRORJ\ WKH theory of reasoned action (TRA) was widely used LQSUHGLFWLQJEHKDYLRUE\DQLQGLYLGXDO¶VDWWLWXGH towards behavior (Fishbein & Ajzen, 1975) and based on which TAM and TPB models have been developed. As shown in Figure 1, TRA argues that behavioral intention (BI) is jointly determined by attitude toward behavior (A) and subjective norm 61 %, LV D PHDVXUH RI WKH VWUHQJWK RI RQH¶V LQWHQWLRQWRSHUIRUPDVSHFL¿HGEHKDYLRU 'DYLV %DJR]]L :DUVKDZ ZKHUHDV$LVGH¿QHG DVDQLQGLYLGXDO¶VSRVLWLYHRUQHJDWLYHIHHOLQJDERXW performing the target behavior (Davis et al., 1989) DQG61UHIHUVWR³WKHSHUVRQ¶VSHUFHSWLRQWKDWPRVW people who are important to him think he should or should not perform the behavior in question”
TAM (Davis, 1989; Davis et al., 1989) was theoretiFDOO\GHULYHGIURP75$EXW7$0LVVSHFL¿FDOO\ meant to explain computer usage behavior (Davis HWDO 7$0LV³DQDGDSWDWLRQRI75$VSHFL¿FDOO\ tailored for modeling user acceptance of information systems” (Davis et al.). TAM focuses on only attitude by ignoring subjective norms intentionally due to SN uncertain theoretical and psychometric status (Davis et al.). This model uses two beliefs of the potential adopter, perceived usefulness (PU) and perceived ease of use (PEOU) of technology as the main determinants of the attitudes toward a new technology. PU is “the degree to which a person believes that using a particular system would enhance his or her job performance” (Davis, 1989), while
Figure 2. Technology acceptance model (TAM) Perceived usefulness (PU)
External variables
Attitude towards using (A) Perceived ease of use (PEOU)
450
Behavioral intention to use (BI)
Actual system use
A Survey on Broadband Internet
PEOU is “the degree to which a person believes that using a particular system would be free of effort” (Davis, 1989). These two beliefs create a favorable disposition or intention toward using the IT that consequently affects its self-reported use (Davis et al., 1989). Comparing Figures 1 and 2, perceived usefulness (PU) and perceived ease of use (PEOU), are antecedents of A while PU and $MRLQWO\GHWHUPLQHXVHU¶V%,ZKHUHDV3(28DOVR KDVDQHIIHFWRQ8([WHUQDOYDULDEOHVLQÀXHQFH XVHULQWHQWLRQVRQO\LQGLUHFWO\E\LQÀXHQFLQJ38 and PEOU or their relative weights. Studies that have investigated self-reported system usage and intended use have found that 38 SOD\V D VLJQL¿FDQW UROH LQ GHWHUPLQLQJ VXFK downstream effects. However, most of the studies, beginning with Davis himself, have not found a direct linkage between PEOU and ICT adoption. Indeed, Davis suggested that, “ease of use operates through usefulness.” Not surprisingly, some research has questioned the overall importance of PEOU in ICT adoption (e.g., Keil et al., 1995). The role of PEOU in TAM, however, remains controversial in that some studies show that PEOU does directly affect either self-reported use or intended ICT use. However, TAM is still somehow incomplete. On one hand, its explanatory power is limited. Typically, TAM can account for 40% of variance in user intentions. Compared with other models, for example, TPB, TAM performs just slightly better (Taylor & Todd, 1995). On the other hand, the relationship within it is inconsistent, which
means maybe TAM is unable to succeed across different contexts. More and more researchers begin questioning the validity of TAM across different contexts.
Theory of Planned Behavior (TPB) To expand the scope of the TRA to become applicable for behaviors or situations in which there is low or no volitional control, Ajzen (1985, 1991) developed the theory of planned behavior. TPB PRGL¿HV75$E\LQFOXGLQJWKHYDULDEOHSHUFHLYHG EHKDYLRUDOFRQWURO 3%& 7KLVPHDVXUHVDSHUVRQ¶V perception of control over performing a given behavior. PBC was theorized to directly predict and explain both intention and behavior and was supposed to explain more variance than the TRA IRUEHKDYLRUVQRWHQWLUHO\XQGHUDSHUVRQ¶VYROLWLRQDO control (Ajzen, 1985). 73%DVVXPHVWKDWDSHUVRQ¶VLQWHQWLRQWRDGRSW ICT is determined by three factors: attitude (A), subjective norms (SN), and perceived behavioral control (PBC). $WWLWXGHGHVFULEHVDQLQGLYLGXDO¶V positive or negative perception towards ICT; subMHFWLYHQRUPVGHVFULEHVWKHVRFLDOLQÀXHQFHWKDW PD\ DIIHFW D SHUVRQ¶V LQWHQWLRQ WR XVH ,&7 DQG perceived behavioral control describes the beliefs about having the necessary resources and opportunities to adopt ICT. As a general rule, the more favorable A and SN with respect to a behavior, and the greater the PBC, the stronger should be an LQGLYLGXDO¶VLQWHQWWRSHUIRUPWKHEHKDYLRUXQGHU consideration (Ajzen, 1991). Intention to adopt ICT,
Figure 3. Theory of planned behavior (Ajzen, 1991) Behavioral Beliefs & Outcome Evaluation
Attitude toward the Behavior
Normative Beliefs and Motivation to Comply
Subjective Norms
Control Beliefs and Perceived Facilitation
Behavioral Intention
Usage Behavior
Perceived Behavioral Control
451
A Survey on Broadband Internet
in return, is expected to affect the actual adoption of ICT (Figure 3). Although intention-based theories have been YHUL¿HG PRVWO\ ZLWKLQ WKH FRQWH[W RI LQGLYLGXDO decision-making, it has been shown that these models could also be applied to strategic decisionmaking at organizational level as well (Harrison, Mykytyn, & Riemenschneider, 1997; Lim, Gan, & Chang, 2002; Riemenschneider, Harrison, Mykytyn, & Pete, 2003). This is because individuals who typically make the decisions such as technology adoption and continuance are the top managers. Accordingly, it is important to investigate the decisions made by those in a position to LQÀXHQFH WHFKQRORJ\ DGRSWLRQ DQG XVDJH IRU DQ entire organization.
IS Continuance Model (Postadoption Stage) Simply studying the decision to adopt a technology is not enough, since it is followed by a process of deeper evaluation through use that may or may not result in continuance. The gap between expectations before adoption and evaluation after experiHQFHLQÀXHQFHVWKLVGHFLVLRQ5HFHQWVWXGLHVVKRZ that people continue using technologies only when they add value to their lives (Carroll, Howard, Peck, & Murphy, 2002). Innovation diffusion theory, in LWV¿YHVWDJHDGRSWLRQGHFLVLRQSURFHVVVXJJHVWV that adopters reevaluate their earlier acceptance GHFLVLRQGXULQJD¿QDO³FRQ¿UPDWLRQ´VWDJHDQG decide whether to continue or discontinue using an innovation (Rogers, 1995). However, these studies
view continuance as an extension of acceptance behaviors (Davis et al., 1989; Karahanna, Detmar, Straub, & Chervany, 1999), and are therefore unable to explain why some users discontinue IS use after accepting it initially (Bhattacherjee, 2001). Bhattacherjee (2001) has shown that circumVWDQFHVRI,6FRQWLQXDQFHGLIIHUVLJQL¿FDQWO\IURP those of initial adoption behavior. Based on the exSHFWDWLRQFRQ¿UPDWLRQWKHRU\ (&7 Oliver, 1980) widely used in the consumer behavior literature to study consumer satisfaction and post-purchase behavior (Churchill, & Surprenant, 1982; Spreng, MacKenzie, & Olshavsky, 1996; Tse & Wilton, %KDWWDFKHUMHH¶V,6FRQWLQXDQFHPRGHODWWHPSWVWRH[SODLQXVHU¶VLQWHQWLRQWRFRQWLQXH using an IS. In this model, intention is positively related to both satisfaction and perceived usefulness. Satisfaction and perceived usefulness are in turn SRVLWLYHO\UHODWHGWRWKHGHJUHHZLWKZKLFKWKHXVHU¶V H[SHFWDWLRQVDERXWWKH,6DUHFRQ¿UPHG )LJXUH 2). The rationale for a direct effect of perceived usefulness on satisfaction is that the perception of usefulness at the postadoption stage can be viewed as perceived performance, which has been VKRZQWREHDVLJQL¿FDQWDQWHFHGHQWRIVDWLVIDFWLRQ in some studies drawing on the expectation-disFRQ¿UPDWLRQWKHRU\PRGHO $QGHUVRQ)RUQHOO Lehmann, 1994; Churchill & Surprenant, 1982; Tse & Wilton, 1988). IS researchers have continued to examine user satisfaction in part because it has been widely adopted as an important determinant of IS success (DeLone & McLean, 1992, 2003; Rai, Lang, & Welker, 2002; Wixom & Todd, 2005; Zviran &
Figure 4. Research model: post-acceptance model of IS continuance Perceived Usefulness
Confirmation
452
Satisfaction
Continuance Intention
Continuance
A Survey on Broadband Internet
Erlich, 2003). Bhattacherjee (2001) was one of the YHU\¿UVWUHVHDUFKHUVWRSURSRVHDQ,6FRQWLQXDQFH PRGHOLQOLQHZLWKWKHH[SHFWDWLRQFRQ¿UPDWLRQ theory. In the marketing literature, satisfaction is considered key to building and retaining a loyal base of long-term consumers. Recently, the proliferation of electronic commerce has further provoked IS UHVHDUFKHUV¶LQWHUHVWLQWKHVWXG\RIVDWLVIDFWLRQLQ the online environment (Devaraj, Fan, & Kohli, 2002; McKinney, Yoon, & Zahedi, 2002; Shim, Shin, & Nottingham, 2002). The D&M IS success model (1992) also has served as a dominant framework for studying user satisfaction. The D&M IS success model suggested that information quality and system quality are two key factors determining user satisfaction. Among diverse theoretical frameworks, expectation con¿UPDWLRQWKHRU\KDVEHHQUHFHLYLQJDJUHDWGHDO of attention in recent IS research (Bhattacherjee, 2001; McKinney et al., 2002; Susarla, Barua, & Whinston, 2003). These studies provided more insights to user psychology and explained user satisfaction formation processes. The research questions addressed in this study are: (1) what are the main motivations underlying broadband Internet initial adoption in organiza-
tions, and (2) what are the factors affecting broadband Internet continuance intention after its initial acceptance. As mentioned earlier the intention and LQÀXHQFHVWRXVHDWHFKQRORJ\FDQEHGLIIHUHQWIRU the near-term and the long-term. Thus, the meaning of the adoption and postadoption may differ. Adoption refers to the initial encounter with a technology by the user, while postadoption refers to the situation in which an individual has used the technology for a while. Therefore, the dependent variables are broadband initial usage in the adoption stage and continuance in the postadoption stage. Understanding continuance in contrast to only initial use is a major feature of this chapter. Following Limayem, Cheung, and Chan (2003), we test an integrated model that explains IS usage at both adoption and postadoption stages. TPB is integrated with the IS continuance model in order to better explain the nature of IS continuance. In a sense, we use the TPB model to explain broadband initial usage in the adoption stage, and the IS continuance model to explain broadband continuance in the postadoption stage (Figure 5). An important assumption of the model is that it applies to individual decisions. Therefore, here we test the theory regarding decisions made by individual
)LJXUH5HVHDUFKPRGHO DGDSWHGIURP/LPD\HPHWDO Post-Adoption Stage Perceived Usefulness
Confirmation
Satisfaction
Continuance Intention
Continuance
Adoption Stage Attitude
Subjective Norms
Behavioral Intention
Initial Usage
Perceived Behavioral Control
453
A Survey on Broadband Internet
PDQDJHUVZKLFKUHÀHFWVWKHRUJDQL]DWLRQDOOHYHO decision-making. At the adoption stage attitude, 61DQG3%&LQÀXHQFHWKHLQLWLDOXVHRIEURDGEDQG Internet. In our context the hypotheses according to TPB are: H1: Adoption intention of broadband Internet is a positive function of attitude towards adopting broadband Internet, SN regarding adoption of broadband Internet, and PBC over adopting broadband Internet H2: Broadband Internet initial usage is determined by its adoption intention H3: PBC is positively associated with broadband Internet initial usage $W WKH SRVWDGRSWLRQ VWDJH FRQ¿UPDWLRQ DQG perceived usefulness are considered as the main GHWHUPLQDQWV RI VDWLVIDFWLRQ &RQ¿UPDWLRQ LV positively associated with perceived usefulness. Both satisfaction and perceived usefulness are essential for the formation of continuance intention. Continuance is in turn determined by broadband continuance intention. The following hypotheses are put forth. H4 &RQ¿UPDWLRQDQGSHUFHLYHGXVHIXOQHVVDUH positively associated with satisfaction H5 &RQ¿UPDWLRQ LV SRVLWLYHO\ DVVRFLDWHG ZLWK perceived usefulness H6: Both satisfaction and perceived usefulness positively affect broadband Internet continuance intention H7: Broadband Internet continuance is determined by broadband Internet continuance intention and its initial usage
RESEARCH METHODOLOGY The survey method was used to test the model. Salkind (2000) suggests that survey research is useful in examining the relationships between sociological variables and taps well into constructs like attitudes, beliefs, and opinions. As suggested E\&KXUFKLOO DOLWHUDWXUHVHDUFKZDV¿UVW carried out within the domain of the constructs
454
to generate sample items. Next, we conducted a pretest with 17 IS faculties and graduate students for purposes of addressing validity and reliability DVSHFWV7KH¿QDOVHWRIPHDVXUHVXVHGLQRXUPDLQ survey is presented in Table 1. This survey utilizes a database from the Singapore 1000 and SME 500 (Dp Information Network, 2003) containing 1500 companies located in Singapore. The key personnel who answered the questionnaire were executives such as MIS manDJHUV GHSHQGLQJRQWKHRUJDQL]DWLRQ¶VVWUXFWXUH because they were likely to have cognizance of their organizations as key decision makers in broadband Internet adoption. Two hundred and seven usable responses were collected, yielding a response rate of 14% (207/1500). This response rate is considered reasonable because the survey was unsolicited and it involved senior management.
DATA ANALYSIS AND RESULTS The Partial Least Squares (PLS) test, an implementation of structural equation modeling (SEM), ZDVXVHGWRGHWHFWVLJQL¿FDQWUHODWLRQVDPRQJWKH variables and to test the validity of the measurement model. The most widely used SEM-based techniques are partial least square (PLS) and LISREL. In this study, PLS was preferred to LISREL for the analysis for several reasons, mainly related to data distribution and sample size.
Testing the Measurement Model The strength of the constructs used for testing the structural model is assessed by examining the relationships between the constructs and the indicators. Such examinations may include the test of internal consistency reliability of the scales and the convergent and discriminant validity of conVWUXFWVPRGHOHGZLWKUHÀHFWLYHLQGLFDWRUV &KZHORV Benbasat, & Dexter, 2001). A common method used for testing internal consistency reliability of LQVWUXPHQWVLV&URQEDFK¶VDOSKDDVVHVVPHQW7DEOH VKRZVWKHFDOFXODWHG&URQEDFK¶VDOSKDVIRUDOO FRQVWUXFWVZKLFKDUHDOODERYH1XQQDOO\¶V criterion of 0.700.
A Survey on Broadband Internet
Table 1. List of measures Measures
Sources
Constructs Attitude ATT1
Effective/ineffective
ATT2
Good/bad
ATT3
Foolish/wise
ATT4
Positive/negative
ATT5
Helpful/harmful
Riemenschneider et al. (2003)
Perceived Behavioral Control PBC1
'LI¿FXOWHDV\
PBC2
Complex/simple
PBC3
Out of /under control
Harrison et al. (1997)
Subjective Norms SN1
Most people who are important approve
SN2
0RVWSHRSOHZKRDUHLPSRUWDQWWKLQNWKH¿UPVKRXOG
Harrison et al. (1997)
Behavioral Intention BI1
Intend to use broadband Internet.
BI2
Certain are plans to use broadband Internet.
BI3
Commitment to use broadband Internet.
Riemenschneider et al. (2003)
Perceived Usefulness PU1
Accomplish tasks more quickly
PU2
Improve job performance of employees
PU3
Improve quality of work
PU4
Make it easier for employees to do jobs
PU5
$OORZ¿UPWRDFFRPSOLVKPRUHZRUN
PU6
(QKDQFH¿UP¶VHIIHFWLYHQHVV
PU7
*LYH¿UPJUHDWHUFRQWURORYHUZRUN
PU8
,QFUHDVH¿UP¶VSURGXFWLYLW\
PU9
%HXVHIXOWRRXU¿UP
Riemenschneider et al. (2003)
&RQ¿UPDWLRQ CRM1
Experience with using broadband Internet was better than what was expected.
CRM2
2YHUDOOPRVWRIWKHH[SHFWDWLRQVIURPXVLQJEURDGEDQG,QWHUQHWZHUHFRQ¿UPHG
7KHEHQH¿WSURYLGHGE\EURDGEDQG,QWHUQHWZDVEHWWHUWKDQZKDWZDVH[SHFWHG
Bhattacherjee (2001)
CRM3 Satisfaction How do you feel about your overall experience of broadband Internet Use: SAT1
9HU\GLVVDWLV¿HG9HU\VDWLV¿HG
SAT2
(Very displeased/Very pleased)
SAT3
(Very frustrated/Very contented)
SAT4
(Absolutely terrible/Absolutely delighted)
Bhattacherjee (2001)
455
A Survey on Broadband Internet
Table 1. List of measures (cont.) Continuance Intention CI1
,IZHFRXOG¿UPZRXOGOLNHWRFRQWLQXHXVHRIEURDGEDQG,QWHUQHW
CI2
$OOWKLQJVFRQVLGHUHG¿UPH[SHFWVWRFRQWLQXHXVLQJEURDGEDQG,QWHUQHWGXULQJ
Limayem et al. (2003)
the next six months.
Bhattacherjee (2001)
CI3
$OOWKLQJVFRQVLGHUHGLWLVOLNHO\WKDW¿UPZLOOFRQWLQXHWRXVHEURDGEDQG,QWHUQHW during the next six months.
Initial Usage/Continuous Usage
IU1/CU1
Very intensively (many hours per day)
Cheung, Chang, & Lai (2000)
IU2/CU2
Very frequently (many times per day)
Limayem et al. (2003)
IU3/CU3
For a great variety of applications (reports, projects, decision making, etc.)
IU4/CU4
Overall a lot
Fornell (1982) suggested three tests for convergent validity, including item reliability, composite reliability, and average variance extracted. Item reliability assesses every indicator loading and WKHORDGLQJ¶VVWDWLVWLFDOVLJQL¿FDQFH7KHUHVXOWVRI the three tests for variables are also presented in Table 2. All the variables have loadings above 0.550 )DON 0LOOHU 1XQQDOO\ VLJQL¿FDQW DW Į WKH FRPSRVLWH UHOLDELOLW\ KLJKHU WKDQ 0.700 (Nunnally, 1978), and the average variance extracted higher than 0.500 (Fornell, 1982). Such results give evidence for convergent validity. 7DEOHSURYLGHVHYLGHQFHIRUWKHFRQVWUXFWV¶ discriminant validity as the square root of the average variance extracted for each construct exceeds the correlation between this construct with any other construct (Fornell, 1982). Constructs have adequate discriminant validity. With assurance of good psychometric properties in the measurement model, we proceed to test the structural model to DVVHVVLWVH[SODQDWRU\SRZHUDQGWKHVLJQL¿FDQFH of its paths.
Testing the Structural Model Since PLS does not generate an overall goodnessRI¿W LQGH[ RQH SULPDULO\ DVVHVVHV YDOLGLW\ E\ examining R 2 and the structural paths (Chwelos et
456
al., 2001). R 2 value may be interpreted in a similar way as it is done in traditional regression analysis. 3DWKFRHI¿FLHQWVDUHHTXLYDOHQWWRWKHVWDQGDUGL]HG beta weights in the multiple regression model. The explanatory power of the structural model could be assessed based on the amount of variance in the endogenous constructs for which the model could account. Bootstrapping techniques were used to obtain the corresponding t-values in order WRDVVHVVWKHVLJQL¿FDQFHRIWKHSDWKFRHI¿FLHQWV Each hypothesis corresponded to a path in the structural model. Figure 6 present the result of WKHVWUXFWXUDOPRGHOVL]HDQGVLJQL¿FDQFHRIWKH SDWKFRHI¿FLHQWVDVZHOODVWKHH[SODQDWRU\SRZHU of the independent variables. At the adoption stage, all three determinants of intention (attitude, subjective norm, and perceived EHKDYLRUDOFRQWURO KDYHVLJQL¿FDQWHIIHFWVRQEHKDYLRUDOLQWHQWLRQZLWKSDWKFRHI¿FLHQWVRI 0.45, and 0.26 respectively. The three constructs explain 57% of the variance in behavioral intention. Regarding initial usage, both behavioral intention and PBC are found to be statistically VLJQL¿FDQW DQG H[SODLQV RI WKH YDULDQFH RI LQLWLDOXVDJH$WSRVWDGRSWLRQVWDJHFRQ¿UPDWLRQ DQGSHUFHLYHGXVHIXOQHVVKDYHVLJQL¿FDQWLPSDFWV RQVDWLVIDFWLRQZLWKSDWKFRHI¿FLHQWVRIDQG 0.26 respectively. The two constructs account for
A Survey on Broadband Internet
Table 2. Psychometric properties of the constructs
Construct
ATT
PBC
SN
BI
PU
SAT
CRM
CI
Item
Loading
t-value
ATT1
0.927
48.023
ATT2
0.948
97.003
ATT3
0.934
58.563
ATT4
0.960
101.751
ATT5
0.922
52.589
PBC1
0.927
57.615
PBC2
0.830
62.807
PBC3
0.895
35.393
SN1
0.961
2.389
SN2
0.955
2.383
BI1
0.965
96.374
BI2
0.979
141.162
BI3
0.973
111.663
PU1
0.816
30.651
PU2
0.882
50.128
PU3
0.814
24.555
PU4
0.870
42.801
PU5
0.914
69.430
PU6
0.897
50.523
PU7
0.822
25.729
PU8
0.885
50.245
PU9
0.820
30.765
SAT1
0.914
47.533
SAT2
0.961
134.887
SAT3
0.926
55.405
SAT4
0.925
59.919
CRM1
0.894
36.124
CRM2
0.906
44.761
CRM3
0.908
53.352
CI1
0.986
195.756
CI2
0.992
297.236
CI3
0.981
77.139
Composite
Cronbach’s
Variance
Reliability
Alpha
Extracted
0.973
0.966
0.880
0.915
0.904
0.783
0.917 0.957
0.912
0.981
0.972
0.945
0.961
0.954
0.737
0.963
0.949
0.868
0.929
0.886
0.814
0.990
0.987
0.972
457
A Survey on Broadband Internet
Table 2. Psychometric properties of the constructs (cont.) IU1 IU
CU
0.914
26.15
IU2
0.939
42.38
IU3
0.909
14.54
IU4
0.902
25.23
CU1
0.906
34.13
CU2
0.939
45.38
CU3
0.809
13.11
CU4
0.912
28.13
RIWKHYDULDQFHLQVDWLVIDFWLRQ&RQ¿UPDWLRQ DOVRKDVDVLJQL¿FDQWHIIHFWRQSHUFHLYHGXVHIXOQHVV ȕ W ZLWKYDULDQFHH[SODLQHG Regarding the antecedents of continuance intention, both satisfaction and perceived usefulness DUHVLJQL¿FDQWZLWKSDWKFRHI¿FLHQWVRIDQG 0.35 respectively, accounting for 22% variance of continuance intention. In addition, continuance LQWHQWLRQ KDV D VLJQL¿FDQW HIIHFW RQ FRQWLQXRXV XVDJHZLWKDSDWKFRHI¿FLHQWRI 0RUHRYHULQLWLDOXVDJHKDVDYHU\VLJQL¿FDQW LPSDFWRQFRQWLQXDQFH ȕ W 7KLVPRGHO
0.922
0.940
0.685
0.918
0.925
0.698
could explain 70% of the variance for continuous usage and 18% of initial usage, as indicated by the R 2 values exceeding 10% criterion level, suggested by Falk and Miller (1992) as an indication of substantive explanatory power. Thus, our research model has predictive validity. The hypotheses were HYDOXDWHG DV SHU VL]H VLJQ DQG VLJQL¿FDQFH RI WKHSDWKFRHI¿FLHQWVDQGDOOWKHK\SRWKHVHVZHUH supported. The next section discusses the research ¿QGLQJVDQGWKHLULPSOLFDWLRQV
Table 3. Discriminant validity of constructs Construct
PU
CRM
SAT
CI
ATT
SN
PBC
PU
0.858
CRM
0.516
0.902
SAT
0.505
0.599
0.931
CI
0.446
0.285
0.355
0.986
ATT
0.560
0.416
0.606
0.582
0.938
SN
0.498
0.307
0.414
0.700
0.586
0.958
PBC
0.454
0.388
0.495
0.602
0.620
0.562
0.884
BI
0.445
0.256
0.373
0.832
0.590
0.694
0.617
BI
0.972
Notes: Diagonal numbers are the square root of the average variances extracted for each construct; off-diagonal numbers are the correlations between constructs.
458
A Survey on Broadband Internet
Figure 6 . Structural model Perceived Usefulness 0.516***
0.359*** 0.266*** R2=0.70 0.254***
Confirmation
Satisfaction 0.174**
0.462*** R2=0.41
Continuous Usage
Continuance Intention R2=0.22
Attitude 0.163*
0.70*** 2
R =0.57 0.263** Subjective Norm
0.450**
Behavioral Intention
R2=0.18
0.263*** Perceived Behavioral Control
DISCUSSION AND IMPLICATIONS In order to better understand complex and dynamic aspects of adoption and usage of broadband InterQHWRYHUWLPHWKLVVWXG\HPSLULFDOO\YHUL¿HVWKH integrated model of TPB with the IS continuance model proposed by Limayem et al. (2003) to examLQHIDFWRUVLQÀXHQFLQJEURDGEDQG,QWHUQHWDGRSWLRQ and postadoption behavior of 1500 organizations in 6LQJDSRUH7KH¿QGLQJVSUHVHQWDVWURQJVXSSRUW for the existing theoretical links of TPB and the IS continuance model. As suggested by the model, LQLWLDOXVDJHZDVIRXQGWRVLJQL¿FDQWO\DIIHFWEURDGband Internet continuance. Results also show that 61DQG3%&PRUHVLJQL¿FDQWO\WKDQDWWLWXGHDIIHFW the broadband Internet adoption decision 7KLVLVFRQVLVWHQWZLWKWKH¿QGLQJVRISUHYLous studies indicating strong support for adoption intentions based on SN (perceived social expectations from customers, suppliers, IS group within WKH ¿UP FRPSHWLWRUV WRS PDQDJHPHQW DQG DOO other employees in order to promote productivity). 2UJDQL]DWLRQVPD\EHLQÀXHQFHGE\WKHLUH[WHUQDO trading partners due to the competitive business environment, as studies in organizational settings have found SN to be an important determinant
Initial Usage
0.261**
of adoption intentions (Harrison et al., 1997; Hartwick & Barki, 1994; Lim et al., 2002; Taylor & Todd, 1995). Similar things can be mentioned about perceived control (perceived obstacles or facilitators to adopting technology such as approval of management, support of employees, and ¿QDQFLDODVVHWV PBC towards broadband Internet adoption UHIHUVWRWKHHDVHRUGLI¿FXOW\RIDGRSWLQJEURDGband Internet in the organization. The stronger an organization believes in its ability to adopt the technology the greater will be its adoption intention. Although cost-effectiveness is often the main factor for organizations and resources such as money and implementation time rank high on most organizaWLRQV¶UHVRXUFHDOORFDWLRQOLVWLWPD\EHSRVVLEOHWKDW the obstacles to the adoption of broadband Internet are overshadowed by its potential usefulness to the ¿UPDQGWKHVRFLDOSUHVVXUHWRIROORZRWKHU¿UPV for competitive advantage. Pressure to keep up with the competition and offer higher quality service to FXVWRPHUVPD\RYHUULGHDQ\REVWDFOHIRUWKH¿UP WRDGRSWEURDGEDQG,QWHUQHW7KH¿QGLQJVRIWKLV study have important implications for governments, ISPs, and other supply-side institutions responsible for enhancing broadband use. Government and 459
A Survey on Broadband Internet
industry are all aiming to obtain online products and services. For this, the role of broadband is pertinent. For the industry, the contribution is that providers will be offered a better argument that demonstrates how broadband diffusion should be HQFRXUDJHG7KLVVWXG\LVDQHPSLULFDOYHUL¿FDWLRQRIWKH/LPD\HPHWDO¶V PRGHOEDVHGRQ TPB and the IS continuance model for strategic decision making; results have indicated that this model provides valuable guidance toward a better understanding of broadband Internet adoption and postadoption behavior. This study also provides empirical support for propositions for integrating these two models, which provides important insights into the complex, dynamic nature of the relationship between broadband adoption and postadoption. At the adoption stage, PBC has the strongest impact on behavioral intention; at the postadoption stage intention is no longer the only determinant of broadband Internet continuance. Initial usage becomes an important factor determining the level of broadband Internet continuance. This suggests that once the users adopt the system, they have a higher tendency for continuous use. Therefore, ISPs and other supply-side institutions should design a plan that encourages usage in the adoption stage and should realize that initial usage, to a very large extent, determines broadband Internet continuance. Many studies have found that past usage beKDYLRUDQGH[SHULHQFHDUHVLJQL¿FDQWSUHGLFWRUVRI later usage behavior (Igbaria, 1990, 1993; Igbaria, Parasuraman, & Baroudi, 1996; Kettinger & Grover, 1997; Venkatesh & Davis, 2000; Venkatesh, 6SHLHU 0RUULV 6SHFL¿FDOO\GXULQJWKH ¿UVW IHZ WLPHV ZKHQ DQ LQGLYLGXDO XVHV DQ ,7 feature, the user engages in the cognitive proFHVVLQJ WKDW LQÀXHQFHV SRVWDGRSWLYH LQWHQWLRQV and, hence, postadoptive behavior (Bargh, 1989, 1994; Ouellette & Wood, 1998). In agreement with %KDWWDFKHUMHH¶VVWXG\ RXU¿QGLQJVLPSO\ that while ignoring preadoption user attitude may not severely impact broadband acceptance among new users, ignoring postadoption user satisfaction can have disastrous consequences for user retention. Perceived usefulness is more crucial
460
for acceptance intention, and satisfaction is more dominant for continuance intention. Consequently, supply-side institutions should adopt a two-fold strategy for maximizing their return on investments in customer training. This strategy involves informing new users of the poWHQWLDOEHQH¿WVRIEURDGEDQG,QWHUQHWXVHDQGHGXcating old users on how to use broadband Internet HIIHFWLYHO\VRDVWRPD[LPL]HWKHLUFRQ¿UPDWLRQDQG satisfaction with broadband Internet usage. Users GLVVDWLV¿HGZLWKEURDGEDQG,QWHUQHWXVDJH GXHWR GLVFRQ¿UPDWLRQRIH[SHFWDWLRQV PD\GLVFRQWLQXH its use, albeit having positive opinions of its usefulness (Bhattacherjee, 2001). As a result, to sustain competitive advantage, ISPs need to understand what it takes to encourage customer loyalty. The current empirical study of broadband Internet was used as an illustrative example to demonstrate the QHHGWRDGGUHVVFRQVXPHUV¶H[SHFWDWLRQVLQXVLQJ broadband Internet.
CONCLUSION This study provides empirical evidence of the appliFDELOLW\RI/LPD\HPHWDO¶V PRGHOEDVHGRQ TPB and the IS continuance model in understanding organizations decision making at the strategic level. A possible limitation of this study concerns the level of analysis. As mentioned earlier, although LQWHQWLRQEDVHGWKHRULHVKDYHEHHQYHUL¿HGPRVWO\ within the context of individual decision-making, it has been shown that these models could be applied to strategic decision-making at the organizational level as well (Harrison et al., 1997; Riemenschneider et al., 2003). This is because individuals who typically make decisions on issues such as technology adoption and continuance are the key executives. In contrast, some scholars believe that organizational behavior is not a simple accumulation of individual behaviors (Schermehron, Hunt, & Osborn, 1998). However, our follow up calls revealed that the decision to adopt and use broadband Internet connectivity is usually made or dominated by a single executive (e.g., MIS manager) because of the unique characteristics associated with the
A Survey on Broadband Internet
broadband Internet technologies. Perhaps just as suggested by Afuah and Tucci (2001), unlike adoption decisions for software applications and other complicated IS technologies, organizational decision on adoption and continuance of Internet technologies can be relatively simple. Therefore, special attention should be paid to key executives such as MIS managers, and information about broadband technologies should be disseminated to them to highlight the importance of its usage. In order to increase the variance explained in the research model, future research should examine the relationships among those components, which could also lead to an increase in the explanation power of the research model such as organizational culture or industry characteristics. The saliency of the cultural element is evident in light of increasing globalization. In this respect, it may need to be addressed not only at the organizational level, but also the national level, although interplay between these two levels cannot be ruled out. The predictive validity of the framework might be enhanced and construct validity might improve as well.
REFERENCES Adams, D.A, Neslon, R.R., & Todd P.A. (1992). Perceived usefulness, ease of use, and usage of information technology: A replication. MIS Quarterly, 16(2), 227-247. Afuah, A., & Tucci C.L. (2001). Internet business models and strategies: Text and cases. Boston: McGraw-Hill Irwin. Agarwal, R. (2000). Individual acceptance of information technologies. In Framing the domains of IT management: Projecting the future through the past 5:=PXG(G &LQFLQQDWL2+ 3LQQDÀH[ Publishing, 85-104. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action-control: From cognition to behavior (pp. 11-39). Heidelberg, Germany: Springer. Ajzen, I. (1991). The theory of planned behavior. Organizational, Behavior and Human Decision 3URFHVVHV 179-211.
Anderson, E.W., Fornell, C., & Lehmann D.R. (1994). Customer satisfaction, market share, and SUR¿WDELOLW\ )LQGLQJV IURP 6ZHGHQ Journal of Marketing, (3), 53-66. Bargh, J.A. (1989). Conditional automaticity: VaULHWLHVRIDXWRPDWLFLQÀXHQFHLQVRFLDOSHUFHSWLRQ and cognition. In S. Uleman & J.A. Bargh (Eds.), Unintended thought (pp. 3-51). New York: The Guilford Press. Bargh, J.A. (1994). The four horsemen of automaticLW\ $ZDUHQHVVLQWHQWLRQHI¿FLHQF\DQGFRQWUROLQ social cognition. In Wyer, Jr. & T.K. Srull (Eds.), Handbook of social cognition (pp. 1-40). Hillsdale, NJ: Lawrence Erlbaum Associates. Bhattacherjee, A. (2001). Understanding informaWLRQV\VWHPVFRQWLQXDQFH $QH[SHFWDWLRQFRQ¿Umation model. MIS Quarterly, (3), 351-370. Carroll, J., Howard, J.A., Peck, J., & Murphy J. $ ¿HOG VWXG\ RI SHUFHSWLRQV DQG XVH RI mobile telephones by 16-22 year olds. Journal of Information Technology Theory and Application, 4(2), 49-61. Cheung, W., Chang, M.K., & Lai, V. (2000). Prediction of Internet and world wide Web usage at work: A test of an extended triandis model. Decision Support Systems, 30, 83-100. Choudrie, J., & Lee, H. (2004). Broadband development in South Korea: Institutional and cultural factors. European Journal of Information Systems, 13(2), 103-114. Choudrie, J., Papazafeiropoulou, A., & Lee, H. (2003). A web of stakeholders and strategies: A case of broadband diffusion in South Korea. Journal of Information Technology, 18(4), 281-290. Churchill, G.A.J.R. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64-73. Churchill, G.A.J.R., & Surprenant, C. (1982). An investigation into the determinants of customer satisfaction. Journal of Marketing Research, 19, 491-504.
461
A Survey on Broadband Internet
Chwelos, P., Benbasat, I., & Dexter, A.S. (2001). Research report: Empirical test of an EDI adoption model. Information Systems Research, 12(3), 304-321. Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. Davis, F.D., Bagozzi, R.P., & Warshaw, P.R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management 6FLHQFH 982-1003. Davis, F.D., & Venkatesh, V. (2004). Toward preprototype user acceptance testing of new information systems implications for software project management. IEEE Transactions on Engineering 0DQDJHPHQW(1), 31-46. DeLone, W.H., & McLean, E.R. (1992). Information systems success: The quest for the dependent variable. Information Systems Research, 3(1), 60-95. DeLone, W.H., & McLean E.R. (2003). The DeLone and McLean model of information systems success: A ten-year update. Journal of Management Information Systems, 19(4), 9-30. Devaraj, S., Fan, M., & Kohli, R. (2002). Antecedents of B2C channel satisfaction and preference: Validating e-commerce metrics. Information Systems Research, 13(3), 316-333. Dutton, W.H., Gillett, S.E., McKnight, L.W., & Peltu, M. (2004). Bridging broadband Internet GLYLGHV 5HFRQ¿JXULQJDFFHVVWRHQKDQFHFRPPXnicative power. Journal of Information Technology, 19(1), 28-38. Falk, R.F., & Miller, N.B. (1992). A primer for soft modeling. Ohio: The University of Akron Press. Fishben, M., & Ajzen, I. (1975). Belief, attitutude, intention, and behavior. Reading, MA: AddisonWesley. Fornell, C.A. (1982). Second generation of multivariate analysis methods (Vol. 1). New York: Praeger.
462
Harrison, D.A., Mykytyn, & Riemenschneider, C.K. (1997). Executive decisions about adoption of information technology in small business: Theory and empirical tests. Information Systems Research, 8(2), 171-195. Hartwick, J., & Barki, H. (1994). Explaining the role of user participation in information system use. Management Science, 40(4), 440-465. Igbaria, M. (1990). End-user computing effectiveness: A structural equation model. Omega, 18(6), 637-652. Igbaria, M. (1993). User acceptance of microcomputer technology: An empirical test. Omega, 21(1), 73-90. Igbaria, M., Parasuraman, S., & Baroudi, J.J. (1996). A motivational model of microcomputer usage. JMIS, 13(1), 127- 143. K a r a h a n n a , E ., D e t m a r, W., St r a u b, & Cher va ny, N.L. (1999). I n for mat ion technology adoption across time: A cross section comparison of pre-adoption beliefs and post adoption beliefs. MIS Quarterly, 23(2), 183-213. Keil, M., Beranek, P.M., & Konsynski, B.R. (1995). 8VHIXOQHVVDQGHDVHRIXVH ¿HOGVWXG\HYLGHQFH regarding task considerations. Decision Support Systems, 13(1), 75-91. Kettinger, W.J., & Grover, V. (1997). The use of computer-mediated communication in an interorganizational context. Decision Sciences, 28(3), 513-555. Lim, J., Gan, B., & Chang, T. (2002). A survey on NSS adoption intention. In Proceedings of the WK+,&66 Limayem, M., Cheung, C.M.K, & Chan, G.W.W. (2003). Explaining information systems adoption and postadoption: Towards an integrated model. ICIS. Limayem, M., & Hirt, S.G. (2003). Force of habit and information systems usage: Theory and initial validation. Journal of the Association for Information Systems, 4(3).
A Survey on Broadband Internet
Lyytinen, K., & Hirschheim, R. (1987). Information V\VWHPVIDLOXUHV $VXUYH\DQGFODVVL¿FDWLRQRIWKH empirical literature. Oxford Surveys in Information Technology, 4, 257-309.
Rai, A., Lang, S.S., & Welker, R.B. (2002). Assessing the validity of IS success models: An empirical test and theoretical analysis. Information Systems Research, 13, 50-69.
McKinney, V., Yoon, K., & Zahedi, F.M. (2002). The measurement of Web-customer satisfaction: $Q H[SHFWDWLRQ DQG GLVFRQ¿UPDWLRQ DSSURDFK Information Systems Research, 13, 296-315.
Riemenschneider, C.K.D., Harrison, A., Mykytyn, J., & Pete, P. (2003). Understanding IT adoption decisions in small business: Integrating current theories. Information & Management, 40(4), 269-285.
Middleton, C.A. (1999). Residential broadband QHWZRUNV 3UHOLPLQDU\¿QGLQJVIURPD&DQDGLDQ information highway trial. In Proceedings of the 9th Workshop on Information Technologies And Systems. Charlotte, NC: WITS. Middleton, C.A. (2002a). Exploring consumer demand for networked services: The importance of content, connectivity and killer apps in the diffusion of broadband and mobile services. In L. Applegate, R. Galliers & J.I. DeGross (Eds.), In Proceedings of the Twenty-Third International Conference on Information Systems (pp. 391-399). Middleton, C.A. (2002b). Who needs a killer app? Two perspectives on content in residential broadband networks. Journal of Research and Practice in Information Technology, 34(2), 67-81. Middleton, C.A. (2003). What if there is no killer application? An exploration of a user-centric perspective on broadband. Journal of Information Technology, 18(4), 231-245. Nunnally, J.C. (1978). Psychometric theory (2nd ed.). New York: McGraw- Hill Book Company. Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitudes. Journal of Information Technology, 18, 267-280. Oliver, R.L. (1980). A cognitive model for the antecedents and consequences of satisfaction. Journal of Marketing Research, 17, 460-469. Ouellette, J.A., & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124(1), 54-74.
Rogers, E.M. (1995). Diffusion of innovations (4th ed.). New York, NY: The Free Press. Salkind, V. (2000). Exploring research (4th ed.). New Jersey: Prentice-Hall, Inc. Sawyer, S., Allen, J.P., & Lee, V. (2003). broadband and mobile opportunities: A socio-technical perspective. Journal of Information Technology, 18(2), 121-136. Schermehron, J.R., Hunt J.G., & Osborn, P.R.N. (1998). Basic organizational behavior (2nd ed.). New York: J. Wiley. Shim, J.P., Shin, Y.B., & Nottingham, L. (2002). 5HWDLOHU:HEVLWHLQÀXHQFHRQFXVWRPHUVKRSSLQJ An exploratory study on key factors of customer satisfaction. Journal of the Association for Information Systems, 3, 53-76. Spreng, R.A., MacKenzie, S.B., & Olshavsky, R.W. (1996). A reexamination of the determinants of consumer satisfaction. Journal of Marketing, 60, 15-32. Susarla, A., Barua, A., & Whinston, A.B. (2003). Understanding the service component of application service provision: An empirical analysis of satisfaction with ASP services. MIS Quarterly, 27(1), 91-123. Taylor, S., & Todd, V. (1995). Understanding information technology usage: A test of competing models. Information Systems Research, 6(2), 144-176. Tse, D.K., & Wilton, P.C. (1988). Models of consumer satisfaction formation: An extension. -RXUQDORI0DUNHWLQJ5HVHDUFK 204-212.
463
A Survey on Broadband Internet
U.S. Department of Commerce. (2002, September). Understanding broadband demand: A review of critical issues. 2I¿FHRI7HFKQRORJ\3ROLF\ Venkatesh, V., & Davis, F.D. (2000). A theoretical extension of the technology acceptance model: Four ORQJLWXGLQDO¿HOGVWXGLHVManagement Science, 46(2), 186-204. Venkatesh, V., Morris, M.G., Davis, G.B., & Davis, F.D. (2003). User acceptance of information WHFKQRORJ\ 7RZDUGDXQL¿HGYLHZMIS Quarterly, 27(3), 425-478. Venkatesh, V., Speier, C., & Morris, M.G. (2002). User acceptance enablers in individual decision making about technology: Toward an integrated model. Decision Sciences, 33(2), 297-316. Wixom, B.H., & Todd, P.A. (2005). A theoretical integration of user satisfaction and technology acceptance. Information Systems Research, 16(1), 85-102. Zviran, M., & Erlich, Z. (2003). Measuring IS user satisfaction: Review and implications. Communications of the Association for Information Systems, 12, 81-103.
KEY TERMS ICT: The term ICT is used to refer generically to the family of related technologies that process, store and transmit information by electronic means. The term IT is narrower and is used to denote computer, software and related technologies not including communications and network technologies, but the boundary between the two is increasingly blurred (OECD, 2004). Broadband: Always on access, at work, at home RURQWKHPRYHSURYLGHGE\DUDQJHRI¿[HGOLQH wireless and satellite technologies to progressively higher bandwidths capable of supporting genuinely new and innovative interactive content, applications and services and the delivery of enhanced public services( Sawyer et al., 2003).
464
Satisfaction:6DWLVIDFWLRQLVLQGLYLGXDOV¶IHHOings of pleasure or disappointment resulting from comparing their perceptions of a product or serYLFH¶VSHUIRUPDQFHWRWKHLUH[SHFWDWLRQOHYHOV Survey: A survey is a way of going from observations to theory validation. The common objective for using this approach is to determine the relationship between constructs as a way of making sense of behavior surrounding and involving IS. In positivist research, surveys are particularly useful in determining the actual values of variables under study, and the strengths of relationships among them (Newsted et al., 1996). Operationalization: The process of measurement development or operationalization involves rules for assigning numbers to objects to represent quantities of attributes (Nunnally, 1978). The operationalization of variables requires the development of a statement or statements that specify the activities or operations used to measure a variable. Content validity: Content validity is representativeness or sampling adequacy of the measure. It refers to the extent to which a test represents the universe of items from which it is drawn (Salkind, 2000). The measurement has content validity when its contents (e.g., questions) adequately represent the construct it measures. Structural Equation Modeling: Structural equation modeling, an integral tool in both managerial and academic research, is a statistical method useful when one dependent variable becomes an independent variable in subsequent dependence relationships (Hair et al., 1995). It provides a straightforward method of dealing with multiple relationships simultaneously while providing staWLVWLFDOHI¿FLHQF\$OOVWUXFWXUDOHTXDWLRQPRGHOing techniques are distinguished by their ability to estimate multiple and interrelated dependence relationships and account measurement error at the same time. It is an important tool for studying causal models (Fornell, 1982; Fornell et al., 1981).
Division II
Europe
466
Chapter XXIX
Broadband Access and Broadband-Based Applications:
An Empirical Study of the Determinants of Adoption Among Italian SMEs Massimo G. Colombo Politecnico di Milano, Italy Luca Grilli Politecnico di Milano, Italy Cinzia Verga Politecnico di Milano, Italy
ABSTRACT Why do some small and medium enterprises (SMEs) adopt Internet broadband technologies (high-speed connection and complementary applications) and others do not? This chapter aims at analyzing the issue through an econometric investigation. Relying on the (thin) previous empirical literature on the topic and focusing on a large and representative sample of Italian SMEs, we analyze the determinants of broadband connection and adoption of complementary applications. Results of the econometric analysis UHYHDOWKDW L DPRQJ¿UPVSHFL¿FFKDUDFWHULVWLFVVL]HDQGWKH¿UP¶VQHHGWRFRPPXQLFDWHDUHPDMRU determinants both of broadband connection and use of complementary applications, while indicators of ¿UPHI¿FLHQF\DQGRIWKHFRPSHWLWLYHSUHVVXUHVXIIHUHGIURPWKH60(SRVLWLYHO\DIIHFWRQO\WKHQXPEHURI XVHGDSSOLFDWLRQVZKLOHQRWH[HUWLQJDQ\LPSDFWRQWKHGHFLVLRQWRFRQQHFW LL DPRQJORFDWLRQVSHFL¿F FKDUDFWHULVWLFVWKHOHYHORIWHOHFRPPXQLFDWLRQVLQIUDVWUXFWXUHSRVLWLYHO\LQÀXHQFHVERWKFRQQHFWLRQDQG applications use, while the presence within the local labor market of young and skilled workforce makes 60(VDGRSWPRUHDSSOLFDWLRQV LLL WLPHVSHFL¿FYDULDEOHVOLNHWKRVHUHODWHGWRWKHDFWXDODQGIXWXUH SULFHRIWKHKLJKVSHHG,QWHUQHWFRQQHFWLRQDIIHFW60(V¶GHFLVLRQWRDGRSWEURDGEDQG7KHVH¿QGLQJV have important implications for suppliers and policy makers.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband Access and Broadband-Based Applications
INTRODUCTION Broadband access to communication networks plays a crucial role for economic development. In particular, it is very important for small and medium enterprises (SMEs), in so far as it provides this type RI¿UPZLWKHI¿FLHQWDQGSHUPDQHQWFRQQHFWLYLW\ to the global market at a price that many SMEs could not previously afford (OECD, 2003). This has generated a policy debate on how to stimulate rapid and widespread adoption of broadband access technologies. However, policy interventions can be effective only if adoption determinants are well understood. 3UHYLRXVVWXGLHVDQDO\]HGWKHPRVWVLJQL¿FDQW IDFWRUVZKLFKLQÀXHQFH60(V¶DGRSWLRQRI,QWHUnet and broadband, highlighting that SMEs may HQFRXQWHUVHYHUHREVWDFOHVPDLQO\UHODWHGWR¿UP VSHFL¿FDQGHQYLURQPHQWDOVSHFL¿FFKDUDFWHULVWLFV (see, among others, Arbore & Ordanini, 2006; Mehrtens, Cragg, & Mills, 2001). In particular, PDQDJHUV¶,7NQRZOHGJHDVZHOODV¿UPVL]HVHHPV WRSRVLWLYHO\DIIHFW60(V¶DGRSWLRQRIEURDGEDQG technologies. For what concerns environmental factors, external pressure on the organization (from customers or trading partners) to adopt the technology, are likely to have positive effects on ¿UPV¶GHFLVLRQ)XUWKHUPRUHWKHJHRJUDSKLFDODUHD ZKHUH¿UPVDUHORFDWHGIXUWKHULQÀXHQFHVDYDLOability of broadband connection and consequently 60(V¶GHFLVLRQWRDGRSW 3ULHJHU Moreover, the adoption by SMEs of broadband connection is likely to have a negligible economic LPSDFWDW¿UPOHYHOE\DQGRILWVHOI,QRUGHUWR JHQHUDWHSRVLWLYHIHHGEDFNVRQ¿UPSURGXFWLYLW\ DQGPRUHJHQHUDOO\RQ¿UPSHUIRUPDQFHVLWKDVWR be associated with the adoption of complementary advanced communications (e.g., virtual private network, VoIP, video-conference) and management (e.g., customer relationship management, supply chain management, human resource, and administration management systems) applications WKDWDOORZ¿UPVWRUDGLFDOO\FKDQJHWKHZD\WKH\ do business. In fact, broadband access like other ICTs, is a general-purpose and enabling technology %UHVQDKDQ 7UDMWHQEHUJ ZKRVHEHQH¿WV can be fully captured only if it is used by adopting
SMEs both to carry out the same transactions or DFWLYLWLHVLQDPRUHHI¿FLHQWZD\DQGDOVRDVDQ instrumental means to generate and develop new transactions and activities (Bertschek & Kaiser, 2004; Preissl, 1995). In turn, for most economic organizations a necessary condition for an effective deployment of the above mentioned applications involves deep changes in their management. It follows that as interesting as the analysis of the determinants of the adoption by SMEs of Internet broadband access is the study of the factors that IRVWHURUKLQGHU¿UPV¶DGRSWLRQRIEURDGEDQGEDVHG DSSOLFDWLRQVLQFOXGLQJ¿UPV¶DELOLW\WRWUDQVIRUP their organizational structure and managerial practices.1 The chapter investigates both aspects: on one hand, we analyze the factors affecting the adoption by SMEs of broadband connection; on the other hand, we investigate the determinants of the adoption of broadband-based applications, once ¿UPVKDYHDGRSWHGDEURDGEDQGFRQQHFWLRQ7KH contribution focuses on Italy, where SMEs traditionally account for most of the wealth produced in the national economic system, and, in line with the extant empirical literature on the topic, gives VSHFLDO DWWHQWLRQ WR ¿UP DQG ORFDWLRQVSHFL¿F factors as fundamental drivers of adoption. After a descriptive illustration of the diffusion of broadband connection and broadband-based applications among Italian SMEs, we study the determinants of their adoption through the estimates of a series of econometric models. For this purpose we take advantage of a new longitudinal dataset composed of 904 Italian SMEs (i.e., number of employees comprised between 10 and 249), that operate in both manufacturing and service sectors (excludLQJSXEOLFDGPLQLVWUDWLRQ¿QDQFHDQGLQVXUDQFH 7KHVDPSOHLVVWUDWL¿HGE\LQGXVWU\VL]HFODVVDQG geographical area so as to be representative of the Italian population of SMEs, and it contains detailed VXUYH\EDVHGLQIRUPDWLRQRQ¿UPVSHFL¿FFKDUDFWHULVWLFVDQGDERXW¿UPV¶DGRSWLRQRIEURDGEDQG connection and broadband-based applications over the period from 1998 to 2005. The analysis highlights a number of interesting ¿QGLQJV$VWRWKHDGRSWLRQRIEURDGEDQGFRQQHFtion, the econometric results highlight that larger
467
Broadband Access and Broadband-Based Applications
SMEs, which have a multiplant structure and belong to business groups are those most likely to be early DGRSWHUV&RQYHUVHO\YDULDEOHVUHÀHFWLQJWKHVNLOO OHYHORIWKHZRUNIRUFH¿UPV¶DJHDQGRSHUDWLRQDO HI¿FLHQF\KDYHQHJOLJLEOHHIIHFWVRQWKHVSHHGRI DGRSWLRQ$PRQJWKHORFDWLRQVSHFL¿FIDFWRUVWKH infrastructural development of the telecommunicaWLRQVQHWZRUNLVWKHRQO\VLJQL¿FDQWGHWHUPLQDQW of adoption. As to broadband-based applications, ¿UPV¶VL]HRUJDQL]DWLRQDOVWUXFWXUHDQGRZQHUship status have the same effects as those shown DERYH 0RUHRYHU TXLWH LQWHUHVWLQJO\ ¿UPV¶ DJH SRVLWLYHO\LQÀXHQFHV60(V¶DGRSWLRQRIEURDGEDQG DSSOLFDWLRQVFRQ¿UPLQJWKDW\RXQJHU¿UPVIDFH lower switching costs in implementing the new management practices and organizational innovations associated with these applications (or that they employ younger and more “IT-familiar” workers). In accordance with this result, the adoption of broadband-based applications is also found to EHGULYHQE\ORFDWLRQVSHFL¿FFKDUDFWHULVWLFVWKDW UHÀHFWWKHDYDLODELOLW\LQWKHORFDOODERUPDUNHWRI younger and more skilled personnel. In the next section of the chapter, we review the extant empirical literature on the determinants of ¿UPV¶GHFLVLRQWRDGRSW,QWHUQHWEURDGEDQGFRQnection and broadband-based applications, with particular emphasis on those studies that focus on SMEs. Then we describe the sample and provide some preliminary descriptive statistics on the diffusion of the phenomena under investigation among Italian SMEs. This will bring us to the empirical analysis: after a brief description of the econometric models and of the dependent and independent variables used, we illustrate the results and the main implications of the econometric estimates. Some summarizing remarks and delineation of future research opportunities conclude the chapter.
BACKGROUND Relying on the extant empirical literature on the GHWHUPLQDQWV RI ¿UPV¶ DGRSWLRQ RI ,&7 FDSLWDO there can be isolated three types of factors which are likely to drive broadband connection and use of broadband-based applications among SMEs.
468
This section is devoted to synthesizing previous studies on the topic and deriving some hypotheses on these determinants.
)LUP6SHFL¿F)DFWRUV Among the factors which are more likely to affect the adoption of broadband connection and broadband-based applications, previous studies on diffusion of innovations (DOI) literature give particular DWWHQWLRQWR¿UPVSHFL¿FFKDUDFWHULVWLFV Organization size is found to be a fundamental driver for innovativeness: slack resources, a large DYDLODELOLW\RIWDQJLEOH LH¿QDQFLDO DQGLQWDQgible (i.e., managerial competencies) capital and a greater ability to access external resources and markets are often advocated to be main reasons for such a positive relationship (Rogers, 1995). In our context, this relationship is strengthened by the VSHFL¿FQDWXUHRI,&7VLQFHODUJHU60(V¿UPV KDYLQJDPXOWLSODQWVWUXFWXUHDVZHOODV¿UPVEHlonging to business groups are more likely to be in need of adopting advanced communication systems DQGDSSOLFDWLRQV7KHVHUHVXOWVDUHFRQ¿UPHGE\ a number of studies. Forman (2005) is the study most similar to the present one, except for the fact that it looks to generic Internet adoption without distinguishing between narrowband and broadband connection. Using a discrete choice econometric approach, Forman analyzes the determinants of Internet adoption and applications use by a large sample RI86VHUYLFHV¿UPVLQ +H¿QGVWKDW¿UP size measured in terms of employees has a scarce relevance in explaining access, while it positively DQGVLJQL¿FDQWO\DIIHFWVDSSOLFDWLRQVDGRSWLRQ$OVR a multi-establishment structure is found to exert a very large and positive impact on both access and adoption of applications. Moreover, the more the employee concentration within a small number of establishments the less is found to be the probability of Internet adoption and applications use. Similar results are highlighted by Arbore and Ordanini (2006). They use a logit model in order to analyze the probability of broadband connection in the year 2003 of a sample of 842 Italian SMEs. Their UHVXOWVFRQ¿UPWKDWVPDOO¿UPVJDLQOHVVDFFHVV
Broadband Access and Broadband-Based Applications
to Internet high-speed technology. This major role of size in affecting the probability of connection is moderated by other factors. In particular, the SUREDELOLW\RIFRQQHFWLRQGHFUHDVHVIRUVPDOO¿UPV located in rural areas except for those SMEs that extensively outsource their ICT activities. &RPSHWLWLYHSUHVVXUHLVDQRWKHU¿UPVSHFL¿F characteristic which may potentially (and actually is found to) affect the “broadband decision.” AcFRUGLQJWRHYROXWLRQDU\WKHRULHVRIWKH¿UP VHH 1HOVRQ :LQWHU ¿UPVZRXOGGHYLDWHIURP “production routines” and introduce some product or process innovation only when their market position is threatened by external parties. Hence, a more intense perceived competition may induce WKH¿UP¶VPDQDJHPHQWWRDGRSW,QWHUQHWEURDGEDQG technologies. Quite in line with this reasoning, Mehrtens et al. (2001) show in their case-studies analysis that external pressure is among the three major factors affecting Internet adoption behavior E\ 60(V WRJHWKHU ZLWK SHUFHLYHG EHQH¿WV DQG organizational readiness. Forman (2005) reaches WKHVDPHFRQFOXVLRQ WKHSUREDELOLW\RI¿UPVWR connect and adopt applications raises with the percentage of competitors within the industry who KDYHGRQHVR$OVRQRWHWKDWDJRRG¿UP¿QDQFLDO performance may even represent an essential prerequisite for investing in new ICT technology given the presence of capital market imperfections VXIIHUHGIURP60(VHVSHFLDOO\IRU¿QDQFLQJWKH purchase of intangible capital goods (see, among others, for the high degree of imperfections characterizing Italian capital markets, Becchetti & Trovato, 2002; Colombo & Grilli, 2007; Fagiolo & Luzzi, 2006). )LQDOO\¿UPV¶,7IDPLOLDULW\DQG¿UPV¶SULRU investments in ICT capital may affect the probability of new (often related) investments in the same area and consequently may increase the likelihood of adoption of broadband connection and use of complementary applications. This evidence is documented by Forman (2005).2
/RFDWLRQ6SHFL¿F)DFWRUV Several studies have highlighted that location matters.
First, broadband connection to Internet is less likely in rural areas as for the consumer segment (see Prieger, 2003; Tookey, Whalley, & Howick, 2006 for contributions that refer to broadband access by population in U.S. and in Scotland, respectively) as for the business one (see again Arbore & Ordanini, 2006). The recent cross-country studies performed by Kim, Bauer, and Wildman DQG *DUFLD0XULOOR FRQ¿UP WKDW broadband penetration is positively correlated with national population density. Allegedly, the socioeconomic conditions and the quality level of the telecommunications infrastructure of the area on which SMEs are located may strongly affect WKHDFWXDO¿UPV¶FDSDELOLWLHVWRFRQQHFWDQGXVH complementary applications. Moreover, also local labor market characteristics may have an impact on D¿UP¶VZLOOLQJQHVVWRXVHEURDGEDQGWHFKQRORJ\ clearly, a great availability in the local workforce of young and IT skilled personnel may incentive ¿UPVWRFRQQHFWDQGHDVHWKHDGRSWLRQRIEURDGband-based applications, while the reverse may GHSUHVV¿UPV¶RSSRUWXQLW\WRGR,7LQYHVWPHQWV Accordingly, an early research conducted by the Federal Communications Commission in 2000 (FCC, 2000) indicated that broadband was less likely to be available in rural and lower-income areas. Garcia-Murillo (2005) in her analysis on the determinants of broadband adoption across several countries found per capita income as a major driver for deployment. Again, the study of Kim et al. (2003) highlights that IT preparedness and income are important determinants of broadband technology deployment in a country. Prieger (2003) in his econometric study on U.S. broadband connection documents that the median level of income of a geographic area has a positive, albeit QRWDOZD\VVLJQL¿FDQWHIIHFWRQWKHDYDLODELOLW\RI broadband services. A positive and more robust impact is found for variables capturing the educaWLRQSUR¿OHRISRSXODWLRQ,QWKLVUHVSHFWWKHVWXG\ FRQGXFWHGE\&DYD)HUUHUXHODDQG$ODEDX0XĔR] (2004) on a sample of 30 OECD countries individuated education level of the population as one of the most consistent factors explaining the level of broadband adoption, followed by the availability and affordability of broadband access.
469
Broadband Access and Broadband-Based Applications
7LPH6SHFL¿F)DFWRUV Prior research in the DOI literature has provided ODUJHHYLGHQFHWKDWWLPHVSHFL¿FIDFWRUVOLNHSULFH of a new technology and its expected variation along time are important drivers of a new technology deployment in an economic system (see, among others, Stoneman, 2001). Ceteris paribus, high prices as well as expectations of a decrease in future prices, slow down the diffusion process. Both factors will be considered in the analysis of the determinants of broadband connection by SMEs.
THE DATASET In this chapter, we consider a sample composed RI ,WDOLDQ ¿UPV 7KH ¿UPV LQFOXGHG LQ WKH sample are small and medium enterprises (i.e., number of employees comprises between 10 and 249) operating in both manufacturing and service VHFWRUV H[FOXGLQJSXEOLFDGPLQLVWUDWLRQ¿QDQFH and insurance). The sample has been developed E\7KLQN7HOLQDQGLWLVVWUDWL¿HGE\LQGXVtry, size class, and geographical area so as to be representative of the Italian population of SMEs. Firms are observed from 1998 to 2005. The dataset contains detailed survey-based LQIRUPDWLRQRQ¿UPVSHFL¿FFKDUDFWHULVWLFV HJ VLQJOHRUPXOWLSODQWVWUXFWXUHZKHWKHU¿UPVEH-
ORQJWRJURXSVRUQRW DQGDERXW¿UPV¶DGRSWLRQRI broadband connection and broadband-based applications. This dataset has been complemented with ¿UPV¶HFRQRPLFDQG¿QDQFLDOGDWD VRXUFH $,'$ information on the socioeconomic characteristics RI WKH DUHD RQ ZKLFK ¿UPV DUH ORFDWHG VRXUFH Tagliacarne Institute and ISTAT) and longitudinal information on price levels of Internet broadband technologies (source: European Commission).3 ,QWKLVFKDSWHUZHGH¿QHEURDGEDQGDFFHVVDV an Internet wired connection via ADSL or other dedicated lines with an upstream speed higher or equal to 256 Kbps (see, among others, Arbore & Ordanini, 2006; OECD, 2002). We identify 15 broadband-based applications ranging from very basic (e.g., e-mail) to advanced (e.g., supply chain management system) Internet use: 1. 2. 3. 4. 5.
VPN (virtual private network) Data and disaster recovery system Local protection system VoIP (voice over Internet protocol) system Video-communication, streaming, or videoconference system 6. E-mail )LOHVKDULQJRU¿OHGLVWULEXWLRQV\VWHP 8. E-learning system 9. CRM (customer relationship management) system 10. SCM (supply chain management) system
Figure 1. Diffusion of Internet broadband connection among Italian SMEs 70.0
broadband diffusion
60.0 50.0 40.0 30.0 20.0 10.0 0.0 1999
2000
2001
2002 year
470
2003
2004
2005
Broadband Access and Broadband-Based Applications
Figure 2. Diffusion of Internet broadband-based applications among Italian SMEs
application count
7 6 5 4 3 2 1 0 1999
2000
2001
2002
2003
2004
2005
year
11. 12. 13. 14. 15.
Co-design system with suppliers and customers E-banking system Internet access system Web site Human resource and administration management systems
Figure 1 shows penetration rates of broadband connection among Italian SMEs: broadband diffusion rate has constantly increased over time, starting from 4.8% in 1999 and reaching 66.5% in 2005. Along with access, broadband-based applications use has increased from an average number of 0.3 applications per SME in 1999 up WRDSSOLFDWLRQVSHU¿UPLQ$OOHJHGO\LI DFFHVVDQGXVHE\¿UPVRIEURDGEDQGWHFKQRORJLHV are sensibly increased since the initial period, these ¿JXUHVDOVRVXJJHVWWKDWPDUNHWVDWXUDWLRQLVVWLOOIDU from materializing. The next sections of the chapter are devoted to the investigation of which factors KDPSHUDQGKLQGHU60(¶VGHFLVLRQWREHFRPHD broadband adopter and applications user.
THE EMPIRICAL ANALYSIS 7KH6SHFL¿FDWLRQRIWKH(FRQRPHWULF Models Our econometric analysis is carried out in two steps. First, relying on the technology diffusion model developed by Karshenas and Stoneman (1993),
we estimate a survival data analysis model of the determinants of the adoption by Italian SMEs of broadband connection. In this case, the observed VSHOO PHDVXUHGLQ\HDUV QHHGHGIRU¿UPVWRDGRSW EURDGEDQGFRQQHFWLRQLVDIXQFWLRQRI¿UPORFDWLRQDQGWLPHVSHFL¿FYDULDEOHV VHH7DEOH 7KHPRGHOLVVSHFL¿HGLQWHUPVRIWKHGXUDWLRQRI a SME not adopting broadband connection during its life, which represents the dependent variable of the model. The basic tool for modeling duration data, given the right-censored nature of the sample LHWKHSUHVHQFHZLWKLQWKHVDPSOHRI¿UPVWKDW did not adopt broadband connection) are duration or hazard models.4 The probability distribution RI GXUDWLRQ FDQ EH VSHFL¿HG E\ WKH GLVWULEXWLRQ function F(t) = Pr (7 W ZKLFK VSHFL¿HV WKH probability that the duration variable T is less than some value t7KHKD]DUGIXQFWLRQLVGH¿QHGDV h(t) = f(t)/S(t), where f(t) is the probability density function and S(t), which is equal to 1 - F(t), is the survivor function. The hazard function may be viewed as the instantaneous probability of adopting, provided that this has not occurred by t. As is frequent in this type of literature, we choose to model the hazard function by a semiparametric approach (Cox, 1972):
hi (t )
h0 (t )exp( ' xit ) ,
(1)
where h0 (t ) is the baseline hazard rate at time t, that is, the hazard rate when all explanatory variables equal zero, xit is the vector of (possibly WLPHYDU\LQJ ¿UPVSHFL¿FORFDWLRQVSHFL¿FDQG
471
Broadband Access and Broadband-Based Applications
WLPHVSHFL¿F H[SODQDWRU\ YDULDEOHV SOXV RWKHU controls and ȕ is the vector of parameters to be estimated. 6HFRQGZHGH¿QHDPHDVXUHRIWKHH[WHQWRI adoption of broadband-based applications based on WKHQXPEHURIDSSOLFDWLRQVDGRSWHGE\¿UPVXSWR time t; then, conditional to connection, we analyze the enabling or hindering factors of applications use through a random effects panel data model. The model is structured as follows (Greene, 2000):
yit
'
xit
it
with
it
vit ui .
(2)
yit LVWKHYDULDEOHWKDWFDSWXUHV¿UPV¶QXPEHURI broadband-based applications at time t and the vector xit includes again most of the explanatory variables used in the survival analysis. The effects of the covariates on the number of broadband-based applications possessed by SMEs are accounted for by the parameter vector ȕ. The error component
ui UHSUHVHQWVWLPHLQYDULDQW¿UPVSHFL¿FHIIHFWV QRWUHÀHFWHGE\WKHLQGHSHQGHQWYDULDEOHVZKLOH vit is the remainder time-varying disturbance. Both components are assumed to be normally distributed with zero means and independently of one another. A summary of the explanatory variables used in the estimation of the econometric models is UHSRUWHG LQ 7DEOH 7KH\ LQFOXGH WKRVH ¿UP ORFDWLRQ DQG WLPHVSHFL¿F YDULDEOHV ZKLFK DUH OLNHO\WRLQÀXHQFH¿UPV¶DGRSWLRQRIEURDGEDQG technologies (see again the section Background) SOXV RWKHU FRQWURO YDULDEOHV 6RPH ¿UPVSHFL¿F WLPHYDU\LQJYDULDEOHVLQGLFDWHGE\VXI¿[t, are one period lagged so as to mitigate possible reverse causality problems. $PRQJ¿UPVSHFL¿FYDULDEOHVZHLQFOXGH¿UP size (Employeest-1 DI¿OLDWLRQWRDEXVLQHVVJURXS (Group), and the presence of more than a plant (Multiplant 7KLVJURXSDOVRLQFOXGHV¿UPV¶DJH Aget),
7DEOH'H¿QLWLRQRIH[SODQDWRU\YDULDEOHV PRQHWDU\YDOXHVDGMXVWHGIRULQÀDWLRQ Variable
Description
)LUPVSHFL¿FYDULDEOHV Employeest-1
1XPEHURI¿UPHPSOR\HHVDWWLPHW VRXUFH $,'$
Group
2QHIRU¿UPVEHORQJLQJWREXVLQHVVJURXSV VRXUFH 7KLQN7HO
Multiplant
2QHIRU¿UPVZLWKDPXOWLSODQWVWUXFWXUH VRXUFH 7KLQN7HO
Employees Growtht-1
3HUFHQWDJHJURZWKRI¿UPHPSOR\HHVEHWZHHQWLPHWDQGWLPHW Employeest-1 – Employeest)/ Employeest-2 (source: AIDA).
2
Aget
1XPEHURI\HDUVVLQFH¿UPVIRXQGDWLRQDWWLPHW
Value Added/ Employeest-1
5DWLREHWZHHQWKHYDOXHDGGHGJHQHUDWHGE\WKH¿UPDWWLPHWDQGWKHQXPEHURI¿UP employees at time t-1 (source: AIDA).
Salaries/ Employeest-1
5DWLREHWZHHQWKHWRWDOVDODULHVSDLGE\WKH¿UPDWWLPHWDQGWKHQXPEHURI¿UPHPSOR\HHVDW time t-1 (source: AIDA).
Cash Flow/ Total Assetst-1
5DWLREHWZHHQWKHFDVKÀRZJHQHUDWHGE\WKH¿UPDWWLPHWDQGWKHWRWDODVVHWVYDOXHRIWKH ¿UPDWWLPHW VRXUFH $,'$
/RFDWLRQVSHFL¿FYDULDEOHV Employee Age
:HLJKWHGDYHUDJHRIHPSOR\HHV¶DJHE\SURYLQFH $YHUDJHLVZHLJKWHGRQWKHQXPEHURI HPSOR\HHV (PSOR\HHV¶DJHLVPHDVXUHGRQDVFDOHIURP \HDUV WR PRUHWKDQ years) (source: ISTAT Italian census, 2001).
Employee Education
:HLJKWHGDYHUDJHRIHPSOR\HHV¶OHYHOHGXFDWLRQE\SURYLQFH $YHUDJHLVZHLJKWHGRQWKH QXPEHURIHPSOR\HHV (PSOR\HHV¶OHYHORIHGXFDWLRQLVPHDVXUHGRQDVFDOHIURP ORZOHYHORI education) to 6 (high level of education) (source: ISTAT Italian census, 2001).
472
Broadband Access and Broadband-Based Applications
Table 1. (cont.) South
2QHIRU¿UPVORFDWHGLQWKH6RXWKRI,WDO\
Average Incomet
Ratio between the provincial income per inhabitant and the national income per inhabitant at time t. Data are available over the period 1991-2001. Missing data have been estimated (source: Tagliacarne Institute database).
Telecommunications Network
Value of the index measuring provincial infrastructural development of telecommunications network in 2000 (source: Tagliacarne Institute database).
7LPHVSHFL¿FYDULDEOHV (only for broadband connection) Pricet
Broadband Internet connection normalized price. The price is the monthly rental charges for 1Mbit/s bitrate. Upload and download bitrates are added to get the total bitrate. Nonrecurring charges are discounted over 3 years and added to the price. Data for years 1998-2000 refer to leased-lines rental price (source: European Commission, Directorate General for Information Society, Telecommunication Tariffs Report, 2000). Data over the period 2001-2004 refer to ADSL connection (source: European Commission, Directorate General for Information Society, Annual Report on Internet Access Costs, 2001, 2002, 2004). Missing data have been calculated basing on the same methodology (source: www.tariffe.it)
Expected Price Changet
Pricet+1 - Pricet, where PricetLVGH¿QHGDVDERYH
&RQWUROYDULDEOHV Sectoral Adoptiont
Nadopt,t/ N¿UPW, where •
for Broadband Adoption, Nadopt,t is the expected within industry cumulated number of
•
for Application Count, Nadopt,t is the expected within industry cumulated number of adopted
adopters and N¿UPWLVWKHZLWKLQLQGXVWU\QXPEHURI¿UPV applications, and N¿UPWLVWKHZLWKLQLQGXVWU\QXPEHURI¿UPVDGRSWLQJDEURDGEDQG connection Sectoral Adoption Changet
(Nadopt,t+1 - Nadopt,t)/ r, where Nadopt,tLVGH¿QHGDVDERYHDQGULVWKHLQWHUHVWUDWH
Geographical Adoptiont
Nadopt,t/ N¿UPW, where •
for Broadband Adoption, Nadopt,t is the expected within region cumulated number of adopters
•
for Application Count, Nadopt,t is the expected within region cumulated number of adopted
and N¿UPWLVWKHZLWKLQLQGXVWU\QXPEHURI¿UPV applications, and N¿UPWLVWKHZLWKLQLQGXVWU\QXPEHURI¿UPVDGRSWLQJDEURDGEDQG connection Year
Value of the index measuring the year: 1=1998, 2=1999, …. 8=2005.
Industry Dummies
7 Industries Dummies: 6HFWRU 2QHIRU6FLHQFH%DVHGPDQXIDFWXULQJ¿UPV 6HFWRU 2QHIRU6FDOH,QWHQVLYHPDQXIDFWXULQJ¿UPV 6HFWRU 2QHIRU6SHFLDOL]HG6XSSOLHUPDQXIDFWXULQJ¿UPV 6HFWRU 2QHIRU7UDGLWLRQDOPDQXIDFWXULQJ¿UPV 6HFWRU 2QHIRU8WLOLWLHVDQG&RQVWUXFWLRQ¿UPV 6HFWRU 2QHIRU7UDGH¿UPV 6HFWRU 2QHIRU2WKHU6HUYLFHV¿UPV
473
Broadband Access and Broadband-Based Applications
the average employee salary (Salaries/Employeest-1), and the ratio between value added and the number of employees (Value Added/Employeest-1). The former two variables capture “IT-familiarity” DQGWKHTXDOLW\OHYHORI¿UPZRUNIRUFHUHVSHFWLYHO\ 2QRQHKDQGZHFODLPWKDW\RXQJHU¿UPVWHQGWR select younger employees, which are more likely to possess good IT knowledge. On the other hand, SMEs characterized by a higher average employee VDODU\DUHOLNHO\WRKDYHHPSOR\HGKLJKHUTXDOL¿HG personnel. The variable Value Added/Employeest-1 PHDVXULQJWKHUDWLREHWZHHQWKH\HDUO\¿UPYDOXH added and the number of employees is a proxy of WKHHI¿FLHQF\OHYHOUHDFKHGE\WKH60()LQDOO\ the variable Cash Flow/Total Assetst-1 is a measXUH RI WKH DYDLODELOLW\ RI ¿QDQFLDO IXQGV ZKLOH Employees Growtht-1LVDQRWKHULQGLFDWRURI¿UP
performance and inversely proxies the degree of competitive pressure faced by SMEs. $VWRORFDWLRQVSHFL¿FYDULDEOHVWKHYDULDEOHV Employee Age and Employee Education capture local labor market characteristics concerning the level of human capital of workforce. Average Incomet and Telecommunications Network catch the overall socioeconomic conditions and the quality level of the telecommunications infrastructure, respectively, of the area on which SMEs are located. A geographical dummy (South) is included in order WRFRQWUROIRUWKH¿UP¶VGHFLVLRQWRORFDWHLQWKH South of Italy, which represents the most economic disadvantaged area in the country. 7LPHVSHFL¿F YDULDEOHV DUH XVHG LQ WKH VXUvival data analysis model. In particular, Pricet and Expected Price Changet represent the hedonic broadband price and the expected price variation
Table 2. Determinants of SMEs adoption of broadband connection 9DULDEOH
&RHI¿FLHQW
Į1
Employeest-1
Į2
Group
0.257 (0.120)**
Į3
Multiplant
0.412 (0.111)***
0.060 (0.015)***
Į4
Employees Growtht-1
0.001 (0.001)
Į5
Aget
0.001 (0.001)
Į6
Value Added/ Employeest-1
0.001 (0.001)
Į7
Salaries/ Employeest-1
-0.001 (0.001)
Į8
Cash Flow/ Total Assetst-1
-0.060 (0.115)
Į9
Employee Age
-0.082 (0.347)
Į10
Employee Education
0.151 (0.594)
Į11
South
-0.001 (0.222)
Į12
Average Incomet
-0.042 (0.082)
Į13
Telecommunications Network
Į14
Pricet
Į15
Expected Price Changet Log-likelihood
0.002 (0.001)* -0.080 (0.039)** -0.002 (0.001) -3075.31
/HJHQG 6LJQL¿FDQFHOHYHOJUHDWHUWKDQ
6LJQL¿FDQFHOHYHOJUHDWHUWKDQ
6LJQL¿FDQFH level greater than 99%. Robust standard errors and number of restrictions in parentheses. Cox proporWLRQDOKD]DUGVPRGHO%UHVORZPHWKRGIRUWLHV1XPEHURI¿UPV QXPEHURIREVHUYDWLRQV &RQWUROYDULDEOHVFRHI¿FLHQWVDUHRPLWWHGIRUVDNHRIV\QWKHVLV
474
Broadband Access and Broadband-Based Applications
Table 3. Determinants of SMEs adoption of broadband-based applications 9DULDEOH
&RHI¿FLHQW
Į0
Constant
Į1
Year
0.504 (0.022)***
-0.431 (3.565)
Į2
Employeest-1
0.536 (0.091)***
Į3
Group
1.162 (0.212)***
Į4
Multiplant
1.025 (0.187)***
Į5
Employees Growtht-1
-0.011 (0.004)**
Į6
Aget
-0.014 (0.006)**
Į7
Value Added/ Employeest-1
0.247 (0.103)**
Į8
Salaries/ Employeest-1
Į9
Cash Flow/ Total Assetst-1
Į10
Employee Age
0.006 (0.006) 0.361 (0.491) -1.868 (0.684)***
Į11
Employee Education
2.746 (1.224)**
Į12
South
1.029 (0.450)**
Į13
Average Incomet
Į14
Telecommunications Network R2 =
2.241 (0.867)** -0.447 (0.167)*** 0.24
/HJHQG 6LJQL¿FDQFHOHYHOJUHDWHUWKDQ
6LJQL¿FDQFHOHYHOJUHDWHUWKDQ
6LJQL¿FDQFH level greater than 99%. Robust standard errors in parentheses. Random effects panel data model. NumEHURI¿UPV QXPEHURIREVHUYDWLRQV &RQWUROYDULDEOHVFRHI¿FLHQWVDUHRPLWWHGIRUVDNHRI synthesis.
over time. Finally, we add to models a set of control variables. They include Sectoral Adoptiont , Sectoral Adoption Changet and Geographical Adoptiont which provide respectively a measure of the within industry diffusion, of the expected change of the within industry diffusion in the interval (t; t+1) and of the within regions diffusion. Moreover, we control for industry (Industry Dummies) in all models and for time of adoption of broadband connection (Year) only in the broadband-based applications model.
Results The results of the econometric analysis are illustrated in Tables 2 and 3. In Table 2, we present the estimates of the survival data analysis model of the hazard rate of adopting broadband connec-
tion. Table 3 reports the estimates of the random effects panel data model on the determinants of adoption of broadband-based applications. Let us ¿UVWIRFXVDWWHQWLRQRQWKHDGRSWLRQRIEURDGEDQG connection by Italian SMEs (see Table 2). The key determinants of adoption appear to be variables WKDWUHÀHFWWKHQHFHVVLW\E\60(VWRFRPPXQLFDWH both Employeest-1 and Multiplant have a positive DQGVLJQL¿FDQWDW LPSDFWRQWKHKD]DUGUDWH WKHFRHI¿FLHQWRIGroupLVSRVLWLYHDQGVLJQL¿FDQW at 95%. Therefore, larger SMEs with a multiplant structure and belonging to a business group are those most likely to be adopters. Among locaWLRQVSHFL¿F FKDUDFWHULVWLFV WKH TXDOLW\ OHYHO RI the telecommunications infrastructure is the only VLJQL¿FDQW GHWHUPLQDQW RI DGRSWLRQ DOWKRXJK DW RQO\ VLJQL¿FDQFH OHYHO &RQYHUVHO\ WKH decision to adopt broadband connection does not
475
Broadband Access and Broadband-Based Applications
seem to be affected by the socioeconomic welfare RIWKHDUHDRQZKLFK¿UPVDUHORFDWHGDQGE\ORFDO labor market conditions in terms of human capital possessed by workforce. Quite unsurprisingly, diffusion of broadband connection is found to be driven by the decline over time of the (hedonic) price of broadband connection. Results on the determinants of adoption of broadband-based applications, conditional to connection (see Table 3), reveal that larger SMEs with a multiplant structure and belonging to a business group are likely not only to be early adopters but also to use more applications: the estimated coef¿FLHQWV RI Employeest-1, Multiplant and Group DUHSRVLWLYHDQGVLJQL¿FDQWDW +RZHYHUWKH analysis reveals that many other factors that were IRXQGWRQRWVLJQL¿FDQWO\DIIHFWEURDGEDQGFRQQHFWLRQDUHLPSRUWDQWGHWHUPLQDQWVRIWKH60(V¶ use of broadband-based applications. In particular, everything else being equal, the number of DSSOLFDWLRQVLVIRXQGWRGHFUHDVHZLWK¿UPV¶DJH DQGWRLQFUHDVHZLWKWKHHI¿FLHQF\OHYHO SUR[LHG by the ratio between value added and number of HPSOR\HHV UHDFKHGE\WKH¿UP7KHQHJDWLYHDVVRFLDWLRQEHWZHHQ¿UPV¶DJHDQGEURDGEDQGEDVHG DSSOLFDWLRQVDGRSWLRQUHYHDOVWKDW\RXQJHU¿UPV often hiring younger people, are more likely to possess in-house valuable IT knowledge that lead them to more extensively use broadband-based applications. The importance of IT competencies LVFRQ¿UPHGE\WKHUHVXOWVFRQFHUQLQJORFDWLRQ VSHFL¿FYDULDEOHV,QSDUWLFXODUDSSOLFDWLRQVXVH E\60(VLVSRVLWLYHO\LQÀXHQFHGE\ORFDWLRQLQ wealthy geographic areas characterized by a labor market with a predominance of young and highly educated workforce: Average Incomet has a posiWLYHDQGVLJQL¿FDQWDW LPSDFWRQWKHQXPEHU RIEURDGEDQGEDVHGDSSOLFDWLRQV WKHFRHI¿FLHQWV of Employee Age and Employee Education are QHJDWLYHDQGSRVLWLYHDQGERWKVLJQL¿FDQWDW VLJQL¿FDQFHOHYHO4XLWHLQWHUHVWLQJO\WKHTelecommunications Network variable and the dummy variable SouthKDYHDVWDWLVWLFDOO\VLJQL¿FDQWLPSDFWRQ the number of applications used by SMEs (99% and 95%, respectively); the former is negative and the ODWWHULVSRVLWLYH$W¿UVWJODQFHWKHVHODWWHU¿QGLQJV may seem counterintuitive if we compare them to
476
those resulting from the analysis on broadband connection, where we highlighted that a high quality level of telecommunications infrastructure helps broadband connection diffusion. In our views, the following line of reasoning may be applied in order to explain such evidence. On the one hand, ¿UPVORFDWHGLQHFRQRPLFGLVDGYDQWDJHGDQGOHVV LQIUDVWUXFWXUHGDUHDVPD\¿QGVHULRXVREVWDFOHV in accessing broadband connection due to scarcity of suppliers and a possible low quality level of the delivered service. Clearly, this may negatively inÀXHQFHWKHLUEURDGEDQd connection behavior.5 On WKHRWKHUKDQGWKH¿UP¶VRSSRUWXQLW\FRVWWRIDFH DQG RYHUFRPH WKHVH GLI¿FXOWLHV GHFUHDVHV DORQJ ZLWKWKHQXPEHURIDSSOLFDWLRQVWKH¿UPGHFLGHV to adopt. This means that once a SME located in less-equipped areas acquires broadband access, LWZLOOGRLWIRUXVLQJDODUJHQXPEHURIVSHFL¿F applications. Finally the analysis highlights that diffusion of broadband-based applications has EHHQLQFUHDVLQJRYHUWLPH WKHFRHI¿FLHQWRIYear LVSRVLWLYHDQGVLJQL¿FDQWDW
CONCLUSION AND FUTURE TRENDS This study is an econometric investigation of the determinants of (a) the diffusion speed of broadband connection, and (b) the number of adopted broadband-based applications among Italian SMEs. For this purpose we have taken advantage of a new longitudinal dataset composed of 904 Italian SMEs (i.e., number of employees comprised between 10 and 249), which operate in both manufacturing and service sectors (excluding SXEOLF DGPLQLVWUDWLRQ ¿QDQFH DQG LQVXUDQFH The sample has been developed by ThinkTel in LW LV VWUDWL¿HG E\ LQGXVWU\ VL]H FODVV DQG geographical area so as to be representative of the Italian population of SMEs, and it contains detailed survey-based information about adoption of broadband connection and broadband-based applications over the period from 1998 to 2005. Data provided by the Thinktel survey has been VXSSOHPHQWHGZLWK¿UPV¶¿QDQFLDOGDWDDQGORFD-
Broadband Access and Broadband-Based Applications
WLRQVSHFL¿FGDWDFROOHFWHGIURPRWKHUSXEOLFDQG private sources. The results of the econometric analysis on broadband connection highlight that, quite unsurprisingly, diffusion is driven by the decline over time of the (hedonic) price of broadband connection. Moreover, key determinants of adoption are structural characteristics of SMEs that affect their need to communicate. Larger SMEs that have a multiplant structure and belong to a business group are those most likely to be early adopters. Among ORFDWLRQVSHFL¿FFKDUDFWHULVWLFVWKHGHYHORSPHQW of the telecommunications network is the only VLJQL¿FDQWGHWHUPLQDQWRIDGRSWLRQ&RQYHUVHO\ DOO YDULDEOHV WKDW SUR[\ ¿UP¶V ³,7 IDPLOLDULW\´ LWV HI¿FLHQF\ DQG SHUIRUPDQFH WXUQ RXW WR EH LQVLJQL¿FDQW GHWHUPLQDQWV RI 60(V¶ EURDGEDQG connection. Turning attention to broadband-based applicaWLRQVZH¿QGWKDW¿UPV¶VWUXFWXUDOFKDUDFWHULVWLFV have the same effects as those shown above. However, there are several factors which are found WRVXEVWDQWLDOO\LQÀXHQFH60(V¶DGRSWLRQRIDSplications, though they do not affect adoption of broadband connection. In particular, the adoption of applications turns out to be higher for younger DQGPRUHHI¿FLHQW¿UPVDVZHOODVIRU¿UPVZLWK high-quality personnel. In addition, the adoption of broadband applications is driven by locationVSHFL¿FFKDUDFWHULVWLFVWKDWUHÀHFWWKHDYDLODELOLW\LQ the local labor market of younger and more skilled individuals. Lastly, a high degree of competitive SUHVVXUHLVIRXQGWRH[HUWDSRVLWLYHHIIHFWRQ¿UPV¶ decision to adopt broadband applications. If we are conscious that our empirical analysis LVRQO\D¿UVWVWHSWRZDUGVDIXOOXQGHUVWDQGLQJRI WKHGHFLVLRQRI¿UPVWRDFFHVVDQGWRXVHEURDGband-based applications and other investigations are needed on the issue, nonetheless we think that WKH¿QGLQJVLOOXVWUDWHGDERYHPD\DOUHDG\VXJJHVW LPSRUWDQWUHÀHFWLRQVIRUERWKEURDGEDQGWHFKQRORgies suppliers and policy makers. As to broadband applications suppliers, they VKRXOG ¿OO WKH VNLOOV DQG FRPSHWHQFLHV JDS RI potential adopters if they want to promote the demand for broadband applications from SMEs. In this respect, the ability to create value for the
customer through customer care, consulting, and training activities may play a key role. In so far as applications suppliers are capable not only to sell products but also to offer to customers those competencies needed to properly use and fully exploit products, this business strategy may exert an LPSRUWDQWDQGVLJQL¿FDQWHIIHFWRQWKHZLOOLQJQHVV of potential adopters to purchase broadband-based applications (and in turn broadband connection). As to policy makers, subsidies to the adoption of broadband connection should be accompanied by more structural and medium-long time horizon policy interventions. First of all, the quality level of telecommunications infrastructure turns out to EHDVLJQL¿FDQWGHWHUPLQDQWRIEURDGEDQGDGRStion, allegedly policy effort should be directed to re-balance the IT infrastructural divide across the country. Secondly, policy makers should target and WU\WR¿OOWKH,7VNLOOVDQGFRPSHWHQFLHVJDSSRWHQtially suffered from SMEs, which in turn prevents adoption of (advanced) broadband applications. In the medium term there is the need for policy schemes favoring employees training activities, the purchase of other supporting services, and the recruitment of skilled personnel. Taking a longer WHUP YLHZ WKH UHVXOWV RI WKLV UHVHDUFK FRQ¿UP that investments in human capital play a crucial role for economic development. While making it easier for Italian SMEs to adopt broadband-based applications, they may enable them to increase WKHLUHI¿FLHQF\LQQRYDWLYHQHVVDQGFRPSHWLWLYHness in international markets. Finally, according to RXU¿QGLQJVLQGXVWULDOSROLF\PHDVXUHVDLPLQJDW UDLVLQJ¿UPVL]HPD\DOVREHKHOSIXOIRULQFUHDVing broadband penetration rate within the Italian industrial system. This study on the determinants of broadband connection and adoption of complementary applications among SMEs raises many new questions for future research. First of all, we (as most of the extant empirical literature on the topic) have not considered possible determinants that may hinder D¿UP¶VZLOOLQJQHVVWRXVHEURDGEDQGWHFKQRORJLHV VXFK DV VHFXULW\ LVVXHV DQG PDQDJHPHQW¶V concerns about a possible increase through the use of the new technology of unproductive activities by employees. Absence of this information may help
477
Broadband Access and Broadband-Based Applications
justify the relatively low amount of total variance explained by our models. Then, it would also be interesting to investigate not only the number of applications adopted by the organizations but also the “intensity” of adoption of these applications ZLWKLQ¿UPV)LQDOO\WKLVLVDUHVHDUFKRQDFFHVV to broadband technologies. The analysis on the effects of broadband connection and applications on ¿UPSHUIRUPDQFHV LHSURGXFWLYLW\DQGJURZWK will be even more interesting. In this respect, since broadband connection is an enabling and general purpose technology, we do not expect it to have any GLUHFWHIIHFWRQD¿UPSHUIRUPDQFH)XUWKHUPRUH the adoption of broadband-based applications is likely to have a positive effect on productivity and JURZWKEXWWKLVLVFRQGLWLRQHGRQWKH¿UP¶VDELOLW\ WRHI¿FLHQWO\XVHWKHDSSOLFDWLRQVWKHPVHOYHV7KH productivity rise caused by applications may pass WKURXJKERWKDQLQFUHDVHLQD¿UP¶VHI¿FLHQF\LQ already existing activities and also the possibility to develop new operations that bring value to the ¿UP,WIROORZVWKDWHVSHFLDOO\IRUZKDWFRQFHUQV more advanced complementary applications, we would expect that much of their potential could only be exploited if managerial and organizational FKDQJHVWRRNSODFHLQWKHDGRSWLQJ¿UP IRUDVLPLODU argument see, among others, Bertschek & Kaiser, 2004; Brynjolfsson & Hitt, 2000).
ACKNOWLEDGMENT Financial support from Thinktel (International Think Tank on Telecommunications) is gratefully acknowledged. The authors are also grateful for useful comments and suggestions to Thinktel committee components.
REFERENCES Arbore, A., & Ordanini, A. (2006). Broadband divide among SMEs: The role of size, location and outsourcing strategies. International Small Business Journal, 24, 83-99.
478
Aron, D.J., & Burnstein, D. (2003). Broadband adoption in the United States: An empirical analysis. Retrieved July 13, 2007, from http://ssrn.com/ abstract=386100 or DOI: 10.2139/ssrn.386100 Becchetti, L., & Trovato, G. (2002). The determinaWLRQVRIJURZWKIRUVPDOODQGPHGLXPVL]HG¿UPV 7KH UROH RI WKH DYDLODELOLW\ RI H[WHUQDO ¿QDQFH Small Business Economics, 19, 291-306. Bertschek, I., & Kaiser, U. (2004). Productivity effects of organizational change: Microeconometric evidence. Management Science, 394-404. Bresnahan, T.F., & Trajtenberg, M. (1995). General purpose technologies “engine of growth”? (NBER Working Paper, W4148). Brynjolfsson, E., & Hitt, L.M. (2000). Beyond computation: Information technology, organizational transformation and business performance. Journal of Economic Perspective, 14(4), 23-48. Cameron, A.C., & Trivedi, P.K. (2005). Microeconometrics: Methods and applications. New York: Cambridge University Press. Cava-Ferreruela, I., & Alabau-Munoz, A. (2004). Key constraints and drivers for broadband development. A cross-national empirical analysis. Paper presented at the 15th European Regional Conference of the International Telecommunications Society, Berlin, Germany. Colombo, M.G., & Grilli, L. (2007). Funding gaps? Access to bank loans by high-tech start-ups. Small Business Economics. Cox, R.D. (1972). Regression models and life tables. Journal of the Royal Statistical Society, 34, 187-220. Fagiolo, G., & Luzzi, A. (2006). Do liquidity conVWUDLQWVPDWWHULQH[SODLQLQJ¿UPVL]HDQGJURZWK" Some evidence from the Italian manufacturing industry. ,QGXVWULDODQG&RUSRUDWH&KDQJH(1), 1-39. FCC. (2000). Deployment of advanced telecommunications capability: Second report. Federal Communications Commission, 00-230. Washington, DC. Available at http://www.fcc.gov/broadband
Broadband Access and Broadband-Based Applications
Forman, C. (2005). The corporate digital divide: Determinants of Internet adoption. Management Science, (4), 641-654. Garcia-Murillo, M. (2005). International broadband deployment: The impact of unbundling. Communications & Strategies, (1), 1-21. Greene, W.H. (2000). Econometric analysis (4th ed.). Prentice Hall International. Karshenas, M., & Stoneman, P. (1993). Rank, stock, order and epidemic effects in the diffusion of new process technologies: An empirical model. The RAND Journal of Economics, 24(4), 503-528. Kim, J.H., Bauer, J.M., & Wildman, S.S. (2003). Broadband uptake in OECD countries. Paper presented at the 31st Conference on Communication, Information and Internet Policy, Arlington, Virginia. Mehrtens, J., Cragg, P.B., & Mills, A.M. (2001). A model of Internet adoption by SMEs. Information & Management, 39(3), 165-176. Nelson, R., & Winter, S. (1983). An evolutionary theory of economic change. Cambridge, MA: Harvard University Press. OECD. (2002). Broadband access for business. In Working Party on Telecommunication and Information Services Policies. Organization for Economic Cooperation and Development (OECD). OECD. (2003). Broadband driving growth: Policy responses. In Working Party on Telecommunication and Information Services Policies. Organization for Economic Cooperation and Development (OECD). Preissl, B. (1995). Strategic use of communication technology diffusion processes in networks and environments. Information Economics and Policy, 7(1), 75-99. Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363.
Rogers, E.M. (1995). Diffusion of innovations. New York: Free Press. Stoneman, P. (2001). The economics of technical diffusion. UK: Blackwell. Tookey, A., Whalley, J., & Howick, S. (2006). Broadband diffusion in remote and rural Scotland. Telecommunications Policy, 30, 481-495.
KEY TERMS Broadband-Based Applications: Software and service applications complementary to an Internet broadband connection. They range from very basic (e.g., e-mail) to advanced applications (e.g., supply chain management system). These latter applications require broadband to work properly. Broadband Internet Connection: Internet wired connection via ADSL or other dedicated lines with an upstream speed higher of equal to 256 Kbps. Competitive Pressure: Degree of competition IDFHGE\D¿UPZKLFKLVUHÀHFWHGLQLWVEXVLQHVV performance. Diffusion of Innovation (DOI) Theory: Theory which aims to describe the patterns of adoption of innovations, explain the underlying dynamics and assist in predicting whether and how a new invention will be successful. Econometric Model of Adoption: A statistical technique that on the basis of economic theory estimates the determinants of adoption of a new product or process. Firm’s Structural “Need to Communicate”: 6WUXFWXUDOFKDUDFWHULVWLFVRID¿UP LHPXOWLSODQW RUDI¿OLDWLRQWRDJURXS WKDWOHDGWRD³QDWXUDO´ increase in its need to communicate with external parties. General Purpose Technology: A highly pervasive technology which gets better over time, keeps lowering the costs of its users, enables the
479
Broadband Access and Broadband-Based Applications
development of other activities, makes it easier to invent and produce new complementary product or processes. 2
IT Familiarity: Amount of competencies and capabilities in information technology practices possessed by individuals. Small and Medium Size Enterprise: Number of employees comprises between 10 and 249.
3
4
5
ENDNOTES 1
480
The analysis of the adoption of broadband applications is also important with regard to the “killer-application” theory. Factors hindering adoption of these applications negatively af-
fect the demand for applications themselves, and consequently the demand for broadband connection (Aron & Burnstein, 2003). See Stoneman (2001) for a wider perspective on the reasons why current use of a technology may encourage further use. A complete description of collected information is presented in Table 1. For a comprehensive treatment of survival analysis, see Cameron and Trivedi (2005). See again results of Table 2, where the coef¿FLHQWRIWKHYDULDEOHSouth was positive but VWDWLVWLFDOO\ LQVLJQL¿FDQW DQG WKH YDULDEOH Telecommunications Network was found to negative affect (albeit at 90%) the probability of SMEs to adopt broadband connection.
481
Chapter XXX
Broadband Diffusion to SMEs in the UK Oluwasola Oni Brunel University, UK Anastasia Papazafeiropoulou Brunel University, UK
ABSTRACT Broadband is a relatively new technology and its adoption in the United Kingdom has been an issue due WRLWVSHUFHLYHGEHQH¿WVIRUEXVLQHVVHVDQGPRUHVRIRUsmall/medium size enterprises (SMEs). In this chapter we argue that previous research focuses on home uses of broadband, particularly for educational purposes with little attention to its adoption by SMEs. We argue that the existing diffusion of innovation theories are inadequate for the study of broadband diffusion and we propose a more sociotechnical approach for that purpose. This study can be useful for SMEs considering adoption of new technologies such as broadband as well as policy makers that seek to apply effective technological adoption policies.
INTRODUCTION 7KH JRYHUQPHQW¶V WDUJHW ZDV IRU WKH 8QLWHG Kingdom (UK) to have the most extensive and competitive market in the G7 by the year 2005 (Ofcom, 2004). The previous adoption, especially within the SME community did not match up with these expectations. According to Ofcom (2004), 68% of UK small/medium size enterprises (SMEs) were connected to the Internet, of which 37% used broadband, 65% used narrowband, 23% used integrated services digital network (ISDN), 32% used a narrowband unmetered service, 21% used a narrowband metered service, and 6% were unsure
of what type of narrowband connection they had. 0RUHUHFHQWO\KRZHYHUWKH¿JXUHVKDYHFKDQJHG dramatically. According to Ofcom (2006), in the last quarter of 2005, 84% of SMEs use the Internet, of which 73% use broadband. The numbers mentioned above provide information on the number of SMEs which have adopted broadband. However, there is no research on factors effecting broadband adoption or whether SMEs use broadband differently than dial up connections. Previous research focuses on home uses of broadband, particularly for educational purposes with little attention to the adoption by SMEs.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband Diffusion to SMEs in the UK
There has been provision of government initiatives aimed at improving broadband adoption by SMEs. In Fife and Pereira (2002), the authors highlight the possibility that government subsidies are an effective way to increase broadband takeXS E\ 60(V $GGLWLRQDOO\ VSHFL¿F JRYHUQPHQW incentives have been made available such that the price of broadband for SMEs should be less of a constraint to take-up than for households (Affuso & Waverman, 2002). The Edinburgh Parallel Computing Center (EPCC, 2002), similarly observed that the UK government sees broadband as the “next leap forward for the Internet” and H[SHFWHG60(VLQSDUWLFXODUWREHQH¿WIURPWKH technology. While we are not assuming that all 60(VZLOOEHQH¿WIURPEURDGEDQG (3&& ZHKLJKOLJKWVRPHSRVVLEOHEHQH¿WVIRU60(VWKDW FKRRVHWRWDNHXSEURDGEDQG2QHRIWKHEHQH¿WVLV the ability to trade and conduct business electronically at a faster rate. Affuso and Waverman (2002) stated that SMEs could potentially experience large transactions costs savings and productivity improvements should they choose to adopt broadEDQG3RWHQWLDOEHQH¿WVRIEURDGEDQGWR60(VDQG WKHJRYHUQPHQW¶VLQWHUHVWLQSURPRWLQJLWPDNHV it worthwhile to look into the different ways that EURDGEDQGFDQEHEHQH¿FLDOWRWKHPDQGIDFWRUV LQÀXHQFLQJLWVHIIHFWLYHXSWDNH As mentioned earlier, previous research focuses on home uses of broadband particularly for educational purposes with little attention to the adoption by SMEs. In addition to the fact that little research has been done on broadband and SMEs, Lee, Sawyer, and Choudrie (2003) highlight the fact that there is little research on how SMEs can adopt new technologies at a faster rate and get full advantage of them. Therefore, there is a need to look at theories that can analyze the phenomenon. The aim of the chapter is to propose a framework for diffusion of broadband to SMEs. We start by looking at the innovation diffusion theory. In previous studies of diffusion of technologies, the innovation diffusion theory as presented by Rogers (1995) has been widely used to understand the reasons behind adoption of innovations. We argue that the diffusion theories are one-sided trying to identify characteristics of the innovation that would
482
make the users adopt the innovation while the perceptions of other stakeholders are not taken into consideration. Thus, we examine the use of social construction of technology (SCOT) as a way to help us examine the different viewpoints involved in the innovation diffusion process. We then propose a framework for the examination of these viewpoints in the case of broadband adoption.
BROADBAND AND SMES 7KHWHUPEURDGEDQGKDVQRHVWDEOLVKHGGH¿QLWLRQ and varies from country to country (Firth & Kelly, 2001). Broadband technology is an umbrella term which covers varying high-speed access technologies including asymmetric digital subscriber line $'6/ FDEOHPRGHPVVDWHOOLWHDQGZLUHOHVV¿[HG (Wi-Fi) Networks. Basically, broadband provides an “always on” and faster Internet connection than dial up. It is an emerging technology that promises to improve Internet use. Previous research has highlighted the lack of content of broadband technology. The slow adoption of broadband has been related to the fact that “killer applications” that can boost the use of broadband are yet to be developed (Carlyle, 2002; Heinzl, 2001; Lessig, 2002). Middleton (2003), in contrast, argues that consumers today gradually ¿QGYDOXHLQEURDGEDQGQHWZRUNVDVWKH\DUHFXUrently deployed. Additionally, a substantial amount of research has been carried out in relation to pricing for broadband services (Clark, 1997; Courcoubetis, Kelly, Siris, & Weber, 1998; Falkner, Devetsikiotis, & Lambadaris, 2000; Yaïche, Mazumdar, & Rosenberg, 2000). The cost of obtaining and maintaining broadband could be a hindrance to its adoption, as it is slightly more expensive than narrowband (Zhang, 2002). This was taken into consideration in South Korea where the government recognized that in order to be successful, broadband access needed to be priced at affordable levels for middle-income households (Choudrie & Lee, 2004). While the issue of cost might be a KLQGUDQFHWKHUHDUHPDQ\EHQH¿WVWKDWKDYHEHHQ related to the adoption and use of broadband, such as productivity (Lee, 2002). Similarly, according
Broadband Diffusion to SMEs in the UK
to the organization for economic cooperation and development (OECD, 2003), small/medium size enterprises (SMEs) in particular could experience growth from an increase in productivity resulting IURP JUHDWHU SURFHVV HI¿FLHQF\ DQG LPSURYHG information exchange. 7KH(XURSHDQ&RPPLVVLRQ GH¿QHVVPDOO and medium size enterprises (SMEs) as enterprises with fewer than 250 employees. Micro enterprises are enterprises with fewer than 10 employees while small enterprises have between 10 and 49 employees with an annual turnover and/or annual balancesheet total not exceeding 10 million euros. Medium enterprises on the other hand have fewer than 250 employees with an annual turnover not exceeding 50 million euros or an annual balance-sheet total not exceeding 43 million euros. SMEs vary in size and nature depending on their area of business. In the same way, their attitudes differ when it comes to the adoption of new technologies. In some cases, SMEs are keen to accept new technologies while in other situations, they are persuaded with the aid of government funding initiatives (European Commission, 2002). SMEs are generally considered to be sceptical when it comes to the adoption of new technologies. As a result of their nature and size, they are unable to spend large sums of money on adoption of technologies to support the running of their business. While SMEs are different in nature IURPODUJHFRPSDQLHVWKHUHDUHDOVRVLJQL¿FDQW differences among SMEs according to size and nature. The decision to adopt certain technologies might depend on these varying sizes and nature. However, Burke (2005) highlighted the fact that not too many researches have examined the size of SMEs as a factor in information systems use. 7KHVWXG\VKRZHGWKDWDVD¿UPJURZVLWFDQEH expected that the need for information systems will be recognized, thus signifying that growth rather than industry type is a more important indicator of ZKHWKHURUQRWD¿UPZLOODGRSWVXFKLQIRUPDWLRQ systems like Internet technologies.
SMES AND THE ADOPTION OF NEW TECHNOLOGIES There are many reasons why SMEs may decide to adopt new technologies. These reasons could be related to the desire for business growth (Levy, Powell, & Yetton, 2002). The reasons for information technology (IT), Internet, and e-commerce adoption by SMEs are complex and interrelated. It is motivated by a mix of management eagerness and the need for better communications (Chappell, Feindt, & Jeffcoate, 2002). The main barriers to IT and e-commerce adoption appear WREHWKHPDQDJHUV¶XQZLOOLQJQHVVWRWDNHULVNV when it comes to technological change (Kalakota & Robinson, 2001). As cited in Levy and Powell (1998) according to Lefebvre et al. (1990), new technology permits 60(VWREHFRPHPRUHÀH[LEOHJLYLQJWKHPWKH DELOLW\WRUHVSRQGPRUHUDSLGO\WRFXVWRPHUV¶QHHGV at the same time improving the quality and variety of their products and increasing production that is WDLORUHGWR¿W+RZHYHUDFFRUGLQJWRDVWXG\FDUULHG out by Levy and Powell, this was not found to be so. In their investigation of the role of information V\VWHPVLQSURPRWLQJRULQKLELWLQJÀH[LELOLW\LQ 60(VWKHLUVWXG\VKRZVWKDWWKH¿UPVDOOH[KLELW UHODWLYH LQÀH[LELOLW\ LQ WKHLU JHQHUDO DSSURDFK Furthermore, information systems do not seem to LQFUHDVHÀH[LELOLW\ UDWKHUWKH\UHLQIRUFHH[LVWLQJ thinking. The reason for this is that SMEs see themselves as having a narrow product range which is XVHGIRUWKHEHQH¿WRIRQHRUWZRFXVWRPHUV7KLV makes them consider the purchase of information V\VWHPVWRLPSURYHHI¿FLHQF\DQGHIIHFWLYHQHVV of current processes rather than considering the capacity of information technology to increase ÀH[LELOLW\DQGLPSURYHFRPSHWLWLYHQHVV 0RVWEDUULHUVWR60(V¶LQIRUPDWLRQWHFKQRORJ\ DGRSWLRQDVLGHQWL¿HGE\/DZVRQ$OFRFN&RRSHU and Burgess (2003), were nontechnical and they suggested some ways in which these barriers may be overcome such as government intervention and industry associations providing information to raise awareness, training, participation in the diffusion process, and working with good quality consultants.
483
Broadband Diffusion to SMEs in the UK
One of the major technological innovations WKDWFDQEHRIEHQH¿WWR60(VLVWKHXVHRIWKH Internet. However, while some SMEs have used DQG EHQH¿WHG IURP WKH XVH RI WKLV WHFKQRORJ\ some others are yet to experience the technology and so are not in a position to decide whether or QRWLWLVEHQH¿FLDOWRWKHLUEXVLQHVV7KLVEULQJVXV to the issue of how SMEs decide whether or not they need to adopt new technologies. As pointed RXWE\%HFNLQVDOHDQG/HY\ 60(V¶GHFLsion whether to adopt the Internet or not lies on WKHSHUFHLYHGEHQH¿WDQGUHODWLYHDGYDQWDJHWKH\ might get. While it might not be possible to reap WKHSUR¿WVRILQYHVWLQJLQQHZWHFKQRORJ\LPPHdiately, the long term effect could turn out to be LPPHQVHO\EHQH¿FLDOWRWKH60($VSRLQWHGRXW by Sandler and Boggs (2001), the use of the Internet could put SMEs on a platform where they are able to compete with larger companies. However, in Saban and Rau (2005) the authors suggest that resources are more of a deterrent to adoption rather than lack of knowledge. In the next section we propose a framework that can be used in the research of broadband adoption by SMEs.
A FRAMEWORK EXAMINING THE DIFFERENT VIEWPOINTS IN THE DIFFUSION OF BROADBAND 5RJHUV GH¿QHVLQQRYDWLRQDV³DQLGHDSUDFtice, or object perceived as new by an individual or other unit of adoption” and diffusion as “the process by which an innovation is communicated through certain channels over time among members of a social system.” Additionally, King, Gurbaxani, .UDHPHU0F)DUODQ5DPDQDQG
484
ing diffusion as a simple linear process (Kautz & Pries-Heje, 1996). It is argued that his view failed WRVXI¿FLHQWO\FRQVLGHUWKHUHODWLRQVKLSEHWZHHQ suppliers and adopters and an active participation of potential adopters in the diffusion process. Beynon-Davies and Williams (2003) highlight the fact that technological diffusion is usually portrayed as a rational process but postulate that diffusion process is similar to broader social movements. Finally, Lyytinen and Damsgaard (2001) showed in their observations that complex technologies will QRWQHFHVVDULO\GLIIXVHLQDVSHFL¿FRUGHU In order to understand and explain how a new technology is adopted, many studies have used the theory of diffusion of innovations developed E\5RJHUV$FFRUGLQJWRKLVZRUNDQLQGLYLGXDO¶V decision to adopt a new technology is a process that occurs over a period of time and consists of a series of actions and decisions rather than an instantaneous act. This theory implies that businesses would decide to adopt an innovation mainly because of its characteristics, thus missLQJRXWRWKHULQÀXHQFHVVXFKDVQDWXUHDQGVL]H of business and background of business owners. Additionally, different views, opinions, and agendas of various groups involved in the process of the adoption of an innovation are not adequately represented in the theory of diffusion of innovation (Papazafeiropoulou, Gandecha, & Stergioulas, 2005). In this chapter we take a similar stand and we try to identify the different viewpoints in the innovation diffusion process; we believe this way we may identify possible gaps in perceptions that may explain reasons for adoption or nonadoption. Thus we propose the use of social construction RIWHFKQRORJ\ 6&27 DVVSHFL¿HGE\3LQFKDQG Bijker (1984), where the role of the various social groups in the production and use of an innovation is of great importance. SCOT has its origin in the VRFLRORJ\RIVFLHQWL¿FNQRZOHGJH 66. With SCOT, the development of a technological artefact is described as alternating between variation and selection, thereby resulting in a multidirectional model that contrasts linear models (Pinch & Bijker, 1984). In SCOT where the introduction of an innovation is faced with objections from differHQWVRFLDOJURXSVRQHVHHNVWR¿QGDVROXWLRQWKDW
Broadband Diffusion to SMEs in the UK
would make the innovation desirable for reasons that would supersede reasons for objections. The point of SCOT is that the successful stages in the development of an artefact are not the only possible ones. Steps involved in SCOT include: • • • •
Identifying the relevant social groups Describing the groups in more detail Identifying the problems each of these groups has with respect to the artefact Identifying several variants of the solution around each of these problems
of the relative advantage would be dependent on the innovation in question (Rogers, 1995). According to Rogers, diffusion scholars found relative advantage to be one of the best predictors of an LQQRYDWLRQ¶VUDWHRIDGRSWLRQ Compatibility is described as the degree to which an innovation is seen as consistent with existing values, past experiences, and needs of the potential adopters. The innovation can either be compatible or incompatible with sociocultural values and beliefs, previously introduced ideas or client needs for the innovation. 1.
This way of describing the developmental process brings out all the different kinds of posVLEOHFRQÀLFWV7KLVPRGHOKLJKOLJKWVWKHPXOWLdirectional character of a technological artefact. 7KHLQWHUSUHWDWLYHÀH[LELOLW\RIDQDUWHIDFWPXVW be shown. In this case, the technological artefact is broadband and there are different social groups (SMEs inclusive) involved in its diffusion. In our model, we propose to extend the innovation diffusion model by identifying the different social groups involved in the diffusion of broadband as described in SCOT. We also use secondary recourses to obtain the views of the different social groups in relation to the technological characteristics of the innovation. Despite the many criticisms of the theory of diffusion of innovations, there are some aspects of the theory that are pertinent in explaining the rate of adoption of an innovation. These are known as the perceived attributes of an innovation. Individual perceptions of these attributes can help to predict the rate of adoption of an innovation (Rogers, 1995). The perceived attributes of an innovation that are important in explaining the rate of its adoption include: relative advantage, compatibility, complexity, trialability, and observability (Rogers). Relative advantage in the theory of diffusion of innovations is described as the degree to which an innovation is perceived as being better than the idea it supersedes. The degree of relative advantage LVXVXDOO\H[SUHVVHGLQWHUPVRIHFRQRPLFSUR¿WDELOLW\VRFLDOSUHVWLJHRURWKHUEHQH¿WV7KHQDWXUH
2.
3.
Sociocultural values and beliefs: an innovaWLRQ¶V LQFRPSDWLELOLW\ ZLWK FXOWXUDO YDOXHV can block its adoption Previously introduced ideas: compatibility of an innovation with a preceding idea can either speed up or retard its rate of adoption. “Old ideas are the main mental tools that individuals utilize to assess new ideas. One cannot deal with an innovation except on the basis of the familiar, with what is known.” Client needs for the innovation
Complexity is described as the degree to which DQLQQRYDWLRQLVSHUFHLYHGDVUHODWLYHO\GLI¿FXOWWR understand and use. Some innovations are quite clear in their meaning while others are not. Trialability is described as the degree to which an innovation may be experimented with on a limited basis. Observability is described the degree to which the results of an innovation are visible to others. This means that the results of some ideas are easily observed and described while other innovations are GLI¿FXOWWRREVHUYHDQGGHVFULEH 5RJHUV The innovation discussed here is broadband. It is constantly proclaimed to be a major improvePHQWDQGDVVXFKEHWWHUWKDQGLDOXS 2I¿FHRIWKH E-Envoy, 2001). Broadband is perceived as better than dial up because of its inherent features such as, speed, always on, and it does not interfere with WHOHSKRQH FRQQHFWLRQV 2I¿FH RI WKH ((QYR\ The most common reason given for consumers to adopt broadband is because it is better than dialup, which is the preceding innovation.
485
Broadband Diffusion to SMEs in the UK
Figure 1. A framework for the examination of the different viewpoints in broadband diffusion to SMEs
SMEs
Technical Complexity
Private Consultants
Independent Groups
Compatibility
Relative Advantage
Broadband Adoption
Government
Vendors
In this study, we will base our model on the information systems variance model (Agarwal & Prasad, 1998; Cooper & Zmud, 1990; Crum, Premkumar, & Ramamurthy, 1996) by examining compatibility, complexity, and relative advantage as the attributes that are of the most importance in information technology (IT) innovations. We nevertheless extend this model by identifying the various social groups involved in the production and use of the innovation. The social groups involved in the diffusion of broadband to SMEs include: the SMEs, the government (Choudrie, Papazafeiropoulou, & Lee, 2I¿FHRIWKH((QYR\ SURIHVVLRQDO associations (Intellect, 2003), independent bodies striving for broadband adoption (BSG, 2004), private consultants (Philpott, 2004), and vendors (BTOpenworld, 2002; Cisilion, 2002). The various social groups that have been idenWL¿HGDFFRUGLQJWRWKHVWDNHKROGHULGHQWL¿FDWLRQ process outlined by Pouloudi and Whitley (1997) are SMEs, vendors, government, private consultants, professional associations, and independent groups. In order to carry out this framework, the views of the various social groups will be exam-
486
Professional Associations
ined concerning the innovation diffusion attributes which are complexity, compatibility, and relative ease of use of broadband. In the next section we use secondary data to analyze the different viewpoints that shape the diffusion of broadband using the UK as an example of a dynamic broadband market.
Compatibility With regards to this attribute of broadband for 60(VYHQGRUVUHSRUWWKHIROORZLQJEHQH¿WVWKDW make broadband compatible with SMEs past experiences, current business practices, and future needs. They proclaim that broadband provides greater download and upload speeds and provides 60(VZLWKWKHÀH[LELOLW\WRFKDQJHH[LVWLQJSUDFtices, target new customers, increase productivity, drive changes in business processes, give greater independence, and allow them to take advantage of associated technologies (BTPlc, 2005; NTL, 2005). SMEs in the IT sector have an understanding of what they require from broadband services and how broadband is compatible with their business. Ac-
Broadband Diffusion to SMEs in the UK
cording to independent groups, this is not the case with SMEs in other sectors (Intellect, 2003). These SMEs do not know how broadband can support or enhance their business activities. They also do not know what broadband service is ideal for them and in some cases have been sold a service that might not be particularly suitable for their nature of business (Intellect, 2005). Private consultants and independent groups argue that there has not been any study or service to determine what kind of broadband service is suitable for SMEs in various sectors, since their communication and technological needs will vary (Arnott, 2005). However, vendors claim that most SMEs who have adopted broadband are quite happy with the service (BT Plc, 2005; Showcase05, 2005; Viatel, 2005;).
Complexity According to groups that promote broadband as well as vendors, broadband is portrayed as easy to LQVWDOODQGLQWHJUDWHLQWR60(V¶H[LVWLQJV\VWHPV (BSG, 2004; BTPlc, 2005). However, it has been UHSRUWHGE\DQLQGHSHQGHQWJURXSWKDW60(V¿QG LW GLI¿FXOW WR DGMXVW WR WKH FKDQJHV ZKHQ XVLQJ broadband (Intellect, 2003). In a study carried out on SMEs that were new to broadband, the time taken to achieve certain tasks increased with the use of broadband suggesting that they did not SDUWLFXODUO\ ¿QG LW HDV\ WR XVH ,QWHOOHFW Independent groups claim the vendors are doing little to educate SMEs on the need and proper use of broadband (Arnott, 2005; BSG, 2004). Furthermore, the organization for economic cooperation and development (OECD) recommended that the government as model users of broadband should demonstrate the potential of broadband services and content, providing demonstration and mentoring for SMEs (OECD, 2005). Broadband may not EH VXLWDEOH IRU VRPH 60(V¶ QDWXUH RI EXVLQHVV They lack the time, information, and knowledge to UHVHDUFKWKHEHQH¿WVRIQHZWHFKQRORJLHVDQGZLOO not know if they have the technological capability to handle any changes to their established business processes (OECD, 2005).
Relative Advantage To further stress the point that broadband is superior to the dial-up, according to Intellect (2003), dialups do not allow SMEs to enjoy the Internet fully. They do not allow SMEs to use the Internet to its full potential as a medium for e-business applications. SMEs fail to see this point because, according to Cisilion (2002), they do not see broadband as integral to helping their business. Independent groups infer that SMEs need to know exactly how EURDGEDQGDGRSWLRQZRXOGEHEHQH¿FLDOWRWKHPLQ monetary terms and would need more information from the government and vendors. All they have heard is the advantage in terms of speed and they are unwilling to adopt broadband on the basis of speed alone (OECD, 2005; Philpott, 2004). Vendors, KRZHYHUFODLPWKDWEURDGEDQGSURYLGHVHI¿FLHQW data back-up, video conferencing, and voice over ,QWHUQHWSURWRFRO 9R,3 ZKLFKSURYLGHVHI¿FLHQW and better communication than dial-up (BTPlc, 2005; Ntl, 2005; Viatel, 2005). It also helps to save money on calls and postage. There have been reports of SMEs susceptibility to spam, virus, hackers, and Internet fraud if they use broadband. Private consultants think this is possibly one reason why they have been reluctant to adopt broadband (Philpott, 2004). Finally, the government mentions growth opportunities as a result of the speed, content, and capabilities of broadband for SMEs that choose to adopt it (DTI, 2004)
CONCLUSION In this chapter we started by identifying the relevant social groups and have made an initial attempt to see their viewpoints. The UK government set out to have the most extensive and competitive broadband adoption in the G7 by 2005 and beyond and sees broadband as the next leap forward for the Internet (BSG, 2004). They expect SMEs in particular to EHQH¿W IURP LWV DGRSWLRQ $FFRUGLQJ WR 2IFRP (2006), more SMEs have adopted broadband in recent times. Nevertheless, there is room for improvement. We argued that despite its usefulness, the diffusion of innovations theory has inherent
487
Broadband Diffusion to SMEs in the UK
limitations; thus, we developed a model based on WKH ¿UVW WKUHH LQQRYDWLRQ DWWULEXWHV $JDUZDO Prasad, 1998; Cooper & Zmud, 1990; Crum et al., 1996). This model uses innovation attributes and takes into consideration the various social groups and views that are related to broadband uptake by SMEs using SCOT. %URDGEDQG KDV EHHQ VDLG WR EH EHQH¿FLDO WR companies and individuals. Communication is enhanced, various processes could be automated and carried out at a faster rate, and it could promote the image of the company (EPCC, 2002; Dutton et al., 2004). Broadband is also seen by governments of different countries as essential for achieving ecommerce goals and providing both social and HFRQRPLF EHQH¿WV WR WKHLU FLWL]HQV ,78 OECD, 2002). Current research in the adoption of broadband technology focuses on broadband services and government policies for home uses, while there is little emphasis on its adoption by SMEs. Focus in the research of broadband has been in various areas such as broadband content, pricing, government policies, educational, and HQWHUWDLQPHQWEHQH¿WV Broadband providers, the government, and other interest groups have continually stressed the importance of broadband but with little results where SMEs are concerned. There are many factors WKDW FRXOG LQÀXHQFH 60(V LQ GHFLGLQJ ZKHWKHU or not they would adopt a new technology such as broadband. In the proposed research, we have considered three main attributes of broadband and have looked at the different social groups and their views about these attributes. Future research will attempt to uncover the SRVVLEOHEHQH¿WV60(VFDQHQMR\ZLWKWKHXSWDNH of broadband and possibly demonstrate this in monetary terms. However, owing to the fact that SMEs and the nature of business carried out can EHTXLWHYDULHGWKHEHQH¿WVRIEURDGEDQGPD\RU may not apply to all SMEs. Different types and sizes of SMEs depending on the sector within which they operate have their unique needs in terms of technological support for their business processes. When determining the usefulness and reason for low broadband adoption to SMEs, these various characteristics will have to be considered.
488
Considering the usefulness and inherent limitations of the diffusion of innovations theory, we have developed a model which includes SCOT as a way to identify different views of broadband adoption. We analyzed the perceived attributes of broadband and the different views of the social groups involved. We argued that the perception gaps between different social groups could explain reasons for adoption RIQRQDGRSWLRQ2XULQLWLDO¿QGLQJVVKRZWKDWLQ the case of broadband adoption and diffusion it seems that there is no consensus as to whether the technology is useful, cheap, or convenient for SMEs to use. There are a number of assumptions taken by the vendors and the government about broadband adoption that seems to not be shared by SMEs. We believe that an in-depth investigation of the views and agendas of all relevant social groups can give us a clearer picture of broadband diffusion process. We have started by identifying the relevant social groups and have made an initial attempt to see their viewpoints. We intend to use primary data to apply this framework in the diffusion of broadband in the UK. We are planning D¿HOGZRUNZKHUHUHSUHVHQWDWLYHVRIWKHLGHQWL¿HG social groups will be asked to offer their opinions, which will be analyzed with the view to getting a better understanding of broadband diffusion. We believe that our research could contribute to practitioners and policy makers as well as researchers LQWKH¿HOGRILQQRYDWLRQGLIIXVLRQ
REFERENCES Affuso, L,. & Waverman, L. (2002). The impact of electronic infrastructure on productivity and growth - a report for the performance and innovation unit. Retrieved July 13, 2007, from http://www. SPJRYXN¿OHVSGI/%6SGI Agarwal, R,. & Prasad, J. (1998). A conceptual and RSHUDWLRQDOGH¿QLWLRQRISHUVRQDOLQQRYDWLYHQHVVLQ the domain of information technology. Information Systems Research, 9(2), 204-215. Arnott, S. (2005, October 26). UK climbs world table in broadband uptake. Computing.
Broadband Diffusion to SMEs in the UK
Beckinsale, M., & Levy, M. (2004). SMEs and Internet adoption strategy: Who do SMEs listen to? In Proceedings of the Twelfth European Conference on Information Systems Turku School of Economics and Business Administration, Turku, Finland. Beynon-Davies, P., & Williams, M.D. (2003). The diffusion of information systems development methods. Journal of Strategic Information Systems, 12, 29-46. BSG. (2004). Third Annual Report and Strategic Recommendations. Retrieved July 13, 2007, from http://www.intellectuk.org/publications/reports/ BSG_3rd_annual_report.pdf BTOpenworld. (2002). BT Openworld shows businesses the route to high-speed cost savings. Retrieved July 13, 2007, from http://www.cisilion. FRPEQHZVBEHQH¿WVKWP BTPlc. (2005). Broadband boost for small and medium business. Retrieved July 13, 2007, from http:// www.btplc.com/Innovation/Strategy/boost/ %XUNH. 7KHLPSDFWRI¿UPVL]HRQ,Qternet use in small businesses. Electronic Markets, (2), 79-93. Carlyle, P. (2002, March 28). Content is king in EDWWOHWR¿QGNLOOHUDSSOLFDWLRQThe Scotsman. Chappell, C., Feindt, S., & Jeffcoate, J. (2002). Best practice in SME adoption of e-commerce. Benchmarking: An International Journal, 9(2), 122-132. Choudrie, J., & Lee, H. (2004). Broadband development in South Korea: Institutional and cultural factors. European Journal of Information Systems, 13(2), 103. Choudrie, J., Papazafeiropoulou, A., & Lee, H. (2003). A Web of stakeholders and strategies: A case of broadband diffusion in South Korea. Journal of Information Technology, 18(4), 281-291. Cisilion. (2002). Small to medium businesses can’t VHHEURDGEDQGEHQH¿WVRetrieved July 13, 2007, from http://www.cisilion.com/bnews_intellect. htm
Clark, D. (1995). Internet cost allocation and pricing. In McKnight, L. & Bailey, J. (Eds.), Internet economics. Cambridge, MA: MIT Press. Cooper, R.B., & Zmud, R.W. (1990). Information technology implementation research: A technological diffusion approach. Management Science, 36(2), 123. Courcoubetis, C., Kelly, F.P., Siris, V.A., & Weber, R. (1998). A study of simple usage-based charging schemes for broadband networks. In Proceedings of the IFIP TC6/WG6.2 Fourth International Conference on Broadband Communications: The Future of Telecommunications. Crum, M.R., Premkumar, G., & Ramamurthy, K. (1996). An assessment of motor carrier adoption, use, and satisfaction with EDI. Transportation -RXUQDO(4), 44. DTI. (2004). Broadband business case. How broadband content can aid the delivery of regional and devolved administration economic strategies. Retrieved July 13, 2007, from http://www.dti.gov. uk/industries/digital_content/broadband-contentbusinesscase.pdf Dutton, W.H., Gillett, S.E., McKnight, L.W., & Peltu, M. (2004). Bridging broadband Internet GLYLGHV 5HFRQ¿JXULQJDFFHVVWRHQKDQFHFRPPXnicative power. Journal of Information Technology, 19(1), 28-38. European Commission. (2002, June 28). Benchmarking national and regional e-business policies for SMEs. Final report of the E-business Policy Group of the European Union, Brussels, Belgium. Elliot, S., & Loebbecke, C. (2000). Interactive, inter-organizational innovations in electronic commerce. Information Technology and People, 13(1), 46-66. E.P.C.C. (EPCC). (2002). SME broadband initiative, the annual review. Retrieved July 13, 2007, from http://www.epcc.ed.ac.uk/overview/publications/newsletters/EPCCnews/EPCCNews47.pdf
489
Broadband Diffusion to SMEs in the UK
European Commission. (2005). The new SME GH¿QLWLRQ 8VHU JXLGH DQG PRGHO GHFODUDWLRQ Brussels: European Commission.
Kautz, K., & Pries-Heje, J. (1996). Diffusion and adoption of information technology. London: Chapman & Hall.
Falkner, M., Devetsikiotis, M., & Lambadaris, I. (2000). An overview of pricing concepts for broadband IP networks. IEEE Communications Surveys, 3(2).
King, J., Gurbaxani, V., Kraemer, K., McFarlan, F., Raman, F., & Yap, F.W. (1994). Institutional factors in information technology innovation. Information 6\VWHPV5HVHDUFK(2), 139-169.
Fife, E., & Pereira, F. (2002). Socio-economic and cultural factors affecting adoption of broadband access: A cross-country analysis. Retrieved July 13, 2007, from http://www.marshall.usc.edu/ctm/ publications/FITCE2002.pdf
Kumar, S., & Swaminathan, J.M. (2003). Diffusion of innovations under supply constraints. OperaWLRQV5HVHDUFK(6), 866.
)LUWK/ .HOO\7 %URDGEDQGEULH¿QJ paper, ITU, Geneva. Retrieved July 13, 2007, from http://www.itu.int/osg/spu/ni/broadband/workVKRSEULH¿QJSDSHU¿QDOGRF Heinzl, M. (2001, June 14). Broadband carriers are hunting for killer apps. Wall Street Journal. ,QWHOOHFW 7KHEHQH¿WVRIEURDGEDQGWR60(V in the ICT sector. Interim conclusions from intellect member research. Retrieved July 13, 2007, from http://www.intellectuk.org/campaigns/broadband/ EHQH¿WV([HFBVXPPDU\B,QWHOOHFWSGI
Lawson, R., Alcock, C., Cooper, J., & Burgess, L. (2003). Factors affecting adoption of electronic commerce technologies by SMEs: An Australian study. Journal of Small Business and Enterprise Development, 10(3), 265-276. Lee, H., Sawyer, S., & Choudrie, J. (2003). Editorial: Broadband Internet and electronic commerce. Journal of Information Technology, 18(4), 227229. Lee, M.R. (2002). Broadband infrastructure development policy perspective of a broadband technology supplier. In OECD Broadband Workshop, Seoul, Korea.
Intellect. (2005). SMEs need guidance in adapting to broadband. Retrieved July 13, 2007, from http:// www.intellectuk.org/press/news/news_170505_ SMEbroadband.asp
Lefebvre, É, & Lefebvre, L.A. (1990). The importance of planning computer acquisitions: The case of small business. Journal of Small Business and Entrepeneurship, 8(3), 56-67.
International Telecommunication Union (ITU). (2003). World telecommunication development report: Access indicators for the information society. Retrieved from http://www.itu.int/ITUD/ict/publications/wtdr_03
/HVVLJ/ -DQXDU\ :KR¶VKROGLQJEDFN broadband? Washington Post.
Kalakota, R., & Robinson, M. (2001). E-business 2.0: Road-map for success. Boston: AddisonWesley Longman. Karahanna, E., Straub, D.W., & Chervany, N.L. (1999). Information technology adoption across time: A cross-sectional comparison of preadoption and postadoption beliefs. MIS Quarterly, 23(2), 183-214.
/HY\0 3RZHOO3 60(ÀH[LELOLW\DQG the role of information systems. Journal of Small Business Economics, 11, 183-196. Levy, M., Powell, P., & Yetton, P. (2002). The dynamics of SME information systems. Small Business Economics, 19(4), 341. Lyytinen, K., & Damsgaard, J. (2001). The role of intermediating institutions in the diffusion of electronic data interchange (EDI): How industry associations intervened in Denmark, Finland, and Hong Kong. Information Society, 17(3), 195. Middleton, C.A. (2003). What if there is no killer application- a user centric perspective on broad-
490
Broadband Diffusion to SMEs in the UK
band. Journal of Information Technology, 18(4), 231-246. 17/ +RZFDQEURDGEDQGEHQH¿W\RXUEXVLness. Retrieved July 13, 2007, from http://business. ntl.com/internet_products/business_broadband/ EXVLQHVVBEHQH¿WVSKS OECD. (2003). Broadband driving growth: Policy responses. Retrieved July 13, 2007, from http:// www.oecd.org/dataoecd/18/3/16234106.pdf OECD. (2005). ICT, e-business and SMEs. Retrieved July 13, 2007, from http://www.oecd. org/dataoecd/32/28/34228733.pdf Ofcom. (2004). The Ofcom Internet and broadband update. Retrieved July 13, 2007, from http://www. ofcom.org.uk/research/consumer_audience_research_int_bband_updt/may 2004/#content Ofcom. (2006). The Ofcom Internet and broadband update. Retrieved July 13, 2007, from www. ofcom.org.uk 2I¿FH RI WKH ((QYR\ UK online: The broadband future. Papazafeiropoulou, A., Gandecha, R., & StergiouODV / ,QWHUSUHWLYH ÀH[LELOLW\ DORQJ WKH innovation decision process of the UK, NHS Care Records Service (NCRS). Insights from a local implementation case study. In Proceedings of the Thirteenth European Conference on Information Systems, Regensburg, Germany. Philpott, M. (2004). Ovum white paper: DSL services for small enterprises. Pinch, T.J., & Bijker, W.E. (1987). The social construction of facts and artefacts: Or how the sociology of science and the sociology of technology PLJKWEHQH¿WHDFKRWKHU,Q%LMNHU:(+XJKHV T.P., & Pinch, T.J. (Eds.), The social construction of technological systems (pp. 18-50). Cambridge, MA: MIT Press. Pouloudi, A., & Whitley, E.A. (1997). Stakeholder LGHQWL¿FDWLRQLQLQWHURUJDQL]DWLRQDOV\VWHPV *DLQing insights for drug use management systems. European Journal of Information Systems, 6(1), 1-14.
Rajagopal, P. (2002). An innovation-diffusion view of implementation of enterprise resource planning (ERP) systems and development of a research model. Information & Management, 40(5), 87. Rogers, E.M. (1995). Diffusion of innovations. New York: Free Press. Saban, K.A., & Rau, S.E. (2005). The functionality of Websites as export marketing channels for small and medium enterprises. Electronic Markets, (2), 128-135. Sandler, M., & Boggs, R. (2001). Small business eCommerce- paving the way for the next generation of opportunity. Retrieved July 13, 2007, from http://netobjects.com Showcase05. (2005). Innovative broadband content. Retrieved July 13, 2007, from http://www.broadbandshowcase.org.uk/results.php?type=keyword Viatel. (2005). Viatel broadband for business. Retrieved July 13, 2007, from http://www.viatel. com/uplds/NetServ_Broadband.pdf Yaïche, H., Mazumdar, R.R., & Rosenberg, C. (2000). A game theoretic framework for bandwidth allocation and pricing in broadband networks. IEEE/ACM Transactions on Networking, 8(5), 667-678. Zhang, B. (2002). Understanding impact of convergence on broadband industry regulation: A case study of the United States. Telematics and Informatics, 19(1), 37-59.
KEY TERMS Broadband: Broadband technology is an umbrella term which covers varying high-speed access technologies including Asymmetric Digital Subscriber Line (ADSL), cable modems, satellite, and Wireless Fixed (Wi-Fi) Networks. Broadband provides an “always on” and provide faster Internet connection than dial up
491
Broadband Diffusion to SMEs in the UK
SMEs: Small and medium size enterprises are enterprises with fewer than 250 employees with an annual turn over not exceeding 10 million euros. Internet: Worldwide system of computer networks. It is a network of networks. Innovation Diffusion: An innovation is an idea, practice, or object perceived as new by an individual or other unit of adoption and diffusion is the process by which an innovation is communicated through certain channels over time among members of a social system.
492
SCOT: Social construction of technology is a response to technological determinism. Social constructivists argue that technology does not determine human action but human action shapes technology. Social groups: Relevant social groups related to the development of a technology that have shared meanings about the technology. Dial-up: Internet access which uses a modem to connect a computer to a telephone line and dials an Internet service provider to establish a link to the Internet.
493
Chapter XXXI
Environmental Drivers of E-Business Strategies Among SMEs Alessandro Arbore Bocconi University, Italy Andrea Ordanini Bocconi University, Italy
ABSTRACT ,QIURQWRIWUDGLWLRQDOLQWHUSUHWDWLRQVRIWKHGLJLWDOJDSEDVHGRQHQGRJHQRXVFRQGLWLRQVRIWKH¿UPVWKLV chapter intends to emphasize the importance that some external pressures may have on the e-business strategy of small and medium enterprises (SMEs). The environmental factors analyzed here are market position, competitive intensity, and institutional pressures. SMEs have been grouped according to their level of e-business involvement, in relation to the number of e-business solutions adopted so far. Three layers are proposed: excluded, tentative, and integrated e-business SMEs. A multinomial logistic regression was used to predict these strategies. A general conclusion is that different models seem to explain H[FOXVLRQDQGLQYROYHPHQW6SHFL¿FDOO\WZRIDFWRUVDPRQJWKRVHDQDO\]HGUHYHDOWREHPRUHVXLWDEOHLQ explaining e-business exclusion. They are the size of a SME and a lack of institutional pressures to adopt. On the other hand, e-business involvement seems to be primarily prompted by a selective competitive environment and not by imitative behaviors, as in the previous case.
INTRODUCTION In this chapter, our general attention is on the adoption and diffusion of e-business solutions among small and medium enterprises (SMEs). 0RUHVSHFL¿FDOO\RXULQWHUHVWLVIRFXVHGRQVRPH external factors that may affect this process, since
we are persuaded that external forces are especially important within small organizations, while largely understated by the literature. We will try to analyze whether these factors do relate to the e-business involvement of an SME. E-business is an umbrella term referring to a wide variety of Internet-based management solu-
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Environmental Drivers of E-Business Strategies Among SMEs
tions. The solutions considered here are: interactive Web sites, e-commerce platforms, e-procurement systems, customer relationship management systems, and telework.
BACKGROUND: SMES AND EXTERNAL FACTORS AFFECTING E-BUSINESS Many researches explain the adoption of e-business solutions in terms of endogenous factors, that is variables which are internal to the organization. Among these factors, different typologies of variDEOHVDUHUHFXUVLYH WKHOHYHORI¿QDQFLDOUHVRXUFHV able to affect any investment decision; the manageULDOFXOWXUHLQÀXHQFLQJWKHSURSHQVLW\WRLQQRYDWH and the organizational readiness, which is relevant for the integration of new technologies (for a review, see Lee, Runge, & Baek, 2001). Conversely, the SRWHQWLDOLQÀXHQFHRIH[RJHQRXVSUHVVXUHVKDYH been largely understated. Within this handbook, a review of internal factors is out of the scope of the current chapter. Instead, our goal is to deepen our understanding about the following environmental drivers: market position, competitive intensity, and institutional pressures. When focusing on SMEs, environmental elements must be considered as especially important because of the high dependency of these organizations from the context: SMEs appear more sensible to external pressures, in terms of both competitive and social “rules of the game” (Fink, 1998). Therefore, important explanatory factors of adoption strategies shall be easily found by studying the environment where SMEs play, rather than focusing only on internal conditions. Above all, since e-business is basically considered an instruPHQWWRLPSURYH¿UP¶VFRPSHWLWLYHQHVV $PLW Zott, 2000), we could expect that the competitive features of the arena where SMEs play require special attention. Starting from the current literature, then, the environmental forces analyzed in this work are: (1) the level of perceived competition (Kuan & Chau, 2001; Riemschneider, Harrison, & Mykytn, 2003); (2) the pressure to be considered technology savvy
494
(Iacovou, Benbasat, & Dexter, 1995; Zhu, Kremer, ;X DQG WKHFRPSHWLWLYHSRVLWLRQRFcupied in the marketplace (Daniel & Grimshaw, 2002; Lal, 1999). 2XUFKDSWHUZLOOEULHÀ\UHYLHZWKHWKHRUHWLFDO antecedents for each of these factors and then we will provide a set of hypotheses.
Competitive Intensity According to the well known industrial organization framework (Andrews, 1971; Porter, 1985), it can be observed that an increasing competitive intensity worsens the balance between opportuniWLHVDQGWKUHDWVUHTXLULQJ¿UPVWRDGRSWPRUHLQQRYDWLYHVWUDWHJLHVWRVXUYLYHDQGPDNHSUR¿WV Within this view, Gatignon and Robertson (1989) found that competitive pressure in the adopter industry has a positive impact on adoption of information and communication technologies. In the same way, Thong and Yap (1995) found that the CEO attitude towards new technologies adoption is positively correlated with the degree of competition faced in the market. Similarly, Premkumar and Roberts (1999) demonstrated that the degree of FRPSHWLWLYHULYDOU\LQWKHDGRSWHU¶VLQGXVWU\DIIHFWV the rate of adoption of digital technologies. Also adopting the more recent resource-based approach (Barney, 1991; Peteraf, 1993), it could be observed that the strength of competition in the factor markets reduces the power of isolating mechanisms sustaining the competitive edge, thus requiring continuous innovation strategies. Within these contexts, innovations lead to resource substitution phenomenon, acting as basic conditions for rent seeking (Malerba & Orsenigo, 1997). Especially here, e-business solutions may contribute to the development of dynamic capabilities sustaining long-term competitive advantages (Teece, Pisano, & Shuen, 1997). This discussion leads to formulate RXU¿UVWK\SRWKHVLV Hp1: The degree of perceived competition increases the intensity of e-business strategy among SMEs.
Environmental Drivers of E-Business Strategies Among SMEs
Institutional Pressures and Technology Legitimacy
might be interpreted as a pre-emptive strategy of the leader. From this point of view, the leader would IHHODQH[WHUQDOSUHVVXUHWREHWKH¿UVWPRYHULQ order not to loose its supremacy. From a theoretical point of view, there is a general consensus on the fact that the adoption of information technologies, by itself, has a marginal GLUHFWHIIHFWRQ¿UPSHUIRUPDQFHZKLOHVLJQL¿FDQW impacts emerge only when such technologies are combined and integrated with other distinctive competencies (see, among the others, Clemons & Row, 1991; Mata, Fuerst, & Barney, 1995; Powell & Dent-Micallef, 1997). A possible implication is that e-business solutions, like other ICTs, would have greater power in consolidating leading positions rather than reduce competitive gaps: the instrumental nature of technology makes e-business strategies viable only when there are some “business strategy” beyond them, thus suggesting that technology FDQQRW¿[DÀDZSURFHVVE\LWVHOIEXWLWLVDEOHWR LPSURYHHYHQVLJQL¿FDQWO\DQHVWDEOLVKHGSURFHVV (Carr, 2001). In other words, there is a “strategic necessity hypothesis” supporting the adoption of ICT technologies, and it is more likely that this necessity is SUHVHQWDPRQJ¿UPVZLWKDVROLGPDUNHWSRVLWLRQ (Clemons & Row, 1991). Following these considerations, SMEs feeling in a leading market position would be more likely to adopt new technologies than “marginal” competitors, which would prefer to exploit existing knowledge and capabilities rather than exploring new possibilities (Leonard-Barton, 1992). Hp3: Being in a leading market position increases the intensity of e-business among SMEs.
$FFRUGLQJ WR WKH LQVWLWXWLRQDO WKHRULHV D ¿UP¶V behavior has to be coherent with norms and social rules requested by their environment (Di Maggio & Powell, 1983; North, 1990). &RQVLVWHQWO\LQRUGHUWRJDLQDFFHVVWRVSHFL¿F resources, to collaborative networks, or to strategic DOOLDQFHV D ¿UP PLJKW EH VXEMHFW WR OHJLWLPDF\ assessment by other social agents (competitors, partners, and other stakeholders). The requirements may be very selective especially for SMEs, since they are not usually perceived as legitimate players due to their lack of resources and capabilities (Grewal, Comer, & Metha, 2001). Although for SMEs the last decision maker is generally the owner-manager, the pressure that she feels from other stakeholders (e.g., employees, customers, suppliers) remains an important determinant of technology adoption. Harrison, Mykytyn, and Riemenschneider (1997) found that similar unwritten norms, maintained by peers and VRFLHW\VWURQJO\LQÀXHQFHWKHLQWHQWLRQWRDGRSW information technologies in small businesses. Along the same line, Lee et al. (2001) posit that 60(V¶PDQDJHUVKHDUDERXWWKHUHODWLYHDGYDQWDJHV of digital technologies largely from the trade press, their friends, business competitors, and peer-social interactions. This would create selective contexts, where it is important to be perceived as technology savvy and where e-business strategies may be driven, among the other things, by a relational need to be reputed as innovative and technology savvy. This leads to our second hypothesis: Hp2: The need for a technology legitimacy increase the intensity of e-business adoption among SMEs.
METHODOLOGY
Market Position
Sampling and Collecting
Given a certain level of competitive rivalry, a leading market position1 may reveal better resources to exploit the potential of the new technologies. This, in turn, would be a further driver for e-business intensity. Moreover, early e-business adoption
The analysis is based on a survey of Italian SMEs. In this case, Italy can be considered as a meaningful ¿HOGWRLQYHVWLJDWHLVVXHVUHODWHGWR60(VJLYHQ the high number of small organizations in most of WKHLQGXVWULHV$UDQGRPVWUDWL¿HGVDPSOHRI
495
Environmental Drivers of E-Business Strategies Among SMEs
SMEs was selected, respecting the breakdown of SMEs among manufacturing and service indusWULHVH[FOXGLQJ¿UPVZLWKOHVVWKDQHPSOR\HHV considered as very small, and with more than 500 employees, considered as large organizations. Data ZDVFROOHFWHGWKURXJKDTXHVWLRQQDLUHDQG¿UPV were contacted by phone using the CATI technique during the period November/December, 2003. $ GUDIW RI WKH TXHVWLRQQDLUH ZDV ¿UVW WHVWHG RQDUDQGRPVDPSOHRI60(V7KH¿QDOYHUVLRQZDVPRGL¿HGDFFRUGLQJO\,QRUGHUWRUHGXFH potential response bias, a unique key informant was chosen: the CEO was selected as the most appropriate respondent. In answering the entire TXHVWLRQQDLUH¿UPVEHFDPHWKHDFWXDOVDPSOH of this study.
Variables The dependent variable of our model is the intensity of e-business adoption among SMEs. More precisely, we measure “e-business intensity” as a multinomial variable (EBUSINT) with three possible levels: •
•
0, for SMEs which do not implement any kind of e-business strategy, and thus excluded IURPHEXVLQHVVWUDMHFWRULHV (;&/8'(' 1, for SMEs which tried to implement one e-business strategy among the following: interactive Web site, e-commerce, e-procurement, customer relationship management, and
•
WHOHZRUN7KLVJURXSRI¿UPVZLOOEHFDOOHG TENTATIVE 2, for SMEs that integrated more than one ebusiness solutions in their processes, and for which e-business could be seen as a normal practice (INTEGRATED)
The intensity of competition perceived by a SME (variable “COMPETITION”) has been PHDVXUHGWKURXJKD¿YHSRLQWVOLNHUWW\SHVFDOH (1= I completely disagree; 5 = I totally agree) for the following sentence: “In the market where we play, competition is extremely strong.” The intensity of technology legitimacy pressure perceived by a SME (“TECHLEGITIM”) has been PHDVXUHG WKURXJK D ¿YH SRLQW OLNHUWW\SH VFDOH (1= I completely disagree; 5 = I totally agree) for the following sentence: “Given the market where we play, it is important for us to be considered as technology savvy.” The relative market position has been measured through a dummy variable (“LEADERSHIP”), as the answer to the following sentence: “Within your VSHFL¿FVHJPHQWRUQLFKHGRHV\RXU¿UPIHHOWREH in a leading market position?” (1 = yes; 0 = no). 6LQFHWKHUDQJHRI¿UPV¶VL]HLQRXUVDPSOHLV remarkable (from 50 to 250 employees), and since EXVLQHVV VL]H DOUHDG\ SURYHG WR EH D VLJQL¿FDQW discriminator between IT adopters and non-adopters among SMEs (Fink, 1998), we controlled for size-effects (variable “SIZE”). To that end, we used the natural logarithm of the number of employees,
Figure 1. The number of e-business solutions adopted by Italian SMEs
2 or more applications = INTEGRATED; 43,6%
0 applications EXCLUDED; 26,1%
1 application = TENTATIVE; 30,3%
Source: I-LAB, Bocconi University
496
Environmental Drivers of E-Business Strategies Among SMEs
considering that the probability to adopt an innovation increases with size, but at decreasing rates. Last, we recognize that service sectors—for their intangible nature—may be more conducive of e-business strategy than manufacturing (LuckingReilly & Spulber, 2001). The dichotomous variable “SERVICE” (0 = manufacturing; 1 = service), then, LVXVHGWRFRQWUROIRUWKLVLQGXVWU\VSHFL¿FLW\
The Analytical Model Multinomial logistic regression (MLR) is useful for situations in which you want to be able to classify subjects based on values of a set of predictor variables. This type of regression is similar to logistic regression, but it is more general because it can be applied when the dependent variable is not restricted to two categories. Since we modeled our dependent variable as a multinomial variable with three levels (excluded, tentative, integrated), we applied the MLR procedure, which is essentially an extension of binary logistic regression to a polytomous multinomial (Agresti, 1999). The method basically proceeds by comparing the effect of factors and covariates on the possibility of being in each of “n-1” categories (in our case, the “excluded” and the “integrated” categories) compared with the baseline category “n” (in our case, the “tentative” category). 7KLVLVFRQFHSWXDOO\HTXLYDOHQWWR¿WWLQJ³Q 1” separate binary logistic models, comparing category 1 with category n, category 2 with n, and so on. Practically, the software usually estimates a simultaneous model which is more statistically sophisticated where, for each independent variable, we have “n-1” parameter estimates, each estimating the effect of a one-unit change in this variable on the log odds of being in category rather than in the baseline category (Long, 1997). In few terms, for each case, there will be n-1 predicted logits, one for each category relative to the reference or baseline category (Menard, 2002). It could be noted that, when multiple classes of the dependent variable can be ranked, as for our case, ordinal logistic regression is sometimes preferred to multinomial logistic regression. Our
choice to adopt MLR is due to the fact that our sample does not fully satisfy the test of parallel lines, which is mandatory to adopt the ordinary logistic regression (Menard). Moreover, MLR allows comparing the effects of independent variables (i.e., the drivers of e-business) across different categories of dependent variables, which is one important feature of our model. In any cases, when the ordered logit model is run adopting a logit as a link function, outcomes are largely similar to those obtainable through the multinomial logistic regression (Pampel, 2000).
FINDINGS 7DEOHVKRZVWKH¿UVWRXWFRPHRIWKHSUHGLFWLYH model for the adoption of e-business solutions among SMEs. The reference category is “tentaWLYH´WKDWLV¿UPVWKDWDGRSWHGRQHEXWRQO\RQH e-business solution. 7KH RYHUDOO PRGHO ¿WWLQJ PHDVXUHV DUH DOO satisfactory.2 We can focus, then, to parameter estimates and discuss our previous hypotheses. 7KH¿QGLQJVDUHUHSRUWHGEHORZ
7KHVL]HRID60(DSSHDUVWREHDVLJQL¿FDQW variable in explaining the leap from being “excluded” to become “tentative,” but it GRHVQRWDSSHDUDVVLJQL¿FDQWLQH[SODLQLQJ a further leap, the one from being “tentative” WREHFRPHDQ³LQWHJUDWHG´HEXVLQHVV¿UP More precisely, a one unit increase in the natural logarithm of the number of employee of a SME (that is, an increase of about 170% in its size), reduces the odds ratio of being an ³H[FOXGHG´¿UP²LQVWHDGRIEHLQJDWOHDVWD “tentative” one—by 34%. S IZE
S IZE **
2.
E xcluded S ME
T entative S ME
Involved S ME
(no ebus ines s s olutions )
(jus t 1 ebus ines s s olution)
(2 or more ebus ines s s olutions )
The lack of competition would not prove to be DVLJQL¿FDQWIDFWRULQH[SODLQLQJHEXVLQHVV exclusion for a SME; on the contrary, the level
497
Environmental Drivers of E-Business Strategies Among SMEs
7DEOH0XOWLQRPLDOORJLVWLFUHJUHVVLRQ PRGHO¿WWLQJ Model Fitting Information Model
-2 Log Likelihood
Chi-Square
df
Sig.
Intercept Only Final
762,955 717,061
45,895
10
0,000
df
Sig.
0 2 2 2 2 2
. 0,036 0,013 0,088 0,029 0,001
Pseudo R-Square Cox and Snell 0,115 Nagelkerke 0,130 McFadden 0,057 Likelihood Ratio Tests
Effect Intercept SIZE TECHLEGITIM COMPETITION SERVICES LEADERSHIP
-2 Log Likelihood of Reduced Model Chi-Square 717,061 723,692 725,818 721,911 724,123 730,237
0,000 6,631 8,758 4,851 7,063 13,176
The chi-square statistic is the difference in -2 log-likelihoods between the final model and a reduced model. The reduced model is formed by omitting an effect from the final model. The null hypothesis is that all parameters of that effect are 0. This reduced model is equivalent to the final model because omitting the effect does not increase the degrees of freedom.
Classification Predicted Observed EXCLUDED TENTATIVE INTEGRATED Overall Percentage
EXCLUDED
TENTATIVE
INTEGRATED
Percent Correct
36 17 22
14 26 18
48 71 124
36,7% 22,8% 75,6%
19,9%
15,4%
64,6%
49,5%
Table 2. Multinomial logistic regression: parameter estimates* Parameter Estimates
EXCLUDED
INTEGRATED
Intercept SIZE TECHLEGITIM COMPETITION SERVICES LEADERSHIP Intercept SIZE TECHLEGITIM COMPETITION SERVICES LEADERSHIP
B 2,947 -0,418 -0,324 0,029 -0,417 0,057 -0,478 -0,093 -0,043 0,250 0,381 0,818
* TENTATIVE is the reference category
498
Std. Error 1,012 0,170 0,121 0,135 0,335 0,301 0,903 0,143 0,112 0,125 0,276 0,258
Wald 8,476 5,999 7,141 0,045 1,551 0,035 0,280 0,423 0,146 3,983 1,901 10,078
df 1 1 1 1 1 1 1 1 1 1 1 1
Sig. 0,004 0,014 0,008 0,831 0,213 0,851 0,596 0,515 0,702 0,046 0,168 0,002
Exp(B) . 0,659 0,723 1,029 0,659 1,058 . 0,911 0,958 1,284 1,463 2,266
95% Confidence Lower Upper Bound Bound . 0,472 0,570 0,791 0,342 0,587 . 0,689 0,770 1,004 0,852 1,367
. 0,920 0,917 1,339 1,270 1,907 . 1,205 1,193 1,642 2,513 3,754
Environmental Drivers of E-Business Strategies Among SMEs
RISHUFHLYHGFRPSHWLWLRQDSSHDUVVLJQL¿FDQW in explaining the leap from “tentative” ebusiness to “integrated” e-business. This UHVXOWZRXOGFRQ¿UPRXU¿UVWK\SRWKHVLV WKH degree of perceived competition does affect the involvement in e-business strategies, even if it does not explain e-business exclusion tout court. More precisely, a one point increase LQ WKH ¿YHSRLQW VFDOH PHDVXULQJ WKH OHYHO of perceived competition increases the odds ratio of being an “integrated” e-business SME—instead of being just “tentative”—by 28.4%.
E xcluded S ME
T entative S ME
Involved S ME
(no ebus ines s s olutions )
(jus t 1 ebus ines s s olution)
(2 or more ebus ines s s olutions )
Low environmental pressures for being WHFKQRORJ\VDYY\SURYHWREHDVLJQL¿FDQW factor in explaining part of the e-business H[FOXVLRQRID60(EXWQRWDVLJQL¿FDQWIDFWRU in explaining the leap from being “tentative” WREHFRPHDQ³LQWHJUDWHG´HEXVLQHVV¿UP Accordingly, the second hypothesis would EHFRQ¿UPHGE\RXUHYLGHQFHEXWRQO\LQ explaining e-business exclusion vs. adopWLRQ7KLVVDPSOHGLGQRWFRQ¿UPDQ\UROH of technology legitimacy requirements in explaining different level of e-business adoption. More precisely, a one point increase in WKH¿YHSRLQWVFDOHPHDVXULQJWKHLQWHQVLW\RI this pressure reduces the odds ratio of being DQ³H[FOXGHG´¿UPLQVWHDGRIEHLQJDWOHDVW “tentative” by 27%. Need f or t ec h. l egitimac *y**
4.
Market l eader
Market leader***
E xcluded S ME
T entative S ME
Involved S ME
(no ebus ines s s olutions )
(jus t 1 ebus ines s s olution)
(2 or more ebus ines s s olutions )
C ompetition**
C ompetition
3.
odds ratio of being “involved” in e-business. ,QFRQ¿UPDWLRQRIRXUWKLUGK\SRWKHVLVWKHQ it seems possible to conclude that a strong market position is positively related to higher levels of e-business involvement. More precisely, when an SME feels to be in a leading market position, the odds ratio of being an “integrated” e-business SME—instead of being just “tentative”—increases by 126.6%.
Need f or t ec h. l egitimac y
E xcluded S ME
T entative S ME
Involved S ME
(no ebus ines s s olutions )
(jus t 1 ebus ines s s olution)
(2 or more ebus ines s s olutions )
A (self-stated) leading market position is highO\VLJQL¿FDQWRQO\LQH[SODLQLQJWKHOHDSIURP being just “tentative” to be an “integrated” HEXVLQHVV¿UP 60(VLQDOHDGLQJPDUNHW position (or feeling so) more than double the
Finally, the role of the industry (namely, services vs. manufacturing) was controlled by the dichotomous variable “services.” Even if the literature already proved the relevance of this factor, our sample produced a Wald statistics below the 10% level RIVLJQL¿FDQFHIRUERWKWKHFRQWUDVWV ³WHQWDWLYH´ vs. “excluded” and “tentative” vs. “integrated”). Considering the two contrasts together, however, the F2 test on the difference between the full model deviance and the baseline model deviance shows DQRYHUDOOVLJQL¿FDQFHDWWKHOHYHO7KHHVWLmates indicate that belonging to a service sector, instead of a manufacturing one, would reduce the odds ratio of being “excluded” by 34%. Similarly, belonging to a service sector would increase the probability of being “integrated” (instead of just “tentative”) by 46%.
IMPLICATIONS AND CONCLUSIONS 7KHYHU\¿UVWFRQFOXVLRQWKDWLVSRVVLEOHWRGUDZ out is that different factors seem to explain the nonadoption of e-business solutions and different levels of e-business involvement. The analysis, in RWKHUZRUGVFRQ¿UPVWKHRSSRUWXQLW\WRNHHSWKH two issues as separated: e-business exclusion and e-business involvement appear, at least partially, to be different phenomena that require different theoretical reasoning and modeling. :LWKVSHFL¿FUHJDUGWRRXUVWXG\WZRIDFWRUV among those analyzed reveal to be more suitable
499
Environmental Drivers of E-Business Strategies Among SMEs
in explaining e-business exclusion of some SMEs (but not to explain different levels of e-business adoption). They are the size of a SME and a lack of institutional pressures/inputs to adopt the innovations in point (see hypothesis two). L ac k o f ins titutional pres s ures to a dopt e bus ines s S ize
E xcluded S ME (no ebus ines s s olutions )
This means that: 1.
When the adoption of e-business is just sporadic, almost casual (tentative, as said), this is generally because such an adoption roots mainly in imitating behaviors or external pressures, and less in the culture of innovation or in the search of competitive advantages. ,WLVDOVRFRQ¿UPHGWKDWWKHODFNRIDGHTXDWH resources (tangibles and intangibles, as approximated by the variable size), may completely impair the adoption of any ebusiness solution, something well-known to the literature of innovations. SMEs in similar conditions need the special attention of e-business supporting policies, keeping in mind that pushing adoption, by itself, is QRWHQRXJKWROHWD¿UPHQMR\WKHEHQH¿WVRI e-business.
The study, then, provides insights on e-business involvement, that is, on a further step to become ¿UPVGHYHORSLQJPRUHHEXVLQHVVLQWHJUDWHGRSportunities. Selec tive c ompetitive environment Leading market p osition
2.
500
Involved SME (more ebus iness solutions )
,QWHUHVWLQJO\ DQG VLJQL¿FDQWO\ WKLV OHDS LV prompted by a selective competitive environment, and not by imitative or induced behaviors, as for the previous, tentative step An integrated e-business involvement, here, seems to be a requirement and an opportunity
to remain competitive. Not by chance, in fact, WKHVH¿UPVDUHWKHOHDGLQJRQHVIHHOLQJWKH necessity to stay on the edge. Competition, then, would create the right conditions for driving the adoption of integrated e-business solutions among SMEs. Procompetitive policies, accordingly, would result in effective (and HI¿FLHQW SURGLJLWDOL]DWLRQSROLFLHVDVZHOO The results also lead to a related consideration: WRGD\¿UPVIDFLQJORZOHYHOVRIFRPSHWLWLRQRUORZ environmental pressures can still afford a low (or nought) e-business involvement and innovation. But how many arenas will have the fortune of maintaining a similar structure in the next future? The evidence tells us that these happy gardens are rarer and rarer, also because of technological changes. The management, here, should ask whether they are getting ready for the upcoming challenges or ZKHWKHULQVWHDGWKH\DUHH[SRVLQJWKHLU¿UPVWR some risky e-business run up. Among the limitations of this study, it must be noted that a possible response bias exists. In fact, ¿UPVZLWKDKLJKHUHEXVLQHVVLQYROYHPHQWPLJKW be more inclined to answer the questionnaire. However, while this would bias the percentage of e-business adopters (statistical frequencies in Table 1), it should not bias the relationships that we discussed in this chapter. Finally, as a suggestion for future research, the implications emerged in this study might be integrated by the analysis of further factors, both endogenous and exogenous, separately for e-business adoption and e-business involvement.
REFERENCES Amit, R., & Zott, C. (2001). Value creation in e-business. Strategic Management Journal, 22, 493-520. Andrews, K. (1971). The concepts of corporate strategy. Homewood, IL: Dow-Jones Irwin. Barney, J. (1991). Firm resources and competitive advantage. Journal of Management, 17(1), 99-120.
Environmental Drivers of E-Business Strategies Among SMEs
Clemons, E., & Row, M. (1991). Sustaining IT advantage: The role of structural differences. MIS Quarterly, 3, 275-292. Cohen, W.M., & Levin R.C. (1989). Empirical studies of innovation and market structure. In R. Schmalensee & R. Willig (Eds.), Handbook of industrial organization (pp. 1059-1107). Amsterdam: North Holland. Daniel, E.M., & Grimshaw, D.J. (2002). An exploratory comparison of electronic commerce adoption in large and small enterprises. Journal of Information Technology, 17, 133-147. Davidsson, P., Hunter, E., & Klofsten, M. (2006). Institutional forces – the invisible hand that shapes venture ideas. International Small Business Journal, 24(2), 115-131. E-business-watch. (2003). The 2nd European Ebusiness Report. European Commission. Fink, D. (1998). Guidelines for the successful adoption of information technology in small and medium enterprises. International Journal of Information Management, 18(4), 243-253. Gatignon, H., & Robertson, T.S. (1989). Technology diffusion: An empirical test of competitive effects. Journal of Marketing, (1), 35-49. Grandon, E.E., & Pearson, J.M. (2004). Electronic commerce adoption: An empirical study of small and medium U.S. business. Information & Management, 42, 197-216. Grewal, R., Comer, J.M., & Metha, R. (2001). An investigation into the antecedents of organizational participations in the business-to-business electronic markets. -RXUQDORI0DUNHWLQJ(7), 17-33. Harrison, D., Mykytyn, P.P., & Riemenschneider, C.K. (1997). Executive decisions about adoption of information technology in small businesses: Theory and empirical tests. Information Systems Research, 8(2), 171-195. Iacovou, C.L., Benbasat, I., & Dexter, A.S. (1995). Electronic data interchange and small organiza-
tions: Adoption and impact of technology. MIS Quarterly, 19(4), 465-486. Lal, K. (1999). Determinants of the adoption of information technology: A case study of electrical DQGHOHFWURQLFJRRGVPDQXIDFWXULQJ¿UPVLQ,QGLD Research Policy, 28, 667-680. Lee, J., Runge, J., & Baek, S. (2001). Adoption of Internet technologies in small business. In ProceedLQJV RI 3DFL¿F $VLD &RQIHUHQFH RQ ,QIRUPDWLRQ Systems (PACIS 2001), Seoul, Korea. Leonard-Barton, D. (1992). Core rigidities and core competencies. Strategic Management Journal, 13, 111-125. Long, J.S. (1997). Regression models for categorical and limited dependent variables. Sage Publications. Lucking-Reiley, D., & Spulber, D.F. (2001). Business to business electronic commerce. Journal of Economic Perspectives, (1), 55-68. Malerba, F., & Orsenigo, L. (1997). Technological regimes and sectoral patterns of innovative actvities. Industrial and Corporate Change, 6, 83-117. Mata, F.J., Fuerst, W.L., & Barney J.B. (1995). Information technology and sustained competitive advantage: A resource-based analysis. MIS Quarterly, 19(4), 487-505. Menard, S. (2002). Applied logistic regression analysis (Rev. ed). Sage Series on Quantitative Applications in the Social Sciences. Newbury Park , CA: Sage. North, D.C. (1990). Institutions, institutional change and economic performance. Cambridge, MA: Cambridge University Press. Pampel, F.C. (2000). Logistic regression. A primer. Sage Series on Quantitative Applications in the Social Sciences, 132. Newbury Park, CA: Sage. Peteraf, M.A. (1993). The cornerstones of competitive advantage: A resource-based view. Strategic Management Journal, 14, 179-192.
501
Environmental Drivers of E-Business Strategies Among SMEs
3ÀXJKRHIW.$5DPDPXUWK\.6RR¿(6
KEY TERMS E-Business: An umbrella term referring to a wide variety of Internet-based management solutions, like a Web site, an e-commerce platform, or a customer relationship management system. Web site: $Q RUJDQL]DWLRQ¶V SUHVHQFH RU DQ LQGLYLGXDO¶VSUHVHQFHRQWKH:RUOG:LGH:HE,WLV
502
a structured collection of multimedia pages linked to each other and very often to pages on other Web sites (hyperlinks). A Web site is hosted on a server computer and, to qualify as a Web site, the server must be available on the Internet 24 hours a day. E-Commerce: According to the American Marketing Association (AMA), an e-commerce strategy incorporates various elements of the marketing mix to drive users to a Web site for the purpose of purchasing a product or service. E-Procurement: A structured information system adopted by some organizations for making business-to-business purchases over the Internet. Such a system can connect companies and their business processes directly with suppliers while managing all interactions between them. E-procurement is also referred to as supplier exchange. Customer Relationship Management (CRM) System: a system combining database and computer technology with customer service to manage at best the relationships with clients. Automated CRM processes are often used to generate automatic personalized marketing actions based on the customer information stored in the system. Telework: According to the European Commission, telework occurs when information and communications technologies (ICTs) are applied to enable work to be done at a distance from the place where the work results are needed or where the work would conventionally have been done. Institutional Pressures: According to the Institutional Theory, institutions work as forces upon individuals and organizations by creating social pressures and restrictions, setting boundaries for ZKDWLVDFFHSWHGDQGZKDWLVQRW6XFKDQLQÀXence can be in the form of normative, coercive, and mimetic pressures (Davidsson, Hunter, & Klofsten, 2006). Normative pressures consist of social pressures on organizations and its members to conform to certain norms. Coercive pressures are often thought of as formal institutions of regulations
Environmental Drivers of E-Business Strategies Among SMEs
or laws but can also be informal expectations on organizations (e.g., technical standards imposed by someone exerting power over another actor, as in a parent-subsidiary relationship). Mimetic pressures represent demands towards imitation of other organizations to cope with uncertainty.
ENDNOTES 1
A leading position for a SME, here, may also UHIHU WR D VSHFL¿F VHJPHQW RU QLFKH RI WKH market.
2
The likelihood ratio test on the entire model is KLJKO\VLJQL¿FDQW WKHLQGHSHQGHQWYDULDEOHV WRJHWKHUGRKDYHDVLJQL¿FDQWHIIHFWRQWKH e-business intensity among SME. With regard to the Nagelkerke pseudo R-square measure, it indicates a satisfactory level above the threshold value of 0.10. Also, the model is able to correctly classify 49.5% of the cases. This is an increase of 13.5% compared to its baseline: also in this case, a lift above 10% is largely considered as satisfactory.
503
504
Chapter XXXII
Exploring SMEs’ Adoption of Broadband in the Northwest of England Boumediene Ramdani University of Manchester, UK Peter Kawalek University of Manchester, UK
ABSTRACT This chapter explores the factors impacting small to medium-sized enterprises’ (SMEs) adoption of broadband. It argues that information and communication technologies (ICTs) are highly differentiated technologies for which there is not necessarily a single adoption model. While most large European companies are connected to broadband, SMEs’ connectivity is lagging behind. The question of why one SME adopts broadband while the other does not is still understudied. Therefore, the purpose of this FKDSWHULVWR¿OOWKLVJDSE\LQYHVWLJDWLQJWKHWHFKQRORJLFDORUJDQL]DWLRQDODQGHQYLURQPHQWDOIDFWRUV impacting SMEs’ adoption of broadband. This chapter starts by highlighting the importance of ICT innovations adoption in general and broadband in particular. Based on the ICT innovations adoption literature, SMEs’ broadband adoption framework will be developed and empirically validated involving nine SMEs’ key decision makers in the northwest of England. Finally, implications for researchers, practitioners, ICTs’ vendors, and policy makers will be discussed.
INTRODUCTION New information and communication technologies (ICTs) such as broadband provide small to mediumsize enterprises (SMEs) with opportunities that are largely unexploited. Although SMEs form a substantial constituent of the global economy and ICT innovations adoption is nowadays economically and strategically feasible for the smallest or-
ganizations (Raymond, 1989; Thong & Yap, 1995), OLPLWHGUHVHDUFKKDVDGGUHVVHGWKHVSHFL¿FVRI,&7 innovations adoption, implementation, and use in the small business context (Brock, 2000; Shiels 0F,YRU 2¶5HLOO\ 7KH EHQH¿WV RI EURDGEDQG DGRSWLRQ DUH ZHOO documented; they include capturing new customers and markets ( Quayle, 2002; Raymond, 2001; Ritchie & Brindley, 2000; Vescovi, 2000), im-
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Exploring SMEs’ Adoption of Broadband in the Northwest of England
proved marketing techniques (Sparkes & Thomas, 2001), and improved relations with business partners (Poon & Swatman, 1999). The question of why one SME adopts broadband while another does not is still understudied. This chapter intends to contribute to the understanding of broadband adoption among SMEs by achieving the following objectives:
• •
7R KLJKOLJKW WKH LPSRUWDQFH RI 60(V¶ adoption of ICT innovations in general and broadband in particular To review the ICT innovations adoption literature 7R GHYHORS 60(V¶ EURDGEDQG DGRSWLRQ framework and empirically validate it To discuss implications for researchers, pracWLWLRQHUV,&7V¶YHQGRUVDQGSROLF\PDNHUV
BACKGROUND: SMES & ICT ADOPTION SMEs are considered as major economic players and a potent source of national, regional, and local economic growth (Taylor & Murphy, 2004). Without a better understanding of the complex processes and differentiating factors that affect ICT innovations adoption level, the drive of ICT adoption and development will not successfully FRQWULEXWHWR60(V¶FRPSHWLWLYHQHVV 0DUWLQ Matlay, 2001). There were 4.3 million small businesses enterprises in the UK at the start of 2005 representing (99%) of all business in the UK and together they accounted for more than half of the employment (58.7%) and turnover (51.1%) in the UK. (SBS, 2006). 7KH(XURSHDQ&RPPLVVLRQ (& GH¿QHVVPDOO businesses based on the number of employees, annual turnover, annual balance sheet total, and level RIDXWRQRP\ (& 0RVWGH¿QLWLRQVRI60(V emanate from the 1971 Bolton Committee Report, ZKLFKGH¿QHVDVPDOO¿UPDVLQGHSHQGHQWRZQHU managed, and with small market share (Simpson & Docherty, 2004). The DTI (2004) categorizes SMEs LQWR PLFUR¿UPVZLWKIHZHUWKDQHPSOR\HHV VPDOO¿UPVZLWK HPSOR\HHVDQGPHGLXP
VL]HG¿UPVZLWK HPSOR\HHV6WRU\ argues that the number of employees is considered to be an appropriate measure of SMEs because of the differences in organizational structures that occur with size. Using number of employees as a PHDVXUHWKH(&DQG'7,GH¿QLWLRQVDUHFRPSDWLEOH and therefore will be used in this study. New ICT provide SMEs with opportunities that are largely unexploited (Brock, 2000; Corso, Martini, Paolucci, & Pellegrini, 2001). It is hard nowadays to imagine SMEs operating without some use of ICT innovations. However, SMEs differ in WKHOHYHORI,&7V¶XVDJH %ODFNEXUQ 0F&OXUH 6RXWKHUQDQG7LOOH\ LGHQWL¿HVWKUHH FDWHJRULHV RI VPDOO ¿UPV ZLWK GLIIHUHQW DWWLWXGH to ICT: •
•
•
SMEs with low-end ICT use—where there is QRWDJRRG¿WEHWZHHQ,&7DQGWKHRZQHU PDQDJHU¶VFRQFHSWRIWKHEXVLQHVV Medium-level ICT users—with more expertise, separate IT and communications systems, open access to company data (network DQG¿OHVVHUYHUV ,7LQSURGXFWLRQHPDLO and a plan for and delegation of the management and routine upgrading of IT High-end ICT users—leading edge and innovative IT use, ICT integrated in the business process, a full digital information and communication system, ICT as a formal responsibility with a dedicated manager (Taylor & Murphy, 2004)
ICT is a broad term used to refer to any technologies from a simple acquisition of hardware to the full implementation of an enterprise resource planning (ERP) system. This study focuses on broadband, which is a term commonly used to describe Internet connections that are “always RQ´DQGWKDWSURYLGHDVSHHGZKLFKLVVLJQL¿FDQWO\ faster than dial-up connections, supporting the delivery of innovative content, applications, and VHUYLFHV (8 %URDGEDQGLVXVXDOO\GH¿QHG on the basis of transmission capacity: faster than the primary rate International Telecommunication Union (ISDN) speeds in excess of 200 kb/s (Federal Communications Commission), and a service with
505
Exploring SMEs’ Adoption of Broadband in the Northwest of England
downstream capacity of at least 256 kb/s (Organization for Economic Cooperation and Development). $VWKHVSHHGHYROYHVWKHGH¿QLWLRQPD\TXLFNO\ become obsolete (EU, 2004). Therefore, for the purpose of this chapter, broadband will refer to a wide range of technologies that have been developed to support the delivery of innovative interactive services, equipped with an always-on functionality, providing broad bandwidth capacity that evolves over time, and allowing the simultaneous use of both voice and data services. (EU, 2004, p. 5) The adoption of broadband is recognized to be one of the most important ICT innovations, since a broadband connection to the Internet is the main SUHFRQGLWLRQWRDFKLHYHWKHEHQH¿WVRI,&7LQQRYDWLRQV¶PDVVLYHDGRSWLRQ $UERUH 2UGDQLQL Mehrtens, Cragg, & Mills, 2001; Prieger, 2003). According to Arbore and Ordanini (2006), the importance of broadband in the business sector is related to the higher potential for data interchange and multimedia applications. This should make the SD\RIIRI,&7V¶LQYHVWPHQWPRUHHYLGHQWIRUVPDOO ¿UPV 6DZ\HU$OOHQ +HHMLQ
In order to meet the objective of becoming the most competitive knowledge-based society in the world by 2010, SMEs are put under pressure to embrace ICTs and e-business techniques (Supurge & Roberts, 2005). The UK government acknowledged there is a slow take-up of e-business techniques among SMEs (UK-Online, 2002). Figure 1 shows SMEs broadband adoption across EU13 countries, where the UK take-up is below the EU13 average. According to Ofcom (2004), SMEs Internet penetration has increased to reach 70%. The proportion of SMEs with broadband has also risen quite dramatically to reach 50% of all 60(VZLWK,QWHUQHWFRQQHFWLRQ2IFRP¶VUHVHDUFK segmented SME consumers into six distinct groups based on overall and relative spending on telecoms, ownership and attitude to new technologies, and the industry or sector in which they operate. The following table describes SME segments. In order to understand the factors impacting SMEs adoption of broadband, the next section reviews ICT innovations adoption literature.
Figure 1. SMEs' broadband adoption across EU13 Countries (EU, 2004) 100 % 90 %
% SMEs connected
80 % 70 % 60 % 50 %
EU13 average*
40 % 30 % 20 %
506
United Kingdom
Note: SMEs defined as enterprises with 10-249 employees. * Arithmetic average Source: European Commission (Eurostat)
Sweden
Finland
Portugal
Austria
Netherland
Luxemburg
Italy
Ireland
Spain
Germany
Denmark
0 %
Belgium
10 %
Exploring SMEs’ Adoption of Broadband in the Northwest of England
7DEOH60(VHJPHQWV VXPPDU\IURP2IFRPSS SME Segment
Description
Top-spend tech (11%)
•
One of the smallest SME segments in terms of size and turnover
•
Largest in terms of size and turnover
•
Knowledgeable about technology
•
Willing to pay more for favored suppliers
High-spend Internet users (9%)
High-spend followers (18%)
Medium-spend nontech (16%)
Low-spend tech (23%)
Low-spend nontech (23%)
•
Second smallest of the SME segments in terms of size and turnover
•
Fairly average in terms of size and turnover
•
More technology-aware than most
•
Their Internet penetration is above average and over half use broadband
•
Average in size
•
Try to keep up with technology
•
Similar to High-spend Internet users in terms of Internet penetration and spending
•
Slower pace compared to High-spend Internet users
•
Average in terms of size and turnover
•
Second smallest proportion of all the companies
/HDVWOLNHO\WRVHHWKHPVHOYHVDVNQRZOHGJHDEOHDERXW¿[HGWHOHFRPV
•
Average overall quarterly spend on telecoms
•
One of the two largest segments in terms of size and turnover
•
Consists of smaller businesses with an average of nine employees
•
Below average Internet and mobile penetration
•
Quarterly spend is also lower than average
•
The second of the largest segments in terms of size and turnover
•
Consists mainly of small businesses with an average of six employees
•
Low interest in telecoms and technology and claim not to try and keep up with developments
•
Low penetration of Internet
•
Lowest spending segment
ICT INNOVATIONS ADOPTION RESEARCH ICT innovations adoption research has become increasingly popular as IT continues its relentless march into almost every aspect of organizational life (Fichman, 2004) and as innovation become an important driver of organizational competitiveness (Hamel, 1998). A recent review of the predictors, linkages, and biases in IT innovation adoption research by Jeyaraj, Rottman, and Lacity (2006) highlights that a rich but diverse body of theoretical and empirical work has accumulated on the adoption and diffusion of ICT innovations. A
large number of theories have been tested, including theory of reasoned action (Fishbein & Ajzen, 1975), innovation diffusion theory (Rogers, 1983), social cognitive theory (Bandura, 1986), diffusion/implementation model (Kwon & Zmud, 1987), technology acceptance model (TAM) (Davis, 1989), theory of planned behavior (Ajzen, 1991), perceived characteristics of innovating (Moore & Benbasat, 1991), tri-core model of IS innovations (Swanson, 1994), innovation diffusion theory for organizations (Rogers, 1995), TAM2 (Venkatesh 'DYLV DQGXQL¿HGWKHRU\RIDFFHSWDQFH and use of technology (Venkatesh, Morris, Davis, & Davis, 2003). Empirically, qualitative (Agarwal &
507
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Table 2. Theoretical models used to examine ICT innovations adoption by SMEs Theory
Sources
technology acceptance model (TAM)
Grandon & Pearson (2004); Igbaria, Zinatelli, Cragg, & Cavaye (1997);
theory of planned behavior (TPB)
Harrison, Mykytyn, & Riemenschneider (1997);
Combined TAM and TPB
Riemenschneider, Harrison, Mykytyn (2003);
TAM2
Venkatesh & Davis (2000);
innovation diffusion theory (IDT)
Cragg & King (1993); Scupola (2003); Iacovou, Benbasat, & Dexter (1995); Mehrtens et al. (2001); Thong & Yap (1995);
resource-based theory
Braun (2002); Caldeira & Ward (2003); Chau (2001); Feeny & Willcocks (1998); Grewal (2001); Iacovou et al. (1995); Mata, Fuerst, & Barney (1995); Mehrtens et al. (2001); Scupola (2003); Thong (2001);
stage theory
Daniel, Wilson, & Myers (2002); DTI (2001); Levy, Powell, & Yetton (2002); Poon & Swatman (1997); Prananto, Marshall, & McKay (2003); Rao, Metts, & Mora (2003);
XQL¿HGWKHRU\RIDFFHSWDQFHDQGXVHRIWHFKQRORJ\
Anderson & Schwager (2003);
(UTAUT)
Prasad, 1997), quantitative (Igbaria & Tan, 1997), and combination of both methods (Thong & Yap, 1994) have been used. Theories used to examine ICT innovations adoption research in small business context are highlighted in Table 2. Although, ICT innovations adoption research tends to address the same research question: what factors facilitate or hinder the adoption and diffusion of IT-based innovations within a population of potential adopters (Jeyaraj et al., 2006), it is essential to understand the factors impacting the adoption of new technologies. According to Fichman (2004), a dominant research paradigm for ICT innovations has emerged. He argues that this dominant paradigm assumed that organizations with greater quantity of the “right stuff” will exhibit a greater quantity of ICT innovations. Jeyaraj et al. (2006) argue that different theories have been used to examine the organizational adoption of ICT innovations. ICT innovations adoption has been examined at different levels of the organization such as the functional units (e.g., Ravichandran, 2000), and entire organizations (e.g., Premkumar, Ramamurthy, & Nilakanta, 1994). Jeyaraj et al. (2006) claim that the challenge of the dominant paradigm has been the lack of integration and understanding
508
of the linkage between individuals and organizational adoption of ICT innovations. Although the unit of analysis in organizational adoption studLHVLVWKHRUJDQL]DWLRQDQRUJDQL]DWLRQ¶VGHFLVLRQ to adopt an innovation is made by an individual within that organization. To understand why an SME adopts broadband while the other does not, this study is an investigation of the innovation (or technological) characteristics, organizational characteristics, and environmental characteristics which impact SMEs adoption of broadband.
TECHNOLOGY-ORGANIZATIONENVIRONMENT (TOE) FRAMEWORK Tornatzky and Fleischer (1990) developed TOE framework to study adoption of ICT innovations. 7KLVIUDPHZRUNLQFOXGHGWKUHHDVSHFWVRI¿UP¶V FRQWH[WWKDWLQÀXHQFHWKHSURFHVVE\ZKLFKLWDGRSWV and implements ICT innovations: technological context, organizational context, and environmental context. The TOE framework has been examined by a number of empirical studies on various ICT innovations (as illustrated in Table 3). From reviewing
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Table 3. Technology-organization-environment (TOE) frameworks used in previous SMEs’ studies Authors
ICT
Technological
Organizational
Environmental
Innovations
Context
Context
Context
Lertwongsatien &
Electronic
Technology Factors
Organizational Factors
Organizational Environment
Wongpinunwatana (2003)
Commerce
3HUFHLYHG%HQH¿WV
(Size, Top management
(Competitiveness)
(EC)
Perceived Compatibility)
Support for E-Commerce, Existence of IT Department)
Scupola (2003)
EC
(&%DUULHUV(&%HQH¿WV
(PSOR\HHV¶,6.QRZOHGJH
Related Technologies
Pressure from Competitors, Buyers, and Suppliers, Role of Government, Technology Support Infrastructure
Kuan & Chau (2001)
Thong (1999)
Fink (1998)
EDI
3HUFHLYHG'LUHFW%HQH¿WV
Perceived Financial Cost,
Perceived Industry Pressure,
(Electronic Data
Perceived Technical
Perceived Government
Interchange)
Competence
Pressure
IS
IS Characteristics
Organizational
Environmental
(Information
(Relative Advantage/
Characteristics (Business
Characteristics
Systems)
Compatibility,
6L]H(PSOR\HH¶V,6
(Competition)
Complexity)
Knowledge)
,7%HQH¿WV
Organizational Culture,
External Environment,
(Information
In-house IT Expertise
Outside Support, External
Technology)
and Resources, IT
Resources
IT
Implementation and Selection Iacovou et al., (1995)
EDI
3HUFHLYHG%HQH¿WV
these empirical studies, it is not clear which factors to include in the TOE framework to study SMEs adoption of broadband. This is because within the three contexts, different factors impact the adoption and implementation of different ICT innovations. In addition, different factors impact the same studied ICT innovation (e.g., e-commerce). Thus, it has been suggested to extend the TOE framework to study other ICT innovations (Chau & Tam, 1997). This study, therefore, investigates the technological, organizational, and environmental factors that impact SMEs adoption of broadband.
Organizational Readiness
External Pressure
Theoretical Framework Based on the TOE framework discussed earlier, Figure 2 presents a conceptual model of SMEs adoption of broadband. The model posits factors that might impact SMEs adoption of broadband.
Technological Context Premkumar (2003) argues that there are very few studies that have examined the impact of WHFKQRORJLFDOFKDUDFWHULVWLFV5RJHUV¶LQQRYDWLRQ diffusion theory for organizations will be used as a theoretical basis for studying the impact of technological factors. Grandon and Pearson (2004)
509
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Figure 2. Theoretical framework for SMEs’ adoption of broadband Technological Context
Relative Advantage
Compatibility Environmental Context
Complexity
Location Trialability Industry Observability Market Scope
Competitive Pressure
External ICTV¶ Support
Government Pressure
SMEs Adoption of Broadband Organisational Context
Top Management Support
Environmental Context Organisational Readiness
ICTs¶([perience
Size
examined the impact of perceived usefulness and perceived ease-of-use and included compatibility DV D VLJQL¿FDQW IDFWRU (DUOLHU VWXGLHV VXFK DV Igbaria et al. (1997) used TAM to examine the impact of relative advantage and complexity on ICT innovations adoption and usage. Relative advantage, compatibility, and complexity have been examined in previous studies and have been VKRZQWREHVLJQL¿FDQW$VDQH[WHQVLRQWKLVVWXG\ LQWHQGV WR H[DPLQH DOO RI WKH ¿YH WHFKQRORJLFDO characteristics.
Organizational Context The characteristics in the organizational context VHHPWREHWKHSULPDU\IRFXVRIPDQ\60(V¶VWXGies (Premkumar, 2003). Top management support
510
has been found to be one of the best predictors of ICT innovations adoption by organizations (Jeyaraj et al., 2006). This factor has also been studied in the small business context (Premkumar, 2003; Guinea, Kelley, & Hunter, 2005). Furthermore, organizational readiness has been shown as an VLJQL¿FDQWRUJDQL]DWLRQDOIDFWRUWKDWLPSDFWVWKH adoption of ICT innovations (Iacovou et al., 1995; Mehrtens et al., 2001). Relevant ICTs’ experience variable have been examined in many studies (Lee, 2004; Lertwongsatien & Wongpinunwatana, 2003). Finally, empirical evidence on the impact of size shows mixed results (Damanpour, 1996; Fink, 1998; Goode & Stevens, 2000; Lertwongsatien & Wongpinunwatana, 2003; Levenburg, Magal, & Kosalge, 2006). Goode and Stevens (2000) study shows that business size, previously the best indicaWRURIWHFKQRORJ\DGRSWLRQZDVQRWVLJQL¿FDQWO\ related to Web adoption.
ICT innovations do not cater to just an internal auGLHQFHEXWDOVRWR¿UP¶VFXVWRPHUVVXSSOLHUVDQG business partner (Premkumar, 2003). Therefore, it is not surprising that environmental characteristics are increasingly being studied in ICT innovations adoption research. The geographical area where SMEs are located can be an obstacle especially for smaller companies (Arbore & Ordanini, 2006). The literature includes mixed empirical results on the impact of industry. On the one hand, it has EHHQDUJXHGWKDWWKHLQGXVWU\LQZKLFKWKH¿UP RSHUDWHV LQÀXHQFHV ,&7 LQQRYDWLRQV DGRSWLRQ (Raymond, 2001). On the other hand, evidence from Levy, Powell, and Yetton (2001) study show that WKHVHFWRUKDVDOLWWOHLQÀXHQFHRQ,&7DGRSWLRQ Other studies have examined the impact of market scope (Daniel & Grimshaw, 2002; Levenburg et al., 2006). Furthermore, the impact of competitive pressure has been examined (Daniel & Grimshaw, 2002; Premkumar & Roberts, 1999). Recent studies (Guinea et al., 2005; Thong, 2001) indicate that external ICTs’ supportLVDVLJQL¿FDQWIDFWRU)LQDOO\ the role of government in the take-up of broadband (government pressure) has been examined by Kuan and Chau (2001) and Scupola (2003).
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Table 4. SMEs’ description &ODVVL¿FDWLRQ Adopters
Firms F1
Industry Manufacturing
Type of Business
Market Scope
No. Employees
Manufacturer of feeders and
International
42
National
40
Local
15
International
200
controls F2
Manufacturing
Manufacturer and supplier of
F3
Retail/Wholesale
Retailer of domestic
DLU¿OWHUV appliances F4
Service
ICTs and Management Services Consultancy
Nonadopters
F5
Manufacturing
Food Manufacturing
Local
12
F6
Retail/Wholesale
Retailer of PCs and
Regional
124
F7
Manufacturing
Manufacturer of paints and
International
160
International
5
International
3
Components
powder coatings F8
Service
Change Management Consultancy
F9
Retail/Wholesale
Retailer of PCs and Components
EMPIRICAL STUDY To empirically examine the impact of TOE factors on SMEs adoption of broadband, case studies are a useful approach because of the contemporary nature of broadband. This approach is considered to be particularly appropriate when the boundaries of the research are not clear, and there is a need to investigate the issue within a real life context, drawing on the views of a number of sources (Yin, 1994). Furthermore, case study research provides a means to review theory and practice iteratively (Levy & Powell, 2003). The case studies employed here are all SMEs located in the Northwest of England. The chosen ¿UPVVDWLVI\ERWKRI'7,DQG(&GH¿QLWLRQVRIDQ SME. This is a convenient rather than a random sample of SMEs that were selected from manufacturing, retail/wholesale, and services industry sectors because they have greater dependence on ,&7V¶ XVDJH *RRGH 6WHYHQV 7R RYHUcome the pro-adoption bias (Rogers, 2003), this study focuses on both adopters and non-adopters of broadband.
The major method of data collection was through fact-to-face semistructured interviews, lasting between one to two hours, with personnel who had been key decision makers regarding ,&7LQQRYDWLRQVDGRSWLRQLQWKH¿UP 0DQDJLQJ 'LUHFWRU RU ,&7 0DQDJHU 60(V¶ NH\ GHFLVLRQ makers were contacted by phone to arrange for an interview. Most interviews were conducted DW WKH ¿UP¶V VLWH 7KH LQWHUYLHZ TXHVWLRQV ZHUH formulated to gather information on the factors that impact the adoption of broadband within the WKUHHFRQWH[WVRI7RUQDW]N\DQG)OHLVFKHU¶V PRGHO7KHTXHVWLRQVFRYHUHGWKH¿UPEDFNJURXQG WKH OHYHO RI ,&7V¶ VRSKLVWLFDWLRQ DQG XVH LQ WKH ¿UPDQGWKHLPSDFWRI72(IDFWRUVRQEURDGEDQG adoption (Appendix). Interview reports were sent to the interviewees subsequently for validation and UH¿QHPHQW0RUHRYHULQIRUPDWLRQZDVJDWKHUHG from secondary sources such as internal company documentation and Web sites.
511
Exploring SMEs’ Adoption of Broadband in the Northwest of England
EMPIRICAL FINDINGS 7KHDQDO\VLVRIWKHHPSLULFDO¿QGLQJVLVVWUXFWXUHG around the three contexts outlined in theoretical framework for SMEs adoption of broadband. Relative advantage, complexity, trialability, top PDQDJHPHQWVXSSRUW,&7V¶H[SHULHQFHORFDWLRQ FRPSHWLWLYHSUHVVXUHH[WHUQDO,&7V¶VXSSRUWDQG government pressure are found to be factors imSDFWLQJ60(V¶DGRSWLRQRIEURDGEDQG
)DQG)H[SUHVVHGWKDWEURDGEDQGZDVGLI¿FXOWWR LQLWLDOO\VHWXSEXWWKH\H[SHULHQFHGQRGLI¿FXOWLHV DIWHUZDUGV0RVW¿UPVH[SUHVVHGWKHLUFRQFHUQV not with implementing the technology, but with prior implementation issues. These issues are: selecting broadband vendor (F3, F5, F6 and F9), switching from narrowband (dial-up connection) to broadband (F6 and F7), and dealing with virus attacks (F7 and F9).
Trialability Technological Context This study found that relative advantage, complexLW\DQGWULDODELOLW\DUHIDFWRUVLPSDFWLQJ60(V¶ adoption of broadband.
Relative Advantage $OWKRXJKWKHSRWHQWLDOEHQH¿WVRIWKHDGRSWLRQRI broadband are well documented in the literature, it has been found that nonadopters (F6, F7, F8 DQG) VHHPWREHXQDZDUHRIWKHVHEHQH¿WV7KH primary reason for broadband adoption among adopters (F1, F2, F3, F4 and F5) is to have a reliable Internet connection. F1, F2 and F3 considered the adoption of e-commerce as one of the main drivers of broadband adoption. Because of the adoption of broadband, F4 and F5 intend to adopt hosted enterprise systems (such as ERP, CRM) in the next 3 years. Some adopters (F1, F2 and F3) claim experiencing productivity gains from broadband adoption.
Many adopters (F2, F4 and F5) have been approached by vendors to freely try broadband for a month before subscribing long-term. F2 and F5 expressed that having adopted broadband on trial EDVLVKDVFRQWULEXWHGWRWKHLU¿QDOGHFLVLRQ+DYLQJ adopted broadband on trial basis shows whether VZLWFKLQJ IURP QDUURZEDQG LV MXVWL¿DEOH ) Nonadopters are willing to adopt broadband on trial basis to be able to observe any difference.
Observability There seem to be an agreement among adopters and nonadopters that many SMEs are nowadays adopting broadband.
Organizational Context In addition to the technological characteristics, this study found that top management support and ,&7V¶H[SHULHQFHDUHDOVRLPSRUWDQWIDFWRUVLQWKH adoption of broadband among SMEs.
Compatibility Top Management Support $OORIWKHLQYHVWLJDWHG¿UPVKDGDQ,QWHUQHWFRQnection. Adopters perceive broadband to be compatible with main aspects of their work. Broadband compatibility does not seem to be an issue with nonadopters.
Complexity It has been found that adopters are comfortable with the complexity level of adopting broadband.
512
It has been found that adopters seem to have stronger managerial commitment towards adopting new technologies. This commitment has been characterized by their enthusiasm in adopting other cutting edge technologies such as enterprise systems (F1, F2 and F3). Compared to the adopters, it was found that most of the nonadopters have less enthusiastic top management.
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Organizational Readiness
Environmental Context
Although this study found that most investigated SMEs are ready to adopt broadband, nonadopters (F6, F7 and F9) do not have the resources to assess whether the adoption of broadband could or could not improve their business performance. Despite perceiving the cost to be high (F3, F5, F8 and F9), PRVWRIWKHLQYHVWLJDWHG60(VDUH¿QDQFLDOO\DQG technologically ready to adopt broadband.
In addition to technological characteristics and organizational characteristics mentioned earlier, this study found that location, competitive pressure, H[WHUQDO,&7V¶VXSSRUWDQGJRYHUQPHQWSUHVVXUH are important factors in the adoption of broadband among SMEs.
ICTs’ Experience 7DNLQJ6RXWKHUQDQG7LOOH\¶V FODVVL¿FDWLRQ into account, adopters can be categorized as highend ICT users (F1, F2 and F3) and medium-end ICT users (F4 and F5). However, most nonadopters (F7, F8 and F9) seem to be on the low-end of ,&7V¶XVDJH7KHVH¿UPVVHHPWREHKLQGHUHGIURP adopting broadband because of their lack of ICT skills and knowledge. Considering Ofcom SME segments discussed earlier, it seems that adopters FDQ EH FODVVL¿HG LQWR KLJKVSHQG IROORZHUV ) F2 and F3) and medium-spend non-tech (F4 and F5), whereas nonadopters can be categorized into low-spend tech (F6) and low-spend non-tech (F7, F8 and F9).
Arbore & Ordanini (2006) claim that one of the main factors for SMEs digital access divide is lack of environmental resources. This does not only refer to the availability of physical infrastructure, but also service providers and partners with expertise. A poor location can be penalizing for very VPDOO¿UPV6RPHRIWKHDGRSWHUV )DQG) DUH located in business parks that provide broadband VHUYLFHV$OWKRXJKPRVWRIWKHLQYHVWLJDWHG¿UPV are urbanely located, F8 and F9 are located in rural areas of the northwest. F7 expressed that broadband connectivity is not available in the area in which their business is located. However, F9 has not adopted broadband even though it is located in a business park that provides broadband connectivity.
Size
Industry
On one hand, traditional technologies tend to be expensive, relatively stable, and complex, and UHTXLUH VLJQL¿FDQW ¿QDQFLDO DQG SHUVRQQHO UHVRXUFHV7KXVRQO\ODUJHU¿UPVPLJKWEHDEOHWR adopt them, as they have the necessary resources, skills, knowledge, and experience (Damanpour, 1996; Montazemi, 1988). On the other hand, many newer technologies are less expensive, require OHVVVXSSRUWLQIUDVWUXFWXUHDQGRIIHU¿UPVDZD\ WRFRPSHWHZLWKODUJHU¿UPV *RRGH 6WHYHQV ,WKDVEHHQIRXQGWKDWPLFUR¿UPV )DQG ) DUHRQWKHQRQDGRSWHUVFDWHJRU\%RWK¿UPV expressed that it is unnecessary to adopt broadband because narrowband connection seem to deliver what they need. Although the size of the adopting ¿UPVLVPRUHWKDQHPSOR\HHV)DQG)DUH nonadopters.
Although Levy et al. (2001) claims that evidence VKRZWKDWWKHVHFWRUKDVDOLWWOHLQÀXHQFHRQ,&7 adoption, the industry of which the SME is a memEHUKDVEHHQIRXQGWREHDIDFWRUWKDWLQÀXHQFHV the adoption of ICT innovations (Levenburg et al., 2006; Raymond, 2001; Yap, 1990). Because it has EHHQVKRZQWKDWWKHLQGXVWU\LVDVLJQL¿FDQWIDFWRU in the adoption of new ICT innovations (Goode & Stevens, 2000), this study concentrated on the three industries that make more use of ICTs. The ¿QGLQJV RI WKLV VWXG\ DUH QRW FRQFOXVLYH VLQFH ¿UPVIURPWKHVDPHLQGXVWU\FDQEHDGRSWHUVDQG nonadopters.
Location
513
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Market Scope It was harder for SMEs to reach broader markets because of their limited resources (Levenburg & Klein, 2006). With the growing use of e-commerce, VPDOOHU¿UPVDUHDEOHWRUHDFKEURDGHUPDUNHWV Because SMEs are operating internationally, there is a clear need for SMEs to adopt broadband in order to reduce costs and improve their customer services. Broadband adopters have expressed their intentions for market expansions. Although some nonadopters (F7, F8 and F9) are operating internationally, they still have no plans to adopt broadband.
Competitive Pressure It has been argued that the use of ICT innovations to respond to competitors, provide enhanced customer service, and improve relationship with suppliers were driving these technologies uptake by smaller businesses compared to their larger counterparts (Daniel & Grimshaw, 2002). Adopters expressed that one of the main reasons for adopting broadband is competitive pressure. F3 argues that if an SME is willing to continue operating, it has to EHXSWRGDWHZLWK,&7V¶PDUNHWRIIHULQJV$OVR F1 stressed that if direct competitors are faster in adopting broadband, they have better chances to increase their market share. Nonadopters do not feel any competitive pressure to adopt broadband. F8 and F9 claim that they are still delivering the same service compared to other broadband adoptLQJ¿UPVLQWKHVDPHLQGXVWU\
External ICTs’ Support Driven by the need for lower costs, faster implementation, easier-to-use applications, and effective use of scarce resources, internal information system development is increasingly moving to an external development and provision model—outsourcing (Ward & Peppard, 2002). Lockett, Brown, and Kaewkitipong (2006) emphasises that the provision of hosted applications, by application service providers (ASPs) on a rented basis is viewed as of particular relevance to SMEs. With the availability
514
of broadband, it has been found that SMEs can feasibly adopt hosted enterprise systems (such as ERP, CRM, e-procurement). F4 is considering to adopt a hosted enterprise system. However, F5 expressed its concern with hosted enterprise systems and stated that it would only invest on client/server in-house application. All of the nonadopters expressed their concerns with security issues and lack of trust in broadband suppliers.
Government Pressure Nonadopters expressed their desire intervention for the government not in terms of subsidies and ¿QDQFLDOLQFHQWLYHVEXWSDUWLFXODUO\LQWHUPVRI awareness creation. It has been found that some adopters (F1 and F3) are aware of DTI initiatives, but most non-adopters (F6, F7, and F9) are unaware of any government initiatives with regard to ICTs adoption.
DISCUSSION OF FINDINGS 7KLVVWXG\VKRZVWKDW7RUQDW]N\DQG)OHLVFKHU¶V (1990) model is a guiding theoretical framework for understanding the factors affecting SMEs adoption of broadband. The application of this framework has contributed to the discussion of the impact of technological, organizational, and environmental factors on the adoption of broadband in small business context. Previous studies have focused on factors affecting other ICT innovations such as e-commerce (Lertwongsatien & Wongpinunwatana, 2003; Raymond, 2001), and EDI (Iacovou et al., 1995; Kuan & Chau, 2001). From Table 5, it can be deduced that ICT innovations are highly differentiated technologies for which there is not necessarily a single adoption model. This study FRQ¿UPV WKDW IDFWRUV LPSDFWLQJ 60(V DGRSWLRQ of broadband are different from factors impacting the adoption of other previously studied ICT innovations. With regard to the technological context, this study found that relative advantage, complexity, and trialability are factors impacting the adoption of broadband among SMEs. As previously estab-
Exploring SMEs’ Adoption of Broadband in the Northwest of England
7DEOH'LIIHUHQWIDFWRUVLPSDFWLQJ60(VDGRSWLRQRIGLIIHUHQW,&7LQQRYDWLRQV Factors
Studies of Factors Impacting Different ICT Innovations IT
Relative Advantage
(Fink, 1998)
Compatability
IS (Thong, 1999)
EDI
E-Commerce
(Iacovou et al.,
(Lertwongsatien &
1995); Kuan &
Wongpinunwatana, 2003;
Chau, 2001)
Scupola, 2003)
(Thong, 1999)
Broadband Current study
(Lertwongsatien & Wongpinunwatana, 2003)
Complexity
(Thong, 1999)
Current study
Trialability
Current study
Observability Top Management
(Lertwongsatien &
Support
Wongpinunwatana, 2003)
Organizational
(Fink, 1998)
Current study
(Iacovou et al.,
Readiness
1995; Kuan & Chau, 2001)
ICTs’ Experience
(Thong, 1999)
(Lertwongsatien &
Current study
Wongpinunwatana, 2003; Scupola, 2003) Size
Organizational
(Thong, 1999)
(Lertwongsatien &
(Arbore &
Wongpinunwatana, 2003)
Ordanini, 2006)
(Scupola, 2003)
Current study;
(Fink, 1998)
Culture IT Implementation
(Fink, 1998)
& Selection Location
(Arbore & Ordanini, 2006) Industry Market Scope Competitive
(Thong, 1999)
Pressure
(Kuan & Chau,
(Lertwongsatien &
2001)
Wongpinunwatana, 2003;
Current study
Scupola, 2003) External ICTs’
(Fink, 1998)
Current study
Support
(Arbore & Ordanini, 2006)
Government
(Kuan & Chau,
Pressure
2001)
External
(Fink, 1998)
Environment External Resources
(Scupola, 2003)
Current study
(Iacovou et al., 1995)
(Fink, 1998)
515
Exploring SMEs’ Adoption of Broadband in the Northwest of England
lished with studying SMEs adoption of other ICTs innovations, relative advantage is shown to be a factor that impacts the adoption of broadband. Lack of awareness among SMEs of broadband potenWLDOEHQH¿WVFDQKLQGHUWKHWDNHXSRIEURDGEDQG Also, complexity is found to be another factor that prevents SMEs from switching to broadband. Providing information about broadband potential, pricing, and support can help increase the take-up of broadband. Unlike previously studied ICT innovations, trialability is found to impact the adopWLRQRIEURDGEDQGDPRQJ60(V,&7V¶YHQGRUVDUH able to demonstrate the complexity level as well as demonstrating what broadband can deliver for SMEs by allowing them to adopt broadband on trial basis. With regard to the organizational context, this study found that top management support and ,&7V¶H[SHULHQFHWREHIDFWRUVLPSDFWLQJ60(V adoption of broadband. Like e-commerce, broadband adoption is determined by how supportive LVDQ60(¶VWRSPDQDJHPHQW7KHLPSDFWRIRUganizational readiness has been found to be weak EHFDXVHWKHDGRSWLRQRIEURDGEDQGLVQRW¿QDQFLDOO\ and technologically demanding. However, what SMEs seem to be concerned with is information with regard to broadband vendors, how to switch from narrowband to broadband, and how to deal with virus attacks. Furthermore, SMEs with more ,&7V¶H[SHULHQFHDUHVKRZQWREHPRUHOLNHO\WR DGRSW EURDGEDQG $OWKRXJK PLFUR¿UPV LQ WKLV study are shown to be nonadopters, the relationship between size and broadband adoption needs to be examined further. With regard to the environmental context, this study found that location, competitive pressure, H[WHUQDO,&7V¶VXSSRUWDQGJRYHUQPHQWSUHVVXUH to be factors impacting SMEs adoption of broadband. Supporting previous study of broadband adoption (Arbore & Ordanini, 2006), location DQGH[WHUQDO,&7V¶VXSSRUWKDYHEHHQIRXQGWREH factors impacting SMEs adoption of broadband. Location can be a deciding factor of whether an SME can or cannot adopt broadband. It can be a limiting factor if an SME is located in a rural DUHD0RUHRYHUH[WHUQDO,&7V¶VXSSRUWFDQSURYH WREHWKHPRVWFKDOOHQJLQJIDFWRUIRU,&7V¶YHQ-
516
dors since SMEs intending to adopt broadband have less concern with hosted applications. The ¿QGLQJV RI WKLV VWXG\ ZHUH QRW FRQFOXVLYH ZLWK regard to industry and market scope. Additionally, external pressure from competitors and suppliers is also shown to be an important factor. Finally, government pressure can increase the take-up of broadband among SMEs by raising their awareness RIXQGHUWDNHQ,&7V¶LQLWLDWLYHV
FUTURE TRENDS The availability of broadband enables more SMEs to adopt hosted applications (such as ERP, CRM, e-procurement) from ASPs. The adoption of these applications may challenge what we already know about the take-up of new ICT innovations among SMEs. Therefore, further research in take-up of these technologies will help understand the broader impact of broadband adoption. Also, it is outside the scope of this study to investigate the impact of other factors on individuals rather than on SMEs. Therefore, further studies may not only whish to differentiate between factors impacting organizations adoption of ICT innovations, but also differentiate between factors impacting individuals within SMEs. This study focused on a limited geographical DUHDZKLFKPDNHVLWGLI¿FXOWWRJHQHUDOL]HWKHUHsults to other UK regions. Although small number of cases investigated has drawn a good picture of the factors impacting the adoption of broadband, a survey of large sample would help generalize the results in the region. Finally, this study focused only on the adoption. To gain a holistic understanding of broadband, implementation process and the impact RIEURDGEDQGDGRSWLRQRQ¿UPSHUIRUPDQFHVKRXOG also be examined.
CONCLUSION This study has established that ICT innovations are highly differentiated technologies for which there is not necessarily a single adoption model. 7KLVVWXG\FRQ¿UPVWKDWGLIIHUHQWWHFKQRORJLFDO
Exploring SMEs’ Adoption of Broadband in the Northwest of England
organizational, and environmental factors impact the adoption of different ICT innovations. Factors impacting SMEs adoption of broadband are different from factors impacting the adoption of other previously studied ICT innovations (e.g., EDI and e-commerce). The major contribution of this study is highlighting these factors, which are relative advantage, complexity, trialability, top PDQDJHPHQWVXSSRUW,&7V¶H[SHULHQFHORFDWLRQ FRPSHWLWLYHSUHVVXUHH[WHUQDO,&7V¶VXSSRUWDQG government pressure. 7KHPDLQLPSOLFDWLRQIRUWKLVVWXG\LVWKDW,&7V¶ vendors can increase broadband rate of adoption by creating awareness and demonstrating the impact RI EURDGEDQG DGRSWLRQ RQ 60(V¶ SHUIRUPDQFH Furthermore, government agencies advocating that ICTs are important and have the potential to contribute to more rapid growth and productivity gains (Pilat, 2003) can help communicate the SRWHQWLDOEHQH¿WVRIEURDGEDQGDGRSWLRQ
REFERENCES Agarwal, R., & Prasad, J. (1997). The role of innovation characteristics and perceived voluntariness in the acceptance of information technologies. Decision Sciences, 28(3), 557-582. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision 3URFHVVHV(2), 179-211. Anderson, J.E., & Schwager, P.H. (2003). SMEs adoption of wireless LAN technology: Applying UTAUT model, In Proceedings of the 7th Annual Conference of the Southern Association for Information Systems. Retrieved July 15, 2007, from http://sais.aisnet.org/sais2004/ Anderson%20&%20Schwager.pdf Arbore, A., & Ordanini, A. (2006). Broadband divide among SMEs. International Small Business Journal, 24(1), 83-99. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Blackburn, R., & McClure, R. (1998). The use of information and communication technologies ,&7V LQVPDOOEXVLQHVVVHUYLFH¿UPVKingston Business School, UK: Small Business Research Center. Braun, P. (2002). Digital knowledge networks: Linking communities of practice with innovation. Journal of Business Strategies, 19(1), 43-54. Brock, J.K. (2000). Information and communicaWLRQWHFKQRORJ\LQWKHVPDOO¿UP,Q'-RQHV(YDQV & S. Carter (Eds.), Enterprise and small business: Principles, practice and policy (pp. 384-408). Harlow, England: FT - Prentice Hall. Caldeira, M.M., & Ward, J.M. (2003). Using resource-based theory to interpret the successful adoption and use of information systems and technology in manufacturing small and mediumsized enterprises. European Journal of Information Systems, 12(2), 127-141. Chau, P.Y.K. (2001). Inhibitors to EDI adoption in small businesses: An empirical investigation. Journal of Electronic Commerce Research, 2(2), 78-88. Chau, P.Y.K., & Tam, K.Y. (1997). Factors affecting the adoption of open systems: An exploratory study. MIS Quarterly, 21(1), 1-24. Corso, M., Martini, A., Paolucci, E., & Pellegrini, L. (2001). Information and communication technologies in product innovation within SMEs – the role of product complexity. Enterprise & Innovation Management Studies, 2(1), 35-48. Routledge, Ltd. &UDJJ3% .LQJ0 6PDOO¿UPFRPputing: Motivators and inhibitors. MIS Quarterly, 17(1), 47-60. Damanpour, F. (1996). Organizational complexity and innovation: Developing and testing multiple contingency models. Management Science, 42(5), 693. Daniel, E.M., & Grimshaw, D.J. (2002). An exploratory comparison of electronic commerce adoption in large and small enterprises. Journal of Information Technology, 17(3), 133-147.
517
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Daniel, E., Wilson, H., & Myers, A. (2002). Adoption of e-commerce by SMEs in the U.K.: Towards a stage model. International Small Business Journal, 20(3), 253-268.
Grandon, E.E., & Pearson, J.M. (2004). Electronic commerce adoption: An empirical study of small and medium U.S. businesses. Information & Management, 42(1), 197.
Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340.
Grewal, R. (2001). An investigation into the antecedents of organizational participation in business-to-business electronic markets. Journal of 0DUNHWLQJ(3), 17.
DTI. (2001). Business in the information age: International benchmarking report. Department of Trade & Industry. Retrieved May, 2003 from: http://www.ukonlineforbusiness.gov.uk/main/resources/publication-htm/bench2001.htm DTI. (2004). Business in the information age: International benchmarking study 2004. Department of Trade & Industry. Retrieved January, 2005 from: http://www2.bah.com/dti2004 EC. (2003). SME user guide explaining the new 60( GH¿QLWLRQ 5HWULHYHG -XO\ IURP http://ec.europa.eu/enterprise/enterprise_policy/ VPHBGH¿QLWLRQLQGH[BHQKWP EU. (2004). Communication connecting Europe at high speed: National broadband strategies. COM. Brussels: European Union. Feeny, D.F., & Willcocks, L.P. (1998). Re-designing the IS function around core capabilities. Long Range Planning, 31(3), 354-367. Fichman, R.G. (2004). Going beyond the dominant paradigm for information technology innovation research: Emerging concepts and methods. Journal of the Association for Information Systems, (8), 314. Fink, D. (1998). Guidelines for the successful adoption of information technology in small and medium enterprises. International Journal of Information Management, 18(4), 243-253. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, and behavior: An introduction to theory and research. London: Addison-Wesley. Goode, S., & Stevens, K. (2000). An analysis of the business characteristics of adopters and non-adopters of world wide Web technology. Information Technology and Management, 1(1/2), 129-154.
518
Guinea, A.O., Kelley, H., & Hunter, M.G. (2005). Information systems effectiveness in small businesses: Extending a Singaporean model in Canada. Journal of Global Information Management, 13(3), 55-79. Hamel, G. (1998). The challenge today: Changing the roles of the game. Business Strategy Review, 9(2), 19-26. Harrison, A.D., Mykytyn, Jr, P.P., & Riemenschneider, K.C. (1997). Excutive decision about adoption of information technology in small business: Theory and empirical tests. Information Systems Research, 8(2), 171-195. Iacovou, C.L., Benbasat, I., & Dexter, A.S. (1995). Electronic data interchange and small organizations: Adoption and impact of technology. MIS Quarterly, 19(4), 465-485. Igbaria, M., & Tan, M. (1997). The consequences of information technology acceptance on subsequent individual performance. Information & Management, 32(3), 113-121. Igbaria, M., Zinatelli, N., Cragg, P., & Cavaye, A.L.M. (1997). Personal computing acceptance IDFWRUVLQVPDOO¿UPV $VWUXFWXUDOHTXDWLRQPRGHO MIS Quarterly, 21(3), 279-302. Jeyaraj, A., Rottman, J.W., & Lacity, M.C. (2006). A review of the predictors, linkages, and biases in IT innovation adoption research. Journal of Information Technology, 21(1), 1-23. Kuan, K.K.Y., & Chau, P.Y.K. (2001). A perceptionbased model for EDI adoption in small businesses using a technology-organization-environment framework. Information & Management, 38(8), 507-521.
Exploring SMEs’ Adoption of Broadband in the Northwest of England
Kwon, T.H., & Zmud, R.W. (1987). Unifying the fragmented models of information systems implementation. In R.J. Boland & R.A. Hirschheim (Eds.), Critical issues in information systems research (pp. 227-251). New York: John Wiley & Sons. Lee, J. (2004). Discriminant analysis of technology adoption behavior: A case of Internet technologies in small businesses. Journal of Computer Information Systems, 44(4), 57-66. Lertwongsatien, C., & Wongpinunwatana, N. (2003). E-commerce adoption in Thailand: An empirical study of small and medium enterprises (SMEs). Journal of Global Information Technology Management, 6(3), 67-83. Levenburg, N.M., & Klein, H.A. (2006). Delivering customer services online: Identifying best practices of medium-sized enterprises. Information Systems Journal, 16(2), 135-155. Levenburg, N., Magal, S.R., & Kosalge, P. (2006). An exploratory investigation of organizational factors and e-business motivations among SMFOEs in the US. Electronic Markets, 16(1), 70-84. Levy, M., & Powell, P. (2003). Exploring SME Internet adoption: Towards a contingent model. Electronic Markets, 13(2), 173-181. Levy, M., Powell, P., & Yetton, P. (2001). SMEs: Aligning IS and the strategic context. Journal of Information Technology, 16(3), 133-144. Levy, M., Powell, P., & Yetton, P. (2002). The dynamics of SME information systems. Small Business Economics, 19(4), 341-354. Lockett, N., Brown, D.H., & Kaewkitipong, L. (2006). The use of hosted enterprise applications by SMEs: A dual market and user perspective. Electronic Markets, 16(1), 85-96. Markus, M. L. and Tanis, C. (2000). The enterprise system experience - from adoption to success. In Zmud, R. W. (Ed.), Framing the domain of IT management: Projecting the future...through the past (pp. 173-207). &LQFLQQDWL 2KLR 3LQQDÀH[ Education Resources.
Martin, L.M., & Matlay, H. (2001). “Blanket” apSURDFKHVWRSURPRWLQJ,&7LQVPDOO¿UPV 6RPH lessons from the DTI ladder adoption model in the UK. Internet Research: Electronic Networking Applications and Policy, 11(5), 399-410. Mata, F.J., Fuerst, W.L., & Barney, J.B. (1995). Information technology and sustained competitive advantage: A resource-based analysis. MIS Quarterly, 19(4), 487-505. Mehrtens, J., Cragg, P.B., & Mills, A.M. (2001). A model of Internet adoption by SMEs. Information & Management, 39(3), 165-176. Montazemi, A.R. (1988). Factors affecting information satisfaction in the context of the small business environment. MIS Quarterly, 12(2), 238-256. Moore, G.C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, 2(3), 192-222. Ofcom. (2004). Strategic review of telecommunications: Phase 2 consultation document (Annex M: Consumer Research). Retrieved July 15, 2007, from http://www.ofcom.org.uk/consult/condocs/ telecoms_p2/tsrphase2/AnnexM.pdf Pilat, D. (2003). ICT and economic growth: EviGHQFHIURP2(&'FRXQWULHVLQGXVWULHVDQG¿UPV OECD. Retrieved July 15, 2007, from http://www. oedc.org Poon, S., & Swatman, P.M.C. (1997). Small business use of the Internet: Findings from Australian case studies. International Marketing Review, 14(5), 385-402. Poon, S., & Swatman, P. (1999). An exploratory study of small business Internet commerce issues. ,QIRUPDWLRQ 0DQDJHPHQW(1), 9-18. Prananto, A., Marshall, P., & McKay, J. (2003). A study of the progression of e-business maturity in Australian SMEs: Some evidence of the applicability of the stages of growth for e-business model. In Hiller, M., Falconer, D., Hanish, J. & Horrocks, S. (eds.), 3URFHHGLQJVRIWKHWK3DFL¿F$VLD&RQIHU-
519
Exploring SMEs’ Adoption of Broadband in the Northwest of England
ence on Information Systems (PACIS). University of South Australia, Adelaide, South Australia. Premkumar, G. (2003). A meta-analysis of research on information technology implementation in small business. Journal of Organizational Computing and Electronic Commerce, 13(2), 91-121. Premkumar, G., Ramamurthy, K., & Nilakanta, S. (1994). Implementation of electronic data interchange: An innovation diffusion percpective. Journal of Management Information Systems, 11(2), 157-186.
Ritchie, R., & Brindley, C. (2000). Disintermediation, disintegration and risk in the SME global supply chain. Management Decision, 38(8), 575583. Rogers, E.M. (1983). Diffusion of innovations. New York: Free Press. Rogers, E.M. (1995). Diffusion of innovations. New York: Free Press. Rogers, E.M. (2003). Diffusion of innovations. New York: Free Press.
Premkumar, G., & Roberts, M. (1999). Adoption of new information technologies in rural small businesses. Omega, 27(4), 467-484.
Sawyer, S., Allen, J.P., & Heejin, L. (2003). Broadband and mobile opportunities: A socio-technical perspective. Journal of Information Technology, 18(2), 121.
Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363.
SBS. (2006). Statistical press release. Retrieved July 15, 2007, from http://www.sbs.gov.uk/SBS_ *RYB¿OHVUHVHDUFKDQGVWDWV60(VWDWVSUSGI
Quayle, M. (2002). E-commerce: The challenge IRU8.60(VLQWKHWZHQW\¿UVWFHQWXU\International Journal of Operations and Production Management, 22(10), 1148-1161.
Scupola, A. (2003). The adoption of Internet commerce by SMEs in the south of Italy: An environmental, technological and organizational perspective. Journal of Global Information Technology Management, 6(1), 52-71.
Rao, S.S., Metts, G., & Mora, M.C.A. (2003). Electronic commerce development in small and medium sized enterprises: A stage model and its implications. Business Process Management, 9(1), 11-32.
Shang, S., & Seddon, P.B. (2002). Assessing and PDQDJLQJWKHEHQH¿WVRIHQWHUSULVHV\VWHPV 7KH EXVLQHVVPDQDJHU¶VSHUVSHFWLYHInformation Systems Journal, 12(4), 271-299.
Ravichandran, T. (2000). Swiftness and intensity of administrative innovation adoption: An empirical study of TQM in information systems. Decision Sciences, 31(3), 691-724. Raymond, L. (1989). Management information systems: Problems and opportunities. International Small Business Journal, 7(4), 44-53. Raymond, L. (2001). Determinants of Website implementation in small businesses. Internet Research, 11(5), 411-422. Riemenschneider, K.C., Harrison, A.D., & Mykytyn, Jr., P.P. (2003). Understanding IT adoption decisions in small business: Integrating current theories. Information & Management, 40(4), 269-285.
520
6KLHOV + 0F,YRU 5 2¶5HLOO\ ' Understanding the implications of ICT adoption: Insights from SMEs. Logistics Information Management, 16(5), 312-326. Simpson, M., & Docherty, A.J. (2004). E-commerce adoption support and advice for U.K. SMEs. Journal of Small Business and Enterprise Development, 11(3), 315-328. 6RXWKHUQ$ 7LOOH\) 6PDOO¿UPVDQGLQformation and communication technologies (ICTs): Toward a typology of ICTs usage. New Technology, :RUN (PSOR\PHQW(2), 138-154. Sparkes, A., & Thomas, B. (2001). The use of the Internet as a critical success factor for the marketing
Exploring SMEs’ Adoption of Broadband in the Northwest of England
of Welsh agri-food SMEs. Twenty First Century British Food Journal, 103(4), 331-347. Story, D. (1994). Understanding the small business sector. London, U.K.: Routledge. Supurge, V., & Roberts, C. (2005). Broadband technology: An appraisal of government policy and use by small-and medium-sized enterprises. Journal of Property Investment and Finance, 23(6), 516-524. Swanson, E.B. (1994). Information systems innovation among organizations. Management Science, 40(9), 1069-1092. Taylor, M., & Murphy, A. (2004). SMEs and ebusiness. Journal of Small Business and Enterprise Development, 11(3), 280-289. Thong, J.Y.L. (1999). An integrated model of information systems adoption in small businesses. Journal of Management Information Systems, (4), 187-214. Thong, J.Y.L. (2001). Resource constraints and information systems implementation in Singaporean small businesses. Omega, 29(2), 143-156. Thong, J.Y.L., & Yap, C.-S. (1994). Engagement of external expertise in information systems implementation. Journal of Management Information Systems, 11(2), 209. Thong, J.Y.L., & Yap, C.S. (1995). CEO characteristics, organizational characteristics and information technology adoption in small businesses. Omega, 23(4), 429-442. Tornatzky, L.G., & Fleischer, M. (1990). The process of technological innovation. Lexington, MA: Lexington Books. UK-Online. (2002). UK online annual report. London: DTI. Venkatesh, V., & Davis, F.D. (2000). A theoretical extension of the technology acceptance model: Four ORQJLWXGLQDO¿HOGVWXGLHVManagement Science, 46(2), 186-204.
Venkatesh, V., Morris, M.G., Davis, G.B., & Davis, F.D. (2003). User acceptance of information WHFKQRORJ\ 7RZDUGDXQL¿HGYLHZMIS Quarterly, 27(3), 425-478. Vescovi, T. (2000). Internet communication: The Italian SME Case. Corporate Communications: $Q,QWHUQDWLRQDO-RXUQDO(2), 107-112. Ward, J., & Peppard, J. (2002). Strategic planning for information systems. West Sussex: John Wiley & Sons. Yap, C.S. (1990). Distinguishing characteristics of organizations using computers. Information & Management, 18(2), 97-107. Yin, R. (1994). Case study research: Design and methods. California: Sage.
KEY TERMS ADSL: Asymmetric digital subscriber line. This is the most used form of DSL. ASP: Application service provider. A company that offers access to applications over the Internet. Broadband: To qualify as broadband, a service must deliver an “always on” service at rates of more than 128 Kbps (DTI, 2004). DSL: Digital subscriber line is a technology for bringing high-bandwidth information to homes and small businesses over ordinary telephone lines. EDI: Electronic data interchange is structured form of data interchange between businesses. Enterprise Systems: “Commercial software packages that enable the integration of transactionoriented data and business processes throughout an organization (and perhaps eventually throughout the entire interorganizational supply chain)” (Markus & Tanis, 2000: 176). The systems include enterprise resource planning (ERP), customer relationship management (CRM), supply chain management (SCM), product life-cycle manage-
521
Exploring SMEs’ Adoption of Broadband in the Northwest of England
ment (PLM) and e-procurement software (Shang & Seddon, 2002). ICTs: Information and communication technologies. A broad term referring to digital technologies such as information networks and software applications as well as telecommunication technologies such as telephony, cable, and satellite. ISP: Internet service provider. A company that provides Internet access services to consumers and businesses.
522
Narrowband: Internet access through a standard modem (dialup speeds can vary between 28.8kbps or 56.6kbps). SMEs: Small and medium-sized enterprises. The Department of Trade and Industry (DTI) differentiates businesses by the number of employees PLFUR¿UP HPSOR\HHVVPDOO¿UP HPSOR\HHVDQGPHGLXP¿UP HPSOR\HHV
Exploring SMEs’ Adoption of Broadband in the Northwest of England
APPENDIX Interview Questions Firm Background: &DQ\RXWHOOPHDERXW\RXU¿UP¶VEDFNJURXQG" (number of employees/ industry/ age) Level of ICTs Sophistication & Use in the Firm: 2. What ICTs does your business have? +RZUHOHYDQWDUHWKHVH,&7VWRWKHEXVLQHVVRI\RXU¿UP" 4. What was the process of selecting these ICTs? 5. What ICTs does your business need in the future? Why? 6. How effective is your business in exploiting new ICTs? 7. What challenges do think SMEs are confronted with in the adoption of new ICTs? Why? Impact of TOE Factors on Broadband Adoption: :KDWWHFKQRORJLFDOIDFWRUVGR\RXWKLQNLPSDFWWKHDGRSWLRQRIEURDGEDQGLQ\RXU¿UP" Why? 9. What is the impact of technological factors on the adoption of broadband? (relative advantage/ compatibility/ complexity/ trialability/ observability) :KDWRUJDQL]DWLRQDOIDFWRUVGR\RXWKLQNLPSDFWWKHDGRSWLRQRIEURDGEDQGLQ\RXU¿UP" Why? 11. What is the impact of organizational factors on the adoption of broadband? (top management support/ organizational readiness/ ICTs experience/ size) :KDWHQYLURQPHQWDOIDFWRUVGR\RXWKLQNLPSDFWWKHDGRSWLRQRIEURDGEDQGLQ\RXU¿UP" Why? 13. What is the impact of environmental factors on the adoption of broadband? (location/ industry/ market scope/ competitive pressure/ external ICTs support/ government pressure)
523
524
Chapter XXXIII
External Pressures for Adoption of ICT Services Among SMEs Andrea Ordanini Bocconi University, Italy Alessandro Arbore Bocconi University, Italy
ABSTRACT This study intends to emphasize the importance that external sources of pressure may have on the level of ICT involvement among small and medium enterprises (SMEs) in Italy. While past research tends to prioritize the role of endogenous conditions for the adoption of information and communication technoloJLHV²WKDWLV¿QDQFLDOUHVRXUFHVRUJDQL]DWLRQFRQGLWLRQVPDQDJHPHQWFXOWXUH²WKHKLJKGHSHQGHQFH of SMEs on their environment requires paying especial attention to external pressures as well. Both competitive and institutional pressures are proposed and tested through an ordinal regression model on DVDPSOHRI60(V7KHUHVXOWVVXJJHVWERWKSROLF\DQGPDQDJHPHQWLPSOLFDWLRQV
INTRODUCTION 7KHGLIIXVLRQRI,&7VROXWLRQVDPRQJ¿UPVLVD key element at the top of management and policy agendas, and it has been widely investigated since the mid 1990s. In particular, drivers and barriers to adoption have been deeply studied. Prior research mainly explained the adoption of ICT in terms of endogenous factors, that is, variables which are internal to the organization. Among these factors, three broad typologies were UHFXUUHQW WKHOHYHORI¿QDQFLDOUHVRXUFHVDEOHWR affect any investment decision; the managerial FXOWXUHLQÀXHQFLQJWKHSURSHQVLW\WRLQQRYDWH DQG
the organizational readiness, which is relevant for the integration of new technologies. These arguments seem particularly consistent in explaining the difference in Internet adoption EHWZHHQVPDOODQGODUJH¿UPVVLQFHWKH\UHÀHFW asymmetric internal conditions largely based on size. Conversely, when considering only small and medium enterprises (SMEs), it should be pointed RXWWKDWWKHVH¿UPVDSSHDUPRUHVHQVLEOHWRH[WHUQDO pressures, in terms of both competitive and social “rules of the game” (Fink, 1998). In this sense, the SRWHQWLDOLQÀXHQFHRIH[RJHQRXVSUHVVXUHVVXFK as competitive pressures or institutional pressures, have been understated, and the empirical contribu-
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
External Pressures for Adoption of ICT Services Among SMEs
tions often provided mixed evidence on the role of such predictors (see, for a review, Grandon & Pearson, 2004; Lee, Runge, & Baek, 2001; Zhu, .UDHPHU ;X This chapter presents and tests a framework, rooted on strategic management and institutional theories, where the role of the environment may assume different meanings: three of them—rivalry competition, leading market position, and international presence—are drawn by strategic management theories and two of them—business group membership, and technology legitimacy—are explained by institutional theory. This model is tested through an ordinal regression methodology in a sample of 285 Italian SMEs. By controlling IRUVL]HDQGLQGXVWU\RXWFRPHVUHYHDODVLJQL¿FDQW association of ICT adoption with the hypothesized drivers, except for international presence. The chapter is organized as follows: an introduction of the theoretical background, an explanation of the formulation of hypotheses, a description of the methodology of the analysis, a presentation RIWKHRXWFRPHVDQG¿QDOO\DGLVFXVVLRQRIWKH results and their implications.
THEORETICAL BACKGROUND The adoption of ICT solutions in the information systems literature has been investigated according to two different theoretical approaches, which have been adapted from more general theories on technology adoption: the technology acceptance model (TAM) (Davis, 1989), and the technologyorganization-environment (TOE) (Thornatzky & Fleischer, 1990). The TAM approach put the emphasis on a set of SUHGLFWRUVODUJHO\HQGRJHQRXVLQWHUQDOWRWKH¿UP According to this model, the adoption of ICT solutions should be driven by the perceived usefulness of the technology, its ease of use, and the cultural orientation toward innovation of the decision maker (Lederer, Maupin, Sena, & Zhuang, 2000). Such an approach inspired several empirical analyses, HVSHFLDOO\IRFXVHGRQWKH¿UVWW\SHRILQIRUPDWLRQ systems innovations, those whose effects end up ZLWKLQWKH¿UP 6ZDQVRQ 7KHNH\SRLQWLV
that the TAM framework was anchored to the idea that the technology adoption lays in the hands of WKH ¿UP¶V GHFLVLRQ PDNHU DQG RI WKH SHUFHSWLRQ about the technology. In this picture, no attention LVGHGLFDWHGWRSRWHQWLDOSUHVVXUHVRULQÀXHQFHV that may come from the environment conditions, RXWVLGHWKH¿UP An alternative theoretical approach employed to investigate the adoption of ICT solutions was SURYLGHGE\WKH72(IUDPHZRUN7KLVLGHQWL¿HG three broad categories of predictors: the features of technology, the organizational readiness of the ¿UPDQGWKHHQYLURQPHQWDOFRQGLWLRQV ,DFRYRX Benbasat, & Dexter, 1995). Although the external dimension was introduced, the subsequent empirical analyses interpreted and treated such environmental conditions in different ways: as competitive SUHVVXUHVDVVRFLDOLQÀXHQFHVDVHQYLURQPHQWDO turbulence, and, in some cases, as a generic perception of the environmental features. According WRWKLVLWLVQRWVXUSULVLQJWR¿QGWKDWWKHDQDO\VHV on the effects of environmental conditions on ICT gave mixed evidences (Kuan & Chau, 2001). For these reasons we will go in deep with the analysis on these external predictors in order to investigate which dimensions are relevant and which effect they have on the process of ICT adoption E\¿UPV7KHVHWRISUHGLFWRUVDQGFRUUHVSRQGLQJ K\SRWKHVHVZLOOEHLGHQWL¿HGWKURXJKERWKVWUDWHJLF management theories, which provide explanations of how different dynamics (rivalry competition, leading market position, and international presence) PD\LQÀXHQFHWKH,&7DGRSWLRQDQGLQVWLWXWLRQDO theories, focused on business group membership and technology legitimacy.
COMPETITIVE PRESSURES FOR ICT ADOPTION Rivalry Competition Economic theories largely recognized the imporWDQFHRIDQRUJDQL]DWLRQ¶VHQYLURQPHQWDOFRQWH[W for innovation and adoption of new technologies (Kimberley & Evanisko, 1981; Kwon & Zmud 1987). Among external forces, a key role is played
525
External Pressures for Adoption of ICT Services Among SMEs
by competition mechanisms, which can be observed in the perspective of strategic management mainstream theories: the industrial organization framework and the resource-based view (Hoskisson, Wan, Yiu, & Hitt, 1999). Following the industrial organization (IO) framework (Andrews, 1971; Porter, 1985), it can be observed that an increasing competitive intensity worsens the balance between opportunities and WKUHDWVUHTXLULQJ¿UPVWRDGRSWPRUHLQQRYDWLYH VWUDWHJLHVWRVXUYLYHDQGPDNHSUR¿WV7KLVKRZever, is only one possible explanation provided by IO literature, since IO theories support both sides of the argument for the question of whether DPRQRSROLVWRUDFRPSHWLWLYH¿UPLQQRYDWHVPRUH Since ICT solutions are tools to make organizations more competitive, we expect that the strength of competitive rivalry may be an important driver for new technology adoption (Utterback, 1974). Within this view, and regarding ICT, Gatignon and Robertson (1989) found that competitive pressure in the adopter industry has a positive impact on adoption of new technologies. In the same way, Thong and Yap (1995) found that the CEO attitude towards new technologies adoption is positively correlated with the degree of competition faced in the market. Similarly, Premkumar and Roberts (1999) demonstrated that the degree of competitive ULYDOU\LQWKHDGRSWHU¶VLQGXVWU\DIIHFWVWKHUDWHRI adoption of digital technologies. In a slightly different perspective, Bhattacharya and Bloch (2004) recently found that in concentrated industries drivers of innovation are mainly endogenous, while fragmented contexts are more dependent from external sources to innovate. Also the resource-based view (Barney, 1991; 3HWHUDI :HUQHUIHOW LGHQWL¿HVDQDVsociation between the degree of competition and WKHLQQRYDWLYHEHKDYLRURIWKH¿UPV$FFRUGLQJWR this view, the strength of competition in the factor markets reduces the power of isolating mechanisms sustaining the competitive edge, thus requiring continuous innovation strategies. Within these contexts, innovations lead to resource substitution phenomenon, acting as basic conditions for rent seeking (Malerba & Orsenigo, 1997). Especially here, then, ICT solutions may contribute to the
526
development of dynamic capabilities sustaining long-term competitive advantages (Teece, Pisano, & Shuen, 1997). These considerations lead to the ¿UVWK\SRWKHVLVRIRXUDQDO\VLV Hp1: The intensity of rivalry competition perceived by a SME positively affects its level of ICT adoption.
Leading Market Position In addition, given a certain level of competitive rivalry, a leading market position1 may reveal better resources to exploit new technology potential. This, in turn, would be a further driver for ICT services adoption. From a theoretical point of view, there is a general consensus on the fact that the adoption of information technologies, by itself, has a marginal GLUHFWHIIHFWRQ¿UPSHUIRUPDQFH)XUWKHUPRUHLW has been pointed out that the causal links between market position and ICT adoption are not clear: VRPHWLPHV¿UPVRFFXS\DOHDGLQJPDUNHWSRVLWLRQ SUHFLVHO\ EHFDXVH WKH\ HI¿FLHQWO\ LPSOHPHQWHG ICT. ,QIDFWVLJQL¿FDQWLPSDFWVHPHUJHRQO\ZKHQ such technologies are combined and integrated with other distinctive competencies (see, among the others, Clemons & Row, 1991; Mata, Fuerst, & Barney, 1995). Recent evidence supports this interpretation (OECD, 2003; Powell & Dent-Micallef, 1997). It appears that digital technologies are really useful only when they leverage on preexisting key resources. A possible implication is that Internet solutions, like other ICTs, would have greater power in consolidating leading positions rather than reduce competitive gaps. Following these considerations, SMEs feeling in a leading market position would be more likely to adopt new technologies than “marginal” competitors, which would prefer to exploit existing knowledge and capabilities rather than exploring new possibilities (Leonard-Barton, 1992). Since ICT solutions can improve competitiveness especially when combined with the appropriate strategic resources (Powell & Dent-Micallef, 1997), we expect that a leading market position positively
External Pressures for Adoption of ICT Services Among SMEs
affects the probability of adoption. This leads to the second hypothesis. Hp2: To be in a leading market position positively affects the level of ICT adoption among SMEs.
the group (Premkumar & Roberts, 1999). These considerations lead to the third hypothesis. Hp3: Business group membership positively LQÀXHQFHVWKHOHYHORI,&7DGRSWLRQ
International Presence INSTITUTIONAL PRESSURES FOR ICT ADOPTION Business Group Membership A further environmental source of pressure to adopt ICT solutions may come from legitimacy UHTXLUHPHQWVLQWKHRUJDQL]DWLRQDO¿HOG This motivation is rooted in the so-called inVWLWXWLRQDOWKHRULHVDVVHVVLQJWKDW¿UP¶VEHKDYLRU has to be coherent with norms and social rules requested by their environment (Di Maggio & Powell, 1983; North, 1990). Past researches reveal that a typical coercive pressure to adopt information and communication technologies is usually exerted by external partners (Grover, 1993; Premkumar & Ramamurthy, 7KHSUHVVXUHUHÀHFWVWKHIDFWWKDW,&7VDUH interdependent technologies whose utility increases as other members adopt (Markus, 1987). A relevant case occurs when an SME is integrated with other organizations because part of a larger business group. In this case, the SME would receive parental pressure to follow corporate strategies aiming at improving communication and coordination within the group (Premkumar & Roberts, 1999). In addition, a group membership might affect the propensity to adopt new technologies by affecting the culture, the organization, and the DFWXDO¿QDQFLDOUHVRXUFHVRIDVPDOOVXEVLGLDU\ A large holding company, for example, may use its size advantage to experiment with innovations (either directly or testing them in peripheral units), then pushing any promising technology to other ¿UPVRIWKHJURXSLQFOXGLQJHYHQWXDO60(V To sum up, whereas an SME is part of a business group, it may receive further pressures to adopt ICT solutions from its parent company, also to improve communication and coordination within
As recently noted by Rogers (2004), most innovative SMEs have higher inclination to export and exploit overseas markets to gain access to new sources of innovation. Other researches indicated WKH OHYHO RI H[SRUW DV D VLJQL¿FDQW SUHGLFWRU IRU the adoption of ICT (Battacharya & Bloch, 2004; Lal, 1999).2 More generally, intensive interactions with foreign counterparts would expose an SME to HQULFKHG NQRZOHGJH ÀRZV ZKLFK PLJKW SURYLGH stronger stimulus to adopt new technologies. For instance, Iacovou et al. (1995) found that small ¿UPVPD\DOWHUWKHLUJHQHUDOSHUFHSWLRQRIKRZWR become more effective via the use of EDI, when they interact with foreign larger organizations. Accordingly, by the extent to which an SME is able to learn from these interactions and absorb IURPRWKHUV¶H[SHULHQFH &RKHQ /HYLQWKDO D VLJQL¿FDQW SUHVHQFH LQ LQWHUQDWLRQDO PDUNHWV might increase the level of ICT adoption. According to this, the presence on international markets, as can be measured by the export activity, exposes an SME to business interactions DQGNQRZOHGJHÀRZVZKLFKPD\SURYLGHVWURQJHU stimulus to adopt new technologies (Rogers, 2004). These considerations lead to a fourth hypothesis. Hp4: A higher level of international presence SRVLWLYHO\LQÀXHQFHVWKHOHYHORI,&7DGRSWLRQ
Technology Legitimacy ,Q RUGHU WR JDLQ DFFHVV WR VSHFL¿F UHVRXUFHV WR collaborative networks, or to strategic alliances, a ¿UPPLJKWDOVREHVXEMHFWWROHJLWLPDF\DVVHVVPHQW by other social agents (competitors, partners, and other stakeholders). The requirements may be very selective especially for SMEs, since they are not usually perceived as legitimate players, due to their lack of resources and capabilities (Grewal, Comer, & Metha, 2001). Our hypothesis, then, is that, in
527
External Pressures for Adoption of ICT Services Among SMEs
some context, the adoption of innovative service solutions, like ICT platforms, may be driven, among the other things, by a relational need to be reputed as innovative and technology savvy. Although for SMEs the last decision maker is generally the owner-manager, the pressure that she feels from other stakeholders (e.g., employees, customers, suppliers) remains an important determinant of technology adoption. In this vein, the issue of legitimacy is essentially a matter of GHFLVLRQPDNHU¶VSHUFHSWLRQ Harrison, Mykytyn, and Riemenschneider (1997), found that similar unwritten norms, mainWDLQHGE\SHHUVDQGVRFLHW\VWURQJO\LQÀXHQFHWKH intention to adopt information technologies in small businesses. Along the same line, Lee et al. (2001) SRVLWWKDW60(V¶PDQDJHUVKHDUDERXWWKHUHODWLYH advantages of digital technologies largely from the trade press, their friends, business competitors, and peer-social interactions. This would create selective contexts where it is important to be perceived as technology savvy and where it could be expected that SMEs are more sensible to ICT adoption, miming strategies of most VXFFHVVIXO¿UPV *UHZDOHWDO :KHQWKH environment entails highly legitimated organizations, that is, where players need to be perceived as technology savvy, SMEs might be forced to adopt these technologies to socially legitimate their role. 7KLVOHDGVWRD¿IWKK\SRWKHVLV
Hp5: Technology legitimacy requirements SRVLWLYHO\LQÀXHQFHWKHOHYHORI,&7DGRSWLRQ )LJXUHVXPVXSRXU¿YHK\SRWKHVHVRQWKH different sources of pressure for SMEs to adopt ICT solutions.
THE ANALYSIS Sampling and Collecting The analysis is based on a survey of Italian SMEs.3 $UDQGRPVWUDWL¿HGVDPSOHRI60(VZDV VHOHFWHGDVUHSUHVHQWDWLYH ZHH[FOXGHG¿UPVZLWK less than 50 employees, considered as very small, and with more than 250 employees considered as large organizations (Zegveld, 1991). Data were collected through a questionnaire. Firms were contacted by phone using the CATI technique during the period November/December ¿UPVDQVZHUHGWKHHQWLUHTXHVWLRQQDLUH becoming the actual sample of this study (response rate = 28.5%). $GUDIWRIWKHTXHVWLRQQDLUHZDV¿UVWWHVWHGRQD UDQGRPVDPSOHRI60(V7KH¿QDOYHUVLRQZDV PRGL¿HGDFFRUGLQJO\,QRUGHUWRUHGXFHUHVSRQVH bias, a unique key informant—the CEO—was selected as the most appropriate respondent.
Figure 1. External drivers for ICT adoption
Environmental Competitive pressures
Environmental Institutional pressures
Rivalry competition Leading market position Business group membership International presence Technology legitimacy
528
ICT services adoption
External Pressures for Adoption of ICT Services Among SMEs
Although the time for collecting data was reasonably short, we also controlled for late vs. early respondent potential bias: for each of the variables, WKHVDPSOHVKRZVQRVWDWLVWLFDOO\VLJQL¿FDQWGLIferences.
Dependent Variables The dependent variable of our model is the level of adoption of ICT service solutions among SMEs. In this way we avoid to consider investment in ICT hardware. Consistently with previous studies (Lee et al., 2001), this variable has been operationalized by measuring the number of ICT service solutions used by a SME among the following: • • • • •
Interactive Web site, that is, not a simple informative Internet site E-commerce E-procurement Telework Customer relationship management (CRM) systems
behavior of the managers: how much pressure do they feel from the competitive environment? Analogously, the relative market position has been measured through a dummy variable (“LEADERSHIP”). It is the answer to the followLQJ VHQWHQFH ³:LWKLQ \RXU VSHFL¿F VHJPHQW RU QLFKHLV\RXU¿UPLQDOHDGLQJPDUNHWSRVLWLRQ"´ (1 = yes; 0 = no). Further, the level of international presence YDULDEOH³(;3257´ KDVEHHQPHDVXUHGWKURXJK the share of export over total sales. The dichotomous variable labelled “GROUP,” then, indicates whether the respondent SME belongs to a business group (1 = yes; 0 = no), that is whether it is independent or controlled by some other company. Finally, the intensity of technology legitimacy pressure perceived by a SME (“LEGITIMACY”) KDVEHHQPHDVXUHGWKURXJKD¿YHSRLQWOLNHUWW\SH scale (1= I completely disagree; 5 = I totally agree) for the following sentence: “Given the market where we play, it is important for us to be considered as technology savvy.”
Control Variables We coded the dependent variable by computLQJWKHVXPRI¿YHGXPP\YDULDEOHVLQGLFDWLQJ the presence of each solution. We obtained, then, a cumulative ordinal variable ranging from 0 (no ICT service solutions adopted by the SME) to 5 (all service solutions adopted).
Independent Variables The set of independent variables contains both categorical and quantitative variables: the former were treated by the ordinal regression model as factors and the latter as covariates. The intensity of competition perceived by a SME (variable “COMPETITION”) has been PHDVXUHGWKURXJKD¿YHSRLQWVOLNHUWW\SHVFDOH (1= I completely disagree; 5 = I totally agree) for the following sentence: “In the market where we play, competition is extremely strong.” We decided to adopt this measure because, in our hypothesis, it is right the perception of competiWLRQ DQGQRWFRPSHWLWLRQE\LWVHOI WRLQÀXHQFHWKH
6LQFHWKHUDQJHRI¿UPV¶VL]HLQRXUVDPSOHLVQRW irrelevant (from 50 to 250), and since business size DOUHDG\SURYHGWREHDVLJQL¿FDQWGLVFULPLQDWRU between IT adopters and non-adopters among SMEs (Daniel & Grimshaw, 1999), we controlled for size-effects (variable “SIZE”). More precisely, we used the natural logarithm of the number of employees, considering that the probability to adopt an innovation increases with size, but at decreasing rates (Battacharya & Bloch, 2004). Last, we recognize that service sectors—for their intangible nature—may be more conducive of ICT adoption than manufacturing (Lucking-Reilly & Spulber, 2001). The dichotomous variable “Manufacturing” (1 = manufacturing; 0 = service), then, is used to control for this industry VSHFL¿FLW\
529
External Pressures for Adoption of ICT Services Among SMEs
Statistical Techniques
OUTCOME
Since our dependent variable was ordinal, the relationship with the independent variables was modeled through an ordinal regression technique. This approach allows for modeling the dependence of a polytomous ordinal response on a set of predictors, which can be factors or covariates, and are assumed to be independent multinomial variables (McCullagh, 1980). The analysis relies on an ordered logit model provided by the statistical software SPSS (version 12.1), which estimates the effects of independent variables on the log odds of having lower rather than higher scores on the dependent variable.
Table 1 shows the results of the model that we XVHG WR WHVW WKH LQÀXHQFH RI H[WHUQDO IDFWRUV RQ ICT adoption among SMEs. 7KHOLNHOLKRRGUDWLRWHVWLVKLJKO\VLJQL¿FDQW DQG WKH QXOO K\SRWKHVHV WKDW DOO FRHI¿FLHQWV IRU independent variables are equal to zero can be rejected. This means that the independent variables, WRJHWKHUGRKDYHDVLJQL¿FDQWHIIHFWRQWKHQXPEHU of ICT solutions adopted by a SME. $OWKRXJK WKH JRRGQHVVRI¿W WHVWV FRXOG EH meaningless, because of the mixed nature of our independent variables (Long, 1997), they show WKDWWKHPRGHO¿WVWKHGDWDYHU\ZHOO DVNQRZQ WKHVHWHVWVLQGLFDWHJRRGQHVVRI¿WZKHQ&KLVTXDUH VWDWLVWLFVDUHQRWVLJQL¿FDQW As for the Nagelkerke pseudo R-square measure, it indicates a moderate satisfactory level of 0.17, which is in the range of acceptable values for explorative researches (0.10-0.20) (Pampel, 2000). Moving to parameter estimates, it will be possible to test and discuss our previous hypotheses. Hp1: First, the model indicates a positive and VLJQL¿FDQW FRHI¿FLHQW IRU WKH OHYHO RI SHUFHLYHG rivalry competition (0.194): an increase of one unit in the scale of perceived competitive pressure reduces the probability to have a lower rather than higher score for each level of ICT adoption by 7KLVRXWFRPHFRQ¿UPVWKH¿UVWK\SRWKHVLV The intensity of rivalry competition perceived by a SME positively affects its level of ICT adoption. In other words, it can be said that rivalry competition is correlated with greater adoption of ICT service solutions. Hp2: Also the perceived leading market position reported by the SME reveals a strong positive association with the level of ICT adoption (0.648), D UHVXOW ZKLFK LV VWDWLVWLFDOO\ VLJQL¿FDQW DW For an SME, to be (or, at least, to feel to be) in a leading market position heavily reduces the probability ratio of “at the most” vs. “at least” for each category of ICT adoption. This indicates that hypothesis 2 is supported as well, and that a leading market position positively affects the level of ICT adoption among SMEs.
K § p (Y d j ) · D ln ¨ j ¦E k Xk ¸ k 1 © p (Y ! j ) ¹
for j = 1 to J - 1
Where: j = threshold levels for dependent variable E = parameter estimates of independent variables The model works by linking the ordinal dependent variable to another latent variable whose values determine what the observed ordinal variable equals. This latent variable has different threshold levels on the basis of the number of categories of WKHGHSHQGHQWYDULDEOHZKRVH¿QDOYDOXHGHSHQG on the fact that the latent variable crosses (or not) a particular threshold level (Menard, 2002). It has to be considered that, since in this case we ran the ordered logit model adopting a logit as a link function, outcomes are similar to those obtainable through the multinomial logistic regression. The preference accorded to the ordered model is motivated by the rank order nature of dependent YDULDEOHVDQGE\WKHJUHDWHUHI¿FLHQF\RIWKHRUGHUHG logit procedure for dependent variables with many possible categories (Pampel, 2000). Through the model, we should be able to predict if higher values of independent variables will increase the probability of obtaining higher values in the ordinal dependent variable.
530
External Pressures for Adoption of ICT Services Among SMEs
7DEOH0RGHO¿WWLQJDQGSDUDPHWHUHVWLPDWHV E
Wald
Sig.
Rivalry Competition
0.194
3.04
*
Leading market position
0.648
8.34
***
Business group membership
0.456
3.98
**
International presence
-0.005
1.57
ns
Technology legitimacy
0.284
9.03
***
Parameter estimates COMPETITIVE PRESSURES
INSTITUTIONAL PRESSURES
Size
0.266
4.10
**
Manufacturing
-0.667
6.37
**
0RGHO¿WWLQJ
Chi-Square
df
Sig.
-2 Log Likelihood
48.291
7
***
3HDUVRQJRRGQHVVRI¿W
1253.3
1378
ns
'HYLDQFHJRRGQHVVRI¿W
810.7
1378
ns
Nagelkerke pseudo R2
0.17
VLJQL¿FDQWDW
VLJQL¿FDQWDW
VLJQL¿FDQWDW QV QRQVLJQL¿FDQW
Hp3: Shifting to the hypothesized pressures from a parent company, the model reveals a positive DQGVLJQL¿FDQWDVVRFLDWLRQEHWZHHQWKHIDFWWKDWD SME belongs to a business group and its level of ICT adoption (0.456), while independent SMEs appear to be less involved with new technologies: this outcome supports Hypothesis 3. Hp4: Contrary to our hypothesis, we found no evidence regarding the role of the export activity of a SME: higher export rates are not proven to affect the level of ICT adoption. Since the parameter HVWLPDWHLVQRWVLJQL¿FDQW+\SRWKHVLVFDQQRWEH VWDWLVWLFDOO\FRQ¿UPHGE\WKLVVWXG\ Hp5: Then, if we consider technology legitimacy pressures, this variable too shows a positive DQGVLJQL¿FDQWDVVRFLDWLRQZLWKWKHOHYHORI,&7 adoption (0.284): within environments where it is important to be considered as technology savvy,
LWLVPRUHOLNHO\WR¿QG60(VZLWKKLJKHUOHYHO of Internet adoption. This evidence supports our Hypothesis 5. Both control variables, on the contrary, are VWDWLVWLFDOO\VLJQL¿FDQWDVH[SHFWHGLQRXUSUHYLous discussion. )LUVW JLYHQ WKH UDQJH RI ¿UPV FRQVLGHUHG D SRVLWLYHDQGVLJQL¿FDQWVL]HHIIHFWVWLOOHPHUJHV WKH analysis reveals that larger SMEs are more likely to exhibit higher level of ICT adoption compared to smaller ones. Second, SMEs in service industries are more likely to have higher level of ICT adoption compared to SMEs in manufacturing industries. This effect was largely expected, too, since value chains in service sectors are generally less fragmented and easier to reengineer.
531
External Pressures for Adoption of ICT Services Among SMEs
IMPLICATIONS AND FUTURE RESEARCH The study presented in this chapter intends to emphasize the importance that external sources of pressure on a SME may have on its level of ICT LQYROYHPHQW:HIRXQGVLJQL¿FDQWIDFWRUVRIERWK competitive and institutional nature. ,QIDFWWKHDQDO\VLVUHYHDOVWKHVLJQL¿FDQFHRI IRXURXWRIWKH¿YHH[SHFWHGGULYHUVDQGVXJJHVWV some implications for both policy makers and 60(V¶PDQDJHUV From the policy perspective, it seems to appear that the efforts to support technology adoption among SMEs might be asymmetric and should EH JUDGXDWHG DFFRUGLQJ WR VSHFL¿F FRQGLWLRQV ,QGHSHQGHQW60(VIRULQVWDQFHVKRXOGEHQH¿WRI higher support than their counterparts belonging to DEXVLQHVVJURXSVLQFHWKHLQÀXHQFHRIDKROGLQJ company seems to correct some structural barriers ¿QDQFLDOFXOWXUDODQGRUJDQL]DWLRQDO IDFHGE\ SMEs in their adoption process. It is also possible to infer that any policy measure affecting the level of competition and the institutional requirements of markets may be relevant in stimulating or impairing ICT investments as well. Obviously, all these considerations should take into account the cost dimension, in order to avoid an indiscriminate support for any service solution to any SME. Other potential implications are for managerial decision makers. The analysis indicates that a SME could avoid Internet adoption only when the reference environment is not particularly selective, both in competitive and institutional terms. This kind of situation, however, may slowly atrophy its business potentiality. The survival path of this kind of SMEs would be linked to their capability to play in stable and protected contexts. Similar situations, however, DUH JRLQJ WR EH PRUH DQG PRUH GLI¿FXOW WR ¿QG and this will inevitably restrict future business RSSRUWXQLWLHVFUHDWLQJIRUWKHVH¿UPVDFRQ¿QHG path of growth. 6LPLODUO\RXU¿QGLQJVVHHPWRFRQ¿UPWKDWWKH participation to interorganizational networks produces pressures to adopt new technologies, perhaps
532
for coercive or legitimacy reasons. To be excluded from these networks might produce analogues HIIHFWVDVEHLQJFRQ¿QHGRQFRPSHWLWLYHFRQWH[WV ZLWKRXW VLJQL¿FDQW SHUVSHFWLYHV 7KHUHIRUH IRU 60(V¶PDQDJHUVWKLVZRXOGEHDIXUWKHUUHDVRQ WR SD\ VLJQL¿FDQW DWWHQWLRQ WR WKHLU QHWZRUNLQJ strategies. From the analysis, no conclusions can be drawn about adoption pressures coming from abroad. It PXVWEHQRWHGWKDWWRDVFHUWDLQPRUHGH¿QLWLYH ¿QGLQJV RQ WKLV SRLQW IXWXUH UHVHDUFK VKRXOG probably limit the evaluation to interactions with international markets having higher levels of ICT adoption than the domestic one. Moreover, a general measure, as in this study, has higher risks to include an inverse causal relationship. A higher cultural predisposition to innovate might drive internationalization, instead of the opposite. Finally, since the study and the questionnaire were designed especially to explore possible impacts of external factors on ICT adoption, internal variables are not explicitly considered in the model, and, obviously, this raises the question of the potential omitted variables. For future research, however, we suggest to include, simultaneously, both internal and external factors. Eventually, further scales of measurement could be tested in order to corroborate the results presented in this study.
REFERENCES Andrews, K. (1971). The concepts of corporate strategy. Homewood, IL: Dow-Jones Irwin. Barney, J. (1991). Firm resources and competitive advantage. Journal of Management, 17(1), 99-120. Bhattacharya, M., & Bloch, H. (2004). Determinants of innovation. Small Business Economics, 22, 155-162 Clemons, E., & Row, M. (1991). Sustaining IT advantage: The role of structural differences. MIS Quarterly, 3, 275-292. Cohen, W.M., & Levin, R.C. (1989). Empirical studies of innovation and market structure. In R.
External Pressures for Adoption of ICT Services Among SMEs
Schmalensee & R. Willig (Eds.), Handbook of industrial organization (pp. 1059-1107). Amsterdam: North Holland. Daniel, E.M., & Grimshaw, D.J. (2002). An exploratory comparison of electronic commerce adoption in large and small enterprises. Journal of Information Technology, 17, 133-147. Davis, F.D. (1989, September). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 319340. Fink, D. (1998). Guidelines for the successful adoption of information technology in small and medium enterprises. International Journal of Information Management, 18(4), 243-253. Gatignon, H., & Robertson, T.S. (1989). Technology diffusion: An empirical test of competitive effects. Journal of Marketing, (1), 35-49. Grandon, E.E., & Pearson, J.M. (2004). Electronic commerce adoption: An empirical study of small and medium U.S. business. Information & Management, 42, 197-216. Grewal, R., Comer, J.M., & Metha, R. (2001). An investigation into the antecedents of organizational participations in the business-to-business electronic markets. -RXUQDORI0DUNHWLQJ(7), 17-33. Grover, V. (1994). Empirically derived model for the adoption of customer-based interorganizational systems. Decision Science, 40(9), 1069-1092. Harrison, D., Mykytyn, P.P., & Riemenschneider, C.K. (1997). Executive decisions about adoption of information technology in small businesses: Theory and empirical tests. Information Systems Research, 8(2), 171-195. Hoskisson, R.E., Wan, W.P., Yiu, D., & Hitt, M.A. (1999). Theory and research in strategic management: Swings of a pendulum. Journal of Management, (3), 417-456. Iacovou, C.L., Benbasat, I., & Dexter, A.S. (1995). Electronic data interchange and small organiza-
tions: Adoption and impact of technology. MIS Quarterly, 19(4), 465-486. Kimberly, J.R., & Evanisko, M.J. (1981). Organi]DWLRQDOLQQRYDWLRQ 7KHLQÀXHQFHRILQGLYLGXDO organizational, and contextual factors on hospital adoption of technological and administrative innovations. Academy of Management Journal, 24, 689-713. Kuan, K.K.Y., & Chau, P.Y.K. (2001). A perceptionbased model for EDI adoption in small business using a technology-organization-environment framework. Information & Management, 38(8), 507-521. Kwon, T.H., & Zmud, R.W. (1987). Unifying the fragmented models of information systems implementation. In R.J. Boland & R.A. Hirschheim (Eds.), Critical issues in information systems research. Wiley. Lal, K. (1999). Determinants of the adoption of information technology: A case study of electrical DQGHOHFWURQLFJRRGVPDQXIDFWXULQJ¿UPVLQ,QGLD Research Policy, 28, 667-680. Lederer, A.L., Maupin, D.J., Sena, M.P., & Zhuang, Y. (2000). The technology acceptance model and the world wide Web. Decision Support Systems, 29, 269-282. Lee, J., Runge, J., & Baek, S. (2001). Adoption of Internet technologies in small business. In ProFHHGLQJVRI3DFL¿F$VLD&RQIHUHQFHRQ,QIRUPDWLRQ Systems (PACIS 2001), Seoul, Korea. Leonard-Barton, D. (1992). Core rigidities and core competencies. Strategic Management Journal, 13, 111-125. Long, J.S. (1997). Regression models for categorical and limited dependent variables. Sage Publications. Lucking-Reiley, D., & Spulber, D.F. (2001). Business to business electronic commerce. Journal of (FRQRPLF3HUVSHFWLYHV(1), 55-68. Malerba, F., & Orsenigo, L. (1997). Technological regimes and sectoral patterns of innovative
533
External Pressures for Adoption of ICT Services Among SMEs
actvities. Industrial and Corporate Change, 6, 83-117. Markus, L.M. (1987). Toward a critical mass theory of interactive media. Communications Research, 14(5), 491-511. Mata, F.J., Fuerst, W.L., & Barney J.B. (1995). Information technology and sustained competitive advantage: A resource-based analysis. MIS Quarterly, 19(4), 487-505. McCullagh, P. (1980). Regression models for ordinal data (with discussion). Journal of the Royal Statistical Society Series, B(42), 109-142. Menard, S. (2002). Applied logistic regression analysis (Rev. ed). Sage Series on Quantitative Applications in the Social Sciences. Newbury Park, CA: Sage. North, D.C. (1990). Institutions, institutional change and economic performance. Cambridge, MA: Cambridge University Press. OECD. (2003). ICT and economic growth. Evidence IURP2(&'FRXQWULHVLQGXVWULHVDQG¿UPV. Paris: OECD Publication Service. Peteraf, M.A. (1993). The cornerstones of competitive advantage: A resource-based view. Strategic Management Journal, 14, 179-192. Porter, M. (1980). Competitive strategy. New York: Free Press. Powell, T.C., & Dent-Micallef, A. (1997). Information technology as competitive advantage: The role of human, business, and technology resources. Strategic Management Journal, 18(5), 375-405. Premkumar, G., & Roberts, M. (1999). Adoption of new information technologies in rural small businesses. Omega, Int. J. Mgmt. Sci., 27, 467-484. 5RJHUV0 1HWZRUNV¿UPVL]HDQGLQQRYDtion. Small Business Economics, 22, 141-153. Swanson, E.B. (1994). Information systems innovation among organizations. Management Science, 40(9), 1069-1092.
534
Teece, D.J, Pisano, G., & Shuen A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509-533. Thong, Y., & Yap, C. (1995). CEO characteristics, organizational characteristics and information technology adoption in small businesses. Omega, Int. J. Mgmt. Sci., 23, 429-442. Tornatzky, L.G., & Fleischer, M. (1990). The processes of technological innovation. Lexington, MA: Lexington Books. Wernerfelt, B. (1984). A resource-based view of WKH ¿UP 6WUDWHJLF 0DQDJHPHQW -RXUQDO (2), 171-180. Zegveld, W. (1991). Technology policy for small and medium enterprises in Netherlands. In A.S. Bhalla (Ed.), Small and medium enterprises. New York: Greenwood Press. =KX..UDHPHU. ;X6 (OHFWURQLF EXVLQHVV DGRSWLRQ E\ (XURSHDQ ¿UPV $ FURVV country assessment of the facilitators and inhibitors. European Journal of Information System, 12, 251-268.
KEY TERMS Information and Communication Technology (ICT) Services: That part of the technology solutions to manage information and communication processes made up with software and other intangibles which are not material assets such like Hardware Small and Medium Enterprises (SMEs): Firms with a number of employees comprised between 20 and 249 Competitive Pressures: Element of the enYLURQPHQWZKLFKLQWHUDFWZLWK¿UP¶VGHFLVLRQV WKDWDUHVSHFL¿FWRWKHFRPSHWLWLYHPDUNHWZKHUH WKH¿UPSOD\ Institutional Pressures: Element of the macroLQVWLWXWLRQDO HQYLURQPHQW ZKHUH WKH ¿UP SOD\V WKDWFDQLQÀXHQFHWKURXJKQRUPVDQGKDELWVWKH ¿UP¶VGHFLVLRQV
External Pressures for Adoption of ICT Services Among SMEs
Technology Acceptance Model (TAM): Theoretical framework developed by Davis in 1989 which identify perceived usefulness, ease of use, and cultural orientation of the decision maker as key drivers of technology adoption Technology Organization Environment (TOE): Theoretical framework developed by Thornatzky and Fleischer in 1990, which identify the features of technology, the organizational readiness RIWKH¿UPDQGWKHHQYLURQPHQWDOFRQGLWLRQVDV key drivers of technology adoption
ENDNOTES 1
2
3
A leading market position for an SME, here, PD\DOVRUHIHUWRDVSHFL¿FVHJPHQWRUQLFKH of the market. It must be noted that these studies considered the export activity as a proxy for competitive intensity. Italy is traditionally considered as an interesting market to investigate issues related to SMEs, given the dominance of small organizations in most of the industries.
Ordinal Regression: Regression model used when the dependent variable is a ranking-order variable (3 is higher than 1 but it is not necessarily three times one).
535
Section IV
Impact on Emerging Applications
Division I
Entertainment Industry
538
Chapter XXXIV
IPTV Business Model Analysis Kate Carney Landow University of Colorado at Boulder, USA Michelle Fandre University of Colorado at Boulder, USA Raghu Nambiath University of Colorado at Boulder, USA Ninad Shringarpure University of Colorado at Boulder, USA Harvey Gates University of Colorado at Boulder, USA Artur Lugmayr Tampere University of Technology, Finland Scott Barker Freelance Editor, Bridgewater, USA
ABSTRACT This chapter focuses on evaluating Internet protocol television (IPTV) business models from different VHUYLFHSURYLGHUVWKURXJKWKHSULVPRIWKH6WDU0RGHO7KHWRROLVEDVHGRQ3RUWHU¶V¿YHIRUFHVDVGHYHORSHGE\0(3RUWHU7KH6WDU0RGHOH[WHQGVWKH¿YH)RUFHVLQWRDVHWRIPHWULFVWRHYDOXDWHFXUUHQW and future business offerings. The Star Model is a simple tool used to identify the strengths and weaknesses of different business models in an appealing geometric shape. To highlight how to use this tool, sample partnership models are analyzed to evaluate the strength of a combined service. This tool will help IPTV service providers and all potential investors to build or identify a sound business model for their target market. The Star Model is explored through multiple case studies in this chapter including CBS, AOL, Google, Sling Media, and YouTube.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
IPTV Business Model Analysis
IPTV BUSINESS MODEL ANALYSIS Growth of broadband has enabled convergence of voice, video, and data, resulting in the evolution of many new and exciting applications. One technology with wide customer appeal is Internet protocol television (IPTV), a service that promises to deliver television content across platforms and geopolitical borders. During the last decade broadband-based access has been one of the fastest growing Internet services throughout the world (Miniwatts Marketing Group, 2006). Supportive regulatory environments and competitive pricing models are a couple of the primary factors that have contributed to the success of broadband-based Internet service. However, there is still a large disparity in growth of broadband among different countries. In some markets, the transition from narrow band to broadband has resulted in additional costs to both consumers and service providers. From the consumer point of view, there is a need for the value of service to offset WKH DGGLWLRQDO FRVW )URP WKH VHUYLFH SURYLGHU¶V perspective, the additional cost in laying out the infrastructure requires a supporting regulatory environment and a lucrative business model that opens the door to multiple services and opportunities. The absence of a killer application—one that can appeal to mass audiences and allow service SURYLGHUVWRHDUQDSUR¿W²LVWKHFRPPRQIDFWRU contributing to slower growth. Television content may be the product providing the greatest incentive for customers to purchase broadband solutions. When television is delivered via IPTV it increases the choices for content available to customers and allows providers to capitalize on the opportunities created in a market when more choices are available to customers. The traditional television market is mostly served by a monopoly service provider, a situation that affects the overall potential growth in a market and also reduces choices for consumers. Furthermore, there has been a lack of innovative applications that enhance the traditional television service. The limitations of traditional television can be overcome with the delivery of content to the market as IPTV. With these developments, television is no longer
bound to a traditional set. Television content is delivered to computers and mobile phones through newer, advanced telecommunication networks. IPTV will change the way consumers purchase entertainment. Content and service providers can now reach increasingly larger audiences through multiple devices. Broadband penetration has increased steadily in European markets due to deregulation and unbundling of services, which has in turn led to competitive pricing from providers. In the United States, broadband penetration has increased because of the increasing number of services providers. In both regions, many vendors have experienced decreasing average revenue per user (ARPU) and increasing customer churn as a result of increase in competition in the market. Adding services like television to the existing broadband access can help these companies retain customers and offset the decrease in revenue (Frost & Sullivan, 2006). IPTV is a technology that has the potential to propel growth of broadband across all markets and can provide an incentive for consumers to use newer, interactive applications. A viewer located in North America will have vast choices of content from a service provider located in China or in any other part of the world. When a television service is provided over broadband, the technology enables the entire world to connect. An IPTV provider must have a mature business model in order to successfully deliver content by additional means. There is now no readily available tool to evaluate a business model. This chapter helps readers evaluate an IPTV business model using the Star Model and its associated metric parameters and provides a readily available tool for business analysis. IPTV has a variety of potential methods of GHOLYHU\ZKLFKVKRXOGEHUHÀHFWHGLQDEXVLQHVV model. This chapter will focus on evaluating speFL¿F,379EXVLQHVVPRGHOVZLWKWKHXVHRIWKH6WDU 0RGHODQH[WHQVLRQRI3RUWHU¶V¿YHIRUFHV 3RUWHU 3RUWHU¶V¿YHIRUFHVDUHDVHWRIPHWULFVWKDW are often used to evaluate business operations. This chapter conducts comparative analysis on a select number of IPTV service providers with the use of the Star Model. The Star Model is useful in its ease of application to each of the methods of
539
IPTV Business Model Analysis
distribution of IPTV. Each case study is analyzed to evaluate the strength of a combined telecom service offering.
BACKGROUND: IPTV BUSINESS MODEL BASICS The broadband marketplace has a competitive nature that will foster growth. Once a company has reached a position in the industry with the appropriate focus on a targeted-audience, the company will be able to continue along a path to improvement based on the feedback from using the Star Model. The company can thus remain a healthy entity for consumers in the current and future telecom market. Overall, the competitive nature of the telecom market also challenges companies, as shown in activity within the European Union (EU). The combined effects of privatizations, EU accession, and continued deregulation in candidate EU countries has led to more competition, network upgrades, greater broadband penetration, and falling DSL access line costs. This portion of the broadband industry has many players competing for many of the same customers and long standing companies to secure a bigger share of the market (IPTV World Forum Eastern Europe, 2006). Globally, the growth of broadband has spurred the development of a variety of Internet-based technologies. Fledging companies must compete with long-standing telecom providers in order to survive. Existing companies have transformed their VHUYLFHV LQ RUGHU WR DGDSW WR FRQVXPHU¶V QHHGV remain competitive, and keep up with the change. Never before have there been so many entertainment choices and technologically advanced gadgets WRLQFUHDVHWKHFRQVXPHU¶VKDSSLQHVVDQGHDVHRI living, not to mention ways in which the consumers can spend their disposable income. Businesses now have the ability to expand their customer base DQGLQFUHDVHSUR¿WVZLWKWKHDGGLWLRQDODYHQXHVRI revenue available. Before service providers leap into the fray, the FRPSDQ\PXVWGH¿QHZKDWLQWHUQHWSURWRFROWHOHYLsion is in the business plan by selecting a method
540
of delivery. As IPTV develops as a technology, service providers and customers are continuing to add granularity to the concept of what IPTV is as a service offering. This is a positive trend as once VSHFL¿F VHUYLFHV DUH GH¿QHG VWDQGDUGV VHWWLQJ organizations will have an easier time authoring XQL¿HGVHFXULW\DQGTXDOLW\RIVHUYLFHEHQFKPDUNV As IPTV becomes more present in the consumer lexicon, this business decision can be framed around the category of service a company offers. Overall, the marketplace manifestation of IPTV for consumers can be described under one comPRQGH¿QLWLRQZLWKDGGLWLRQDOJHQHUDOFDWHJRULHV (IPTV Monitor, 2006). Generally, internet protocol television is a service delivered over a network infrastructure utilizing internet protocol (Wikipedia, 2006a). In this chapter, we will present case studies DGGUHVVLQJWKHVHIRXUFODVVL¿FDWLRQV • • • •
User generated content Value added services Broadcast Internet
In addition to this key business decision, a modeling tool is needed to effectively review the stability of telecom companies that are in an evolving market. By limiting the discussion in this FKDSWHUWRDERYHIRXUPRGHOVRIH[LVWLQJSURYLGHU¶V operating business models, the ease of use of the Star Model will be quickly revealed.
ANALYSIS OF IPTV BUSINESS MODELS The tool presented in this chapter is based on PorWHU¶V¿YHIRUFHVDQG,QIR'HY UHFRPPHQGDtions for open telecommunications markets. Other resources used to develop this analysis included: census counts that determined the population size in the targeted market pool, Value Line Report (Value Line, 2006), which is an independent analysis of business and industry reporting on earnings, debt, and stock prices, and subscriber information that detailed the number of subscribers to wire line, mobile, and entertainment services. Analysis
IPTV Business Model Analysis
DOVRFRQVLGHUHGDFRPSDQ\¶VUHYHQXHJHQHUDWLRQ inherent costs, dependence on IPTV, pricing model between the three players (providers, consumers, and middleware), technical support burden, and the transferability of a business mode from one particular country to another. Much of the information used to create a Star Model is publicly available, aiding in creating a consistent method of analysis when comparing companies. When fully H[SORUHGWKH¿QDO6WDU0RGHOKDVSDUDPHWHUV The next two subsections highlight the methods used to develop the tool and the metric parameters to build a model in greater detail.
Methods Used to Create the Star Model The Star Model is loosely based on 3RUWHU¶V)LYH Forces. Developed in 1979 by Michael Porter, the Five Forces group microeconomic forces within WKHPDUNHWSODFHLQWRWKHIROORZLQJ¿YHFDWHJRULHV bargaining power of customers, the bargaining power of suppliers, the threat of substitute products, the threat of new entrants, and industry rivalry (Porter, 1998). This idea provided companies with a method of analyzing a market in order to predict what forces affect a product offering, resulting in DQDQDO\VLVRQKRZWKHPLFURHFRQRPLFLQÀXHQFHV ZRUN6LQFH3RUWHU¶VPRGHOGHEXWRWKHUWKHRULHV have been presented that also attempt to understand market forces and apply knowledge gained from such analysis to improve business models. In PRGLI\LQJ3RUWHU¶VZRUNVFKRODUVKDYHSURSRVHG further categorization of the forces within a market to anywhere from six (Carr, 2005) to eight forces (Downes, 1997). Each of these models provides VLJQL¿FDQWTXDOLWDWLYHLQIRUPDWLRQWRFRPSDQLHV requiring snapshots of the marketplace in which a product is expected to launch. As with other scholars, these authors found YDOXHLQXVLQJ3RUWHU¶VIRUFHVWRDQDO\]HWKH,379 market by categorizing the microeconomic forces acting within a marketplace to allow examination of potential market opportunities. In prior research, LWKDGEHHQGLI¿FXOWWRXQGHUVWDQGWKHVWUHQJWKVDQG weaknesses of an IPTV product offering beyond claims made by companies. With IPTV in a major
F\FOHRIGHYHORSPHQWWKHPDUNHWSODFHUHÀHFWHGD rapidly changing pricing of products and the lack LQVWDQGDUGVFUHDWHGWRGH¿QHZKDW,379LV1RW only does the standard consumer want to know how WKH\FDQEHQH¿WIURP,379EXWPRUHLPSRUWDQWO\ telecom providers need to know how they can SUR¿WIURPWKLVJURZLQJWHFKQRORJ\&RPSDQLHV expecting to have successful launches of IPTV need a quantitative way to gauge how all of the forces acting within a market could affect the success or failure of an IPTV launch. With a challenging market, there is intellectual opportunity to create an analysis tool that provides quantitative data in a dimensional form. The result is a set of standardized questions that combines 3RUWHU¶VEDVLFFRQFHSWVZLWKRWKHULGHDVDQGDVsigns metrics. The metrics are provided with a set of standardized questions that are answered with “yes” or “no,” and then translated into a number. When the data is entered into common graphing tools and each line of questioning is assigned to an arm of a radial graph, a star appears as a model for viewing the quantitative data. The clear geographic shape results in illustration for management teams and any potential investors. As the questions are standardized, the data gathered over time is also consistent. This allows users to compare alternatives in developing their IPTV platform. The arms of the Star Model will represent different dimensional data when applied to either their business model development than when the company considers partnership agreements. In the next section, we explain the organizational aspects of the star metrics.
Major Metrics of the Star Model With the Star Model, the way in which IPTV manifests in the market is being analyzed. To represent Star Model ratings, the basic concept of 3RUWHU¶VIRUFHVLVRYHUODLGRQWRPDUNHWSDUDPHWHUV 6HOHFWHGFRPSDQLHVDUHUDWHGE\DVNLQJVSHFL¿F questions, resulting in a parameters scale of 1 to 10. The major points are considered in relation to minor points—with the minor parameters rated on a base scale of 1 to 5 (labeled in resulting graphs from “A” to “E”).
541
IPTV Business Model Analysis
The questions used to rate company performance in the major metrics were inspired by the ¿QDQFLDO VWDWHPHQWV UHTXLUHG IURP FRPSDQLHV RSHUDWLQJ XQGHU WKH MXULVGLFWLRQ IRU 6(& ¿OLQJV :LNLSHGLD F ,Q IRUPXODWLQJ WKH VSHFL¿F questions stock investment research materials, such as the Value Line (Value Line, 2006) were referenced in order to reduce the number of questions to a number that was applicable for use in D VLPSOH DQDO\VLV $V 3RUWHU¶V IRUFHV LQÀXHQFH WKHPDMRUPHWULFVWKHTXHVWLRQVZHUHUH¿QHGWR remain consistent with that established methodolRJ\RIJDXJLQJKRZWKH¿YHPDMRUPHWULFVDIIHFW a market segment. When answering questions considered in the case study, an answer “yes” is equated to numeric one and is considered positive strength for the company. However, a negative response “no” is rated as numeric value zero and is considered possible weakness for the company. Data gathered to answer questions for parameters is available from publicly accessible sources, company press releases, census data, and other like documentation. ,QWKHQH[WVHFWLRQZHH[SODLQWKHEDVLFGH¿QLtions of each parameter and how those parameters affect the value chain. All questions for major and minor parameters are listed at the end of this chapter. 1.
2.
3.
542
Power Against Suppliers: This metric will allow each company to be rated at its current level of bargaining power with respect to the suppliers of the end products. The power of supply affects the costs of producing a product DQGZLOOEHUHÀHFWHGLQWKHYDOXHFKDLQ Power Against Customer Bargaining: This PHWULFHYDOXDWHVWKHFXVWRPHU¶VEDUJDLQLQJ power with an IPTV service provider. The questions for this parameter also discover costs the company passes on to consumers DQGLIWKHFRPSDQ\¶VVHUYLFHWRFXVWRPHUVLV unique. Power Against New Entrants: As part of this metric, companies are evaluated based on their vulnerability to new entrant competition. Questions will also help evaluate the
4.
5.
level of resistance the company poses to new entrants. Power Against Substitution: This metric will help evaluate companies based on the ease of customers substituting a product offering with another competing product. The parameter evaluates the market power of the company with respect to the product being RIIHUHG 6SHFL¿F TXHVWLRQV WDUJHW LI WKHUH are high start-up costs for entrants, whether nor not there are exit barriers for companies divesting from the market and if the product can be replicated by others. Power Against Competition: This metric will be used to evaluate the level of competition that the company faces in the overall market, otherwise known as rivalry within the market. This metric evaluates whether the company is operating under healthy competitive environment. Questions relate whether the company holds the patents and intellectual property rights for goods and services.
Minor Metrics of the Star Model 7RFRPSOHWHWKH6WDU0RGHO¿YHPLQRUSDUDPHWHUV interpret how other market forces interact with major parameters. Almost any IPTV business PRGHOZLOOKDYHDVLJQL¿FDQWSRUWLRQRIWKHSURGXFW development that will be based on branding and presentation of content and while concurrently protecting the intellectual property and public perception of the company. Additionally, in a competitive market, protecting internal operational decisions and information on future offerings is important. To enable business leaders to properly analyze these internal company forces and external subtle market forces that will affect product offerLQJV¿YHDGGLWLRQDOPLQRUSDUDPHWHUVDUHDGGHGWR the Star Model. These minor parameters explore PRUHVSHFL¿FDUHDVLQZKLFKDFRPSDQ\VHFXUHV SURGXFWDQGWKHPVHOYHVIURPRXWVLGHLQÀXHQFHV Providing companies with these additional parameters encourages leaders to make internal strategic decisions that will play a key role in the success of a company, thus leading to increased marketability of that content.
IPTV Business Model Analysis
The questions used to rate company performance in the minor metrics were inspired by theories like SWOT (strengths, weaknesses, opportunities, and threats) analysis (Wikipedia, 2006d). Utilizing a foundation like SWOT enabled IRUPXODWLRQRITXHVWLRQVWDUJHWHGWRZDUGV¿QGLQJ NH\DUHDVRIDFRPSDQ\¶VSHUIRUPDQFHWKDWUHTXLUH attention in order to garner success in the IPTV market. For the arms of the star gauging customer EHKDYLRUWKHTXHVWLRQVZHUHUH¿QHGWRXQGHUVWDQG how a product impacts the customer buying decision. For example, one of the questions concerns HDVHRIXVH6WDWHGVLPSO\LIDFXVWRPHUFDQ¶WHDVily understand how to use a product, he is likely WREHUHÀHFWHGLQDFRPSDQ\¶VUHSRUWVRQWURXEOH calls or disconnect numbers. To develop the arm RIWKHVWDUJDXJLQJWKHLQÀXHQFHRIUHJXODWLRQLQ WKHPDUNHW,QIR'HY¶VDUWLFOHVRQRSHQWHOHFRPmunications markets (infoDev, 2004) inspired WKHTXHVWLRQV7KHTXHVWLRQVZHUHIXUWKHUUH¿QHG when referenced with the Internet world stats usage and population statistics and cross-referenced with the World Fact Book (Miniwatts Marketing Group, 2006). In building the security questions consideration was given to how well a company protects content offered to customers and how well a company protects customer information. The forces considered for these parameters include the following: a.
b.
Customer Loyalty: Stable Customer Base. This metric evaluates steps taken by DFRPSDQ\WRLQÀXHQFHDQGHQFRXUDJHFXVtomer loyalty. A higher rating will indicate WKH FRPSDQ\¶V DELOLW\ WR UHWDLQ LWV H[LVWLQJ customer base. Additionally, the parameter LVDOVRLQGLFDWLYHRIWKHFRPSDQ\¶VDELOLW\WR utilize existing customers while adding new VXEVFULEHUV)LQDOO\WKHUDWLQJDOVRKDVLQÀXence on the ability of the company to face competition from potential new entrants. Product Usability: Interactivity Between Product and Customer. This parameter UHÀHFWVDFRPSDQ\¶VDELOLW\WRHQYLVLRQUHquirements while providing user-friendly interactive features to its customers. The parameter also evaluates incremental ARPU
c.
d.
e.
generated from existing customers. Higher UDWLQJ LQÀXHQFHV WKH FRPSDQ\¶V EDUJDLQing power against customers. Additionally, differentiated services also result in higher bargaining power against suppliers. Secured Solution: Enabling Security within the Company and its Services. The objective of secured services is not only to protect the legitimate users while accessing the network but also to keep the malicious users away from any potential disruption of the services. Content is one of the most important elements for IPTV services. Availability of secured infrastructure for delivery encourages access to premium contents and attracts incremental subscribers. Market Regulation: Business Operation within a Regulated Market. This factor includes clarity of regulatory environment in the target country. An example of this type of regulation can be found in the United .LQJGRP¶V2I¿FHRI&RPPXQLFDWLRQV 2IFRP 2IFRP KDV LGHQWL¿HG PHGLD content transmitted over the Internet that is outside the scope of its current regulatory structure (Ofcom, 2006). However, this is FRQWUDU\WRWKH(XURSHDQ&RPPLVVLRQ¶VODWest proposal—“Television without Frontiers” Directive. As per the proposal, the regulatory provisions need to be extended to Web sites and online services which provide streaming audiovisual digital content to consumers (Burbridgea, 2006). The uncertain regulatory environment will have high impact on planning development of newer services. Additionally, increase in regulation of content will also result in increased cost to suppliers. Customer Satisfaction: Immediacy of Satisfaction. 7KLV SDUDPHWHU UHÀHFWV WKH RSHUDWLRQDO HI¿FDF\ RI WKH FRPSDQ\ ZKLOH handling problems reported by customers. A higher satisfaction index results in increased PDUNHWVKDUHLQÀXHQFLQJWKHFRPSDQ\¶VDELOity to attain higher bargaining power against suppliers. Additionally, the parameter also KDV LQÀXHQFH RQ SRZHU DJDLQVW FRPSHWLQJ vendors.
543
IPTV Business Model Analysis
$VLQWKHSULRUVHFWLRQVSHFL¿FTXHVWLRQVGLVcover the practices of a company pertaining to each of these parameters. Each set of questions focuses on market-generic questions, better enabling a company to capture how business modeling may change as a company enters another market that may have different forces—for example, different regulatory structures. The Star Model in Figure 1 combines the metrics, major and minor parameters, and superimposes them on the company in the circle. Conceptually, the model shown in Figure 1 allows us to view how far a company is extended into the market and whether or not they are sucFHVVIXO(DFKRIWKH¿YHPDMRUSRLQWVWUDQVODWHVLQWR an arm of the star, and each minor point translates into the connection point between the points of the star. Because the star is a geometric shape, an arm of the star can illustrate the direction that a company has taken in the market. While reviewing Figure 1, the Star Model illustrates why the major parameters are rated 1-10, and the minor parameters are rated 1-5. In considering the offset, Figure 1 reveals that to complete a business analysis, the star is overlaid onto the target market. In general, a company will never capture 100% of the market. 7RDGMXVWWKH6WDU0RGHOWRUHÀHFWWKHUHDOLW\RIWKH market, the minor parameters are rated on a scale
Figure 1. Generic star demonstration
544
of 1-5, as these are the areas in which a company is most likely to miss capturing in a targeted area of the market. Our case studies utilize this tool to gauge the size of the arms in an effort to compare its connection points to others that have been added to the same 6WDU0RGHO)RUHDFKRIWKHFDVHVWXGLHVLGHQWL¿HG LQWKLVFKDSWHUDQDQDO\VLVRIPHWULFVIRUWKH¿YH major parameters is summarized in the section Analysis of IPTV business models. The ratings for all 10 parameters, including major and minor ones, are shown in the Star Model included along with each of the case studies.
CASE STUDIES When entering the market as an IPTV provider, a business must have content to sell and the technology to distribute the product. However, in many cases, content providers do not have the brand name IRUGLVWULEXWLRQWKDWZRXOGSURYLGHWKHPRVWHI¿FLHQW delivery of the product. Other businesses may have WKHDELOLW\WRSURYLGHHI¿FLHQWGHOLYHU\EXWQRWWKH resources to legally provide and protect content. In such cases, a company often decides to partner with another company that can provide the resources WKH¿UVWFRPSDQ\PD\QRWKDYH7KHFDVHVWXGLHV
IPTV Business Model Analysis
for IPTV partnerships include a description of the FRPSDQLHV²¿UVWE\LQYHVWLJDWLQJHDFKSDUWQHUDQG then discussing the resulting relationships. The studies presented in this chapter cover: • • •
The value added service of Sling Media The user generated network of YouTube/ Google The partnership model involving the broadcast and Internet companies CBS/Google and AOL/Warner Bros.
Users and Disruptors ,QDQHIIRUWWRWDUJHWVSHFL¿FDXGLHQFHVFRPSDQLHV LQWKH,379PDUNHWFDQXVHWKHRSHUDWLRQDOHI¿FLHQFLHVLQKHUHQWWRWKHWHFKQRORJ\WRWKHFRPSDQ\¶VDGvantage. Companies providing alternative services that take advantage of these technologies in unique manners, either in the form of content delivery or enabling technology, are considered wild cards. Wild card products, such as “boutique TV,” have become popular. By focusing on a smaller audience, companies can choose targeted advertising, VSHFLDOW\FRQWHQWDQGRWKHUPRUHUH¿QHGWDFWLFV to generate revenue. Enabling technologies, like the Sling Media products, allow the television experience to be time and place shifted to suit the FRQVXPHU¶VOLIHVW\OH7KHVHSURGXFWVFDQGLVUXSW the IPTV market by encouraging innovation and competition as well as take away some of the market share not captured by leading companies. Although these companies are not covered in this chapter in our concluding Star Model, these companies DUHLPSRUWDQWWRQRWHDVLQÀXHQFHVLQWKHPDUNHW :LOGFDUGVOLNH6OLQJ0HGLDDUHEULHÀ\GLVFXVVHG by summarizing their core competencies and how they disrupt the IPTV market.
Sling Media 6OLQJ0HGLD¶VVLJQDWXUHSURGXFWLVWKH6OLQJER[ Unlike a set top IPTV box, which restricts viewers to their living room, this product enables consumers to access TV programs anytime and anywhere over the Internet without the requirement of additional monthly subscription fees. Available in
key markets from Canada to Hong Kong, Sling 0HGLD¶V SURGXFW IDPLO\ VXSSRUWV XVHUV ZLVKLQJ to connect their PCs to their basic cable or HD subscription television services. Conceptually, the core line-up of Sling devices act like an IPTV set top box (STP) with one notable exception. Unlike a traditional IPTV STP that moves IP to a television set, the Slingbox moves television to the personal home computer network. This allows the device to be located anywhere, place-shifting content over the Internet to consumers. In addition to the Sling Box, Sling Media has also announced development of the SlingCatcher (Sling Media, 2007a). The SlingCatcher platform of solutions will place-shift content from networked home 3&¶VWRWHOHYLVLRQV The market for place-shifting devices proved to be $5 million within 6 months of the Slingbox launch, and it is expected to grow in double-digit numbers in the next 24 months (Frost & Sullivan, 2005). During the last week of September 2006, Sling Media announced a new product lineup that includes enhanced customer support. Sling Media has priced entry level Slingbox devices at approximately $179.99 (Sling Media, 2006b), encouraging growth in the market with an attractive entry-level price. Sling Media has attracted several investors to its business (Sling Media, 2006a), including the highSUR¿OH86FRQWHQWSURYLGHUV(FKR6WDU6DWHOOLWH //&DQG/LEHUW\0HGLD6OLQJ0HGLD¶V6HULHV$ and Series B investment rounds have provided Sling Media with the capital to maintain sustainability of the Slingbox product in the IPTV market. These W\SHVRIUHODWLRQVKLSVDUHDOVRKHOSLQJWRUHGH¿QH many of the carrier business models, particularly since the service will be expanded to devices other than PCs. While the Slingbox product can be used with any television service, it promotes network neutrality in promoting platform personalization devices. While contributing to network neutrality, Sling is paving the way towards building QOS standards for service by developing technology that optimizes connections and bandwidth utilization. The technology is called “SlingStream” (Sling Media, 2005), with beta testers reporting
545
IPTV Business Model Analysis
VLJQL¿FDQW JDLQV LQ UHFHLYHG TXDOLW\ RI YLGHR from the service. According to Sling, SlingStream monitors bandwidth and makes adjustments in the stream to compensate for low or high surges. As a result, the video quality delivered to the viewer is constantly optimized. While the above notes are impressive and aid in presenting a strong Star Model from Sling Media, a key strength of the Sling Media operations is in the attention paid to research and developPHQWDQGWRWKHXVHUFRPPXQLW\6OLQJ0HGLD¶V presence at the 2007 International CES (CES, 2007) provided an opportunity for the company leaders to promote several new advancements in the Sling Media product line; including that of the Clip+Sling (Sling Media, 2007b). The Clip+Sling is a tool for Slingbox customers to share content via an interactive portal with others. As a whole, advancements like Clip+Sling add to the strength of Sling Media against competition in the market by further encouraging product innovation and continuing to take away some of the market share not captured by other companies.
YouTube Founded in February of 2005 and purchased by Google in October of 2006, YouTube is a video sharing service that accepts content from private users and from corporate content creators (Wikipedia, 2006e). Based off of a Web site, the YouTube Service allows users to upload, view, and share video clips free of charge. The popularity of YouTube is fast growing; YouTube reports 100 million videos viewed daily (YouTube, 2006c). The video content ranges from short independent ¿OPV WR VSRUWV FOLSV FRYHULQJ FDWHJRULHV OLNH sports, travel, music, and pets. Under the YouTube general use policy (YouTube, 2006b), users can upload and share home movies and other personal videos for noncommercial use. According WR6HDUFK(QJLQH:DWFK¶V&OLF=1HZV 061 YouTube gained one of the largest percentages of the video-sharing audience at 43% (Hitwise, 2006). In addition to viewers, the Web site has managed to attract the attention of some major content owners. The Web site currently posts clips contributed
546
by major studios, with past collections including teasers for the NBC Fall 2006 season of The Of¿FH (YouTube, 2006d). YouTube users can expect to see more NBC content on the site, because in July 2006 YouTube and NBC formally announced (YouTube, 2006a) a strategic partnership. In this partnership, YouTube provides the online audience and NBC provides content. Because of this strong YouTube identity, Google announced as part of the purchase plan that the YouTube service would be maintained as a brand within the Google operations (Google, 2006d). Before the Google purchase, YouTube demonstrated an awareness of DRM and privacy that RWKHUFRPSDQLHVGLGQRWKDYH,QDQRGWR1DSVWHU¶V late 1990s blow-up, YouTube set a 10-minute maximum on video length to prevent piracy and protect DRM. However, users often worked around the restriction by splitting the videos and posting them as smaller segments, each shorter than 10 minutes. This practice was not a proven solution, DVVHHQLQODZVXLWVDJDLQVW
IPTV Business Model Analysis
sure on YouTube to develop stronger methods of protecting content offered on their site by partners like NBC and to increase its brand awareness in the marketplace.
Star Models for Sling Media and YouTube The tables below contain metrics for Sling Media and the YouTube brand service. Figure 2 is a diaJUDPRIWKH6WDU¿JXUHH[SUHVVLQJWKHPHWULFVLQD visual fashion for easy analysis. Although YouTube and Sling Media offer different products, the Star Model allows direct comparison of both companies based on how each company implements the business plan using the same metrics. The star provides a visual example of how YouTube and Sling Media compare. Although the companies offer dissimilar products, the star allows a method of comparing how each company performs against the same metrics. When companies are considering partnerships, this method provides
DQHDV\ZD\WR¿QGFRPSOHPHQWDU\UHODWLRQVKLSV and potential areas of weakness. The next section analyzes a partnership model to provide a case study on this use of the Star Model. At the end of the Google section, Figure 4 will demonstrate a comparison of the YouTube brand compared with Google, and the resulting relationship from the YouTube acquisition.
CBS CBS is an expert in content creation. As one of the oldest broadcasters in the United States, CBS has the experience in maintaining a network through extreme technical changes. From radio to the original black-and-white broadcasts and from the launch of cable television to today, CBS has survived. In March of 2006, CBS held the highest season-to-date ratings over the other primetime networks at 8.2/13 shares and rating (Nielsen Media 5HVHDUFK $SSO\LQJWKH1LHOVHQGH¿QLWLRQ
Figure 2. Sling vs. YouTube Star Model
547
IPTV Business Model Analysis
Table 11. Star Model Ratings for Sling Media, YouTube, and Google/YouTube venture Ranking Metric Parameter
Sling Media
YouTube
Google and YouTube
1: Power Against Suppliers
7
5
7
2: Power Against Customer Bargaining
7
5
7
3: Power Against New Entrants
6
5
7
4: Power Against Substitution
6
6
8
5: Power Against Competition
7
4
8
A: Customer Satisfaction
4
2
4
B: Customer Loyalty
4
4
4
C: Secured Solution
2
3
5
D: Market Regulation
3
2
4
E: Customer Satisfaction
3
2
4
Major Parameter
Minor Parameter
this rating translates into 9,036,400 U.S. viewers (Wikipedia, 2006b). CBS offers audiences a wide variety of news and entertainment programming. CBS supports these offerings with product enhancements. As an example, CBS has been broadcasting its primetime SURJUDPPLQJLQKLJKGH¿QLWLRQ +' VLQFH6HStember 2002. Since 2001, CBS has made the secondary audio program (SAP) available to stations to provide Spanish language and descriptive language feeds for soap operas and other programs. Another effort by the company to provide additional services is the “Always On” initiative, which is meant to make CBS news and programming information
548
available to all markets at all times. In Colorado, 'HQYHU¶V&%6&KDQQHO KWWS FEVGHQYHUFRP announced its own initiatives by partnering with Comcast for a video on demand (VOD) service that KDVWKHVWDWLRQ¶VUHFHQWQHZVFDVWVDYDLODEOHDWDQ\ time of the day. These efforts are part of how CBS is facing the challenge of adapting its product to the new frontier of video content, which is available to anyone at any time, anywhere. CBS leadership H[HPSOL¿HGWKHFRPSDQ\FRPPLWPHQWWRFRPSHWing in the new media frontier by participating in WKH DQQRXQFHPHQW RI 6OLQJ 0HGLD¶V &OLS6OLQJ (Sling Media, 2007b).
IPTV Business Model Analysis
In partnering with other companies like Google and Comcast, CBS can focus on its core competencies of content creation and rely on other companies to provide backbone and distribution. This business model is similar to that used by local stations that work with CBS. In partnering with companies such as AOL, Google, and Yahoo, CBS does not have to invest in an IPTV infrastructure. Furthermore, CBS can maintain copyright control over its mateULDODQGFUHDWHDQGVHOOFRQWHQWDVLWVHHV¿WWR¿OO its brand identity. This arrangement allows CBS to take advantage of IPTV as a revenue stream in order to supplement its traditional content creation business while minimizing sunk costs associated with creating a new physical network. CBS has VSHFL¿FDOO\SRVLWLRQHGLWVHOIDVWKHSULPDU\YLGHR partner for Google. According to Nielsen metered rankings (Sullivan, 2006), the Google partnership has clear attractions because Google ranks as the top search engine by capturing 49.2% of search WUDI¿F Capturing this market early with a strong partner like Google will give CBS an edge in becoming the premier provider of network television delivered via IPTV. This move, in turn, will help the company reduce costs on branding and customer education as IPTV gains in popularity. At the conclusion of this section, we will compare KRZ&%6¶V6WDU0RGHOFKDQJHVZKHQLQÀXHQFHG by various partners.
Google Video Google has launched its new Web site (www.video. JRRJOHFRP IRUYLHZLQJYLGHR¿OHV7KH:HEVLWH DOORZVXVHUVWRXSORDGYLGHR¿OHVRQLWVVHUYHUV This feature is very popular among users who wish to share their home videos with others. In addition, Google also allows content owners such as CBS and other content creators to upload copyrighted content onto the Web site. In the past, popular American sports programming providers, like the NBA, participated in this service (NBA, 2006). For customers, Google charges a fee to download this content in its entirety and prevents the illegal copying of the content. Under this service, popular shows such as Survivor, Brady Bunch, and selected
movie titles are available. All the downloaded video ¿OHVFDQEHSOD\HGRQO\YLDWKHSURSULHWDU\*RRJOH Video Player or iPod video player (Google, 2006a). +RZHYHU*RRJOH¶VFROODERUDWLRQZLWK$SSOHDQG its popular iPod product, is a positive step towards its marketability. Given that Apple has reported that its iTunes music store has reached 15 million videos purchased and downloaded, the relationship with Apple provides important visibility for Google to the market (Apple, 2006), which could lead WRJUHDWHUUHYHQXHIURP*RRJOH¶VYLGHRYHQWXUH Adding YouTube to the Google collection should only increase the depth of demographics Google attracts as a whole, further establishing Google as a go-to source for online video. In contrast to the iPod, Google offers a neutral platform for downloading material. The company reports on its Web site (video.google.com) that Google and some of its services are accessible to users on a variety of platforms, including all Web browsers and a number of mobile devices such as FHOOSKRQHVDQG3DOP3'$¶V7KHYLGHRVRQ*RRJOH Video can be streamed through Macromedia Flash Player 7.0+. This plug-in is widely available across several platforms and browsers (Google, 2006b). Google and AOL have also joined together with Intel to provide entertainment services to users via WKH,QWHO¶V9LLYWHFKQRORJ\,QWHO9LLYWHFKQRORJ\ has been designed to make it simple for companies, such as Google, to offer compelling new media platforms incorporating digital rights management and search technologies that work on open LQWHURSHUDELOLW\VSHFL¿FDWLRQV7KLVFRPELQDWLRQ provides consumers with the best use of Google VHDUFKHQJLQHDQG9LLY¶VDELOLW\WRRIIHUDQH[FLWLQJ new TV experience (Intel, 2006).
AOL and Warner Brothers The collaboration between AOL and Warner Bros. Domestic Cable Distribution (http://www.timewarner.com) offers a new form of broadband network. The network, called In2TV, will allow consumers to stream full-length episodes from favorite TV shows. Fans can enjoy special interactive features, including:
549
IPTV Business Model Analysis
• • •
Sing-alongs with favorite TV show themes The Ultimate TV Quiz: Fans can test their TV trivia skills AIM interviews with cast members
,Q79 ZLOO DOVR SURYLGH $2/¶V DGYHUWLVHUV with a compelling video inventory for in-stream broadband advertising, as well as opportunities for sponsorships and accompanying banner ads. An advertisement 15 or 30 seconds long is limited to a total of 1 to 2 minutes within each 30-minute episode in comparison to 8 minutes advertising on broadcast television. AOL also has its own channels with customized programs for streaming music videos and movie previews, but with 15-30 seconds of advertising (Byrne, Rowe, & Sarfaty, 2006). According to a CNET Report, AOL's video search engine has access to 15,000 pieces of content, with an average playback length of 3 minutes. AOL allows users to play back the content with technology already installed on PCs. This means that users do not have to download any new software (Mills & Olsen, 2005). It is important to note that this is a different DSSURDFK WKDQ *RRJOH¶V VHUYLFH *RRJOH DOORZV content to be downloaded, but this content can be played back on a proprietary video player. However, Google has an advantage in that it allows viewers WR XSORDG WKHLU FRQWHQW DQG GLVWULEXWH LW $2/¶V advantage is in the search-video market with its advertising capabilities and breadth of content. The company has strong ties with Hollywood through LWVSDUHQWFRPSDQ\¶V:DUQHU%URVVWXGLRVDQGLW can leverage to outdo video assets from Google and Yahoo. AOL is also selling and running 15second advertisements that run before video clips. ,QFRPSDULVRQ*RRJOH¶VPDQDJHPHQWWHDPLVVWLOO developing their plan on placing commercials before or adjacent to video, but has continually DVVXUHGLQYHVWRUVRIPDQDJHPHQW¶VFRPPLWPHQW to seek new revenue (Google, 2006c).
2.
legal peril. For instance, Google had to race to remove links from downloadable copyrighted content on its video search service, including The Matrix Revolutions and episodes of the Family Guy cartoons (Borland & Mills, 2005). Copyright protection risks have become VHYHUHDIWHUWKH866XSUHPH&RXUW¶VUXOing held companies liable for copyright content piracy that takes place on their networks (Google, 2006b). To minimize this risk, Google Video has deactivated links to copyrighted foreign content, such as Indian movies.
In looking at the IPTV business model for these partnerships, we rated their performances in the PDUNHWXVLQJ¿YHSDUDPHWHUV7DEOHVDQG 14 at the end of this chapter demonstrate the data recorded. From these metrics, this is the resulting Star Model. It is important to note that the ratings were produced from research on public records, GHPRJUDSKLFVDQG¿QDQFLDOGDWDVHUYLFHVVXFKDV Value Line and Yahoo! Finance. The Star Model was created as a way to easily reference the analysis completed on a company. The illustration should be used for reference on many contributing market factors in the current market rather than the caSDELOLW\RIWKHFRPSDQ\RYHUDOO7KH6WDU0RGHO¶V SXUSRVHLVWRUHFRUGWKHFRPSDQ\¶VSHUIRUPDQFHDW WKDWSRLQWLQWLPHDQGUHFRUGDQ,379SURYLGHU¶V stance as a quick snapshot in time. When Figure 3 is compared to Figure 4, the ability of the Star Model to model how different partnerships affect metrics is demonstrated. Figure 4 diagrams YouTube and Google as a separate brands, and analyzes the Google/YouTube acquisition under the assumption Google management practices will be dominant.
FUTURE TRENDS The Resulting Relationships Business Summary: 1.
550
The capability to text search into copyrighted audio and video puts the search providers in
Any IPTV business model can be evaluated using the Star Model provided in this chapter. Whether a company is looking to start an IPTV service, build on existing services, or partner with another
IPTV Business Model Analysis
Figure 3. CBS and resulting relationships
Figure 4. Google, YouTube, and resulting relationships
company, the Star Model can help study the business model that will succeed in the long term. As viewed in the Google and YouTube example, the Star Model can also aid in exploring strengths and weaknesses of acquired companies. This tool provides management with concise information on which to base decisions concerning business plan adjustments required for the new ventures to succeed. Additionally, the Star Model allows an IPTV provider to create a unique business model of its own, using proven models implemented by multiple providers.
will enable companies to better compete in the growing telecommunications industry. Stated more simply, the Star Model shows companies how to thoroughly evaluate their business and adapt to growth of the industry. By framing information in a familiar geometric shape, the Star Model clearly documents strengths and weaknesses for any business model. The Star Model also offers companies a method of monitoring both their own changing needs as well as consumer behavior. The resulting data offers companies the information necessary to adapt products and practices required to compete in the marketplace. We conclude that a business model in the IPTV industry can work for a company that invests in a partnership, a head-end middleware solution, or a major telecom provider with existing infrastructure. When discussing offerings that matter, we have looked at a select number of companies that KDYHEHFRPHSUR¿WDEOHSOD\HUVLQ,3797KURXJK the Star Model, which is an extended version of 3RUWHU¶V¿YHIRUFHVPRGHOUHDGHUVKDYHWKHDELOLW\WR evaluate the business models of an IPTV provider. Additionally, through case studies presented in this chapter, readers are provided with examples on how a Star Model can be used to evaluate different IPTV business models.
CONCLUSION The objective of this chapter is to offer companies and telecom professionals a basic strategic tool based on a common data model. With the built-in capacity to visualize comparable data, the Star Model allows businesses to record past performance and to utilize that information to maintain projected growth trajectories. The information gathered from this model can assist company leaders in developing a more effective business model by improving core business practices that
551
IPTV Business Model Analysis
This chapter developed the Star Model for a company to use in order to evaluate current and future business in an appealing geometric model. By using the Star Model, service providers can attain a more mature business model and be better prepared to succeed in the IPTV market. The chapter provides a service provider, or any potential investor, a powerful and easily understood tool to build a strong and mature business model within a short time frame. This will result in many healthy new entrants to the market, further promoting investment and competition in the marketplace. In turn, having a healthy marketplace with attractive product offerings will further encourage global penetration of broadband.
REFERENCES Apple Computer. (2006). iTunes music store downloads top one billion songs (Apple Press Release). Retrieved July 16, 2007, from http://www.apple. com/pr/library/2006/feb/23itms.html Borland, J. & Mills, E. (2005). Now playing on Google: Matrix, family guy. IPTV. Retrieved July 16, 2007, from http://www.cnet.com.au/software/ internet/0,239029524,240055663,00.htm Burbridgea, C. (2006). IPTV: The dependencies for success. Computer Law & Security Report, 22, 409-412. Retrieved July 16, 2007, from doi:10.1016/ j.physletb.2003.10.071 Byrne, J.R., Rowe, S., & Sarfaty, R. (2005, November). AOL and Warner Bros. announce In2TV, new broadband network on AOL.com. Time Warner. Retrieved July 16, 2007, from http://www.timewarner.com/corp/newsroom/pr/0,20812,1129725,00. html Carney, K., Fandre, M., Nambiath, R., & Shringarpure, N. (2006, April). IPTV: Business model DQDO\VLV IURP 3RUWHU¶V ¿YH IRUFHV SHUVSHFWLYH 8QSXEOLVKHG 0DVWHU¶V &DSVWRQH %RXOGHU &2 University of Colorado, Boulder. &DUU1* 7KHSXEOLFZDQWV\RXUSUR¿WV Forrester Magazine. Retrieved July 16, 2007,
552
from http://www.forrester.com/magazine/articles.003/03.responsibility.pdf &(6 -DQXDU\ :RUOG¶VODUJHVWWHFKQRORJ\ tradeshow attracts record international attendance. Retrieved July 16, 2007, from http://www.cesweb. org/press/news/rd_release_detail.asp?id=11228 Downes, L. (1997). Technosynthesis: Beyond Porter (Premier Issue). Context Magazine. Diamond Management & Technology Consultants, Inc. Retrieved July 16, 2007, from http://www. contextmag.com/setFrameRedirect.asp?src=/current/archive.asp Frost & Sullivan (2005). Market segment analysis H[HFXWLYHVXPPDU\PDUNHWUHVHDUFK¿QGLQJQRUWK American consumer DVR and space-shifting devices markets. Retrieved July 16, 2007, from http://www.frost.com/prod/servlet/report-document.pag?docid=6A20-01-01-02-02 Frost & Sullivan (2006, January 13). IPTV markets in Europe—telecom operators strategies. Retrieved January 13, 2006 from http://www.frost. com/prod/servlet/report-brochure.pag?id=B62501-00-00-00 Google. (2006a). Google video support, downloading basics. Retrieved July 16, 2007, from http://video.google.com/support Google. (2006b). Google video help center, downloading videos/downloading basics/what video viewer does Google use? (Google Video). Retrieved July 16, 2006, from http://video.google.com/support/bin/answer.py?answer=27998&topic=8700 *RRJOH F ,QYHVWRU UHODWLRQV ZKDW¶V QHZ Retrieved July 16, 2007, from http://investor.google. com/whatsnew.html Google. (2006d). Press center. Google to acquire YouTube for $1.65 billion in stock. Retrieved July 16, 2007, from http://www.google.com/press/pressrel/google_youtube.html Hitwise Press Center. (2006). Hitwise data shows overall visits to video search sites up 164%. Retrieved July 16, 2007, from http://www.hitwise.
IPTV Business Model Analysis
com/press-center/hitwiseHS2004/videosearch. php InfoDev. (2004). InfoDev telecommunications handbook. Retrieved July 16, 2007, from http:// www.infodev.org/en/Publication.22.html Intel. (2006, January). Intel teams with Google, AOL. Light Reading. Retrieved January 06, 2006, from http://www.lightreading.com/document. asp?doc_id=86457 •
Cross-referenced with Intel News Releases from Intel Corporate Site, Press Room. Intel, Google Announce Plans to Bring Video Search Technology to the Living Room on New Intel® Viiv™ Technology Platforms, January 2006. Retrieved February 2, 2007 from http://www.intel.com/pressroom/archive/releases/20060105corp_n.htm
And • “AOL and Intel Join Forces to Bring AOL On–Demand Entertainment to the Living Room” January 2006. Retrieved February 2, 2007, from http://www.intel.com/pressroom/ archive/releases/20060105corp_g.htm ,379 0RQLWRU %XVLQHVV FRQ¿GHQFH LQ IPTV: Few illusions, high hopes (Issue 1). Accenture, Written in co-operation with the Economist Intelligence Unit. Retrieved July 16, 2007, from http://www.accenture.com/xdoc/en/industries/ communications/iptv/research/iptv_monitor1.pdf IPTV World Forum Eastern Europe. (2006, June 22-23). Delivering live and on-demand content over broadband. Hungary. Retrieved July 16, 2007, from http://www.iptv-easterneurope.com/ Law.com. (2006, August 10). The recorder. YouTube lawsuit is latest internet copyright case for Weil *RWVKDOSDUWQHU.RE\ODU];HQLD35HWULHYHG-XO\ 16, 2007, from http://www.law.com/jsp/article. jsp?id=1155114327773 Mills, E., & Olsen, S. (2005, June). AOL launches video search service. CNET. Retrieved July 16, 2006, from http://news.com.com/AOL+launches+v ideo+search+service/2100-1032_3-5770777.html
Miniwatts Marketing Group. (2006, March 31). Internet world statistics: Usage and population statistics. Internet usage statistics - the Big Picture: World Internet users and population stats. Retrieved July 16, 2007, from http://www.internetworldstats. com/america.htm#us •
Cross-referenced with The World Factbook: Rank Order-Population. Retrieved July 18, 2006, from https://www.cia.gov/cia/publications/factbook/geos/xx.html#People
MSN. (2006). MSN joins the video sharing fray. Retrieved July 16, 2007, from http://www.clickz. com/showPage.html?page=3623470 NBA. (2006). Google taps NBA to launch video marketplace. Retrieved July 16, 2007, from http:// www.nba.com/news/google_060106.html Nielsen Media Research. (2006). Retrieved July 16, 2007, from Zap2it.com Ofcom. (2006). Draft annual plan 2006, 2007, Building on competition and innovation. Retrieved July 16, 2007, from http://www.ofcom.org. uk/about/accoun/reports_plans/annual_plan0607/ statement/statement.pdf Porter, M.E. (1998). Competitive strategy: Techniques for analyzing industries and companies (Original work published 1980). New York: Free Press. Sling Media. (2005). Sling media releases new VRIWZDUHDQG¿UPZDUHIRUDZDUGZLQQLQJ6OLQJER[ (Press release). Retrieved July 16, 2007, from http:// us.slingmedia.com/object/io_1157566427362. html Sling Media. (2006a). About us/investors. Retrieved July 16, 2007, from http://us.slingmedia. com/page/pg_79.html Sling Media. (2006b). Sling US home page. Retrieved July 16, 2007, from http://us.slingmedia. com/ Sling Media Press Release (2006, September). Press release: Sling Media's next generation Slingbox
553
IPTV Business Model Analysis
family delivers. Retrieved July 16, 2007, from http:// us.slingmedia.com/object/io_1159468208005. html Sling Media. (2007a, January). Sling media announces SlingCatcher (Press release). Retrieved July 16, 2007, from http://us.slingmedia.com/object/io_1168286861787.html Sling Media. (2007b, January).: Sling media announces clip and sling (Press release). Retrieved July 16, 2007, from http://us.slingmedia.com/object/io_1168395718976.html •
Cross-referenced with Press Release: Sling Media and CBS Announce Beta Test of New Delivery System for Video. January 2007. Retrieved February 20, 2007 from http://us.slingmedia.com/object/io_ 1168395380643.html
YouTube. (2006a). Broadcast yourself! (Press Releases). NBC and YouTube announce strategic partnership. Retrieved July 16, 2007, from http:// youtube.com/press_room_entry?entry=c0g5NsDdJQ YouTube. (2006b). Broadcast yourself! Terms of use. Retrieved February 2, 2007, from http://youtube.com/t/terms YouTube. (2006c). Broadcast yourself! YouTube home page, fact sheet. Retrieved July 16, 2007, from http://youtube.com/t/fact_sheet YouTube. (2006d). Director videos. 7KH 2I¿FH Pam & Jim Moments. Credited to NBC. Retrieved July 16, 2007, from http://youtube.com/ ZDWFK"Y 2L;K;MEN
KEY TERMS Sullivan, D. (2006, January 24). Nielsen NetRatings search engine ratings. Retrieved July 16, 2007, from http://searchenginewatch.com/reports/article. php/2156451 Tribune Media Services Inc. (2006). Week of March 20, 2006 through March 26, 2006. Retrieved July 16, 2007, from http://tv.zap2it.com/tveditorial/tve_ main/1,1002,272%7C%7C%7Cweekly,00.html.
Business Model: A tool highlighting the relationship between a set of elements in order to provide business logic to a company. Complete model may include statements on targeted customer segments, pricing models, and revenue models. Churn: The rate at which contractual customers choose to leave a supplier.
Value Line. (2006). About value line (Home page). Retrieved July 16, 2007, from http://valueline. com/
Content Aggregators: Companies that use their resources in order to develop a library of content to sell to consumers.
Wikipedia. (2006a). IPTV. Retrieved July 16, 2007, from http://en.wikipedia.org/wiki/IPTV
Content Providers: Companies or corporations that provide content to consumers.
Wikipedia. (2006b). Nielsen ratings. Retrieved July 16, 2007, from http://en.wikipedia.org/wiki/ Nielson_rating
Internet Protocol (IP): A data level protocol utilized to transport information across a dataswitched network.
:LNLSHGLD F 6(&¿OLQJ5HWULHYHG-XO\ IURPKWWS HQZLNLSHGLDRUJZLNL6(&B¿Oings
Internet TV: (1) Internet content that is disSOD\HGWKURXJKDFRQVXPHU¶VWHOHYLVLRQWKURXJKD variety of vendors; (2) Television content delivered over the Internet.
Wikipedia. (2006d). SWOT. Retrieved July 16, 2007, from http://en.wikipedia.org/wiki/SWOT Wikipedia. (2006e). YouTube. Retrieved July 16, 2007, from http://en.wikipedia.org/wiki/YouTube 554
Internet Protocol Television (IPTV): Digital television content delivered through IP based networks.
IPTV Business Model Analysis
Metrics: Quantitative and qualitative data that is organized and tabulated in an effort to analyze trends.
telecommunications provider or consumer who relies on telecommunications to transfer important information.
Multimedia: Voice, video, or data that is displayed on a variety of platforms.
Star Model: A competitive analysis model used by a company to display the strengths and ZHDNQHVVHV RYHU WLPH GHYHORSHG IURP 3RUWHU¶V Five Forces.
Porter’s Forces: 0LFKDHO3RUWHU¶V)LYH)RUFHV GLVVHFWIDFWRUVWKDWLQÀXHQFHFRPSHWLWLRQDQGHFRQRPLFLQÀXHQFHVDPRQJSOD\HUVLQDPDUNHW Security: The strategy for protecting a business from ill-will, harm, or loss. A necessity for any
Telco: (1) A company that provides telecommunications services, such as data, phone, or subscription television; (2) Common carrier, also known as an incumbent or local exchange carrier.
555
IPTV Business Model Analysis
APPENDIX Major Parameters The tables below list the metrics and questions used in the Star Model while rating major parameters for the companies selected in the case studies. Table 1. Metrics for major parameter [1. Power Against Suppliers]___________________________________ Metric Question____________________________________________________________ 1.
Does the supplier company enjoy majority market share?
2.
Does the supplier company have a global presence?
3.
Would the cost of changing suppliers of materials used to make the product have a negligible impact on the business? Is the cost of switching suppliers negligible?
,VWKHSUR¿WPDUJLQORZUHVXOWLQJLQJUHDWHUPRWLYDWLRQIRUVZLWFKLQJVXSSOLHUV",VWKHUHDFRVW VDYLQJVEHQH¿WWRVZLWFKLQJVXSSOLHUV" 5.
Does the company have access to substitute product from market reducing dependency from current suppliers?
6.
Does the company have access to alternative suppliers with capability to deliver similar or higher TXDOLW\UDZPDWHULDOVUHTXLUHGIRUWKH¿QDOSURGXFW"
7.
Does the company buy product in huge volume from a supplier?
8.
Is the IPTV industry important for the supplier?
9.
Is the product from the supplier easily available in the market from multiple vendors?
10. Has the supplier already made investments for the product (For example: customization of manufacture to requirements)?
Table 2. Metrics for major parameter [2. Power Against Customer Bargaining]___ _______________________ Metric Question__________________________________________________________ 1.
Is the product offered by the company unique and not replaced by substituting product?
2.
Does the company provide a lot of a la carte pricing models to customers?
556
IPTV Business Model Analysis
3.
Is there a lock-in period or switching cost for the services being provided?
4.
Is the service being provided critical and does it have high bonding with customer?
5.
Had the company kept the price of the service the same without any lucrative package during last 6 months?
6.
Does the company deliver unique content as part of the service?
7.
Does the company deliver unique interactive contents as part of the products?
8.
Can the customer buy smaller quantities than advertised?
9.
Does the company hold monopoly status for delivery of the product?
10. Is the cost of exit high for customer?
Table 3. Metrics for major parameter [3. Threat of New Entrants]_________________________________ Metric Question_________________________________________________________ 1.
Does the company offer a proprietary product?
2.
Does the company offer a product or service which cannot be duplicated with other technology?
3.
Is there a new market potential for the product?
4.
Is the company reliant on stable, matured product technology?
5.
Does the company participate in international standards bodies in developing the product? (Longevity of product in market)
,VWKHSURGXFWSDUWRIWKHFRPSDQ\¶VFRUHSDFNDJLQJ"
7.
Is the company enjoying economies of scale with respect to R&D, manufacturing, and marketing while offering the product?
8.
Is the cost of entry high resulting in fewer new entrants into the market?
9.
Is government policy advantageous to the company and discouraging to new entrants?
10. Will the cost of switching/attracting new customers discourage any new entrant from gaining market share?
557
IPTV Business Model Analysis
Table 4. Metrics for major parameter [4. Power Against Substitution]________________________________ Metric Question_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
,VWKHSURGXFWGLI¿FXOWWREHUHSOLFDWHGE\FRPSHWLWRUV"
2.
Has the company protected its intellectual property in the solution or product?
,VWKHUHDKLJKLQLWLDOFRVWWRSDUWLFLSDWHLQWKLVFRPSDQ\¶VVROXWLRQ"
4.
Is there a switching cost to the customer?
5.
Do industry analysts predict growth in the industry?
6.
Does the company have a brand plan?
7.
Does new technology offer new choices for consumers?
8.
Are there exit barriers?
9.
Does the company enjoy brand recognition and customer loyalty?
10. Is government policy advantageous to the company?
7DEOH0HWULFVIRUPDMRUSDUDPHWHU [5. Power Against Competition]_ _______________________________ Metric Question____________________________________________________________ 1.
Does the company enjoy monopoly status?
2.
Is the company the sole probable supplier to its end customers?
3.
Is the product offered by the company unique and cannot be replaced by a competing product?
4.
Does the company hold patents and rights for the products being delivered?
5.
Does the company participate in standards bodies?
'RHVWKHPDUNHWIRUFRPSDQ\SURGXFWKDYHOHVVWKDQWKUHHVLJQL¿FDQWDQGDFWLYHSOD\HUV"
'RHVWKHFRPSDQ\KDYHVLJQL¿FDQWPDUNHWVKDUHDKHDGRILWVFRPSHWLWRUV"
8.
Is there a high cost of entry for the new player?
558
IPTV Business Model Analysis
9.
Does the company have a unique advantage with respect to economies of scale and is it recognized as a leader in the market?
10. Is the technology delivered by the company proprietary?
Minor Parameters The tables below list the metrics and questions used in the Star Model while rating minor parameters for the companies selected in case studies. Table 6. Metrics for minor parameter [A. Product Usability]_____________________________________ Metric Question____________________________________________________________ 1.
Does the company use transaction information as a feedback tool to improve product?
2.
Does the company have customer-loyalty programs?
3.
Do customers use the product more than expected?
4.
Does the company provide incentives for referral?
5.
Does repeat use lower the cost of the investment for the customer?
Table 7. Metrics for minor parameter [B. Customer Loyalty]_____________________________________ Metric Question____________________________________________________________ 1.
Does the product have a loyal customer base?
2.
Is ARPU increasing for the company?
3.
Is the company experiencing subscriber growth?
4.
Does the company have unique IPTV features that attract customers?
,VWKHSURGXFWZHOOWKRXJKWRXW"$UHGLUHFWLRQVDQGRIIHULQJVFOHDUDQGXQGHUVWDQGDEOHRQ¿UVWWLPH use?
559
IPTV Business Model Analysis
Table 8. Metrics for minor parameter [C. Secured Solution]_________________________________ Metric Question________________________________________________ 1.
Is the security solution by the company open per industry standard?
2.
Does the company participate in copyright-protection programs?
3.
Does the company deliver unique content to customers?
4.
Is the company able to trace all transactions from the customer?
5.
Is the company able to trace any intrusion and restrict piracy?
Table 9. Metrics for minor parameter [D. Market Regulation]________________________________ Metric Question_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1.
Is an independent agency in charge of regulation (not part of government?)
2.
Is this a free-economy market controlled by market forces and not regulation?
3.
Is this an open market, according to ITU rankings?
4.
Is the regulatory process transparent? Please see InfoDev (2004)
5.
Does the company practice due diligence in participating in regulation?
Table 10. Metrics for minor parameter [E. Customer Satisfaction]______________________________ Metric Question_________________________________________________________ 1.
Is there a transaction? (Money, information, time)
2.
Is there a response-time commitment from the company?
3.
Does the company allow multiple mechanisms of support: e-mail, phone, chat, and so forth?
4.
Are all support issues closed with minimal number of iterations from customer?
5.
Are all support issues closed with no repeat request?
560
IPTV Business Model Analysis
Table 12. Major Parameter Ratings for Google Video
Metric Parameter
Power Against Supplier
Commentary
The value of this metric is just below the half-way mark, partially because of the symbiotic relationship between content owners and Google. Google has to depend on the content providers for its success;
5
KRZHYHUWKHFRQWHQWRZQHUVDOVRQHHG*RRJOH¶VZRUOGZLGHSUHVHQFHDQGSRSXODULW\WRVHOOWKHLUFRQWHQW
Power Against
Google has a medium rank for this metric. Although consumers have access to unique content from CBS,
Customer Bargaining
they also have the option of buying content from competitors such as AOL. In addition, consumers have ÀH[LELOLW\RIEX\LQJFRQWHQWWKDWWKH\FDQVHOHFWLYHO\FKRRVH
5
Power Against
The value of this metric is high. While Google has an immediate competitor, AOL, Google has a strong
New Entrants
brand image that they must not alter. For entry in this competitive market, a new entrant must have a
5
Power Against Substitution
reputation equal to Google and AOL. Google enjoys a majority market share right now.
Google enjoys the advantage of being a popular brand. It also protects the intellectual property and copyrighted content through proprietary means; thus, the value of this metric is high.
7
Power Against Competition 7
The rivalry among the competing partnerships of Google/CBS and AOL/Warner Bros. is low because both players have an agreement. Both players distribute content owned by CBS. In addition, Google has a 5% stake in AOL.
561
IPTV Business Model Analysis
Table 13. Major Parameter Ratings for AOL Metric
Commentary
Parameter Power Against Suppliers
AOL has to depend on the content providers for its success. However, the content owners also QHHG$2/¶VZRUOGZLGHSUHVHQFHDQGSRSXODULW\WRVHOOWKHLUFRQWHQW,QDGGLWLRQ$2/KDV
5 Power Against Customer Bargaining 5
made considerable investment in this sector. AOL is rated at the half-way mark for this metric. Though the consumers have access to unique content from CBS and other creators, they also have the option of buying content from FRPSHWLWRUVVXFKDV*RRJOH7KHFRQVXPHUVDOVRKDYHÀH[LELOLW\RIEX\LQJVHOHFWHGFRQWHQW
Power Against New Entrants 4 Power Against Substitution 6
The value of this metric is close to the half-way mark. AOL has shown how partnering can have great advantages. New entrants are likely to follow this successful business model. AOL enjoys the advantage of being a popular brand. It also protects the intellectual property and copyrighted content through proprietary means. In addition, it has strong ties with Hollywood and thus content is available comparatively easily; thus, the value of this metric is high.
Power Against
The Power against Competition is low for AOL is low because Google still has a majority
Competition
market share. Google also just invested $1 billion for a 5% stake in AOL. This investment
3
DOORZV*RRJOHXVHUVWRDFFHVV$2/¶VSRSXODUFRQWHQW%RWKWKHSOD\HUVGLVWULEXWHFRQWHQW owned by CBS.
Table 14. Major Parameter Ratings for CBS/Google and AOL/Warner Brothers
Ranking
CBS and Major Parameter
CBS
CBS and Google
AOL& Warner Bros.
Power Against Supplier
6
8
7
Power Against Customer Bargaining
6
8
8
Power Against New Entrants
5
7
9
Power Against Substitution
6
9
9
Power Against Competition
5
6
5
562
563
Chapter XXXV
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry Stanford L. Levin Southern Illinois University Edwardsville, USA John B. Meisel Southern Illinois University Edwardsville, USA Timothy S. Sullivan Southern Illinois University Edwardsville, USA
ABSTRACT This chapter describes the far-reaching effects of broadband Internet access on the motion picture indusWU\,W¿UVWSURYLGHVDVXPPDU\RIWKHHIIHFWVRQWKHLQGXVWU\¶VEXVLQHVVPRGHO WKH,QWHUQHW SDUWLFXODUO\ when combined with broadband connections) provides a new window for the movie studios to utilize in releasing their product. It next examines the ways that legal, political, and cultural environments are DOUHDG\LQÀXHQFLQJWKHLQGXVWU\¶VVHDUFKIRUDQHZEXVLQHVVPRGHOWRUHSODFHWKHROG)LQDOO\ZHGUDZ on lessons from the music industry to predict how the industry will ultimately incorporate broadband technology into a new business model. The authors believe that the motion picture industry provides an excellent case study of broadband’s effects on a mature industry.
INTRODUCTION The United States motion picture industry, often referred to as Hollywood, is facing and will continue to face shocks to its existing business model and vertical relationships due to technological innovations in computer hardware and software
applications associated with the Internet. Two major developments have led to the entry of a new distribution technology: the online delivery of movies to consumers. First, personal computers have become increasingly powerful and the storage capacity on a hard disk has increased VLJQL¿FDQWO\6HFRQGFRQVXPHUVDUHLQFUHDVLQJO\
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
upgrading their Internet connections from dialup to broadband and new consumers are initially adopting a broadband Internet connection. These developments create shocks to the existing ways of conducting business for the movie industry similar to the threats experienced by the music industry ZLWK WKH SRSXODUL]DWLRQ RI ¿OH VKDULQJ VRIWZDUH supplied by Napster. These shocks originate in both the internal and external environments of the industry. Internally, WKHLQGXVWU\PXVWGHFLGHKRZWR¿WWKHUHOHDVHRI YLGHRRQGHPDQG¿OPVLQWRLWVH[LVWLQJVHTXHQFH of distribution channels, including theater release and video store release. The position of the Internet in the distribution sequence should be affected by factors similar to those that affected the placement of video rental and purchase in the distribution sequence 20 years ago. The external political, legal, and economic environments confronting Hollywood are causing the movie companies to DGMXVW WR LVVXHV VXFK DV SLUDF\ DQG ¿OH VKDULQJ facilitated by peer-to-peer networks as a result of the Internet, the lowering of reproduction costs of digital content, and the lowering of distribution costs of that content over the Internet. Related noteworthy characteristics of the Internet that have given rise to the challenges facing content providers in general are its end-to-end architecture and its openness, blurring jurisdictional boundaries. The resulting issues are analogous to what the music industry has experienced in its adjustment to the LPSDFW RI WKH ,QWHUQHW RQ WKH PXVLF LQGXVWU\¶V business model. The chapter is organized as follows. In the second section, the existing distribution channels for and revenue sources from motion pictures are GHVFULEHGDQGIDFWRUVWKDWZLOOLQÀXHQFHWKHSRVLtion of the Internet in that sequence are analyzed. Then, the legal, cultural, political, and economic environments are examined, including piracy and ,QWHUQHW¿OHVKDULQJDQGWKHYDULRXVHIIRUWVWKDWWKH industry has undertaken to confront these issues. 7KH¿QDOVHFWLRQRIWKHFKDSWHUZLOOH[DPLQHWKH lessons that the motion picture industry can learn IURP WKH PXVLF LQGXVWU\¶V H[SHULHQFH ZLWK WKH Internet.
564
BACKGROUND: DISTRIBUTION CHANNELS During its lifespan, a motion picture is distributed through various channels, or “windows.” In typical chronological order, these include theater release, DVD release (both rental and sales), home pay-perview (PPV), pay cable television (such as Home %R[ 2I¿FH >+%2@ DQG 6KRZWLPH DQG ¿QDOO\ standard cable and free broadcast television. Figure 1 summarizes the progression of a motion picture through various windows and indicates the new roles played by the Internet. Windowing allows the motion picture industry to price discriminate. By price discriminating, the industry forces those consumers who are most eager to see a new movie and those who want the best experience (in terms of video and sound quality) to pay the highest price. This results in KLJKHULQGXVWU\SUR¿WVWKDQUHOHDVLQJWKHPRYLH into all windows simultaneously. In particular, the consumers that are most eager to see a movie with the highest possible quality pay the highest price (at a theater). Consumers willing to wait some number of months (and settle for a smaller screen and, typically, lower-quality sound) pay lower prices (DVD and home PPV). Ultimately the most patient consumers pay essentially nothing (broadcast/basic cable). 7KH¿UVWZLQGRZWKHDWHUVRIIHUVWKHKLJKHVW quality experience in terms of video and sound quality. But, in part because it offers a superior experience (and an experience that is relatively expensive to provide), theater releases are priced the most expensively. North American sales of movie tickets totaled just under $9 billion in 2005, down about 5% from 2004 (McBride, Grant, & Marr, 2006). Worldwide sales were approximately $23 billion. The theater window accounts for less than a quarter of motion picture industry revenue (McBride et al., 2006). Decades ago, movie ticket sales accounted for the vast majority of motion picture industry revenue, but the opening of other windows has made ticket sales a less important window in terms of total revenue. By placing movies in the theater window, exclusively, for a number of months
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
Figure 1. Distribution windows
Motion Picture Studios
Windows Window #1 Movie Theaters ($9 billion in 2005)
Window #2 DVD Release ($23 billion in 2005)
Brick and Mortar DVD Sales
Internet DVD Sales
Window #3 Internet sales and rentals
Window #4 TV (including PPV, premium networks, standard cable, and over-the-air broadcasts)
Brick and Mortar DVD Rentals
Internet DVD Rentals
the industry is able to charge the highest prices to consumers who utilize this window. The second window, DVD sales and rentals, now represents the largest portion of motion picture industry revenue. Combined U.S. revenue from sales and rentals totals over $23 billion and accounts for about half of industry revenue (McBride et al., 2006). While home rentals are important, they have been passed, in terms of revenue, by home sales. U.S. DVD sales total about $16 billion per year (Marr, 2006) with Wal-Mart having the largest share of these sales with about 40% of the market (McBride & Marr, 2006). A new component in home rentals is the Internet rental company. These include Filmcaddy (owned by Blockbuster), Wal0DUWFRP RZQHGE\:DO0DUW DQG1HWÀL[FRP 7KHVHVHUYLFHVFKDUJHDÀDWPRQWKO\IHHW\SLFDOO\ about $20 per month. This entitles users to a set number of movies (higher fees allow a higher numbers of movies). Users request movies from the Internet site and they are sent as they become DYDLODEOH8VHUVFDQNHHSWKHPRYLHVLQGH¿QLWHO\
and subsequent movies are delivered when the previous movies are returned. Other revenue comes from much smaller sources. One relatively small window, in terms of revenue, but one that is expected to grow in coming years, is home PPV. Home PPV differs from standard cable movie channels in that customers must pay an extra fee per movie. At the present time this window represents only about 1% of U.S. revenue (McBride et al., 2006). The motion picture industry earns revenue in this window through revenue sharing agreements with the cable companies (roughly 50-50). This window should increase in importance as video on demand (VOD) capabilities improve. VOD allows users to watch the movie as desired, pausing and rewinding as with a DVD. However, the actual use of the service has stalled as customers must purchase or rent additional equipment to take full advantage of this ability. Most remaining revenue comes from selling broadcast rights to premium movie channels (such as HBO and Showtime) and networks that
565
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
are part of most basic cable packages (including the major over-the-air networks). The immediate effect of the Internet is to create a new means of distribution—a new window. The viability of this window, at least for the time being, has been limited by technology. But as access to broadband increases, the distribution of movies over the Internet will become more important. Where will this new window be placed in the life cycle of a motion picture? Owen and Wildman DQDO\]HWKHLQGXVWU\¶VWLPLQJRIWKHZLQGRZV (not including the Internet) and point to a number of considerations that determine the relative timing of each window. Three of these considerations are especially relevant for the Internet window—the amount viewers are willing to pay to utilize the window, the number of additional viewers who will use the window, and the ease of copying broadcasts that utilize the window. All else equal, a window will be timed earlier the more that viewers are willing to pay when YLHZLQJ LQ WKDW ZLQGRZ &RQVXPHUV¶ ZLOOLQJness to pay for Internet broadcasts is likely to be considerably less than a theater, slightly less than their willingness to pay for PPV broadcasts, but slightly more than they would be willing to pay for a rental. The main drawback of Internet broadcasts is that, while top-of-the-line computer systems can be quite impressive, standard computer setups have screens and sound that are far inferior to even a standard television set (let alone wide-screen home theater systems). One potential advantage RI WKH ,QWHUQHW EURDGFDVW LV ÀH[LELOLW\²WHOHYLsion broadcasts cannot be stored on a laptop and viewed, for example, on an airplane. However as on-demand video systems (such as Tivo) become more sophisticated, they may eventually match WKHÀH[LELOLW\RIWKH,QWHUQHWEURDGFDVWV,QWHUQHW broadcasts do have the advantage over DVD rentals in convenience (there is no need to make a return trip to the video store) and in the speed at which they can satisfy consumer demand compared to the rental services that rely on the mail. All else equal, a window will be timed earlier the more additional viewers are available in that window. The number of viewers gained with Internet broadcasts is smaller than DVD and cable. It is
566
true that the number of U.S. households with the capability for Internet access nears the 80% with cable or satellite service (Angwin & Pasztor, 2006). But of these Internet households, over half have slow, dial-up access. While most U.S. households have the option of signing up for broadband, about 50 million households currently do so (Reuters, 2006). This is a crucial consideration, as feature ¿OPVKDYH¿OHVL]HVWKDWFDQEHVHYHUDOKXQGUHG WLPHVODUJHUWKDQDW\SLFDOPXVLF¿OH :HUEDFK 2005). Of these just over half utilize a cable modem, with most of the rest using digital subscriber lines or DSL (Grant, 2006). This suggests that wherever the Internet window initially resides, it is likely to be moved forward in time as broadband access increases. This occurred during the past 15-20 years as the VCR/DVD window has been pushed earlier at the same time as the number of households with VCRs, and then DVDs, increased. All else equal, a window will be timed earlier the harder it is to copy the broadcast in that window. Copy protection technology (and the efforts to evade it) continue to evolve, but, currently, streaming video is about as hard to duplicate as DVDs but easier than PPV, which is to say that it is beyond the level of expertise (and patience) of the typical consumer but still a potential source of concern. Estimates are that more than half of peer-to-peer users are swapping video (not all of it illegally). %RRWOHJJHGFRSLHVRI¿UVWUXQPRYLHVIRUH[DPSOH typically arrive on the Internet 24 hours after their U.S. theatrical release (Mathews, 2002a). These are typically low-quality copies made from hand-held cameras (although the industry is concerned that, DVWKH\VZLWFKIURP¿OPUHHOVWRGLJLWDOGLVWULEXtion to theaters, illegal digital copies may appear). Of more immediate concern for the industry is the potential illegal distribution of digital video that is redistributed after being viewed by authorized users (as happened with music). The industry has responded with antipiracy encryption technology and legislative efforts to combat this unauthorized distribution. If motion pictures become easy to copy and distribute over the Internet, this will have a large impact, especially on home video revenue and home PPV and pay TV revenue and will encourage the industry to delay the Internet
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
ZLQGRZVLJQL¿FDQWO\ VHHWKHVHFWLRQVRQSLUDF\ and lessons from the music industry). Based on these considerations, the Internet window is likely to be late in the short run but is likely to move up in time as technology improves—eventually coinciding with the DVD release. There are a number of ventures that currently offer Internet movie downloads. Two of these, Movielink and CinemaNow, have offered online movies for several years but have failed to catch on. However, two recent entrants have better name recognition and may revolutionize this industry. $PD]RQFRP¶V 8QER[ UHFHQWO\ EHJDQ RIIHULQJ movies from most major studios. Users have the option of “purchasing” the movies, and watching them as many times as they wish, whenever they wish, or “renting” the movies, in which case they must be watched within 30 days of the download (Vascellaro, 2006). Another potentially major player, Apple, has recently added over 75 Disney movies to its iTunes site and plans to add more in WKHQHDUIXWXUH :LQJ¿HOG 0DUU 2WKHU companies considering entering this industry include Wal-Mart, Blockbuster, Comcast, Time Warner, and AT&T (Vascellaro). Currently, the studios seem to be delaying their Internet release until several months past the rental/purchase window. However, as more households gain access to broadband, the Internet broadcast window should move forward, perhaps until it is simultaneous with the rental/sales window. This prediction, however, is subject to one important caveat—the ability of the studios to create and enforce anti-piracy measures. The studios will not be foolish enough to release easily-copied digital versions of their movies only to see unauthorized versions competing with DVDs sitting on the shelf. Finally, as more homes obtain broadband Internet access through cable, the line between the home PPV, pay TV, and Internet windows will become blurred. The lines are already blurring as people use their computers to watch DVDs. As more households obtain their Internet access through the cable lines that run to their television, Webenabled televisions are likely to seem less exotic to typical households. With a browser-enhanced television, viewers could, in addition to visiting
standard Web sites for text and audio, visit sites such as iTunes and Unbox and view video on a ODUJHVFUHHQ$SSOH¶VQHZL79GHYLFHFDSDEOHRI television playback and having a storage capacity of up to 150 hours of video, may begin a general PRYHPHQWLQWKLVGLUHFWLRQ :LQJ¿HOG0F%ULGH 9DVFHOODUR :LQJ¿HOG 0DUU If this should happen, the cable companies may lose their status as middlemen between the viewer and the studios. Viewers will watch movies over the Internet accessed through the cable, but the cable companies will have no more of a role than an Internet service provider does today when a computer user visits a news or sports Web site. As a result, cable companies might attempt to partner with the studios to hold their position. In fact, cable companies are already becoming active providers of Internet service and telephone service. The distinction between these three is likely to disappear in the near future. Another outcome might occur if wireless Internet becomes more popular. In this case satellite companies (such as DirectTV) might enter the market as ISPs, and home PPV, pay TV and Internet access might migrate to a new wireless window.
EXTERNAL ENVIRONMENT Generally speaking, piracy of content involves WKH XQDXWKRUL]HG XVH RI DQRWKHU SDUW\¶V SULYDWH property and can take two forms: 1.
2.
Physical piracy of content contained on, for H[DPSOHPXVLFFRPSDFWGLVNV &'V RU¿OP DVDs or videocassettes. Digital piracy of content in cyberspace.
Technological advancement, such as audio tape recording and video tape recording, has threatened the music and movie industries before and will continue to do so in the future. There is no question that in a world with the Internet and digital compression techniques, much more copyrighted content is theoretically susceptible to digital piracy. While piracy of the physical manifestation of content, such as bootleg CDs of live performances and
567
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
FRXQWHUIHLW&'VRIRI¿FLDOO\UHOHDVHGPXVLFKDV been a continuing concern of the music industry and content industries in general for some time, digital piracy is a relatively recent (starting in 1999) and growing concern. A 2006 study commissioned by the movie industry estimated that piracy costs the movie industry $6.1 billion dollars annually (McBride & Fowler, 2006). In general, there are a host of activities that content companies have undertaken to confront the issue of digital piracy, or what they argue to be piracy, that involve two main objectives: 1. 2.
Limiting the unauthorized copying of content. Limiting the Internet redistribution of the unauthorized content.
In particular, the movie industry has made efforts in the legal, cultural, political, and economic environments to try to achieve these objectives. Table 1 provides a list of the key market stakeholders who facilitate digital piracy. Legal actions by the music and movie industries have been directed and continue to be directed at each stakeholder. In addition, technological efforts via digital rights management strategies to enhance copyright protection are being developed. These are directed at the machines and software that are used in the digital piracy process, including computers and computer networks at home, work, and school, DVD players, CD burners, and reproduction software (for example, the now shut-
GRZQ6WXGLRV¶'9';FRS\ FDPFRUGHUVDQG computers in peer-to-peer networks. Efforts to protect movie studios from digital piracy include protections by technology, protections by law, protections through market responses, and protections by culture. One concern is that copyrighted content, such as a movie, will be taken (“ripped”) off a purchased DVD and converted LQWRDGLJLWDO¿OHRQDFRPSXWHUKDUGGULYHDQGRU copied to a recordable DVD with a DVD burner or recorder. In either case, it is possible that the FRQWHQWFRXOGHQGXSE\,QWHUQHW¿OHVKDULQJLQ the hands of an unauthorized user, leading to a violation of the copyright laws, as has happened with music. Movies on DVDs are protected by an encryption code, using a standard called the content scrambling system (CSS). In turn, these encrypted DVDs will only play on devices licensed by Hollywood. Fortunately for the industry, this behavior has generally been accepted as appropriate by society since the advent of DVD distribution in 1994, in contrast to the negative reaction by consumers recently to similar efforts by music companies to start to protect music CDs by encryption. The means to circumvent this code and digitally pirate DVDs exist at this time but generally are regarded by the public as not worth the trouble and aggravation for the money saved. Nevertheless, the movie companies see these efforts at circumvention by FRS\LQJ DQG E\ ¿OH VKDULQJ IDFLOLWDWHG E\ SHHU to-peer networks as having a disturbing parallel WRWKHPXVLFLQGXVWU\¶VH[SHULHQFHDQGWKHUHIRUH
Table 1. Facilitators of digital piracy Key stakeholders
6RIWZDUHFRPSDQLHVWKDWFUHDWHDQGGLVWULEXWHVRIWZDUHWKDWFDQEHXVHGWRH[FKDQJH¿OHVRQSHHUWRSHHU networks on the Internet.
•
,QWHUQHW:HEVLWHVWKDWSURYLGHGLUHFWRULHVRIXQDXWKRUL]HGFRQWHQWDYDLODEOHIRU¿OHVKDULQJ Internet service providers, universities and colleges, and businesses that provide access, especially high-speed access, to the Internet.
6WXGHQWVRQFDPSXVHPSOR\HHVDWZRUNDQGLQGLYLGXDOVDWKRPHZKRWUDGHVKDUHXQDXWKRUL]HG¿OHV
•
Individuals that make an original unauthorized copy of copyrighted content or that distribute pre-released copyrighted content.
568
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
seek additional legal, political, and technological protection for their intellectual property.
Legal Environment Efforts to protect movies from piracy are proceeding along several legal fronts. The biggest self-perceived threat to the industry is the availability of movies for free from use of decentralized peer-to-peer networks. Customers are able to GRZQORDGPRYLH¿OHVIURPPHPEHUVRIWKHSHHUWR peer community. In 1999, Napster introduced the PXVLFZRUOGWRWKHSRZHURI¿OHVKDULQJWKURXJKD version of a centralized peer-to-peer network that utilized central servers to facilitate the trading RI ¿OHV DPRQJ WKH PHPEHUV RI WKH FRPPXQLW\ Under a more decentralized architecture, content is stored on peer machines located on the edges of the Internet that communicate with each other directly without relying upon company-provided, FHQWUDOO\ORFDWHG:HEVHUYHUV6SHFL¿FDOO\SHHU to-peer networks today have evolved to enable users to share content among members directly without the use of a central server directory that was required with Napster. For a description of the differences in the architecture of centralized peer-to-peer systems and decentralized peerto-peer systems, see Alexander (2002). Besides trading copyrighted and public domain content, SHHUWRSHHUWHFKQRORJLHVKROGVLJQL¿FDQWSURPLVH for improving search capabilities on the Internet DQG LQFUHDVLQJ HI¿FLHQF\ IRU GDWD VWRUDJH DQG distribution. Popular decentralized peer-to-peer software ¿UPVDQGWKHLUORFDWLRQVKDYHLQFOXGHGWKHIROORZing: Kazaa Media Desktop, owned by Sharman 1HWZRUNV /LPLWHG LQFRUSRUDWHG LQ WKH 3DFL¿F Island nation of Vanuatu but headquartered in Australia; eDonkey, owned by Metamachine, Inc., based in New York; BearShare, owned by Free Peers, Inc., based in Miami; Blubster, owned by Optisoft, SL, based in Madrid, Spain; Imesh, RZQHGE\L0HVK/WGEDVHGLQ,VUDHO DQG;ROR[ RZQHG E\ ;ROR[ %9 EDVHG LQ WKH 1HWKHUODQGV (Grow, 2002). Additional peer-to-peer networks of note are StreamCast Networks, based in Tennessee that distributes Morpheus, and Grokster,
incorporated in the West Indies but owned by $PHULFDQV6KDUPDQ¶V.D]DD0HGLD'HVNWRSZDV once the largest of these systems with an estimated 2.48 million simultaneous users each day but now eDonkey has reportedly over 2.5 million users daily (Borland, 2004b). Now that the original Napster, as well as former peer-to-peer systems Madster (originally known as Aimster) and Audiogalaxy, all of which are or were American based companies, have been legally vanquished, the motion picture industry and other content providers are attempting legal action to halt the newer decentralized peer-to-peer software companies that, in their view, contribute indirectly to copyright infringement by individuals DQGEHQH¿WIURPLW¿QDQFLDOO\WKURXJKIRUH[DPSOH advertisements included with the software. The SRVW1DSVWHU YHUVLRQV WKDW IDFLOLWDWH ¿OH VKDULQJ HQDEOH D XVHU WR XVH D VHDUFK FRPPDQG WR ¿QG content he desires from the hard drives of other peer users in the community. In fall 2001, the Recording Industry Association of America (RIAA) and the Motion Picture Association of America (MPAA) sued StreamCast and Grokster, alleging indirect copyright infringement, in federal court in Los Angeles. The plaintiffs added Sharman to the suit, which challenged but lost its attempt to protect itself from U.S. law on jurisdictional grounds. A distinguishing aspect of the suit is that none of these companies used FHQWUDOVHUYHUVWRFRQWUROWKHWUDQVIHURI¿OHVDV did Napster. In April 2003, a federal judge ruled that StreamCast and Grokster were not liable for copyright infringement since the software had substantial, legal uses and the software companies were unable to reasonably control the uses to which the software is applied. In August 2004, an DSSHDOVFRXUWXSKHOGWKHIHGHUDOMXGJH¶VGHFLVLRQ In December 2004, the Supreme Court agreed to hear an appeal of this case and a decision favorably to the movie industry was made in summer 2005 (Metro-Goldwyn-Mayer Studios Inc. et al. vs. Grokster, Ltd., et al.). There were interesting legal issues in the Grokster case for the Supreme Court to consider, including jurisdictional issues, since two of the companies are not headquartered in the United
569
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
States; the “potential for substantial noninfringing use” for these peer-to-peer networks, a rule established by the Supreme Court in 1984 in the famous Sony v. Universal City Studios case that made use of videocassette recorders legal since they could be used for a substantial noninfringing use, time-shifting a television program; and the distinction between direct and secondary (indirect) infringement. The last issue was perhaps the most fundamental question in the case for it would determine the liability of companies that produce and distribute software that enables (actively induces?) others to violate copyright protections. Unfortunately, the decision did not clarify the boundaries of the Sony safe harbor but rather imported from patent law the standard of active inducement. A producer of a new technology with both infringing and noninfringing uses who behaves, according to the Supreme Court, with a ³FOHDUH[SUHVVLRQRURWKHUDI¿UPDWLYHVWHSVWDNHQ to foster infringement” is liable for infringement engaged in by direct infringers. This is somewhat at odds with a March 2002 case involving Kazaa in which a Netherlands appeals court concluded (and this decision was upheld by the Dutch Supreme Court) that the liability for copyright infringement lies with the users of the software and not with the software distributor (Grow, 2002). Legal examples and precedents abound illustrating the global nature of digital piracy and the potentially contradictory messages they send to U.S. courts, users, and content companies. For instance, in March 2004 in Canada, a federal judge UXOHG WKDW XSORDGLQJ DQG GRZQORDGLQJ ¿OHV LV legal (Borland, 2004a). On the other hand, digital content companies won an international victory in a trial that commenced in November 2004 that pitted the major record companies in Australia against Sharman Networks alleging indirect copyright infringement. In December 2005, Sharman closed down access to its Web site in Australia (Ferguson, 2005). A second legal front is aimed at individual infringers in an effort to discourage or eliminate the use of these sharing services by individuals. One result of this effort is a legal battle between movie and music companies and Internet service
570
providers (ISPs). Content companies attempted to halt piracy by customers who use an ISP for access to the Internet by utilizing the ISP as an enforcement agent and/or a collection agent for the content provider. The ISPs rebelled against being used for these functions. Under the Digital Millennium Copyright Act of 1998, ISPs are responsible for UHPRYLQJ D VXEVFULEHU¶V DFFHVV WR XQDXWKRUL]HG content utilizing equipment operated by the ISP by taking that content off their network. But peerto-peer networks rely on users at the edge of the network to store the content and not on servers within the network. 7KLVGLIIHUHQFHOHGWRFRQÀLFWEHWZHHQFRQWHQW companies and ISPs. A controversial case involved a request in July 2002 by the RIAA to Verizon for VSHFL¿FVXEVFULEHULQIRUPDWLRQIRUDQLQGLYLGXDO DOOHJHGO\HQJDJHGLQPDVVLYHPXVLF¿OHVKDULQJ XVLQJ .D]DD¶V SHHUWRSHHU VRIWZDUH 9HUL]RQ¶V Internet unit balked at the request, arguing that the legal process that the music companies were LPSRVLQJ ZRXOG YLRODWH DQ ,QWHUQHW VXEVFULEHU¶V privacy and due process rights. The case centered on the legal procedures that a content company must follow when requesting the identity of an ,63¶VFXVWRPHU$86'LVWULFW&RXUWMXGJHUXOHG in favor of the content companies in January 2003, requiring Verizon to turn over the name of the ¿OHVKDUHU,QHIIHFWWKLVGHFLVLRQPDGHLWOHJDOO\ easier, using provisions in the Digital Millennium Copyright Act, for content companies to proceed ZLWKOHJDODFWLRQDJDLQVWLQGLYLGXDO¿OHVKDUHUV However, in December 2003, an appeals court RYHUWXUQHGWKHORZHUFRXUW¶VGHFLVLRQDQGLQ2Ftober 2004 the Supreme Court rejected hearing an appeal of the case. This means that the content companies must follow a more cumbersome and FRVWO\OHJDOSDWKWRJHWDFFHVVWRWKHQDPHVRI¿OH sharers from ISPs, thus slowing down the pace of litigation. Not to be detoured, music companies KDYHVWLOO¿OHGWKRXVDQGVRIODZVXLWVDJDLQVWLQdividuals, a strategy that movie companies have since followed. For example, in November 2004 the 03$$¿OHGDQLQLWLDOEDWFKRIRYHUODZVXLWV against individuals. The problem for the movie industry from this aggressive step is the potentially VLJQL¿FDQWSXEOLFUHODWLRQVEDFNODVKWRVXFKOHJDO
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
HIIRUWV GLUHFWHG DW LQGLYLGXDOV 6XLQJ RQH¶V RZQ customers is a risky endeavor and, perhaps, unfair given the disparity in access to legal resources and the somewhat random nature in selection of the prosecuted infringers. Working in the opposite direction, it is possible that peer-to-peer companies will instead sue content providers and argue that content companies are engaging in anticompetitive behavior by denying them access to copyrighted content concertedly in situations in which the peer-to-peer companies would willingly develop digital stores that compensate copyright holders. Sharman Networks made this counter argument in the Grokster case but its allegations were dismissed on legal standing grounds and thus were not analyzed from an economic perspective. In September 2006, the ¿OHVKDULQJGLVWULEXWRU/LPH:LUHVXHGWKHUHFRUG industry for antitrust violations in an attempt to drive it out of business (McBride, 2006). Even though music and movie distributors relentlessly sue peer-to-peer networks for indirect copyright infringement, new versions of peer-topeer technology continue to evolve. The peer-toSHHUFRPSDQ\%LW7RUUHQWLVDGHFHQWUDOL]HG¿OH VKDULQJ WHFKQRORJ\ WKDW PD\ SURYH GLI¿FXOW WR legally challenge. BitTorrent has decentralized the search functionality of the process of peer-to-peer exchanges and only acts as a delivery protocol. Choi (2006) predicts that as peer-to-peer networks decentralize the control and discovery functions, the effectiveness of pursuing secondary liability is approaching a dead-end.
Cultural Environment Content companies have turned to moral suasion and warnings to discourage what they view as digital piracy. Content providers do not understand why people, and in particular young people, view the unauthorized, anonymous, and widespread VKDULQJ RI GLJLWDO ¿OHV DFURVV WKH ,QWHUQHW DV DQ acceptable form of behavior in society when, in contrast, taking a DVD or CD from a retail store without paying would almost universally be FRQVLGHUHG LOOHJDO DQG LPPRUDO 2I FRXUVH ¿OH sharing does not involve entering a store to take
property that you do not own, but rather involves JLYLQJ D ¿OH ZKLOH UHWDLQLQJ SRVVHVVLRQ RI WKH ¿OH\RXUVHOIWRDQRWKHUSHUVRQWKDW\RXPD\KDYH purchased yourself or downloaded from someone else. It has generally been viewed as acceptable in society to share property that you own with IDPLO\PHPEHUVDQGFORVHIULHQGV6KDULQJD¿OH ZLWKRWKHUVGRHVQRWVHHPOLNHVWHDOLQJWRWKH¿OH sharing community. $SSDUHQWO\WKH,QWHUQHWKDVDOWHUHGWKHGH¿QLtion in practice, although not in law, of what is acceptable behavior. A form of moral outrage was directed toward music companies for bundling inferior songs with desirable songs onto a CD and then selling the package but not the individual VRQJV7KLVEXQGOLQJSUDFWLFHMXVWL¿HGLQWKHH\HV RIPDQ\PXVLFORYHUVXVLQJD¿OHVKDULQJQHWZRUN to get access to only the song(s) desired. However, a similar rationale does not exist for a movie, but PRYLH¿OHVKDULQJLVRQWKHULVHHVSHFLDOO\ZLWK the growth of high-speed Internet connections, so there is something more fundamental at work that has changed cultural norms. A user who downloads DQLOOLFLW¿OHJHWVLWIRUIUHHEXWDWWKHVDPHWLPH participation in a peer-to-peer network generally requires a user to play both a host and client relationship. The former role causes the user to incur costs by the use of bandwidth and computing resources ZKHQKHVKDUHV¿OHV WKXVSDUWLFLSDWLRQLQVXFKD community is not free. Compared to the original, widely used dial-up connections to the Internet, high-speed networks DUHPRUHFRQGXFLYHWR¿OHVKDULQJHVSHFLDOO\ODUJH YLGHR ¿OHV *LYHQ WKLV IDFW FRQWHQW FRPSDQLHV sent letters to university presidents with a plea to VWRSXQDXWKRUL]HG¿OHVKDULQJRQWKHLUKLJKVSHHG university networks by their users, in particular students, of these networks. A similar request has been made to large corporations, addressing HPSOR\HH LOOHJDO ¿OH VKDULQJ RQ FRUSRUDWH QHWworks. Eventually, these warning letters provide the grounds for copyright infringement lawsuits directed toward the universities and corporations themselves if they do not take meaningful steps to VWRSWKH¿OHVKDULQJ 0DWKHZVF 8QLYHUVLWLHV have tried to respond, apparently ineffectively, to the letters and have “closed off the portals used
571
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
E\ ¿OHWUDGLQJ VHUYLFHV LQVWDOOHG VRIWZDUH WR limit how much bandwidth each student can use, DQG GLVFLSOLQHG VWXGHQWV ZKR VKDUH PHGLD ¿OHV But nothing, so far, has proved entirely effective” (Harmon, 2002). Some universities, such as George Washington University and the University of Southern California, pioneered the practice of including a subscription to an online music service as a part of the bundled services (e.g., cable TV) they provide to residential students and by fall 2006 there were more than 90 universities involved in this practice (Ali, 2006; McBride, 2004). Public service announcements, identifying the economic impact of digital piracy on the industry, especially on the lower-wage employees in the movie industry, directed at movie theater audiences is another attempt to change public opinion about digital piracy (Film industry to try, 2002).
Political Environment For Hollywood, a more comfortable and timetested approach to protecting content by reducing piracy is working for change in the political arena. Recent legislative efforts have included limiting the publication and distribution of software code, forcing a more secure design of consumer electronics equipment, potentially disrupting the operation of peer-to-peer networks, increasing the possibility of bringing federal civil cases against individuals for copyright infringement, making it a federal crime to make a video recording in a movie theater, and making it illegal to create technology that has the effect of intentionally inducing individuals to break copyright laws. The 1998 Digital Millennium Copyright Act extended copyright protections into the digital and Internet eras and can be viewed as a major legislative victory for content companies. This legislation attempted to balance the concerns of content companies and users of that content with respect to digital copyright issues. 7KHDFW¶VDQWLFLUFXPYHQWLRQSURYLVLRQPDNHVLWLOlegal under most circumstances to create software that can undo (“crack”) the encryption code that protects copyrighted work. This act has led to controversial legal suits, alleging the illegality of UHYHUVHHQJLQHHULQJDQGWKHLGHQWL¿FDWLRQRIÀDZV
572
in encryption codes, by content companies in order to enforce the new anti-circumvention provision of the law. Their goal is to stop the copying and subsequent distribution of unauthorized DVDs for whatever reason. Since 1998, with the innovation and widespread adoption of peer-to-peer software, content companies have tried unsuccessfully to address with legislation new issues that have arisen. For example, a proposed bill in 2002 was aimed at requiring consumer electronics equipment providers to add anti-copying features to their machines (“digital media devices”) to provide a second line of defense WKH¿UVWOLQHEHLQJWKHHQFU\SWLRQRIWKHFRQWHQW on a DVD) against copyright infringement. This effort has led to a bitter dispute between technology and electronics interests and content companies. It also has led to a continuing countervailing effort by consumer groups to push for legislation to codify consumer personal use rights, such as time-shifting, space-shifting, and back-up copies that, according to groups such as DigitalConsumer. org, this failed legislation threatened. In 2004, in direct response to the trial and appeals courts unsettling Grokster legal decisions, the content companies unsuccessfully backed legislation, called the Inducing Infringement of Copyrights Act, which made it illegal for anyone to intentionally induce individuals to engage in a copyright violation, such as trading unauthorized PRYLH¿OHV7KLVOHJLVODWLRQZDVDLPHGDWKROGLQJ peer-to-peer software companies responsible for XQDXWKRUL]HG¿OHVKDULQJ A bill introduced in 2002 aimed to grant legal protections from criminal and civil liability for copyright holders to disrupt publicly accessible peer-to-peer networks where they suspect unauthorized distribution is taking place. This was another failed attempt to stop or discourage individuals from downloading unauthorized content from the Internet. Individual users of such a QHWZRUN FRXOG ¿QG WKHLU FRPSXWHUV KDFNHG LQWR and they could be stopped from utilizing such a network. This type of legislation illustrates the EDWWOHRIFRQÀLFWLQJULJKWVWKHULJKWWRSULYDF\E\ computer users, and the right to protect their content from infringement by the content companies
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
Table 2. Key dates
•
Mid-2001:6HYHQPDMRUPRYLHFRPSDQLHVDQQRXQFHGSODQVIRUWZRVHSDUDWHRQOLQHMRLQWYHQWXUHV0RYLHÀ\DQG0RYLHV com, to provide video-on-demand movies over an Internet platform.
•
December 2001: The Department of Justice launched a preliminary antitrust investigation into the competitive effects of
•
April 2002: Fox Studios quits its venture with Disney, thus ending Movies.com. For details, see Mathews and Orwell
WKHVHMRLQWYHQWXUHVSDUDOOHOLQJDVLPLODULQYHVWLJDWLRQRIWKHPDMRUPXVLFFRPSDQLHV¶,QWHUQHWMRLQWYHQWXUHV (2002). •
September 2002:,QWHUWDLQHUDQLQGHSHQGHQWRQOLQHGLVWULEXWRURIIHDWXUH¿OPV¿OHGDQDQWLWUXVWVXLWDJDLQVWWKUHHPRYLH
•
November 2002: Movielink commenced operations.
•
April 2003: )R[ 6WXGLRV OLFHQVHG LWV FRQWHQW WR &LQHPD1RZ D ULYDO RQOLQH GLVWULEXWLRQ VRXUFH LQ )R[¶V ¿UVW ,QWHUQHW
FRPSDQLHVDQG0RYLHOLQNIRUPHUO\NQRZQDV0RYLHÀ\
distribution deal. •
July 2003: Walt Disney licensed its content to Movielink.
•
June 2004: The Department of Justice ended its investigation of Movielink by concluding that the joint venture does not have anticompetitive effects or harm consumers.
•
June 2004: Starz Encore, a cable network, and RealNetworks introduced a rival online movie rental service.
•
August 2004:1HWÀL[WKHPDLOUHQWDOPDUNHWOHDGHUDQQRXQFHGWKDWLWZLOOHQWHUWKHRQOLQHPRYLHUHQWDOPDUNHWLQ
•
April 2006: Movielink begins to sell movies online.
•
Summer 2006: Peer-to-peer software providers, Wurld Media and BitTorrent, negotiate deals with various movie studios
•
Summer 2006: Apple Computer, Amazon, and AOL enter the online movie download business.
•
Summer 2006: CinemaNow announces a burn-to-DVD service enabling a purchased online movie to be viewed on a
WRGLVWULEXWHDXWKRUL]HGPRYLHVXVLQJ¿OHVKDULQJWHFKQRORJ\
television set.
that will be surely played out in the future in the legislative arena. It is highly likely that legislation will continue to be introduced at the request of content companies that increases opportunities for prosecution and penalties for the participants in digital piracy. This includes individuals that make original unauthorized copies of movies, companies that produce and distribute peer-to-peer software regardless if it has non-infringing uses, and individuals that engage LQXQDXWKRUL]HG¿OHVKDULQJ2QWKHRWKHUKDQG consumer groups are likely to push for legislation that will make it legal to circumvent a technology that blocks copying if it is done to accomplish a lawful, fair use purpose such as to make a backup copy of a movie for personal use.
Economic Environment Video-on-demand is a distribution innovation in the economic environment that promises to alter the existing set of vertical relationships that exist in the industry. Digital cable and the Internet are emerging distribution vehicles that transmit movLHVGLUHFWO\WRFRQVXPHUVIURPWKH¿OPGLVWULEXWRU bypassing middlemen such as pay television movie channels and video rental stores, in a form that ofIHUVDXQLTXHFRPELQDWLRQRIEHQH¿WV1RPDUJLQDO WUDQVSRUWDWLRQ FRVWV YLHZLQJ RQ WKH FXVWRPHU¶V own schedule, no late charges, and the ability to SDXVHUHZLQGDQGUHSOD\DYLGHRGXULQJD¿[HG SHULRGRIIHUVDVHWRIEHQH¿WVIRUPRYLHYLHZHUV that should cause video-on-demand to have a huge SRWHQWLDOIRUJURZWK7KH¿YHPRYLHFRPSDQLHV that formed the joint venture Movielink that began operations in fall 2002, responded to this techno-
573
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
logical innovation by offering their own online movie service in competition with several small LQGHSHQGHQWRQOLQHGLVWULEXWRUVRIWKHLQGXVWU\¶V movies. Initially, Movielink offered to rent, for a ¿[HGSHULRGGRZQORDGHG¿OPVIURPWKHVWXGLRV¶ OLEUDULHVIRUDUHQWDOFKDUJHSHU¿OPRIEHWZHHQ $1.99 and $4.99. In addition, the joint venture FDQ EH YLHZHG DV WKH LQGXVWU\¶V OHJDO HFRQRPLF response to their biggest threat, the availability of IUHHXQDXWKRUL]HG¿OPVRYHU¿OHVKDULQJSHHUWR peer networks on the Internet. 7DEOHLGHQWL¿HVNH\GDWHVLQWKHHYROXWLRQRI the video-on-demand joint venture and its emerging online competitors. Movielink is the most recent attempt by the movie production companies to integrate forward LQWRWKHGLVWULEXWLRQQHWZRUNIRU¿OPV3DVWDWWHPSWV to alter the vertical relationship by the production companies attempting to integrate forward and to bypass distribution middlemen had not been well received by the antitrust authorities. In the ODWH VLQWKH3DUDPRXQWFDVHWKH¿YHOHDGing movie companies were required to divest over ¿UVWUXQPRYLHWKHDWHUVORFDWHGLQFLWLHVZLWK populations over 100,000, at a time when movie theaters were essentially the only distribution winGRZDYDLODEOH$FFRUGLQJWRWKHFRXUW¶VDQDO\VLV collective horizontal market power in the movie business was facilitated by the vertical ownership of the theaters (Litman, 1998). A structural solution was deemed by the U.S. Supreme Court to be necessary in order to restore competition in the market IRU¿UVWUXQPRYLHV7RGD\EHFDXVHWKHDWHUVDUH only one of several important distribution windows IRU¿OPVWKDWOHVVHQVFRQFHUQRYHUDQ\H[HUFLVHRI market power by the movie production companies from such vertical integration (Movielink) into only one of the several distribution windows. In 1980, the Department of Justice blocked the plans of four leading motion picture distributors to integrate forward into the cable pay television window by creating their own movie-programming network, called Premiere, to compete with HBO. The movie companies felt that this was a legitimate business response for, in their view, HBO was exerting buying market power over them. From an antitrust perspective, two troubling aspects of the
574
Premiere joint venture were the exclusive window WKDW LW ZRXOG UHFHLYH EHIRUH WKH ¿OPV ZRXOG EH available to competitive pay television channels such as HBO and the collective determination by the four companies of the prices of licenses for the VKRZLQJRIWKH¿OPV7KHYHQWXUHFROODSVHGWKH same year, once the government issued a preliminary injunction against Premiere. Not surprisingly, independent online movie distributors are unhappy with the movie comSDQLHV¶HQWUDQFHLQWRWKHLUOLQHRIEXVLQHVV2QH VXFK LQGHSHQGHQW ,QWHUWDLQHU ¿OHG DQ DQWLWUXVW VXLWLQ6HSWHPEHUDJDLQVWWKH¿OPXQLWVRI AOL Time Warner, Sony, and Vivendi Universal, and Movielink, charging that these companies had attempted to monopolize the online video-onGHPDQG PDUNHW 0DWKHZV E 6SHFL¿FDOO\ Intertainer charged, in part, that these companies had reneged on existing licensing agreements and had delayed negotiating new agreements while they formed their own competitive venture. This suit KLQJHVLQSDUWRQWKHDFFHSWHGPDUNHWGH¿QLWLRQIRU the rental service that Movielink provides, for only then can one assess the state of competition. $QDUURZPDUNHWGH¿QLWLRQZRXOGDUJXHIRUDQ Internet video-on-demand market that includes companies such as CinemaNow, Alwaysi, SightSound, RealNetworks, Movielink, and, formerly, Intertainer. These companies rent authorized ¿OPVGLUHFWO\WRFRQVXPHUVRYHUWKH,QWHUQHW,W LV ,QWHUWDLQHU¶V VWUDWHJ\ WR GH¿QH WKH PDUNHW LQ this narrow fashion to increase the market share RI0RYLHOLQN7KLVGH¿QLWLRQKRZHYHULVFOHDUO\ too narrow, for consumers can gain instant access to videos on their own schedule and in their own home through the services of digital cable television providers such as Charter, Comcast, Cox, Cablevision, and Time Warner Cable. Thus, at least two competing delivery platforms can be used to provide video-on-demand and each is a close VXEVWLWXWHIRUWKHRWKHU$¿UVWOHVVRQLQPDUNHW GH¿QLWLRQLVWKDWLWLVJHQHUDOO\DPLVWDNHWRGH¿QH DPDUNHWDURXQGVSHFL¿FGLVWULEXWLRQPHWKRGWHFKnologies. For example, two different distribution PHWKRGVDWHOHSKRQHFRPSDQ\¶VZLUHRUDFDEOH WHOHYLVLRQFRPSDQ\¶VFDEOHFDQSURYLGHDFFHVVWR
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
the Internet. Each competes with the other in the same market for Internet access. ,QWKHPRYLHLQGXVWU\¶VVHTXHQFHRIGLVWULEXWLRQ windows, video-on-demand is currently placed after the renting and sales of movies in retail outlets such as Blockbuster and before pay-per-view and pay television channels such as HBO. One can argue that both of these adjacent windows encompass VXI¿FLHQWO\ VLPLODU FKDUDFWHULVWLFV VR WKDW WKH\ are close substitutes for video-on-demand, further EURDGHQLQJWKHGH¿QLWLRQRIWKHPDUNHW2QHFKDUacteristic of video-on-demand is that it eliminates the trip to the video store, but so do pay-per-view and pay television channels. Services that use the mail to rent DVDs to the home also reduce transportation costs to the customer (Wayner, 2002). Another characteristic of video-on-demand is the ability of the customer to watch a movie on his own schedule. If one purchases or rents a movie, the VDPHÀH[LELOLW\LVDYDLODEOHWRWKHFRQVXPHU2Q the negative side, video-on-demand services have a limited selection compared to what is available in video stores. Fowler (2002), for instance, predicts WKDW WKH VHW RI YLGHRRQGHPDQG ¿OPV DYDLODEOH for viewing will be limited for at least the next ¿YH\HDUV$QRWKHUGUDZEDFNIRUWKH,QWHUQHWGLVtribution method is the time it takes to download D¿OP7KXVWKHWKUHHZLQGRZVWUDQVDFWLRQVDW retail outlets, video-on-demand, and pay-per-view and pay television channels, possess similar characteristics but in varying proportions. This leads to the conclusion that Movielink is a competitor in a market encompassing at least these three distribution alternatives. This means that there exist both intra-window competitors and inter-window competitors for Movielink. Consequently, any concern about monopolization or anticompetitive practices by Movielink is overstated. Moreover, in contrast to the Premiere case, the joint venture does not have an exclusive window for distribution RILWVPHPEHUV¶¿OPVDQGHDFKFRPSDQ\LVVHWWLQJ LQGHSHQGHQWO\LWVRZQSULFHWROLFHQVH¿OPVDQG UHOHDVHGDWHVIRU¿OPV 7KHPRYLHLQGXVWU\¶VJHQHUDOUHVSRQVHWRWKH growth of online distribution is to embrace it only to the extent that they can control the technology. As was detailed in the legal section, companies that
have utilized peer-to-peer online distribution technologies (such as Napster, Aimster, and Grokster) have been successfully sued for facilitating copyright infringement. As Currah (2006) notes, the industry prefers a centralized server-client online architecture in which they can closely control the ÀRZ RI FRQWHQW )XUWKHUPRUH &XUUDK SUHGLFWV that if a true peer-to-peer decentralized distribution system is to be implemented in the industry, ZKLFKKHWKLQNVZLOOEHXOWLPDWHO\EHQH¿FLDOWRWKH copyright owners; it will have to originate with companies on the creative and economic fringes of the movie industry. A trend that could signal an emerging threat to incumbent movie and music distributors is the creation and distribution of video through Web sites such as MySpace, Revver, Grouper, Bolt, and YouTube (recently purchased by Google). These sites initially offered home videos and clips recorded from television and viewed on a personal computer. For some uploaded content, there have EHHQFRS\ULJKWLQIULQJHPHQWQRWL¿FDWLRQVE\FRS\right owners resulting in a take-down by the Web sites of disputed video, following the procedures outlined in the Digital Millennium Copyright Act of 1998. In response to the concern about copyright infringement, YouTube has signed licensing deals with CBS Television, UMG, Warner Music Group, and SonyBMG. Also, an interesting development is that user-generated content appears to capture the interest of Web audiences. It should be noted that these increasing sources of digital content provide the consumer with an increasing variety of choices in selecting content to consume. Two other market developments may also aid WKH HIIRUW WR UHGXFH XQDXWKRUL]HG ¿OH VKDULQJ Some ISPs and universities are limiting the amount of bandwidth that is available to its subscribers/ VWXGHQWV 6LQFH ¿OP ¿OH VKDULQJ SODFHV D ODUJH demand on bandwidth, particularly compared to music, bandwidth supply reductions could reduce WKHQXPEHURIXQDXWKRUL]HG¿OHVWUDGHG$VHFond market response involves private companies WKDW LQWURGXFH IDNH ¿OHV WKDW LV VSRR¿QJ LQWR peer-to-peer networks. This is intended to lead to frustration on the part of users who are searching for a particular song or movie. Such efforts
575
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
are attempts to demonstrate the quality control problems associated with use of these networks. These market responses lead to increased costs to FRQVXPHUVRI¿OHVKDULQJ
THE FUTURE: LESSONS FROM THE MUSIC INDUSTRY The music industry has made strategic adjustments but continues to adjust to the shocks it faced from the combination of improvements in digital technology and the growth of broadband over the Internet (see, for example, Meisel & Sullivan, 2002). The music industry made a mistake initially by concentrating LWVHIIRUWVRQ¿JKWLQJWKHQHZGLJLWDORQOLQHWHFKQROogy and relying on legal and political victories to protect its old business model of relying for revenue extensively on selling physical copies of music. It is still recovering from this focus. Because of the GLI¿FXOW\RIGLJLWDOYLGHRGLVWULEXWLRQFRPSDUHG to digital music distribution, related primarily to WKHVL]HRIWKH¿OHVUHODWLYHWRWKHGRZQORDGVSHHG the movie industry has had more time than did the music industry to prepare for the new online distribution channels and to develop new business responses. But, as broadband connections accelerate within the United States and worldwide, the window of opportunity is closing. In late 2001, the major music companies commenced online provision of music over the Internet through two joint ventures, Pressplay and MusicNet. These online music stores and other recent entrants have over time become more responsive to consumer demand, including features such as a wider selection, the unbundling of albums to allow individual song purchase, and increased portability. Subscription services exist today, as well as opportunities to purchase downloaded music on D RQHWLPH EDVLV $SSOH¶V L7XQHV GLJLWDO PHGLD store is the market leader in an expanding market of competitors providing digital music downloads and streaming services. Online download prices are likely to continue to fall from the $.99 benchmark established by Apple, thus discouraging piracy. Another positive development is emerging cooperative alliances between the music companies and
576
new technology companies such as Snocap and Mashboxx to utilize the Internet as a decentralized distribution platform in authorized ways. Peer-topeer online music stores promise to combine the GLVWULEXWLRQHI¿FLHQFLHVRIWKH,QWHUQHWZLWKOHJDOO\ sanctioned trading activity. 7KHPRYLHFRPSDQLHV¶VWDUWXSRI0RYLHOLQN in late 2002 was a similar initial rental strategy by movie companies to meet the expectations of consumers in the Internet age. Early indications suggest, however, that the motion picture companies KDYHH[SHQGHGPRVWRIWKHLUUHVRXUFHVRQ¿JKWLQJ the effects of the Internet in the political and legal environments instead of competing economically in the market with digital delivery, basically replicating the failed strategy followed by the music companies initially. However, there is preliminary evidence that the movie companies have reached DQLQÀH[LRQSRLQWDQGDUHUHYLVLQJWKHLUVWUDWHJ\ to try to tap the potential of the Internet. Another factor that gives the movie industry more leeway in its response to the threat of the Internet to its existing business models is portability. Part of the attraction of digital music is that it can be easily converted into formats that are portable (CD players and MP3 players). This is important because users want to have their music wherever they are, whether at home, at work, or in their automobiles. Users are less likely to watch a movie at work, in their automobile, or while jogging than they are to listen to music. A related difference involves the environment to experience the content. Consumers are willing to experience music utilizing a computer, while that LVOHVVOLNHO\WUXHIRUDPRYLH:DWFKLQJD¿OPRQ a computer screen or on a hand-held device will never be as attractive as in front of a big-screen television set. However, the wired or wireless conQHFWLRQRIWKHVHGHYLFHVWRKLJKGH¿QLWLRQODUJH screen televisions in the future is on the horizon. The threat of digital piracy will likely shift the focus of content companies to distribution channels where the threat is the smallest and where they have the greatest opportunity to add value for consumers. For music, this would mean more revenue derived from concert performances, the window least susceptible to piracy, and less revenue
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
from CDs, the window most susceptible to digital piracy. For movies, this would mean increasing emphasis on revenue generated in upgraded theDWHUV SRWHQWLDOO\ ,0$; WHFKQRORJ\ TXDOLW\ IRU ¿UVWUXQPRYLHV New online distribution platforms such as YouTube are complicating the adjustment process to a new equilibrium in the industry. Consumers are very excited with the opportunity to create their own content and to upload copyrighted content to Web sites for sharing with others in which they have a sense of community. The content companies need to balance the positive aspect of the promotional value their copyrighted content receives from this publicity with the potential negative effect of a reduction in copyright income. In the digital era, ownership of content will be altered. As digital rights management technology advances, licenses associated with digital content will change with respect to freedoms such as on what devices one can play the content, the number, if any, of copies one can make of the content, and the ability to distribute the content. Licenses will vary from unrestricted licenses, that is, unfettered freedom, to a rental system limiting the right to play the content one time, a much narrower concept of freedom. One can expect a continuing battle between companies that produce the hardware to play the content, companies that produce the software that instructs the hardware how to play the content, and the companies that produce the content. The only thing certain, based on the music LQGXVWU\¶VH[SHULHQFHLVWKDWFRQVXPHUVZLOOEHD minor stakeholder in crafting the decision-making UXOHVEXWWKH\ZLOOH[HUWDPDMRULQÀXHQFHRYHU the outcome.
Angwin, J., & Pasztor, A. (2006, August 5). Weaker reception: Satellite TV growth is losing altitude as cable takes off. The Wall Street Journal, A1. Borland, J. (2004a, March 31). Judge: File sharing legal in Canada. Retrieved July 17, 2007, from http://Cnetnews.com Borland, J. (2004b, October 12). Kazaa loses P2P crown. New York Times Online Edition. Choi, B.H. (2006). The Grokster dead-end. Harvard Journal of Law & Technology, 19(2), 393-411. Currah, A. (2006). Hollywood vs. the Internet: The media and entertainment industries in a digital and networked economy. Journal of Economic Geography, 6, 439-468. Digital Millennium Copyright Act of 1998. (October 28, 1998). Public Law No. 105-304, 112 Stat. 2680. Ferguson, I., (2005, December 6). Sharman cuts off Kazaa downloads in Australia. Retrieved July 17, 2007, from c/net news.com Film industry to try movie theater, video messages to combat piracy. (2002, August 21). The Wall Street Journal Online. Fowler, G.A. (2002, March 5). Video on demand is an idea whose time has come—almost. The Wall Street Journal, R10. Grant, P. (2006, August 17). Cable industry may need to spend heavily on broadband upgrades. The Wall Street Journal, A1. Grow, B. (2002, December 9). Dutch treat. The Wall Street Journal Online Edition.
REFERENCES
Harmon, A. (2002, November 27). Students learnLQJ WR HYDGH PRYHV WR SURWHFW PHGLD ¿OHV New York Times Online Edition.
$OH[DQGHU3- 3HHUWRSHHU¿OHVKDULQJ The case of the music recording industry. Review of Industrial Organization, 20(2), 151-161.
Litman, B.R. (1998). The motion picture megaindustry (pp. 64-69). Boston: Allyn and Bacon.
Ali, S. (2006, April 6). Becoming part of the solution. The Wall Street Journal, B5.
Marr, M. (2006, September 2). In Hollywood, the SLFWXUH EOXUV IRU VWXGLR SUR¿WV The Wall Street Journal, A1.
577
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
Mathews, A.W. (2002a, April 26). The studios strike back. The Wall Street Journal, B1. 0DWKHZV$: E6HSWHPEHU 2QOLQH¿OP outlet intertainer names studios in antitrust suit. The Wall Street Journal Online Edition. Mathews, A.W. (2002c, October 24). Movie, music ¿UPVWRZDUQFRPSDQLHVRQZRUNHUSLUDF\The Wall Street Journal Online Edition. Mathews, A.W., & Orwall, B. (2002, April 29). Fox studios quits venture plan with disney for movies. com. The Wall Street Journal Online Edition. McBride, S. (2004, August 23). Stop the music! The Wall Street Journal Online Edition. 0F%ULGH6 6HSWHPEHU /LPHZLUH¿OHV countersuit against music-industry group. The Wall Street Journal Online Edition. McBride, S., & Fowler, G. (2006, May 3). Studios see big rise in estimates of losses to movie piracy. The Wall Street Journal, B1. McBride, S., Grant, P., & Marr, M. (2006, January 4,). Movies may hit DVD, cable simultaneously. The Wall Street Journal, B1. McBride, S., & Marr, M. (2006, October 9). Target, a big DVD seller, warns studios over download pricing. The Wall Street Journal, A1. Meisel, J.B., & Sullivan, T.S. (2002). The impact of the Internet on the law and economics of the music industry. Info, 4(2), 16-22. Owen, B.M., & Wildman, S.S. (1992). Video Economics. Cambridge, MA: Harvard University Press. Reuters News Service. (2006, July 27). U.S. broadband subscriptions rise. The Wall Street Journal, D4. Vascellaro, J.E. (2006, September 8). Amazon GHEXWV GRZQORDG VHUYLFH IRU ¿OPV VKRZV The Wall Street Journal, B3. Wayner, P. (2002, September 23). The packaging of video on demand. New York Times Online Edition.
578
Werbach, K.D. (2005). The implications of video P2P on network usage. New York: Video Peer to Peer, Columbia Institute for Telecommunication. :LQJ¿HOG1 0DUU0 6HSWHPEHU Apple computer aims to take over your living-room TV. The Wall Street Journal, B1. :LQJ¿HOG10F%ULGH6 9DVFHOODUR-( September 6). Apple, Amazon to unveil movie downloads. The Wall Street Journal, D1.
KEY TERMS Digital Millennium Copyright Act: A 1998 U.S. law that governs the copying of materials, LQFOXGLQJ ERRNV PXVLF DQG YLGHR ,W VSHFL¿HV circumstances under which online providers can seek a safe harbor from copyright infringement claims by copyright holders. Distribution window: The time period during which a movie is made available in particular form (i.e., movie theaters, DVD, PPV). Free broadcast television (Free TV): Television programs that are broadcast over the air and require no subscription fee. Motion Picture Association of America (MPAA): An industry trade group composed of the major motion picture companies operating in the U.S. (so-called “studios”). Pay-per-view (PPV): A system where home viewers pay an extra fee to a cable company to have a movie shown on their television. The times are pre-set by the cable companies (as opposed to video on demand). Peer-to-peer network: A group of users utilizing common software (and perhaps an Internet VLWH WRVKDUH¿OHVRYHUWKH,QWHUQHW Piracy: The unauthorized used of another SDUW\¶VSULYDWHSURSHUW\ Price discrimination:$V\VWHPZKHUHD¿UP charges a higher price to those consumers who most
The Impact of the Internet on the Law and Economics of the United States Motion Picture Industry
desire the good or service. Price discrimination is often accomplished by offering the good or service ¿UVWWRFXVWRPHUVZKRSD\WKHPRVW Recording Industry Association of America (RCIAA): An industry group composed of major recording companies operating in the US (so-called “labels”).
Video-on-demand (VOD): A system where home viewers pay an extra fee to a cable or Internet company to have a movie shown on their television or computer. The viewer selects the time (as opposed to pay-per-view).
579
Division II
Health Industry
581
Chapter XXXVI
Broadband for Health In Developing Countries Aradhana Srivastava Participatory Research in Asia (PRIA), India
ABSTRACT This chapter highlights the major issues in the use of broadband technologies in health care in developing countries. The use of Internet technologies in the health sector has immense potential in developing countries, especially in the context of public health programs. Some of the main uses of information and communication technologies (ICT) in health include remote consultations and diagnosis, information dissemination and networking between health providers, user groups, and forums, Internet-based disease VXUYHLOODQFHDQGLGHQWL¿FDWLRQRIWDUJHWJURXSVIRUKHDOWKLQWHUYHQWLRQVIDFLOLWDWLRQRIKHDOWKUHVHDUFK and support to health care delivery, and administration. The technology has immense potential, but is also constrained by lack of policy direction, problems with access to technology, and lack of suitable infrastructure in developing nations. However, given its crucial role in public health, comprehensive efforts are required from all concerned stakeholders if universal e-health is to become a reality.
INTRODUCTION We live today in the age of information—the age of the global village when distance and physical barriers no longer pose hurdles to communication. The virtual world of the Internet has opened up avenues for unlimited information storage, transfer, and sharing, which is being utilized increasingly by each one of us to meet personal and professional needs. Information technology has also brought revolutionary changes in global trade, commerce, business, and professional service delivery. Increasing technological outreach and facilitation of information exchange has also laid bare before us the glaring disparity that exists in
the world today between nations in terms of economic and social development. Never before was the difference between the “developing” and the “developed” so visible. Grim pictures of poverty, famine, disease, and strife from all over the world reach our desktops in no time—and the truth is for us to see even as it unfolds. Globalization and a world culture seem but utopian ideals when we look at the deeply divided world—whether it is income, health, education, or infrastructure—the difference persists. Information and communication technologies (ICT) ranging from the humble radio to broadband Internet help bring clarity to the picture and are the most effective agents for reaching out and bridging the development divide.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband for Health in Developing Countries
Information technology is in fact now visibly impacting the nature of development programs across the world. One of the most fundamental EHQH¿WVRIWHFKQRORJ\VXFKDVWKHFRPSXWHUVDQG Internet emerged in its ability to facilitate connectivity in remote and disparate regions, often impoverished and resource poor communities, where communication was a major hurdle to development efforts (Chandrasekhar & Ghosh, 2001). The achievements are indeed encouraging and ICT is beginning to emerge as an indispensable arm of global progress. Within this broader perspective of global develRSPHQWDQG,&7¶VSODFHLQLWWKLVFKDSWHUIRFXVHVRQ the health care sector in developing countries, more VSHFL¿FDOO\SXEOLFKHDOWKSURJUDPVDQGGHOLYHU\ of health care services to all. The chapter aims to highlight the various issues regarding the use of Internet, particularly broadband technologies in the public health sector, its potential, constraints, and possible future scenario in developing countries. The chapter is theoretical in nature, drawing upon current academic literature and news reports. Herein lies its limitation—lack of objective factual analysis to support the discussion. The suggested future course of direction in this context would be an objective analysis to validate the discussion presented here, possibly through a data-based intercountry comparison. The chapter has been divided into six subsections including the introduction—the second section outlines the context of health and development in developing countries, the third section discusses the role of broadband in health care, the fourth section lists out the various avenues for the use of Internet in health care in developing countries, WKH¿IWKVHFWLRQLVDGLVFXVVLRQRQWKHFRQVWUDLQWV to large-scale adoption of Internet technology in poorer nations, and the sixth is the concluding section with a discussion on the future scenario of Internet for health care in developing countries.
HEALTH IN DEVELOPMENT Health is universally recognized as a primary social goal and a prerogative of the welfare state.
582
However, health remains a constantly neglected sector in developing nations where scarce resources compete with often more urgent alternate needs such as defense and industrialization. In the era of globalization, however, health is being viewed not as a compulsory social expenditure on part of welfare states but a constructive investment in human capital with implications on the future long-term economic productivity and health of nations (World Bank, 1993). In other words, it is no longer a purely consumption expenditure but a productive investment. Healthier populations are economically more productive, hence it makes economic sense to invest in health. This change in the global outlook on health has resulted in increased emphasis on health planning and expenditure on health programs by governments and aid agencies DOLNH :+2¶V FDOO IRU ³+HDOWK IRU $OO´ :+2 1978) has been the driving slogan behind all global development initiatives in the health sector. Health encompasses not just “absence of illness” but a feeling of wellbeing in the broader sense, which is ZK\:+2GH¿QHVKHDOWKDVDVWDWHRI³FRPSOHWH physical, mental, and social wellbeing.” Health in the developmental context relates to basic human survival and the need for a physically healthy life IRUDOOIUHHIURPGLVHDVHDQGLQ¿UPLW\ The distance that developing nations need to cover in terms of health status can be gauged by VRPHNH\¿JXUHV/LIHH[SHFWDQF\DWELUWKLVRQH of the most comprehensive indicators of health status of populations. According to UNDP (2006), high income nations have an average life expectancy of 79 years while low income nations have an average life expectancy of 59 years, which is a JDSRI\HDUV8QGHU¿YHPRUWDOLW\VWDWLVWLFVLV a good indicator of child health and also points to the poor health status of children in low income FRXQWULHV²WKHXQGHU¿YHPRUWDOLW\UDWHIRUKLJK income nations being just 7 per thousand while for low income countries it is as high as 120 per thousand. HIV prevalence is 1.8% in low income countries as compared to 0.4% in high income countries (UNDP, 2006). On the institutional side, public expenditures on health as a proportion of total GNP are much higher in countries with high human development, ranging between 6-9%, while
Broadband for Health in Developing Countries
in countries with low human development they normally range from 1-3%. Even within the developing world, disparities in health status are glaring. Sub-Saharan Africa, for example, is perhaps the ZRUOG¶VPRVWXQGHUGHYHORSHGUHJLRQLQWHUPVRI most health indicators (UNDP, 2006). Adoption of a time-bound approach to addressing the major global health concerns in the form RIWKH8QLWHG1DWLRQ¶V0LOOHQQLXP'HYHORSPHQW Goals has resulted in enhanced efforts on the parts of governments and global development agencies ZRUNLQJLQWKH¿HOGRISXEOLFKHDOWK,QWKLVVFHQDULR of earnestness towards reaching the health-related goals and being able to demonstrate the gains made through such efforts, Internet technology has emerged as a key instrument aiding public health programs in a number of ways. In the next sections, the focus is on how Internet technology has emerged as a powerful tool in public health programs, the enormous possibilities it offers, and the hurdles in its path towards universal acceptance and adoption as a key component of all development programs.
BROADBAND APPLICATIONS IN HEALTH There are a number of ways in which information WHFKQRORJ\VSHFL¿FDOO\WKH,QWHUQHWUHQGHUVLWVHOI invaluable to the health sector. Numerous popular Web sites on health have ensured that information on all aspects of health, such as diseases and risk factors, treatments, health care products, diets, and lifestyles, is just a click away for the lay person. Internet has increasingly become more and more specialized in information collection, storage, management, and dissemination, and it is this which lends it indispensable to all public health efforts (Srivastava, 2005). As the conduit IRULQIRUPDWLRQÀRZLWHQDEOHVWKHSROLF\PDNHUV and planners to make informed decisions on health policy and infrastructure in order to maximize equity in public health. It helps providers in orienting service provision towards the needs and demands of the population being served. It also enables remote consultation and diagnosis, which is very
useful in regions lacking specialized health care personnel. The networking function of Internet helps health care providers, health workers, scientists, students, patients, and high-risk groups to interact and exchange relevant information, and for knowledge-sharing on current research and clinical breakthroughs. It is also being used widely in targeting, disease surveillance, and monitoring of public health programs. Perhaps one of the most relevant areas in which information technology is extremely useful is generation of awareness—whether it is through simple radio and television programs, or specialized online information, education and communication (IEC) materials and courses, information technology enables people to improve their quality of life, minimize health risks through adequate preventive measures, and to reach the right source for help when in need (Accenture, Markle Foundation, and UNDP, 2001). ICT can also be used as an instrument to ensure transparency and accountability in delivery of public health VHUYLFHVWKURXJKHIIHFWLYHSHRSOH¶VSDUWLFLSDWLRQLQ governance (Chandrasekhar & Ghosh, 2001). The use of computers and narrowband Internet in health services is well established. However, VRPH VSHFL¿F UHTXLUHPHQWV LQ WKH KHDOWK VHFWRU make broadband or high-speed Internet data technologies the most appropriate for the purpose. This is especially appropriate for clinical applications. These include the ability to transmit large amount of data (such as video images of symptoms, data, or surgery) at high speeds with no corruption or data loss during transmission. Continuous network availability with no transmission breaks LVDOVRLPSHUDWLYHWRWKHVPRRWKDQGWLPHO\ÀRZ of information. Networks with high bandwidth ensure successful data transfer and low response time which is so crucial when the information WUDQVPLWWHGFRXOGPHDQVDYLQJVRPHERG\¶VOLIH Thus, the advantages of broadband in clinical applications include more timely health care, greater access to specialist opinion, and greatly reduced travel time and costs (NOIE, 2002). There are several studies available now on e-health applications in the developed countries, including broadband technologies. Telemedicine is one of the most widely discussed and debated
583
Broadband for Health in Developing Countries
among the many applications of broadband in health (for example, see Ferrante, 2005; Hisle, 2000; Wootton, 1996, 2001). There are studies RQH[LVWLQJSURJUDPVDVZHOODVSRWHQWLDOEHQH¿WV from proposed ones (such as Jennett et al., 2005). One of the main areas for evaluation and impact analysis has been home-based telemedicine and telecare (for example, Guillen, 2002). Research has also focused on evaluations of rural health care delivery using Internet protocol networks (Janes et al., 2005). $XVWUDOLD¶V1DWLRQDO2I¿FHIRUWKH,QIRUPDWLRQ (FRQRP\ 12,( LGHQWL¿HGWKHIROORZLQJ key clinical applications of broadband: a.
b.
c.
d.
584
Remote consultation or telemedicine: This is a broad term referring to remote consultations directly between physicians and patients or referral of cases for opinion between physicians and specialists. A form of remote consultation is telepsychiatry, which facilitates psychiatric consultations of patients with remote specialists using videoconferencing. Remote surgery: This refers to surgery performed using robotics commanded by physicians in remote locations. Though initiDWHGLQLWLVVWLOO\HWWR¿QGZLGHVSUHDG application. Teleradiology and telepathology: TeleUDGLRORJ\ LV WKH WUDQVPLVVLRQ RI ;UD\ LPages across broadband using shared image software, which enables two physicians to view the image almost instantaneously and therefore greatly facilitates second or specialist opinion for remotely located practitioners. Similarly, telepathology enables transmission of microscope images from miles away. Remote monitoring: This is a breakthrough in home/hospice care of ageing persons. It enables chronically ill elderly persons to be monitored from their homes using small wireless devices which constantly transmit data on vital signs and symptoms. This has the multiple advantages of reducing costs of
e.
f.
hospital care, enabling timely response to change in symptoms, and improving quality of life of the older persons. Teaching, research, and training: Besides the use of video conferencing and transmission of surgical procedures online, other breakthroughs in the use of broadband technologies for training include simulated surgery environment for realistic surgery training (providing students with “virtual patients” for practice) using very high end broadband network. Online databases and patient records: Faster access to online databases has speeded up diagnosis and enabled effective treatment. Maintenance of patient records online has also been greatly speeded up and expanded using broadband technology.
The main hurdle in the spread of broadband data technology is its high infrastructure requirement which leads to high capital costs. However, this may not be the case in developed countries. Developed countries spend a large proportion of WKHLU *'3 RQ KHDOWK DQG IRU WKHP WKH EHQH¿WV RIUHGXFHGDFFHVVFRVWVFRXOGOHDGWRVLJQL¿FDQW HFRQRPLFEHQH¿WVDQGRXWZHLJKWKHFRVWVLQYROYHG in setting up broadband based health care infraVWUXFWXUH 12,( %HVLGHVPRQHWDU\EHQH¿WV WKHUHDUHDOVRWKHVRFLDOEHQH¿WVRIJUHDWHUDFFHVV to health for all and improved quality of life. Yet, governments have been slow in providing for broadband technologies on a large scale. Spending RQ³QHZ´WHFKQRORJ\LVGLI¿FXOWWRMXVWLI\ZKHQ there are pending “core” commitments, such as expenditure on more hospitals and beds. Therefore such initiatives have often invited criticism about expenditure priorities (NOIE, 2002). Some impediments to popular acceptance of broadband have included reservations about quality and reliability of broadband services, lack of standardization leading to incompatibility, and lack of interoperability between broadband networks (Committee on Enhancing the Internet, 2000).
Broadband for Health in Developing Countries
AVENUES FOR THE USE OF INTERNET-BASED TECHNOLOGIES IN HEALTH CARE IN DEVELOPING COUNTRIES In developing nations, broadband technologies are relatively nascent and less extensive as compared to developed countries. India is a typical example of a developing country just opening its doors to broadband technology. The scenario of broadband technology in India is only beginning to take shape. Since its launch in 2005 using ADSL technology, broadband covers about 300 cities with 1.5 million connections (Basu, 2006). India is one of the largest emerging global markets of broadband technology, but with a 700 million strong rural population base which the global players in Internet technology cannot ignore. Intensive research and development is going into developing technologies to suit Indian conditions at “Indian prices” with major global IT players like Intel, Nokia, and Motorola entering the fray. Wireless Internet technologies and WiMax are also being introduced into the Indian market (DQC News Bureau, 2005; Parthasarathy, 2006). Extending connectivity into rural areas is also a priority and companies like Intel and Motorola are working with the government for extending Internet connectivity to all rural areas of the country. At present, broadband penetration in India is quite low and the government is treading cautiously on higher spectrum allocations to major service providers; thus, it will be some time before broadband technologies establish themselves on the Indian scene ZHOOHQRXJKWRLQÀXHQFHGHYHORSPHQWVWUDWHJLHV (Basu, 2006). Research has highlighted several major initiatives through which broadband and other Internet-based applications are beginning to make an impact on health care delivery in developing FRXQWULHV6RPHRIWKHPDMRU¿HOGVLQZKLFK,&7 KDVLQÀXHQFHGSXEOLFKHDOWKDUHDVIROORZV
Information Dissemination and Networking In health, information often means life saved. Hence the importance of generating awareness on basic
preventive health care is obvious in regions with poor, illiterate, and largely ignorant populations, who suffer a high disease burden simply for want of basic health information. Information dissemination therefore features high on the agenda of public health programs in developing countries, whether run by governments or agencies like WHO and UNICEF. Hoardings, radio programs, television spots, and now the Internet—every means is being utilized to the maximum possible extent to take public health messages to the people. The ICT for Health Digital Opportunity Task (DOT) Force team study by UNDP in collaboration with Accenture and the Markle Foundation reports “communitylevel application of ICT to support informational initiatives” to be the most effective approach in educating the public on preventive health as well as recognition and treatment of symptoms of various commonly occurring ailments (Accenture, Markle Foundation, and UNDP, 2001). Networking through the Internet also enables people to enter into virtual communities where they can form user groups, support groups, or discussion forums. Such groups provide platforms for sharing experiences and local knowledge on issues of concern. A major advantage with using virtual portals for communication is that it can ensure anonymity, and thus help interaction and knowledge sharing even on sensitive subjects which may carry social stigma and which people may otherwise be hesitant to discuss—for example, those relating to sexually transmitted diseases including HIV/AIDS. This is a very important point with respect to health (Srivastava & Noznesky, 2005). 1HWZRUNLQJ EHQH¿WV QRW RQO\ HQGXVHUV EXW also health workers and activists, especially in developing countries, since most of them function in remote environments. They are able to keep tab of the current developments in health research and treatment breakthroughs which would be useful in their work. Interaction and sharing of experiences FDQKHOSWKHPOHDUQIURPHDFKRWKHU¶VH[SHULHQFHV and thus add to their knowledge base and also seek expert advice in emergencies or in special cases. For an example of a case study, see Wootton et al. (2004). Networks also help create pressure groups when it gives a common platform to health activists, 585
Broadband for Health in Developing Countries
patient groups, service providers, and concerned people who can then forge alliances and together bargain for a better deal in health policy and planning. Thus, in this respect, ICT emerges as an important tool for public policy advocacy and effective governance as well (Chandrasekhar & Ghosh, 2001).
Remote Consultation, Diagnosis, and Treatment Advances in telemedicine offer promising potential IRUVLJQL¿FDQWLPSURYHPHQWLQKHDOWKFDUHGHOLYery. In fact the technology has been no less than a boon for many poor families living in remote impoverished regions who cannot afford even a simple dispensary visit. Telemedicine also enables sharing knowledge among physicians themselves regarding new methods of treatments, and caseVSHFL¿FFRQVXOWDWLRQIRUH[SHUWRSLQLRQWKHUHE\ ensuring more effective treatment. Connectivity via the Internet and satellites has facilitated remote consultations and is increasingly being used in developing countries that may lack necessary medical facilities to transmit patient information to medical practitioners and specialists in other countries who then advise on treatments to be given. For example, nurses in Gambia are sending digital camera images of symptoms through the Internet to doctors in nearby towns for diagnosis and treatment (Zanker, 2001). Telemedicine has also been used effectively for post-tsunami trauma treatment and rehabilitation projects for affected coastal communities by leading medical institutes in India (Gogia & Surwade, 2005). The technology, however, remains underutilized in developing countries due to the many constraints in terms of access to high bandwidth for transmission of complex data and images and the high cost of setting up the necessary infrastructure for the same.
Support to Health Research Networking of libraries and other information agencies increases access to the latest information by students and researchers. This includes current publications in books and journals as well
586
DV RQOLQH GDWDEDVHV 'DWD JHQHUDWHG LQ WKH ¿HOG by health workers through disease surveillance and treatment activities is being transmitted for further analysis to remote research institutes. Health research is also supported through online collaborations in research between researchers in different institutes. Several voluntary organizations and external aid agencies like the WHO, World Bank, and UNDP have set up discussion forums for greater exchange of information and experience in the development sector. The WHO supported Health Internetwork (HIN) program, while aiming at a broader level at strengthening policy making, also piloted networking of libraries and medical colleges in the Indian states of Orissa and Karnataka to facilitate exchange of health information for research purposes (Srivastava, 2005). ICT-enabled training mechanisms are being used increasingly to deliver training to health workers. This is another ¿HOG LQ ZKLFK ,QWHUQHW WHFKQRORJLHV DUH JDLQLQJ popularity rapidly.
Disease Surveillance and Targeting of Health Interventions Disease surveillance is vital to better prediction, preparedness, and management of all public health programs, especially epidemics with a geographical pattern surpassing regional boundaries. Such diseases often require regional and subregional cooperation in monitoring and surveillance. Research institutes and agencies across the world are coming together through technical collaborations to set up early warning systems for predicting the expansion of epidemics of a global severity, such as malaria and tuberculosis. Such a system is technology intensive and demands a high cost burden, which necessitates collaborative measures. 'DWD JHQHUDWHG IURP ¿HOGOHYHO KHDOWK LQIRUPDtion systems regarding transportation processes, pathogens, vectors, hosts and physical, and socioecological environment is integrated through bio-mathematical modeling to predict epidemic movements and likely periods of expansion. This system can also integrate remote sensing data on environmental and meteorological factors into the
Broadband for Health in Developing Countries
models. The Integrated Disease Surveillance and Response (IDSR) project in Africa and Eastern Mediterranean countries is one such collaborative disease surveillance structure involving several international agencies (WHO, 2003). Data from disease surveillance can be integrated with other relevant information using GIS tools to help identify high-risk groups and their geographical location in order to improve targeting of public health programs. Targeting is particularly useful in resource-crunch situations when the scale of the problem is too large and the resources available ZRHIXOO\LQVXI¿FLHQW7DUJHWLQJWKHQKHOSVIRFXV the resources where they are most in need. For example, an ADB-assisted project on prevention and education on HIV/AIDS in the Greater Mekong Subregion encompassing the borders of China, Vietnam, Laos, Cambodia, and Thailand has as one of its components a GIS-based network for monitoring interventions for vulnerable populations and direct targeted interventions for high risk groups and interstitial populations (Srivastava & Noznesky, 2005).
Health Care Delivery and Administration Internet tools have proven to be a boon to health workers operating in resource-poor settings with lack of transport and communication infrastructure as well as lack of trained medical personnel. Setting up centralized systems of data collection and management by medical institutes and agencies working in public health extension in remote resource-poor areas helps in logistics support for distribution of drugs, medical and drug supply chain management, and greater responsiveness to medical emergencies. Such systems are generally based on local area networking with a central server connecting computers in remote locations and a centralized database management system. The Electronic Medical Record (EMR) is a well documented example of such an application being used successfully in regions with poor connectivity, infrastructure, and very few doctors, where expansion of HIV treatment services, as well as follow-up of treatments and coordination of drug
supplies, is a major challenge. The EMRs help in logistics management, disease monitoring, and surveillance as well as remote consultation on case histories. Some successful EMR applications have been well documented, such as in Haiti and Kenya (Fraser et al., 2004; Siika et al., 2005). In India the use of hand-held computing devices called personal digital assistants (PDAs) to transmit data directly from source to higher levels for more prompt action and better follow-up of cases was experimented with in a government collaborative project in Rajasthan and later in a World Bank supported project in Nalgonda district in the state of Andhra Pradesh. PDAs were provided to auxiliary nurse midwives (ANMs) who visited each house in their jurisdiction monthly to track demographic data and check on maternal and child health through immunization and counseling services (Roy Sastry, 2003). Internet-based networks in remote rural areas can in fact greatly facilitate the delivery of centralized health services, especially during postemergency relief operations such as epidemic control in WKHDIWHUPDWKRIF\FORQHVÀRRGVDQGHDUWKTXDNHV Such systems are being set up in some states in India, such as Andhra Pradesh and Maharashtra. In a country where 70% of the people reside in villages, it is important to envision Internet as a tool for knowledge sharing among rural communities. There are several initiatives in India to provide village level Internet connectivity to ensure greater DFFHVVWRWKHEHQH¿WVRI,QWHUQHWE\WKHUXUDOIRON (Chandrasekhar & Ghosh, 2001). Internet technologies are thus opening up many frontiers and blazing new trends in health care delivery in developing countries. However, there are also some constraints which pose challenges to widespread adoption of broadband and Internet technologies in the health development sector.
CONSTRAINTS TO INTERNET TECHNOLOGIES IN THE HEALTH SECTOR IN DEVELOPING COUNTRIES At present the biggest constraint to global penetration of Internet technology is the digital divide in 587
Broadband for Health in Developing Countries
terms of technical capability for as well as access to Internet technologies (World Bank, 2006). This divide exists not only between the developed and developing countries but within developing countries as well. Access constraints within developing countries typically include unequal physical access to computers, Internet, and related facilities, poor literacy and inadequate technical skills to access WKHLQIRUPDWLRQDQGPDNHIXOOXVHRILWDQG¿QDOO\ lack of economic capability to keep up with the innovations in technology leading to the failure to PDNHIXOOXVHRIWKHEHQH¿WVRIQHZHUDQGPRUH HI¿FLHQWWHFKQRORJLHV,QFRXQWULHVOLNH,QGLDZKHUH DERXWKDOIWKHFRXQWU\¶VSRSXODWLRQGRHVQRWSRVsess even basic literacy, universal digital literacy seems but a dream, and ranks obviously low on the development priorities of the government. Language constraint is a problem affecting utilization of broadband technologies in largely non-English speaking populations (Chandrasekhar & Ghosh, 2001). Development of software in the vernaculars is progressing, albeit slowly with limited reach. A large part of this investment must be undertaken by the already overburdened state, since the market is unlikely to service these needs. This itself is a severe constraint on its development. Development of Internet-enabled health services is thus likely to impose a high cost burden on developing economies, the bulk of which is initially more likely to fall on governments, since the market may not be willing to support initial costs with longer gestation periods for recovery. Health remains an essentially social good and hence basically a state responsibility. While the potential EHQH¿WVRI,&7LQKHDOWKVHFWRUDUHWRRODUJHWREH completely ignored, the resource constraints are also very real and intimidating. Some possible options which can be suggested for adoption by governments include cost recovery through taxation of information technology-based commercial services and selective investment in the health sector in developing countries on connectivity to KHDOWKZRUNHUVZKRFDQWKHQFRQYH\WKHEHQH¿WVWR DODUJHQXPEHURI¿QDOEHQH¿FLDULHV,QWKLVZD\WKH outreach can be maximized with minimum input cost (Chandrasekhar & Ghosh, 2001).
588
Some of the basic challenges before full-scale adoption and establishment of Internet technologies in the public health sector are discussed in detail below, with special reference to the Indian scenario: 1.
2.
Creating a favorable policy environment: (KHDOWKLVDQHPHUJLQJ¿HOGLQGHYHORSLQJ countries and there are many pressing policy matters and politico-legal issues remaining to EHGH¿QHG*UHDWHUFODULW\DQGSROLF\GLUHFtion is required on a number of aspects, especially with regard to the use of Internet-based health services. Public policy issues include setting up guidelines to ensure protection of patient identity in medical records, patient privacy, and anonymity in online discussion groups, and security of patient data being stored and transmitted online. Regulating access by different types of users to different resources on the network is also important, as the information may in many cases be private and sensitive in nature. Governments also need to set up requisite standards and guidelines to ensure authenticity of medical information provided on Web sites and circulated in online discussion forums. Greater clarity in the use of e-health technologies for clinical purposes requires determination of legal status of e-consultations in case of any eventuality leading to litigation, acceptance of e-consultations and e-payments as valid modes of health care delivery, and sorting out intellectual property issues in information being shared on the Internet for educational or research purposes. Access to technology: The major problem in universalization of Internet use in India, as in other developing countries, is lack of access to Internet itself. This means the use is restricted to the literate, largely urban and economically better off classes, who are not necessarily the population most in-need of Internet-based health services. While more than 700 million people live in the villages, the majority of them cannot afford computers and many villages simply do not have the
Broadband for Health in Developing Countries
3.
4.
infrastructure to put computers or Internet in place. Within a region different social groups may have differential access to Internet technologies due to gender bias and social exclusion factors as well. Hence e-health has a limited reach where it matters the most. Infrastructure support: Internet technology is heavily dependent on the availability of certain supporting infrastructure, such as regular power supply and communication lines. Broadband, especially for health, requires high-speed high capacity networks in order to support applications such as telemedicine and telepsychiatry. But what happens in regions with little or no power supply, lack of all-weather approach roads and poor telephone connectivity? Poor infrastructural support or the complete absence of it is a common problem in remote rural areas in developing countries, which also happen to be the regions with poorest health status, and hence most in need of enhanced health care interventions. This is one of the reasons why WKHEHQH¿WVRI,QWHUQHWKDYHQRWEHHQDEOHWR SHQHWUDWHVRPHRIWKHZRUOG¶VQHHGLHVWDUHDV in sub-Saharan Africa and South Asia. Skilled Personnel: Skilled personnel are required to set up Internet-based support V\VWHPVLQSURJUDPVDQGDOVRWRWUDLQ¿HOG level health functionaries to use their computer-based tools which feed information into the system. The sector, therefore, cannot do without a minimum level of literacy and skilled manpower. Though India is fortunate in the sense that technically trained personnel are available to administer information technology-enabled services, this is not the scene in many other developing countries. In many African nations, for example, there are very few technically skilled persons in WKH¿UVWSODFH7KRVHZKRGRH[LVWDOVRSUHIHU moving out for better opportunities abroad; hence, these nations suffer a high attrition rate RIVXFKSHUVRQVOHDYLQJDGH¿FLWRIWHFKQLFDO manpower at home.
CONCLUSION: THE WAY AHEAD Broadband data technologies are beginning to PDNHVLJQL¿FDQWIRUD\VLQWKHKHDOWKVHFWRULQGHveloping countries. But their advent is recent and they are yet to realize their full potential. In fact, JLYHQWKH:RUOG%DQN¶VODWHVW¿JXUHVRIEURDGEDQG penetration at 0.6 connections per thousand persons (World Bank, 2006), the path is long indeed. We have seen the many ways in which they can facilitate health care delivery right from program planning to targeting, implementation, and documentation of implementation outcomes. In spite of the encouraging results, IT-based initiatives remain sporadic and at the experimental stage for DQXPEHURIUHDVRQV/DFNRIWHFKQLFDOO\TXDOL¿HG personnel, poor supporting infrastructure, and high per unit costs are some of the most pertinent questions in this regard. Cost-effectiveness is perhaps the most imporWDQWIDFWRULQÀXHQFLQJZLGHUDGRSWLRQRIEURDGEDQG technologies in health care in low income countries. :LUHGEURDGEDQGQHWZRUNVDUHGHSHQGHQWRQ¿EHU optic cable networks and hence costly investments. Setting up a technology and infrastructure-intensive Internet network indeed can be a considerable H[SHQGLWXUHIRUVRPHRIWKHZRUOG¶VSRRUHFRQRPLHV ,QVXFKDVFHQDULRWKHEHQH¿WVRIWKHXVHRI,QWHUQHW technologies need to be demonstrably greater than the input cost, only then can they be sustainable in the long run. Expenditure on information technology in developing countries is growing rapidly and forms a high proportion of GDP, but, in absolute terms, the quantum of investment is very low (LBJ School of Public Affairs, 2006). In order to ¿UPO\HVWDEOLVKORQJWHUPFRVWJDLQVLQWKHXVHRI information technology, substantial research into WKHFRVWHIIHFWLYHQHVVRI,QWHUQHWDQGVSHFL¿FDOO\ broadband based interventions needs to be carried out in developing countries. The huge amount of initial investment required for setting up the basic infrastructure necessary for Internet-based services implies substantial public spending, especially in the initial stages. This calls for advocacy and opinion-building at the global level to sensitize nations towards the considerable gains in terms of health outcomes per unit cost of
589
Broadband for Health in Developing Countries
investment in information technology infrastructure. Of course, Internet technology needs to be part of the larger picture of development, especially in nations with a poor record of public spending on KHDOWK7KHEHQH¿WVKDYHEHHQGHPRQVWUDWHGDQG successful models of sustainable use of ICT in public health in developing nations are there for all to see. Yet without the initiative through a centralized SROLF\IUDPHZRUNDQGVXVWDLQHG¿QDQFLDOVXSSRUW from national governments, a universal e-enabled health sector is far from being a reality. Around the world, developing countries are committed to achieving the MDGs in time. For the purpose enhanced budgetary commitments towards public health have been pledged by them; thus the environment is favorable for the emergence of Internet and broadband technology as fully integrated component of public health programs. Last but not least, local communities need to be VHQVLWL]HGWRZDUGVWKHEHQH¿WVRI,QWHUQHWWHFKQRORgies so that they are in a position to demand and utilize such facilities to the maximum extent. This is extremely important for new technologies to be viable in the long run. New technologies always take some time to establish themselves, but nevertheless a committed leadership and community support ensures the sustainability of such efforts, which then in due course deliver the desired results. It is DSHRSOH¶VWHFKQRORJ\WDNLQJLQIRUPDWLRQOLWHUDOO\ WRDSHUVRQ¶VGHVNWRSFRQQHFWLQJSHRSOHDQGFRPmunities as never before, and opening an enormous world of possibilities for information generation and utilization. Avenues have been opened and in time broadband technologies are expected to HVWDEOLVKWKHPVHOYHV¿UPO\LQWKHKHDOWKVHFWRULQ developing countries.
REFERENCES Accenture, Markle Foundation, and UNDP. (2001). &UHDWLQJDGHYHORSPHQWG\QDPLF²¿QDOUHSRUWRI the digital opportunity initiative (section on ICT for health). Retrieved July 19, 2007, from http://www. opt-init.org/framework/pages/2.2.1.html
Basu, I. (2006). Indian broadband market could explode. Spacemart. Retrieved July 19, 2007, from http://www.spacemart.com/reports/Indian_Broadband_Market_Could_Explode_999.html Chandrasekhar, C.P., & Ghosh, J. (2001). Information and communication technologies and health in low income countries: The potential and the constraints. Bulletin of World Health Organization, 79(9), 850-855. Committee on Enhancing the Internet for Health Applications: Technical Requirements and Implementation Strategies, Computer Science, and Telecommunications Board, National Research Council. (2000). Networking health: Prescriptions for the Internet (Executive Summary). Retrieved July 19, 2007, from http://newton.nap.edu/execsumm_pdf/9750.pdf '4&1HZV%XUHDX :,0$; %URDGEDQG Internet for masses. Retrieved July 19, 2007, from http://www.dqchannels.com/content/mirror/105041901.asp Ferrante, F.E. (2005). Evolving telemedicine/ehealth technology. Telemedicine Journal and E-Health, 11(3), 370-83 Fraser, Hamish S.F. et al. (2004). An information system and medical record to support HIV treatment in rural Haiti. BMJ, 329, 1142-1146. Gogia, S.B., & Surwade, M.R. (2005). Health care technology aspects of disaster planning—based on the post-tsunami experiences. Retrieved July 19, 2007, from http://openmed.nic.in/1774/01/telmedtsunami.pdf Guillen, S. et al. (2002). User satisfaction with home telecare based on broadband communication. Journal of Telemedicine and Telecare, 8(2), 81-90. Hisle, M. (2000, Summer). Universal broadband communications creates new options for health care networks. Journal of Healthcare Information and Management, 14(2), 71-84. Janes, R. (2005). Few rural general practitioners use the Internet frequently in regard to patient
590
Broadband for Health in Developing Countries
care. New Zealand Medical Journal, 118(1212), U1380. Jennett, P. et al. (2005). Delivery of rural and remote health care via a broadband Internet protocol network—views of potential users. Journal of Telemedicine and Telecare, 11(8), 419-24. LBJ School of Public Affairs. (2006). Understanding the contexts for health information technology in developing countries (White Paper). Retrieved July 19, 2007, from http://www.wcit2006.org/ Healthcare/media/whitepaper/hit.pdf
WHO. (1978). Declaration of Alma Ata. Retrieved July 19, 2007, from www.who.int/hpr/NPH/docs/ declaration_almaata.pdf WHO, CDC, USAID, SARA, UNFIP. (2003). The implementation of integrated disease surveillance and response in the African and Eastern Mediterranean regions (a synthesis report). Retrieved July 19, 2007, from http://www.cdc.gov/idsr/focus/surv_ sys_strengthening/doc_idsr_implement.pdf World Bank. (1993). World development report. New York: Oxford University Press.
NOIE (National Office for the Information Economy). (2002). Broadband in health: Drivers, LPSHGLPHQWV DQG EHQH¿WV 5HSRUW SUHSDUHG IRU Broadband Advisory Group, Govt. of Australia). Retrieved July 19, 2007, from http://www.dcita. gov.au/communications_and_technology/publications_and_reports/2002/august/broadband_in_ KHDOWKBGULYHUVBLPSHGLPHQWVBDQGBEHQH¿WV
Wootton, R. (2001). Recent advances: Telemedicine. BMJ, 323, 557-560.
Parthasarathy, A. (2006, October 12). Twin tracks to broadband business. The Hindu. Retrieved July 19, 2007, from www.hindu.com/seta/2006/10/12/ stories/2006101200621500.htm
Wootton, R. et al. (2004). Prospective case review of a global e-health system for doctors in developing countries. Journal of Telemedicine and Telecare, 10(S1), 94-96.
Roy Sastry, C.L. (2003). India health care project: An application of IT in rural health care at grass root level. Information Technology in Developing Countries, 13(1). Retrieved July 19, 2007, from KWWS ZZZLLPDKGHUQHWLQHJRYL¿SMXQDUticle4.htm
Zanker, C. (2001). The global digital divide problems and solutions. Paper presented at UNI Seminar for Trade Union Representatives from Africa, Asia and Latin America. Retrieved July 19, 2007, from http://www.input-consulting.com/ download/berlin-dd-eng.pdf
Siika, A.M. et al. (2005). An electronic medical record system for ambulatory care of HIV-infected patients in Kenya. International Journal of Medical Informatics, 74(5), 345-355.
KEY TERMS
Srivastava, A. (2005, May). ICT in Indian health scenario: Accelerating delivery of health services, i4d, 12-15. Srivastava, A., & Noznesky, E. (2005, December). Better intervening HIV/AIDS with ICT. i4d, 1921. UNDP. (2006). Human development report 2006 (Tables). Retrieved July 19, 2007, from http://hdr. undp.org/hdr2006/pdfs/report/HDR_2006_Tables.pdf
World Bank. (2006). Information and communication for development—global trends and policies. Washington: World Bank. Wootton, R. (1996). Telemedicine: A cautious welcome. BMJ, 313, 1375-77.
ADSL Technology: The asymmetrical digital subscriber line (ADSL) technology is the technology which enables transmission of Internet data at faster rates than a regular modem using existing systems of ordinary telephone lines. Its range is limited to within about 5 km. Alma Ata Declaration, 1978: The Alma Ata Declaration was the outcome of the International Conference on Primary Health Care held in 1978 in Alma Ata, Kazakhstan. This declaration led to
591
Broadband for Health in Developing Countries
the adoption of the primary health care approach as the key public health strategy by all WHO member countries.
storing, and analyzing spatially referenced data which can also be used for resource planning and management purposes.
E-Health: The term “e-health” has been used in this chapter to denote all applications of Internet technology in health care, whether for commercial or for developmental purposes.
PDA: The personal digital assistant (PDA) is a hand-held computing device with a number of functions including palm-top Internet access. Though commercially used as a personal assistant, it is being used by development personnel working LQWKH¿HOGWRUHFRUGGDWDHOHFWURQLFDOO\DQGRIWHQ DOVRWRWUDQVPLW¿HOGLQIRUPDWLRQHOHFWURQLFDOO\WR a remote server.
EMR: The electronic medical record (EMR) is simply a medical record in electronic format, either captured directly on a computing device or created by entering patient data onto a computer. GIS: Geographic information system (GIS) is the application of computer programs for capturing,
592
Telemedicine: Telemedicine refers to the exchange of information between patient and health care provider through electronic media for consultative, diagnostic, or treatment purposes.
593
Chapter XXXVII
Improving Health Services via Advanced ICT Networks Peter Farr Peter Farr Consultants, Australasia Isabelle Ellis Charles Darwin University, Australia John Royle Reliance Consulting Pty Ltd, Australia
ABSTRACT This chapter describes an innovative broadband initiative that connects a group of general practices, medical specialists, hospitals, and other health providers in rural areas of Australia through a managed virtual private network. It provides secure connectivity for a variety of mission-critical healthcare delivery applications—for example, transmission of pathology and radiology test results direct to clinicians. The medical practices involved are small-medium enterprises (SMEs) and the key aspects of ICTs for them are the impact on costs, productivity, and customer service. The formal evaluation process examined WKHGRPDLQVRIDSSURSULDWHQHVVHI¿FLHQF\DQGFRVWHIIHFWLYHQHVV%HLQJWKH¿UVWVXFKKHDOWKQHWZRUN of its kind in Australia, the project encountered challenges and, by overcoming these, has been guiding JRYHUQPHQWSROLF\LQUHVSHFWWRHKHDOWK,QLWLDOO\IXQGHGIURP0DUFKYLDD&RPPRQZHDOWK*RYHUQment grant, the GoldHealth network moved into a sustainable mode in July 2006. This chapter provides insights into GoldHealth and should be a useful guide to any similar broadband network initiatives for the health sector elsewhere in the world.
INTRODUCTION The term e-health has become widely prevalent since originating in the year 2000. The stakeholders considered to be the users or targets of e-health YDU\ZLWKDUDQJHRIGH¿QLWLRQV6RPHGH¿QLWLRQV emphasize applications for providers and organiza-
tions—particularly those stressing electronic data exchange for clinical and administrative purposes. Others emphasize provision of information, education, and services to consumers, including patients and “citizens.” A small number clearly identify e-health with consumer health informatics. e-health is widely proposed to be a solution for the lack of medical and other healthcare specialists
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Improving Health Services via Advanced ICT Networks
in rural, remote, and outer metropolitan regions of many countries, including Australia. However, WRUHDOL]HWKHSRWHQWLDOEHQH¿WVWKHLQIUDVWUXFWXUH needed to provide many of the services requires VXI¿FLHQW EDQGZLGWK WR DOORZ WUDQVIHU RI ODUJH GDWD¿OHV$IIRUGDEOHEURDGEDQGWUDQVPLVVLRQLV therefore seen as a key enabler and change agent for secure, functional, and equitable participation in e-health activities. Conversely, having broadband does not necessarily mean that health IT penetration will automatically follow (Farr, October 2006). Exchanging highly sensitive personal and health information requires an acceptable legal and privacy framework that ensures that such information can be transferred securely to authorized users for approved purposes (National Health Information Management Advisory Council, 2001). This requires a secure broadband connection. Increasingly, healthcare providers are implementing a secure broadband connection to not only help improve their daily business processes, but to enable better collaboration in patient care, continuing professional education, and so forth. Although these challenges have been recognized for a relatively long period of time in the literature, there has been a huge diseconomy of scale when individuals, institutions, and medical practices have considered establishing electronic health information management systems, as, for example, in the Australian healthcare system where each institution and general practice operates separately. Rural residents in Australia tend to have poorer health status and less access to health services (Simmons & Hsu-Hage, 2001). The cost of healthcare, the lack of choice, and the lack of access to health information is of concern to consumers (Bourke, ,WLVDOVRUHFRJQL]HGWKDWWKHUHLVDGLI¿FXOW\ in attracting and retaining medical, allied health, and nursing staff in rural hospitals, remote community health services, and so forth. Against this background, the potential to improve access to healthcare has been used by governments as a vehicle to encourage adoption of broadband in urban, rural, and remote Australia. For example, in 2004, the “Broadband for Health” initiative funded broadband Internet access to general practices and Aboriginal Community
594
Controlled Health Services across Australia. DeVSLWHWKHORJLFDOEHQH¿WVRILPSOHPHQWLQJHKHDOWK systems for rural and remote areas, there have been GLI¿FXOWLHVLQGLIIXVLQJWKHVHVHUYLFHV A variety of change management models and methods have been proposed for implementing e-health, including: interaction design theory (Coiera, 2002); informatics, which has been the label put on sociotechnical systems design by a number of authors (Coeira, 2004; Hersch, 2002); and some more loosely put together frameworks such as that proposed by Kaur, Forducey, Smith, and Schneideman-Miller (2005) who propose that there are basic elements of organizational change, “strategy, structure, people and processes, all of which need to be addressed for a balanced approach to change.” Leonard (2004) points out that much of the research around implementing new technologies is “detailed and outline creative solutions and management theories, but have done little to facilitate the adoption of technology in healthcare.”
REFERENCE SITE OBJECTIVES The primary objective of the (DVWHUQ*ROG¿HOGV Regional Reference Site (EGRRS) was to test, PHDVXUHDQGGHPRQVWUDWHWKHEHQH¿WVRIKHDOWKFDUH providers working in rural and remote zones having access to high-speed, continuous, higher-quality broadband connectivity through which they may effectively and securely access a range of relevant information and communications applications, that is, e-health. Prior to the EGRRS rollout in early 2005, the majority of medical participants had either no Internet connection or only dial-up Internet from their practice or home, and this would have been typical for medical practices in other regions of Australia at that date.
DEMAND AGGREGATION CONCEPT The EGRRS broadband initiative provided a catalyst for organizations and individuals in the
Improving Health Services via Advanced ICT Networks
community to aggregate their demand for telecommunications services, potentially creating competition among providers and investment that will lead to greater sustainability and parity with metropolitan prices and conditions, hence VXSSRUWLQJWKH$XVWUDOLDQ*RYHUQPHQW¶V1DWLRQDO Broadband agenda. EGRRS had strong support in advance from key stakeholders across the region LQFOXGLQJWKH(DVWHUQ*ROG¿HOGV0HGLFDO'LYLVLRQ of General Practice (EGMDGP), a large number of individual health service providers, and other local bodies including the City of Kalgoorlie-Boulder, the local Chamber of Commerce, and a number of mining companies. The EGMDGP (http://www.egmdgp.com.au/) is one of 122 divisions of general practice in Australia. The EGMDGP was formed as a result of the &RPPRQZHDOWK *RYHUQPHQW¶V *HQHUDO 3UDFWLFH Strategy developed in 1992. This has resulted in greater involvement of GPs in health policy, planning, and delivery of health services at the local and regional levels, working in collaboration with other health agencies and managing a range of health service programs. Doctor members of the EGMDGP cover one RI$XVWUDOLD¶VODUJHVWUHJLRQV DURXQGWKHVL]HRI Texas), stretching from the town of Wiluna in the north to Esperance in the south, from the South Australian border in the east to Southern Cross in the west, with the city of Kalgoorlie-Boulder as the main center of administration. With a mix of metro-like, rural, as well as very remote geographies, the provision of quality health services is a challenging task. Prior to EGRRS, some key issues were: •
Distance. Due to the remoteness of healthcare practices in this region, communication of patient and medical information was infrequent and cumbersome between patients and health workers, as well as between healthcare practices themselves. Many patients in the region move regularly through communities and may see a number of different health practitioners for management of their healthFDUH7KLVFDXVHVDGLI¿FXOW\LQHVWDEOLVKLQJ normal support structures for patients.
•
•
•
•
Cost. Transport costs of medical equipment and patient visits in this region are high due to the great distances involved. Voice calls and fax alleviate some of the necessity of a direct visit, but when it comes to radiology images and other data that require large amounts of storage, direct communication is necessary, incurring a transmission cost. Service levels. There are less stable links between patients and GPs due to both the transient nature of the population and the turnover of doctors in the region. Digital patient records that are centrally stored in WKHGRFWRU¶VSUDFWLFHVDQGHDVLO\DFFHVVLEOH could greatly improve the continuity of patient care. However, due to the dispersed nature of healthcare workers in the region, coordination and advancement of the medical community as a whole is affected, with the participants in the healthcare supply chain (GPs, medical specialists, hospitals, pharmacies, etc.) operating independently of one another. Security. Healthcare practices need to be esSHFLDOO\GLOLJHQWDERXWSURWHFWLQJFRQ¿GHQWLDO patient data. Compromized medical data can expose a practice to error-prone diagnoses, OHJDOOLDELOLW\DQGORVVRIFRQ¿GHQFHIURP the patient. Consumer access to information. Rising expectations of up-to-date and comprehensible medical/health information.
In 2004, the Australian Government's Department of Health and Ageing (DoHA) invited telecommunications service providers to submit tenders for: •
•
The supply of EGRRS Broadband Services to participants in the EGRRS broadband initiative, to be paid for by DoHA for the Reference Site Period Provision of an EGRRS continuation plan under which participants may continue to acquire EGRRS broadband services for a further period of up to two years beyond the end of the reference site period at the same rates and on similar or better terms to
595
Improving Health Services via Advanced ICT Networks
those offered to DoHA for the reference site period. A network with managed Quality of Service was sought that could handle all potential applications as well as providing essential security and SULYDF\UHTXLUHGE\WKHVSHFL¿HGKHDOWKFRPPXnity. The successful tenderer for the network was IP Systems Pty Ltd http://www.ipsystems.com. au/industry_health_award.html.
Links Between the Reference Site and Other Australian Government Initiatives 7KHUHIHUHQFHVLWHIRUPHGDSDUWRIWKH'R+$¶V Broadband for Health program which draws together a number of government initiatives including the Access to Broadband Technology Initiative, to support whole-of-sector broadband connectivity. The Broadband for Health Program is a $60 million Australian Government program to provide broadband Internet access to GPs and aboriginal community controlled health services nationwide. The program is a key component of the $XVWUDOLDQ*RYHUQPHQW¶VUHYLVHG+HDOWKConnect LPSOHPHQWDWLRQVWUDWHJ\DQGUHSUHVHQWV'R+$¶V contribution to the national broadband strategy managed by the Department of Communications, Information Technology, and the Arts. HealthConnect aims to allow healthcare providers to securely transfer clinical information such as electronic referrals, hospital discharge summary reports, prescriptions, imaging and pathology requests, and results, and aims to create shared electronic health record repositories, guided by the iterations of National E-Health Transition Authority (NeHTA) ZLWK UHJDUGV WR VWDQGDUGV DQG VSHFL¿FDWLRQV IRU interoperability.
SCOPE OF THE PROJECT The physical locations of EGRRS participants were linked together by a Virtual Private Network, with GLIIHUHQWIRUPVRIWHFKQRORJ\ RSWLFDO¿EHU'6/
596
and satellite) providing the link from each site into the core network. The participants in the project comprised approximately 81 sites made up of 22 medical practices (health points) and 39 home points in rural Western Australia. Also involved were the (*0'*3RI¿FHVDW.DOJRRUOLHDQG(VSHUDQFHWKH Kalgoorlie and Esperance Hospitals, a pathology laboratory, a radiology service, three KalgoorlieBoulder pharmacies, a University Rural Clinical School, and the Royal Flying Doctor Service. The scope of the project included these “Core” applications: Fast Internet access (at the surgery DQG DW GRFWRUV¶ KRPHV IRU PHGLFDO LQIRUPDWLRQ and other purposes, secure e-mail (at the surgery DQGDWGRFWRUV¶KRPHV DFFHVVWRMedicare Online Claiming; VOIP telephony (both entry level and full IP telephony), desktop videoconferencing and boardroom videoconferencing, remote after hours access from home or a hospital to the surgery practice management system, and other applications UHVLGLQJRQHDFKSUDFWLFH¶VVHUYHU Also included were additional medical applications which broadly fell into four categories: (1) online education and peer support, (2) clinical therapeutic guidelines, resources and information services, (3) clinical applications, and (4) new technologies and small business tools relevant to healthcare providers.
INDEPENDENT EVALUATION For the formal evaluation, DOHA appointed a multidisciplinary team led by Peter Farr Consultants Australasia. Their task was to thoroughly examine WKUHHGRPDLQV HI¿FLHQF\FRVWHIIHFWLYHQHVVDQG appropriateness, and the ability of the network to be replicated elsewhere. The methodology used was a pragmatic or “mixed methods” methodology, involving the collection and analysis of extensive quantitative and qualitative data. Relevant information for the evaluation was obtained from a very wide range of sources over a period of 16 months. The use of controlled test scenarios was an innovative method used in the evaluation. Volun-
Improving Health Services via Advanced ICT Networks
teer doctors participated in realistic exercises that provided contextual use of the various e-health technologies. This enabled the doctors to envision the future, and consider how various technologies could be adapted or appropriated (Carroll, 2002). Examples included desktop videoconferencing and e-mailing images and other attachments. (The scenarios used related to dermatology and obstetrics cases). The test scenarios also served to help technically test the technology infrastructure and its appropriateness. Recognizing that the gains or losses that might accrue from the reference site are likely to vary according to the intended target or user, the nature RIWKHDSSOLFDWLRQDQGWKHVSHFL¿FWHFKQRORJLFDO FRQ¿JXUDWLRQ WKDW LV XVHG D WKUHHGLPHQVLRQDO model linking health work, users and target groups, and technologies was developed and is depicted in Figure 1. The executive summary of the project evaluation is available—see Department of Health and
Ageing (2007a), together with a write up of the lessons learnt—see Department of Health and Ageing (2007b).
GOVERNANCE It was never considered feasible for a regional health information network of the scale of EGRRS to operate informally. Accordingly, the overall governance structure in place during the EGRRS project period comprised (Peel, 2006): DoHA (the sponsor and funder), working cooperatively with the EGMDGP, EGRRS advisory committee, and EGRRS stakeholder reference group The network contractor (IP Systems) The evaluator (Peter Farr Consultants Australasia).
•
• •
ch no lo gi es
Figure 1. Three-dimensional model for the reference site evaluation
Te m un ic at ion s
Specific medical applications
Video
Internet
E-mail etc
Public health education
Medical education / professional development
Voice comms
Administration, Practice Management System, etc.
Practice Principals & other Doctors
In fo rm at io n
Practice Admin. Staff
an d
Co m
Patients
Clinical
Evaluation Perspectives
Community (Society)
Focus of General Practice and Specialist Work
597
Improving Health Services via Advanced ICT Networks
NETWORK SOLUTION The technology foundation of the reference site was the Cisco Medical Grade Network which is designed to meet the rigorous security and performance requirements of healthcare organizations. The Cisco Medical Grade Network can be tailored to suit all types of healthcare organizations, from large scale hospitals to the smallest rural healthcare practice. Utilizing cutting edge network technology, a robust and scalable network infrastructure can be built to support various medical and applications (Cisco, 2005). For EGRRS, the network consisted of a mix of Cisco routers and switches deployed over more than 80 sites across the region. For IP telephony, a mixture of Cisco IP phones and Cisco analogue telephone adapters were deployed at medical practices and the EGMDGP (Cisco, 2005).
EDQGFRQQHFWLRQVXVLQJDPL[RIRSWLFDO¿EHUDQG DSL from a carrier (Telstra) and two-way satellite from PanAmSat. IP Systems installed its own Cisco routers in Kalgoorlie so that the group of VLWHVLQWKH(DVWHUQ*ROG¿HOGVFRXOGEHURXWHGWR its national backbone network. In this phase, onsite training was conducted in regard to Internet, e-mail, IP telephony, role of the help desk, and so forth. In Phase 2, IP Systems connected in further sites and revisited the Phase 1 sites for refresher WUDLQLQJ DQG DGGUHVVLQJ VSHFL¿F SUREOHPV DQG issues. The EGRRS project team considered that using the staged approach to rollout of network and applications would give time for users to develop skills for each type of technology or application prior to the next “new thing” being deployed. A network schematic is shown in Figure 2.
Network Rollout
ADDITIONAL MEDICAL APPLICATIONS
IP Systems installed the majority of the network points in March-April 2005, the endpoint broad-
The value of broadband is principally in two WKLQJV²WKH¿UVWWKDWLWLV³DOZD\VRQ´DQGVHFRQG
Figure 2. Diagrammatic representation of EGRRS
Source: IP Systems (2006).
598
Improving Health Services via Advanced ICT Networks
via the convenient/quick/cheap access it gives to relevant content and relevant applications. The suite of additional e-health applications that formed part of EGRRS addresses the latter consideration and was vital in increasing the “value proposition” for GoldHealth. Through an expression of interest process the EGMDGP invited software vendors, education institutions, and other interested parties to propose applications to be hosted on the network. A list of “additional” applications was selected and implemented. Training and support was provided by the suppliers of the additional applications and the EGMDGP project staff. The GoldHealth Additional Applications portal was launched at the end of October 2005, the rationale being that an applications portal would assist the participants in seamlessly accessing the whole range of applications (Farr, October 2006).
HEALTHCARE VALUE The DoHA is quoted (Cisco, 2005) as being very optimistic in advance of the rollout about the potenWLDOEHQH¿WVWKDWWKHUHIHUHQFHVLWHFRXOGSURYLGH "Broadband is a key enabler for e-health applications such as HealthConnect, a system of electronic health records with cross-jurisdictional support in Australia. This type of cutting edge technology has not been experienced by participants in the past. Participants are gradually increasing their knowledge of the technology and potential for associated applications. As people become more aware of the capacity of the infrastructure, there are many ideas coming forward from the participants themselves to innovate the way they currently communicate." At its core, the EGRRS aimed to bridge the geographical distances that separate healthcare practices within the large area. The ability to transmit digital data at high speeds across such distances vastly improves the ability of the healthcare practitioner to share information and receive peer support. Participants in the reference site were able to use e-mail, access the Internet at broadband speeds, and make voice calls via IP telephony. Sites had the ability to access clinical and educational
information directly from the source, thus keeping themselves up-to-date with best practices. Additionally, they were able to: • • • •
Receive radiology and pathology reports electronically Transmit clinical images electronically to specialists for review Transmit and receive specialist reports without the need for scanning Connect to the Medicare Online Claiming to DFFHVVVHUYLFHVVXFKDVLQVXUDQFHDQGEHQH¿WV information, claims, and processing, as well as immunization and organ donor registry lookups.
Through the various channels of communication available on the network, collaboration between sites was made easier, fostering a sense of community among the caregivers. As an addiWLRQDOEHQH¿WWKLVFDQDOVRFRQWULEXWHWRWKHDELOLW\ to retain and attract more healthcare workers to the region. The use of IP telephony allowed on-net voice calls to be made at a reduced cost compared with conventional telephony. Since IP calls travel as digital data along the high-speed network, quality DQG¿GHOLW\LVNHSWDWDKLJKOHYHOXVLQJ4R66WDII members from any of the sites are able to make voice calls without incurring any long distance charges. With the VPN, staff members who move between sites had a convenient way to stay connected. Routine technical maintenance and support could also be done remotely, thus saving the time and cost of physical travel. Video conferencing has also provided healthcare practitioners within the region the ability to conduct and receive training remotely. A government spokesperson is quoted in a publicly available case study (Cisco, 2005) as saying: "For the most part, many of the health professionals involved in the project are minimal users of IT. There are some advanced users, or 'champions,' which are important people in terms of creating acceptability and understanding of the technology. Desktop videoconferencing capability will enable a practitioner to conduct remote consultations,
599
Improving Health Services via Advanced ICT Networks
undertake education, and attend professional meetings without the need to travel." One important aspect of EGRRS is its ability to scale incrementally to future growth. Additional hardware and software can be added without the QHHG IRU D GUDVWLF UHFRQ¿JXUDWLRQ +HQFH WKLV network becomes the seed from which further expansion and added services can be incorporated in the future. Additional applications which may be explored include backup services and disaster recovery—potentially important for small-medium enterprises (SMEs).
DISCUSSION 2YHUDOOWKHSURMHFW¶V¿QGLQJVFRQ¿UPWKHNQRZOedge gained from research into similar projects involving the integration of new technologies. Broadly, the research suggests that diffusion of new technologies generally follows an “S” shape over time, with low take-up in the early months/years, then an increase in the rate of take-up, when network HIIHFWVVWDUWWREHFRPHVLJQL¿FDQWDQGDSSDUHQWWR the group of people for whom the technology is intended. This is consistent with the innovation adoption curve of Rogers (1995). Leonard (2004) has reported that once the decision to adopt a technology has been made, regardless of the industry, the amount of work required by the business increases until the new system is working effectively. How much the workload increases, and how long this effect lasts, has been OLQNHGWR¿YHFULWLFDOVXFFHVVIDFWRUV 1. 2. 3. 4. 5.
600
The amount of resistance to change and experience in using technology The amount of training before and during the transition The amount of buy in, or contribution from stakeholder groups The level of effective reporting on outcome measures during and after implementation The level of effectiveness in dealing with “breaks” (i.e., setbacks)
An aspect that was certainly to be considHUHGEHQH¿FLDOZDVWKHGHYHORSPHQWRIVWURQJHU relationships between participants (and with the EGMDGP) during the course of the project through project-related activities and events, including camaraderie-building social events. Members of the project visited participants many times which generally involved extensive travel. They also made a big effort to balance work priorities with social activities for the project. This balance is important in service delivery for health networks, and should not be underestimated. The major technical hurdles encountered were the technical problems affecting IP telephones and the satellite connected services. An aspect that did not reach initial expectations for practitioners in rural and remote areas was onnet videoconferencing. Along with the voice over IP, this technology required the implementation of quality of service, which comes at a cost. Training and support also were issues, compounded by staff turnover at medical practices. Limitations also arose from the relatively low OHYHO RI ,&7 FRQ¿GHQFH RI PDQ\ SDUWLFLSDQWV DW the commencement of the project. The executive summary of the project evaluation is available—see Department of Health and Ageing (2007a), together with a write up of the lessons learnt—see Department of Health and Ageing (2007b).
SUSTAINABILITY The initial set-up funding included both capital funding to establish the GoldHealth VPN and its associated services as well as the operational fundLQJDQGVXSSRUWIRUDVXI¿FLHQWSHULRGWRHQDEOH EGRRS to reach a viable level of operation and VHUYLFH)XQGLQJRIWKHSURMHFWE\'R+$¿QLVKHG as scheduled in June 2006. Farr and Papandrea (2006) have found that the various factors that contribute to sustainability for community ICT facilities can be grouped into these three key interdependent dimensions:
Improving Health Services via Advanced ICT Networks
Figure 3. Diagrammatic representation of the sustainability model for community ICT facilities Financial Resources
SUSTAINABILITY Community Empowerment & Socio-Economic Impact
• •
Financial resources Community empowerment and socioeconomic impact (I¿FLHQWRSHUDWLRQVDQGVXSSRUWV\VWHPV
The interrelationship between these three key dimensions is illustrated in Figure 3. Each of these three dimensions is critical to the viability of a managed health network. Also, because of the interrelationship between the GLPHQVLRQV LQVXI¿FLHQW DWWHQWLRQ WR DQ\ RQH RI them will be likely to have an adverse impact on the other two. It is the nature of ICT projects that there are economies of scale via aggregation and other means. If there are too few active participants over a very dispersed area, it will probably be uneconomic to offer or maintain a sophisticated service in the EGRRS style.
THE FUTURE With respect to “testing” and “demonstrating” the EHQH¿WVWKLVLVDQRQJRLQJDFWLYLW\IRU*ROG+HDOWK which is the name the EGRRS project was carried forward under from July 2006. The great majority of participants signed up from July 2006 to continue with a network connection (IP Systems, 2006; Siegmund, 2006). Continuity of governance arrangements has occurred as a result of putting in
Efficient Operations & Support Systems
place in July of 2006 a GoldHealth advisory committee and a stakeholders reference group. 6SHFL¿FLPSDFWVRI(*556RQFRQWHPSRUDQHous policy considerations or decisions include the aspect that the project has helped to inform the general development of change management around HKHDOWK7KHUHDUHDQXPEHURIVSHFL¿FRXWFRPHV that DoHA is believed to have taken forward from a policy perspective as a result of this project, including the Security Awareness and Conformance Report for the Broadband for Health program, the managed health network grants which have been VLJQL¿FDQWO\LQIRUPHGE\WKH(*556H[SHULHQFH especially considerations of change management and governance of such projects (Peel, 2006). Wider issues around broadband infrastructure and access for the health sector, especially in remote areas, have also been informed by this project and are being followed through at the national broadband policy level. The local division of general practice was well VDWLV¿HGZLWKWKH(*556SURMHFW³,WKDVEHHQVR well received by participants that it is now able to be sustained as a viable network under the name of GoldHealth,” said the CEO of the EGMDGP. “The GPs and practices throughout the region have embraced the concept and recognized that there DUHVLJQL¿FDQWEHQH¿WV-XVWDVLPSRUWDQWO\*3V throughout the region have realized that e-health is a part of the future of medicine and primary
601
Improving Health Services via Advanced ICT Networks
care and GoldHealth is at the forefront of it” (IP Systems, 2006). It is considered that health communities are needed from which lessons can be learned in terms of standards generally and in particular those needed for electronic pharmaceutical prescriptions, hospital discharge summaries, a standard e-health record, and interoperability issues. The Eastern *ROG¿HOGVKHDOWKSURYLGHUFRPPXQLW\LVLGHDOO\ placed to provide such a reference point. A central agency that can support network users is also a key success factor in projects of this nature and this involvement, when at a detailed level rather than just a bystander role, can help to inform policy.
CONCLUSION The establishment of managed health networks LQ UXUDO DQG UHPRWH FRPPXQLWLHV UHÀHFWV WKH pursuit of the social policy objective of providing healthcare providers and clients with access to ICT services that are indispensable to their well-being at reasonable and affordable prices. Evidence from EGRRS shows that such a managed health network can cut down a feeling of isolation and “falling on the wrong side of the digital divide.” It can lead to the development of many new skills along with long-term employment opportunities, economic development, and a greater ability to cope with change. The medical practices and pharmacies involved in EGRRS are SMEs and key aspects of ICTs for them from a project of this nature are the impact on costs, productivity, and customer service. EGRRS has proven that a VPN can be created to suit the majority of a wide range of healthcare service providers. The network has been leveraged to deliver DSSOLFDWLRQVLQDPHDQLQJIXOWDLORUHGDQGÀH[LEOH way that can continue to grow with the changing needs of the participants and the network. Partly from this evidence, GoldHealth was a joint winner of the “Best Regional Communications Solution” at the Australian Telecommunications Users Group 2006 awards. )XQGLQJ RI WKH SURMHFW E\ 'R+$ ¿QLVKHG LQ June 2006 and, on a self-standing basis, the great
602
majority of health point sites continued as part of the GoldHealth network beginning in July of 2006. The managed VPN which is well bedded down enables the local division of general practice to have control, to decide who can be part of the network. It also controls the applications that are made available, and can broker the commercial conditions between the service providers and the participants.
REFERENCES Bourke, L. (2001). Rural health consumers of health issues. Australian Journal of Rural Health, 9(1), 1-6. Carroll, J., Howard, S., Vetere, F., Peck, J., & Murphy, J. (2002, January). Just what do the youth of today want? Technology appropriation by young people. In 3URFHHGLQJVRIWKHWK+DZDLL InternationalConference on System Sciences (HICSS-34), Maui, Hawaii. Institute of Electrical and Electronics Engineers, Inc. Cisco. (2005). Connected health in western Australia with Cisco medical grade network infrastructure. Retrieved July 19, 2007, from http://www. cisco.com/web/strategy/docs/healthcare/KalgoorlieCaseStudy.pdf Coiera, E. (2002). Interaction design theory. International Journal of Medical Informatics, (00), 1-18. Coiera, E. (2004). Four rules for the reinvention of healthcare. British Medical Journal, 328, 11971199. Department of Health and Ageing. (2007a). (DVWHUQ JROG¿HOGV UHJLRQDO UHIHUHQFH VLWH GHYHOopment phase report summary. Retrieved December 28, 2007, from http://www.healthconnect. gov.au/internet/wcms/publishing.nsf/Content/ B2D8D97206D4CCB4CA256F18004FBF7F/ $File/EGRRS%20Executive%20Summary%20F INAL.pdf Department of Health and Ageing. (2007b). (DVWHUQ JROG¿HOGV UHJLRQDO UHIHUHQFH VLWH GHYHO-
Improving Health Services via Advanced ICT Networks
opment phase lessons learnt. Retrieved December 28, 2007, from http://www.healthconnect. gov.au/internet/wcms/publishing.nsf/Content/ B2D8D97206D4CCB4CA256F18004FBF7F/ $File/EGR RS%20Lessons%20Lear nt%20%20FINAL.pdf (DVWHUQ*ROG¿HOGV0HGLFDO'LYLVLRQRI*HQHUDO 3UDFWLFH7KH(DVWHUQ*ROG¿HOGV5HJLRQDO5HIHUence Site (EGRRS). Retrieved Jul 19, 2007, from http://www.egmdgp.com.au/Eastern%20Goldf ields%20Regional%20Reference%20Site/eastHUQBJROG¿HOGVBUHJLRQDOBUHIHKWm Farr, P. (2006, October). Bandwidth connectivity and online content. Innovative strategies for the education and health sectors. Paper presented at ICT WA 2006 Conference, Perth, Australia. Retrieved July 19, 2007, from http://www.ictwa. org.au/program/Peter%20Farr%20ICTWA%202 7Oct06.pdf Farr, P., & Papandrea, F. (2006). Sustainability of community online access centers. In Priessl, B., & Muller, J. (eds.), Governance of communications networks—connecting societies and markets with IT (pp. 165-186). Heidelberg, Germany: PhysicaVerlag. Hersh, W.R. (2002). Medical informatics: Improving health care through information. Journal of the American Medical Association, 288, 1955-1958. IP Systems. (2006, July 17). As good as gold—rural GPs prepared to pay for technology which provides better healthcare connections than doctors have in Australian cities (Media Release). Retrieved July 19, 2007, from URL: http://www.ipsystems.com. au/industry_health_rural.html Kaur, K., Forducey, P.G., Smith, L., & Scheideman-Miller (2005). Organizational effectiveness of a telerehabilitation system in Oklahoma. Telemedicine Journal and E-Health, 10(Suppl. 1), S29. Leonard, K.J. (2004). Critical success factors relating to healthcares adoption of new technology: A guide to increasing the likelihood of success. Electronic Healthcare, 2(4).
National Health Information Management Advisory Council. (2001). Health online: A health information action plan for Australia (2nd ed). Canberra, Australia: Department of Health and Aged Care. Peel, N. (2006, April 5). GPs making a differHQFHLQWKHHDVWHUQJROG¿HOGV Paper presented at CHIK Health-E-Nation Conference, Melbourne, Australia. Rogers, E.M. (1995). Diffusion of innovations (4th ed.). New York: The Free Press. Siegmund, A. (2006, May). Connecting rural clinicians. Paper presented at ATUG 2006 Regional Conference. Retrieved July 19, 2007, from www. atug.com.au/ATUG2006RegionalConf/DrAndrewSiegmundPresATUG2006RegConf.pdf Simmons, D., & Hsu-Hage, B. (2001). Determinates of health, disease, and disability. In D. Wilkinson (Ed.), Rural health in Australia (pp.79-93). Melbourne: Oxford University Press.
KEY TERMS Virtual Private Network (VPN): A hybrid network that includes both public and private facilities. The user leases a bundle of circuits and FRQ¿JXUHVWKH931RQDQDVQHHGHGEDVLVVRWKDW VRPHWUDI¿FWUDYHOVRQWKHSULYDWHOHDVHGQHWZRUN DQGVRPHWUDYHOVRQWKHFRPPRQFDUULHU¶VSXEOLF network. E-Health: AQHPHUJLQJ¿HOGLQWKHLQWHUVHFWLRQ of medical informatics, public health, and business, referring to health services and information delivered or enhanced through the Internet and related technologies. Technology Diffusion Curve: The innovation DGRSWLRQFXUYHRI5RJHUVLVDPRGHOWKDWFODVVL¿HV adopters of innovations into various categories based on the idea that certain individuals are inevitably more open to adaptation than others. It is also called diffusion of innovations theory.
603
604
Chapter XXXVIII
Remote Patient Monitoring in Residential Care Homes: Using Wireless and Broadband Networks Tanja Bratan Brunel University, UK Malcolm Clarke Brunel University, UK Joanna Fursse Brunel University, UK Russell Jones Chorleywood Health Center, UK
ABSTRACT The UK’s National Health Service (NHS) is undergoing great reform. Driven by a demand for higher quality health care provision, information and communication technologies (ICT) are increasingly being used as tools to realize this change. We have investigated the use of remote patient monitoring (RPM), using wireless and broadband networks, in three community care homes between July 2003 and January 2006. The aim of the project was to determine for what conditions and in which setting the RPM was most useful and to establish an organizational and clinical infrastructure to support it. Evaluation of WKHSURMHFWGHPRQVWUDWHGFOLQLFDOEHQH¿WVVXFKDVWKHHDUO\GHWHFWLRQRIFDUGLDFHYHQWVDOORZLQJSURPSW intervention and routine monitoring of other conditions. A change in work practices resulted in a more collaborative approach to patient management and led to an increase in communication between health care professionals from different sectors, as well as the establishment of protocols for seeking advice. Technically, the equipment largely met the users’ needs. In conclusion, the monitoring proved a useful tool for the management of chronic diseases and has great potential to contribute to the reform of the NHS.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Remote Patient Monitoring in Residential Care Homes
INTRODUCTION 7KH8.¶V1DWLRQDO+HDOWK6HUYLFH 1+6 KDVEHHQ and continues to be subject to major change. Driven by an increase in expectations, together with an aging population and the availability of new medical technologies (Department of Health, 2006), the government presented its vision for reform of the NHS in the NHS Plan in 2000. The main themes of this plan were to: (1) develop a service that would offer prompt and convenient care; (2) enable rapid access to diagnosis and treatment in modern facilities; and (3) give patients the choice over the time, place, and personnel involved in their treatment (Department of Health, 2001). Central to the plan was the creation of the National Program for IT (NPfIT), the largest IT program in the world. A national data spine and a national broadband network (N3) were designed to connect health care providers to a central secure system (Calkin et al., 1999) and support many e-health initiatives. These initiatives include the electronic patient record, repeat prescription, choose and book, as well as KHOSZLWKSUR¿OLQJFOLQLFDOJRYHUQDQFHDQGUHXVH of data (Department of Health, 2002). HKHDOWKLVDQXPEUHOODWHUPDQGFDQEHGH¿QHG as “the application of information and communications technologies (ICT) across the whole range of functions which, one way or another, affect the health of citizens and patients” (Maheu et al., 2001). This ranges from medical applications such as telemedicine, remote patient monitoring, and electronic patient records to telecare and beyond to tools that empower patients such as health Web sites. The objectives of this chapter are to: • • •
Inform the reader about the potential of ICTs to reform health care Demonstrate the application of ICTs in form RIDVSHFL¿FHKHDOWKFDVHVWXG\ Discuss future trends in the area
The case study, e-Vital, was a feasibility and market validation project providing remote patient monitoring (RPM). The UK element of the project investigated the use of RPM in two residential
care homes and one nursing home. The work was novel in that it exploited new forms of technology, wireless and broadband networks, to provide the communication infrastructure to small health care organizations.
BACKGROUND As the number of elderly people in society continues to grow, so do the health care costs associated with this section of the population and the need to ¿QGDPHDQVIRUHI¿FLHQWDQGHIIHFWLYHSURYLVLRQ of health care. Continued aging of the population LVLQHYLWDEOHGXULQJWKH¿UVWKDOIRIWKLVFHQWXU\DV the relatively large number of people born after the Second World War and during the 1960s baby boom become older (British National Statistics: Ageing, 2005). Chronic disease is more prevalent among the elderly, with almost 75% of the over 65 year-olds suffering from at least one chronic disease, while nearly 50% have two or more (Calkins, Boult, & Wagner, 1999). Many require frequent medical attention, both in a clinical environment and at home. As a result of cost-issues, over-crowded hospitals, and the preference of elderly people to remain in their normal environment, there has been a trend to move away from hospital-based health care to home-based health care (Maheu, Whitten, & Allen, 2001). The UK government has been actively promoting home-based health care as part of its program to move services into the community (Wistow, 2000). It has recently made available 80 million pounds for preventative technologies over a 2 year period (TeHIP, 2005). It is speculated that with the provision of home care services, patients can live in their usual environment for longer, thus DYRLGLQJWKHKRWHOFRVWVRIKRVSLWDO WKHSDWLHQWV¶ own care-givers can provide no-cost nursing; and the actual costs of primary care are often lower than the equivalent service provided from hospital (Hersh et al., 2002).
Remote Patient Monitoring 530FDQEHGH¿QHGDVWKHPRQLWRULQJRISK\VLological measurements in a setting other than a 605
Remote Patient Monitoring in Residential Care Homes
hospital, using ICTs to transfer data over geographical distances. It has the potential to offer a convenient and cost-effective solution for the problems described. Information such as blood pressure, respiration, temperature, weight, blood oxygen saturation (SpO2), glucose monitoring, SHDNH[SLUDWRU\ÀRZDQGUHPRWH(&*PRQLWRUing can be collected and transmitted (Bratten & Cody, 2000). Currently, most RPM takes place in WKHSDWLHQW¶VRZQKRPH There are three main reasons for remote patient monitoring: (1) the prevention of hospital admissions through early detection of deterioration, (2) prompt hospital admission where indicated, and (3) the provision of hospital-like services in the home, which might facilitate keeping patients at home and out of hospital or early discharge from hospital and the follow-up care in the home. The aim of each is to triage cases so that resources can be allocated according to need and that patients receive the most appropriate form of care for their condition at the time they need it. %HQH¿WVLQFOXGH •
•
•
Early discharge of patients from hospital, as it can provide an alternative to hospital monitoring Help in identifying deterioration early and thus either expedition of hospitalization or prompt treatment resulting in a reduced number of unnecessary admissions, and enabling of prompt emergency response (Bratan et al., 2005) Allowing for health care professionals to look after a greater number of patients as they FDQGRVRIURPWKHLURI¿FHV HJ'HPHULV 2004)
There are three distinct types of RPM: (1) chronic disease management which might also include other regular monitoring for example during pregnancy, (2) acute monitoring, and (3) ambulatory monitoring. Diagnostics, which have a different purpose, for example, Holter ECG monitoring, will not be covered in this chapter.
606
1.
2.
3.
Chronic and other disease management: This is currently the most common form of RPM. Measurements are taken on a periodic or regular basis, for example, daily, and data is forwarded to the data store usually without processing. The purpose is to GHWHFW D VLJQL¿FDQW FKDQJH LQ D SDUDPHWHU over a period of time, which may be an early indicator of deterioration and would indicate that intervention in therapy may be required. The most important aspect of designing such D VHUYLFH LV WR GH¿QH WKH FOLQLFDO UHVSRQVH and the intervention in the pathway of care, DQGLPSRUWDQWO\KRZLWPD\¿WZLWKLQWKH existing organizational structures. The most commonly measured data includes blood pressure, weight, and oxygen saturation of the blood (SpO2). Data are frequently measured by the patients themselves. Chronic heart failure (CHF) and chronic obstructive pulmonary disease (COPD) are the most commonly monitored diseases. Acute: The purpose is to monitor a patient in situations where there might be a sudden or rapid deterioration in condition which must receive prompt medical intervention. This can include acute exacerbation of a chronic disease, cardiac conditions, stroke, or post operative care. Data from vital signs will have to be sent on a near continuous basis and would include ECG, SpO2, temperature, and blood pressure. The requirement in design is to ensure that the clinical response is capable of rapid intervention and would include provision of professional medical personnel close to hand. Ambulatory: In this form of monitoring, the goal is to identify a critical and sudden change in the vital sign in an otherwise reasonably healthy person, for example a prolonged ST segment change (a part of the ECG wave) in an ischemic patient that might be the early indicator of imminent myocardial infarction. This requires that the signal be monitored continuously and algorithms be used to detect an event. The system must be capable of sending an alarm, and, ideally, data representative of
Remote Patient Monitoring in Residential Care Homes
Figure 1. Patient being monitored with the telemonitor
the event. The system should allow the patient mobility and be designed to operate within a range of environments, such as within the home or the community, and so is likely to be based on wireless communication.
E-VITAL CASE STUDY The e-Vital project established a monitoring service in three community care homes (two residential homes and one nursing home) and investigated its use in this environment. Community care homes are characterised by having a large proportion of dependent and semidependent residents who often suffer from multiple chronic diseases and frequently require medical attention. They also have residential staff members who are either mediFDOO\TXDOL¿HG LQWKHFDVHRIWKHQXUVLQJKRPH or have some medical training (in the case of the residential care home). RPM was seen as supporting the staff members in the management of their residents. They are responsible for their care and well-being. With training, they are able to apply and use the monitoring system to gather data for
onward transmission to a health care professional to provide support and advice. 7KH ¿UVW SKDVH RI WKH H9LWDO SURMHFW ZDV WR evaluate the practical feasibility and acceptability of the monitoring. The second phase determined the impact of the monitoring on clinical outcomes, organizational and human factors, and technical issues. This chapter discusses the outcomes of both phases, although it focuses on the second phase.
Set-Up The three community care homes within Watford and Three Rivers Primary Care Trust (PCT) to the northwest of London were equipped with a telemonitor from RGB Medical Devices, S.A/Spain. The telemonitors were able to measure several parameters including 7-lead ECG, blood pressure, oxygen saturation (SpO2), heart rate, temperature, and respiration, and were designed to support monitoring of conditions that commonly occur in a community care home setting, including cardiac and pulmonary diseases. As such, they were to be operated in a nonclinical environment by nonmedical personnel. If one of the residents gave cause for concern in their condition, physiological measure-
607
Remote Patient Monitoring in Residential Care Homes
Figure 2. The network architecture in the pilot sites
ments would be recorded and securely transmitted over the Internet to the data server. Residential staff members would then contact the health care professional to inform them of the situation and that data was available to be viewed, which was via a secure Web site. Figure 1 shows a patient being monitored. The overall network architecture for the system is shown in Figure 2. Monitoring devices were connected to the network in each community care home using a wireless local area network (WLAN EJ WRJLYHIUHHGRPDQGÀH[LELOLW\RISODFLQJ the devices according to the needs of the resident or staff. Recordings were most often made in the seclusion of the bedroom, although some routine measurements such as blood pressure were made in communal areas when convenient for the resident. The wireless network connected through an access point to an asymmetric digital subscriber line (ADSL) broadband router to provide access to
608
the public Internet. Data was transferred through a virtual private network (VPN) connection to the data server. The data was viewed by medical staff at the associated health center or local hospital through their connection to NHSNet, the regional KHDOWKQHWZRUNRIWKH8.¶V1DWLRQDO+HDOWK6HUYLFH Being a private managed network, this allowed no external access, and it was necessary to place the server outside of NHSNet. The community care homes were privately run and therefore not allowed access to NHSNet. Although innovative in its use of technology and designed as a feasibility study, it was important not to concentrate solely on the technical aspects but to also consider the processes, interactions, and impacts on the various stakeholders of the system. Failure to consider this has often resulted in unsuccessful implementations of information systems in the past (see Lyytinen & Hirschheim, 1987; Sauer, 1993).
Remote Patient Monitoring in Residential Care Homes
Process of Care The community care homes normally have a contract for care with the local general practitioner (GP) and the residents of each home are registered with the local health center. Routine health care, such as reviewing medication, chronic care follow up, and dealing with acute illness, is provided by the GP visiting the home on a regular basis. The GP would also make unscheduled visits to the home to deal with urgent problems. We have used the term community care home to refer to all institutions providing residential and in some cases also nursing care. The homes having individual bedrooms but share communal and living facilities and all meals are provided. However, the type of resident in such homes may vary from the very well, but usually aged, through to high dependency, with the latter referred to as QXUVLQJKRPHVDQGKDYLQJVHYHUDOZHOOTXDOL¿HG and experienced nursing staff. This population of the homes are characterized by individuals who become ill relatively frequently and whose health, often during an illness, can deteriorate rapidly. As a result, they require a relatively high level of medical intervention. In the case of sudden deterioration, there will be great concern over the well-being of the resident, and it is normal practice to seek medical assistance from the local GP on an emergency basis. However, help may often be sought out of hours for the GP service or at busy periods, and is not immediately available. This often results in the person being sent to the hospital by ambulance, often with accompanying stress and ordeal. One of the main aims of the project was therefore to reduce these unnecessary admissions to the hospital. As part of this, we investigated for which conditions and in which setting the monitoring would be most useful.
carer, one nurse, and the technical director. We developed a theoretical model and framework for factors important for successful implementation of an ICT project in health care and in particular related to e-health. This was used to underpin the design of the project and to direct the interview questions.
Evaluation: Findings from the Literature It has been suggested that the evaluation of e-health projects is highly complex. Research teams tend to underestimate the technical and organizational complexity of the task, which leads to problems VHOHFWLQJ DQG LQVWDOOLQJ HTXLSPHQW DQG GLI¿FXOties in interpreting the data collected (May et al., 2002). Wootton and Hebert (2001) state that before evaluating a telehealth project, an understanding of what constitutes success is necessary. Listed below are evaluation criteria adapted from the authors. It is emphasized, however, that telehealth projects must be considered in relative, not absolute terms, that is, “success cannot be judged in isolation” (Wootton & Hebert, 2001, p.6). This means that success depends on the perspective and on the available alternatives. If no other health care is available for instance, telehealth will always be preferable, even if it is clinically inferior to conventional health care. Success can also depend on the perspective adopted, such as that of the clinician, the patient, the health care provider, or society as a whole, and so forth. Consequently, there is “no single criterion for success,” but the most relevant criteria should be applied (May et al., 2002). As an overall guideline, it is suggested that success in telehealth can be measured in the extent to which it has contributed to the provision and maintenance of a health service (Hailey, 2001). :RRWWRQDQG+HEHUW LGHQWL¿HGWKHIROORZing indicators for success in e-health projects:
Design of the Project A qualitative evaluation was used to assess the success of the monitoring in each home. This was based on semistructured interviews with key members of staff: two doctors, four managers, one
• • •
Routine operation High activity levels &OLQLFDOHI¿FDF\ Cost-effectiveness
609
Remote Patient Monitoring in Residential Care Homes
• • • •
$GHTXDWH¿QDQFLQJ QRVSHFLDOIXQGLQJDUrangements required) Acceptance by clinicians Acceptance by patients Improved access to health care, particularly in rural and remote areas Reduction of travel
Although these apply to service programs rather than pilot services, many indicators are still useful for evaluating pilots and can give an indication of the “maturity” of the project. In the next section, e-Vital will be compared to the above indicators, DQG RWKHU ¿QGLQJV IURP WKH HYDOXDWLRQ ZLOO EH SUHVHQWHG7KH¿QGLQJVDUHGLYLGHGLQWRFOLQLFDO organizational, human, and technical issues.
Findings from the Case Study Clinical Outcomes Thirty residents were monitored for various condiWLRQVDQGLQGLYLGXDOVSHFL¿FFDVHVRIUHVSLUDWRU\ disorders, cardiac problems, high blood pressure, diabetes, and renal problems. Generally, the monitoring was used to investigate residents who were unwell, and in most cases the outcome was that no
Figure 3. Existing clinical pathway
Hospital
Resident Enters Home
Emergency
Clinical Organization
Residential Nursing Home
Clinical Services Given
Resident Registered
Local Surgery
610
immediate action was required. However, in these cases it did reassure the staff whether a resident needed to be seen by a doctor or not. ,QWKUHHFDVHV DVLJQL¿FDQWFDUGLDFHYHQW (two asymptomatic myocardial infarctions and one pericardial effusion) was found and permitted a prompt response, and, in one case, controlled admission to hospital. We are unable to comment on the outcome of such patients without RPM; however, response time can be critical for the outcome for the patient. We are also unable to quantify how many admissions may have been avoided, as the effect of RPM is not always direct. But use and experience of the technology, improved awareness, and closer professional relationships also impact on how patients are managed in an RPM environment. Staff members learn to be more proactive, take increased responsibility, and become more autonomous (Aas, 2000) and this will have an impact on admissions. E-Vital demonstrated that RPM in residential homes has value. We envisioned that it could reduce unnecessary admission to hospital, saving cost and discomfort for the resident. In the event, our system diagnosed several cases of silent myocardial infarction, which otherwise would have gone undiagnosed and untreated. Incidence of this type is to be expected in the environment of the comPXQLW\FDUHKRPHZLWKLWVVLJQL¿FDQWSRSXODWLRQ of elderly residents having existing disease, and with increased experience of use we would expect such systems to have greater value. Furthermore, we would also expect such a system in the future to include the capability to perform routine daily measurements on patients to manage residents with chronic disease.
Patient care pathways were used as a methodology to plan the introduction of the technology. The pathway before the introduction of the RPM system shows how residents giving concern were normally immediately sent to hospital (Figure 3). Many such admissions were unnecessary and resulted in residents spending long periods waiting in the accident and emergency department. This could cause distress to the resident and their
Remote Patient Monitoring in Residential Care Homes
Figure 4. E-vital pathway of care Resident Enters Home Possible Emergency Residential Nursing Home
Telemonitor Routine Monitor
Clinical Services Given
Resident Registered
Emergency
Hospital
Local Surgery
family, and might also lead to the unnecessary occupation of hospital beds. Introducing RPM into the pathway, as shown in Figure 4, allowed staff members to triage cases and so determine those needing admission to hospital or those who could safely remain in the home. In this system only the information moved, not the patient or the GP.
Organizational Factors The effects of introducing technology to an organization are well known (Aas, 2001) and must be well planned and considered. Staff members must adopt new working practices, be introduced to and trained on new equipment, learn new skills, and relate to new people or in new ways. In this project, staff members in the community care homes were expected to have more direct communication with staff at the health center regarding medical conditions and their management. They were to apply medical monitoring devices to take physiological measurements, which were, for many, new skills. Staff members in the health center had to access medical data from which decisions had to be taken without having direct access to the patients and on types of data to which they may not previously have been accustomed. New diagnostic and interpretive skills may have been required and access to expert skills and support found.
Other changes, such as daily routine could be affected. This might include learning how best to manage a situation such as when to review data in the busy schedule of a morning clinic, how to liaise with the staff of the community care home to elicit physical symptoms and condition, how to manage the patient, and learning the limits and capabilities of the system so that each health care professional felt safe and secure in later decisions. Users reported that they found their time expenditure “manageable,” but commented that they considered the support and encouragement from higher management to be critical and felt that the RPM system was more valuable when supported by their superiors. Meeting user requirements is one of the most important factors for a successful information system (Currie & Brown, 1997). The technical director therefore arranged two meetings with the staff before implementation and two during implementation in order to receive feedback and comments on the RPM system. This user involvement in setting up and conducting the trial proved WR EH YHU\ EHQH¿FLDO DQG FUHDWHG WKH QHFHVVDU\ trust and understanding between the medical and care staff as well as the research team. The system also built upon the existing relationship between each home and the GP.
611
Remote Patient Monitoring in Residential Care Homes
The time required to train the staff to use the equipment was relatively short and built upon existing skills. The staff members were quick to recognize the potential advantages of the system to improve patient care. Initially, key members RIWKHVWDIIZHUHLGHQWL¿HGDQGWUDLQHGXVLQJWKH concept of the champion (see the next section). Once established, training for other members of the staff was arranged. Although the staff members were shown how to access the data on the server in order to check the data was available, we had to stress that it was not their role to interpret the data; rather, that they should inform the GP that data was available and seek advice on management. Some confusion and concern about this role was found. We determined that ensuring each individual had a clear understanding of their role and how, when, and where to seek support was very important. Legal issues were encountered. During a routine inspection by the Commission for Social Care Inspection (CSCI) of one of the homes, it was decided that the staff did not possess the necessary skills to operate the RPM system. Use of such equipment was not considered normal. This decision was subsequently reversed after lengthy discussions.
Human Factors Communication between the staff of the community care homes and the health care professionals was key, as a greater level of communication than QRUPDOZDVUHTXLUHG:HWKHUHIRUHWRRNVSHFL¿F actions to develop the relationship between the SDUWLHV&RQ¿GHQFHLQWKHXVHRIWKHV\VWHPE\WKH home staff was important, and thus was affected by technical problems experienced early in the project. Staff members were sometimes reluctant to report problems, assuming that they were major and insurmountable. However, often, they were due to simple issues and quickly resolved. Having a clinical champion at each site was YLWDO:HWULHGWR¿QGDSHUVRQZKRZDVHQWKXVLDVWLF DERXWWKHPRQLWRULQJDQGKDGVXI¿FLHQWDXWKRULW\ WR LQÀXHQFH RWKHU PHPEHUV RI VWDII 7KH KRPH where the monitoring proved most successful was
612
the one with the most enthusiastic champion. The FKDPSLRQGHPRQVWUDWHGWKHEHQH¿WVRIWKHPRQLWRULQJWRWKHKRPH¶VPDQDJHPHQWDQGHQWKXVLDVP convinced others to use it. The response to using the system was generally very positive. Most members of staff embraced the system readily once they had been trained in its use. This can be explained by a shared interest in providing the highest level of care for the residents. However, resistance to the RPM system was also encountered. In this case, one of the GPs was uncooperative and this appeared to be as a result of DUHOXFWDQFHWRFKDQJHZRUNSUDFWLFHV6SHFL¿FDOO\ the GP refused to use a Web browser and insisted on the data being printed out and faxed. Good project management was essential. This LQFOXGHGKDYLQJFOHDUO\GH¿QHGSKDVHVRIIXQFWLRQDO XVHUUHTXLUHPHQWVDQGV\VWHPVSHFL¿FDWLRQSURcurement, implementation, training, and support %H\QRQ'DYLHV &RQÀLFWVZLWKWKHVXSSOLHUV were not encountered. This was attributed to the trust that had been formed between the technical director and supplier in the early stages of the project and had been built on an existing relationship.
Technical Issues The monitoring equipment chosen met most of the needs of the users. Overall, use was considered to be “easy” or “very easy” by all the community care home staff interviewed. This included setting up the monitor, attaching the sensors to the patient, taking a measurement, and transmitting the data. Data transmission was found to be unreliable on occasion, which was because of a number of wireless dead spots. This was resolved by extending the wireless network in the homes. Data access and presentation were considered acceptable, although suggestions for changes were made. Although the telemonitor was considered easy to use, some commented that it was cumbersome to move around due to its size and weight. We observed that the staff could quickly and easily become demoralized in use of the equipment folORZLQJWHFKQLFDOSUREOHPV,WFRXOGWDNHVLJQL¿FDQW effort to restore trust. We would advocate expend-
Remote Patient Monitoring in Residential Care Homes
ing effort early in a project to ensure reliability of equipment, training, and familiarizing users with the equipment and establishing good support mechanisms.
Comparison Against the Indicators for Success The project may be evaluated against the indicators for success of Wootton and Hebert (2001), as shown in Table 1. Five of the eight indicators for success were met. However, the project was designed as a pilot, and so was not integrated into routine health care or ¿QDQFLDOO\VXVWDLQDEOH,WVKRZHGJRRGDFFHSWDQFH and good clinical results, which were the objectives for the project.
Comparison of Sites The nursing home differed from the residential homes by having nursing staff. They were better able to recognize situations when monitoring could be of value. They were also often able to interpret the data and take responsibility for managing the condition with the support of the distant doctor. In contrast, the staff in the residential homes were
hesitant to make recordings and would delay contacting the doctor. In part, this could be explained by the cumbersome size of the equipment and reluctance to set it up in a resident room unless there was clear need. The role of the home would also affect usage. The nursing home, with its greater proportion of high dependency patients and nursing staff, had a greater need to use the system and opportunity for use was better recognized. This resulted in more than 70% of total use being carried out in the one nursing home.
BROADBAND AND RPM In any RPM application, there are several important aspects of the network characteristics that ZLOO LQÀXHQFH SHUIRUPDQFH DQG EHKDYLRU 0RVW important is bandwidth as this ultimately dictates which applications will be possible and the measure of their success. Furthermore, it is essential to consider the asymmetric bandwidth of most current implementations of broadband, as RPM applications generally will demand the greater bandwidth in the upstream direction, and where LW LV PRVW OLPLWHG RIWHQ E\ D VLJQL¿FDQW GHJUHH
Table 1. Indicators for success Indicator
Achieved
Routine operation
No
Comment The project was a pilot and did not become a routine part of health care delivery.
&OLQLFDOHI¿FDF\
Yes
Cost effectiveness
Not evaluated
$GHTXDWH¿QDQFLQJ QRVSHFLDOIXQGLQJ
No
arrangements required):
The project was funded by the European Union, and therefore relied on external funding.
Acceptance by clinicians
Yes
Acceptance by patients
Yes
Improved access to health care
Yes
Reduction of travel
Yes
613
Remote Patient Monitoring in Residential Care Homes
(e.g., 512 kbps upstream, 8 Mbps downstream). The high bandwidth of broadband will support well a large variety of RPM applications including not only simple data transmission but also streaming of vital signs, video conferencing, and good access to Web resources for the user. RPM applications may also need to be permanently connected with their server, so that the remote server PD\FRQ¿JXUHWKH530RUFKHFNLWVVWDWXVDWDOO times. Most RPM applications will tolerate delay and jitter in delivery of packets, except for video conferencing. All RPM applications will demand low error rates. Broadband provides excellent support for RPM over other residential connection methods such as dial up and delivers qualities that match well to the desired level of quality—it has inherent high speed, low delay, is reliable, and is continuously DYDLODEOHZLWK¿[HGFRVW,WZLOOVXSSRUWDOOW\SLFDO RPM applications, which have modest bandwidth needs that are well within the upstream bandwidth limitations, except for high quality video conference applications. One aspect of normal residential broadband needs to be discussed. Usually broadband will be implemented by providing the user with a single dynamically allocated IP address each time they connect to the ADSL service, with the address being retained for the duration of the connection. A NAT router will be installed to provide connection for any computers located in the premises. Each of these computers is allocated a “private” address that may not be directly connected from the public Internet; the computers may only connect outwards to the public. This is done to “share” the single IP address. The NAT router maintains a mapping between the internal IP address of the computer and the external public IP address whenever a connection is made. This mapping is retained as long as data is transmitted or received, but is lost if there is no activity for some period (necessary in case a computer should be switched off). Although the connection can be easily restored by the internal computer, the route from the remote server will be lost and cannot be restored. Steps therefore must be taken to ensure that the TCP/IP connection between the RPM and the distant server
614
is maintained against loss of connection through the reset of the route mapping in any NAT routers, HDVLO\DFFRPSOLVKHGE\NHHSDOLYHWUDI¿F Using the NAT router has the advantage of allowing many computers to be connected using the same broadband connection, within the limits of sharing the bandwidth between applications. Furthermore the internal private network may WKHQLQFRUSRUDWHDZLUHOHVV/$1WRDGGÀH[LELOLW\ and convenience and can be well protected from outside intrusion from the public Internet by the NAT router.
FUTURE TRENDS While early detection and intervention should prevent unnecessary admissions, treatment, and expense, at present there is still some debate over WKHFRVWEHQH¿WVWKDWFRXOGEHDFKLHYHGIURPXVing remote patient monitoring technology. While VRPHSXUSRUWVLJQL¿FDQWVDYLQJVWKURXJKWKHXVH of RPM, it has been suggested that 15% of all home visits in the UK could be replaced by RPM WHFKQRORJ\VDYLQJ
PLOOLRQLQWKH¿UVW\HDU (Celler et al., 1999). Others argue that the scale of the projects implemented to date—most projects have not continued past the initial pilot stage and have not been integrated fully into the healthcare system (May et al., 2003)—has meant that it has EHHQLPSRVVLEOHWRWUXO\DVVHVVWKHFRVWEHQH¿WV that such a service would achieve. This has created somewhat of a paradox, with health professionals reluctant to invest in a technology that has so far failed to prove its cost effectiveness. Yet for the technology to effect the cost savings, there is a need to reach economies of scale and a critical mass of users. In order to combat this, the UK government has recently put out to tender for three demonstrator projects which will cover a population base of over one million each. It is hoped that this will provide the evidence-base required for the cost effectiveness for RPM. However, despite this uncertainty, the need for RPM looks promising. It is to be expected that as use and experience increase, so will the capabilities of the systems. At present, current solutions
Remote Patient Monitoring in Residential Care Homes
provide the basic functionality to support simple management of chronic disease. Systems are wired in the home to the telephone and take a limited range of measurements (blood pressure, oxygen VDWXUDWLRQEORRGJOXFRVHZHLJKWSHDNÀRZ RQD daily basis. However, the increasing availability of extra functionality such as ECG, temperature, and continuous oxygen saturation will allow systems to monitor the same patient in the home during an acute phase of an illness, provided that the clinical infrastructure to support this is available. Wireless transmission of alarms and data would allow monitoring of patients with conditions such as angina, who, although relatively well, may deteriorate very rapidly. Such systems would support well the work in community care homes. Systems are getting more powerful by enabling UHPRWHUHDOWLPHFRQ¿JXUDWLRQE\WKHFOLQLFLDQVR that reliable, sophisticated alarms can be set and PRGL¿HGEDVHGRQWKHFRQGLWLRQRIWKHSDWLHQW'DWD could be recorded and sent automatically based on SUHFRQ¿JXUHGVFKHGXOHV7KLVPLJKWPHDQWKDW a system monitoring a person with angina would send an alarm if ST depression for last 10 minutes and 10 seconds of the ECG. If there was cause for FRQFHUQWKHFOLQLFLDQFRXOGFRQ¿JXUHWKHGHYLFH to send the ECG every few minutes until recovery was seen or action taken. Such systems do exist in prototype and pilot form. An example is the Telecare project (Clarke, 2004), which produced a system having all the above capabilities that GPRS to provide wireless connectivity for ambulatory monitoring. The system would also manage blood pressure and weighing scales for chronic disease, and fall detectors, smoke alarms for simple in home monitoring, so WKDWDVLQJOHXQL¿HGDUFKLWHFWXUHFRXOGVXSSRUWWKH full range of monitoring modalities. Work also must be undertaken on how these systems become integrated into mainstream health care delivery. This would include how and where data are managed, how sectors of health care inWHUUHODWHZKLFKSDWLHQWVPLJKWPRVWEHQH¿WDQG best strategies for use of equipment and medical intervention.
CONCLUSION This chapter has discussed acceptability and feasibility issues of an RPM system piloted in three community care homes in the UK. This included detailed consideration of clinical, organizational, and human factors as well as technical issues. A number of factors were found to be critical for the success of the system. The project demonstrated feasibility and was well accepted by patients and VWDII&OLQLFDOEHQH¿WVLQFOXGHGHDUO\GHWHFWLRQRI cardiac events, allowing prompt intervention, and routine monitoring of other conditions, particularly cardiac and respiratory chronic diseases. The monitoring system allowed staff to triage cases and determine the most appropriate action. The evaluation showed that the monitoring proved very effective in the population of a high dependency QXUVLQJ KRPH EXW VKRZHG IHZHU EHQH¿WV LQ WKH relatively well residents of a residential care home. 7KHQXUVLQJKRPHDOVREHQH¿WHGIURPKDYLQJD clinical champion. The new approach required increased communication and collaboration between different sectors of the health care system involved in the care of the residents of the homes. Closer relationships between the staff at the residential care homes and the staff at the health centers fostered a team approach that was important for successful use of the technology. At the same time, staff gained a greater sense of autonomy. Rather than immediately requesting a visit or calling an ambulance in case of a problem, they were able to do more for the patients. This did require that the staff of the community care homes learn to seek support when reporting and requesting assistance for a problem and that the staff at the health centers respond appropriately. Patient pathways may be used to analyze how to best redesign the system to support the new method of working. In this case, the telemonitor enables the movement of information rather than of the patient or health care professional and allows cases to be triaged remotely. The continuous involvement of users prior to and during the implementation of the system, as well as the reliance on existing relationships between the homes and the health
615
Remote Patient Monitoring in Residential Care Homes
centers, created a positive atmosphere between the participants in the project. This allowed a relatively smooth introduction of the system. Both the support from management of the homes and local clinical champions played an essential role in establishing the project, and its continued use. Wireless networks were key to the success of the pilot, as they allowed use of the telemonitors everywhere in the large old buildings of the care homes without the need for rewiring. Broadband enabled speedy data transmission. The equipPHQWFKRVHQZDVIRXQGWRODUJHO\PHHWWKHXVHUV¶ needs. The overall conclusion is that the wireless and broadband based RPM system offers much potential for improving and providing excellent health care delivery, both in terms of reducing unnecessary hospital admissions and expediting admission where indicated. It is expected that such mobile health care applications will become an integral part of health care once the current challenges have been resolved.
REFERENCES Aas, I.H.M. (2001). A qualitative study of the organizational consequences of telemedicine. Journal of Telemedicine and Telecare, 7, 18-26. Beynon-Davies, P. (1999). Human error and information systems failure: The case of the London ambulance service computer-aided dispatch system project. Interacting with Computers, 11, 699-720. Bratan, T., Clarke, M., Paul, R., & Jones, R. (2005). Evaluation of practical feasibility and acceptability of home monitoring in a residential home setting. Journal of Telemedicine and Telecare, 11(S1), 29-31. Bratten, R.L., & Cody, C. (2000). Telemedicine applications in primary care: A geriatric patient pilot project. 0D\R &OLQLFDO 3URFHHGLQJV 365-368. Brennan, S. (2005). The NHS IT project, the biggest computer program in the world…ever! Oxon: Radcliffe Publishing. 616
British National Statistics: Ageing. (2005). Retrieved July 21, 2007, from http://www.statistics. gov.uk./cci/nugget.asp?ID=949 Calkins, E., Boult, C., & Wagner, E. (1999) New ways to care for older people. Building systems based on evidence. New York: Springer. Celler, B.G., Lovell, N.H., & Chan, D.K.Y. (1999). The potential impact of home tele-care on clinical practice. Medical Journal of Australia, 171, 518-521 Clarke, M. (2004). A reference architecture for RPM. In L. Bos, S. Laxminarayan & A. Marsh (Eds.), Medical and care compunetics: Volume 103 studies in health technology and informatics (pp. 374-380). Amsterdam: IOS Press. Currie, G., & Brown, A.D. (1997, October-December) Implementation of an IT system in a hospital trust. Public Money and Management, 69-75. Demeris, G. (2004). Electronic home healthcare: Concepts and challenges. International Journal of Electronic Healthcare, 1(1), 4-16. Department of Health. (2001). The NHS plan: A progress report. Retrieved October 14, 2006, from http://www.dh.gov.uk/en/PublicationsAndStatistics/Publications/AnnualReports/DH_4064826 Department of Health. (2002). Developing the information systems – national service frameworks: A practical implementation in primary care (Home Page). Retrieved July 21, 2007, from http://www. dh.gov.uk/assetRoot/04/06/06/42/04060642.pdf Department of Health. (2006). Health reform in England; update and commissioning framework. Policy & Strategy Directorate, Department of Health. Retrieved July 21, 2007, from www.dh.gov. uk/assistroot/04/13/72/27/04127227.pdf European Commission. (2006). What is e-health? Information Society. Retrieved July 21, 2007, from http://europa.eu.int/information_society/eeurope/ ehealth/whatisehealth/index_en.htm Hailey, D. (2001). Some successes and limitations with telehealth in Canada. Journal of Telemedicine and Telecare, 7(S2), 73-75.
Remote Patient Monitoring in Residential Care Homes
Hersh, W., Helfand, M., Wallace, J., Kraemer, D., Patterson, P., & Shapiro, S. et al. (2002). A systemDWLFUHYLHZRIWKHHI¿FDF\RIWHOHPHGLFLQHJournal of Telemedicine and Telecare, 8, 197-209. Lyytinen, K., & Hirschheim, R. (1987) Information V\VWHPVIDLOXUH $VXUYH\DQGFODVVL¿FDWLRQRIWKH empirical literature. Oxford Surveys in Information Technology, 4, 257-309. Maheu, M., Whitten, P., & Allen, A. (2001). Ehealth, telehealth and telemedicine-a guide to start-up and success. San Francisco: Josey-Bass. May, C.R., Harrison, R., MacFarlane, A., Williams, T., Mair, F., & Wallace, P. (2003) Why do telemedicine systems fail to normalize as stable models of service delivery? Journal of Telemedicine and Telecare, 9(S1), 52-26. May, C.R., Williams, T.L., Mair, F.S., Mort, M.M., 6KDZ17 *DVN/ )DFWRUVLQÀXHQFLQJ the evaluation of telehealth interventions: Preliminary results from a qualitative study of evaluation projects in the UK. Journal of Telemedicine and Telecare, 8(S2), 65-67. Sauer, C. (1993). Why information systems fail: A case study approach. Henley-on-Thames: Alfred Waller Ltd. TeHIP. (2005). The impact of e-health and assistive technologies in health care. The E-Health Innovation Professionals Group of IHM, ASSIST and BCS HIF. Retrieved July 21, 2007, from www. health-informatics.org/tehip/tehipstudy.pdf
Wootton, R., & Hebert, M.A. (2001). What constitutes success in telehealth? Journal of Telemedicine and Telecare, 7(S2) 3-7.
KEY TERMS Community Care Home: Facility in which residential and sometimes nursing care is provided to disabled or elderly people. Chronic Disease: A disease that lasts for a long time (at least three months) and cannot be prevented or cured. E-Health: The use of emerging information and communication technology, especially the Internet, to improve or enable health and health care, thereby enabling stronger and more effective connections among patients, doctors, hospitals, payers, laboratories, pharmacies, and suppliers. ICTs: Information and communications technology, a broad subject concerned with technology and other aspects of managing and processing information. NHS:7KH8.¶V1DWLRQDO+HDOWK6HUYLFH Remote Patient Monitoring: The monitoring of physiological measurements in a setting other than a hospital, using ICTs to transfer data over geographical distances. Wireless Network: A computer network that is not connected by wires but by radio frequencies.
Wistow, G. (2000). Home care and the reshaping of acute hospitals in England – an overview of problems and possibilities. Journal of Management in Medicine, 14(1), 7-24.
617
Division III
Social Impact
619
Chapter XXXIX
Social Consequences of Broadband Access in Japan Kenichi Ishii University of Tsukuba, Japan
ABSTRACT In Japan, both the cheapest wired broadband services and the most advanced 3G mobile phone services are widely available. Because of recent procompetitive policy drives such as the “e-Japan policy,” the Japanese broadband market has become very competitive. While the digital divide has narrowed in recent years in terms of Internet access, a divide still exists with regard to Internet usage. Comparison between narrowband and broadband users demonstrates that broadband services currently are used mainly for entertainment. Unlike wired Internet use, mobile Internet is not used for information-gathering activities. Results do not support the media substitution effect of the Internet. Mobile Internet use VLJQL¿FDQWO\DQGSRVLWLYHO\FRUUHODWHVZLWKVRFLDOL]LQJZLWKIULHQGVZKHUHDVWKHZLUHG,QWHUQHWXVHGRHV not correlate with socializing. Experience of past policies suggests that customer orientation will be a key factor in the success of the “U-Japan” policy.
INTRODUCTION
BACKGROUND
Transition to broadband access has been rapid since 2000. As of February 2006, the household penetration rate of broadband access reached 41%, and approximately 70% of Internet users access the Internet via a broadband connection. The ,QWHUQDWLRQDO7HOHFRPPXQLFDWLRQV8QLRQGH¿QHV broadband as a technology providing transmission capacity that is faster than primary rate ISDN (1.5 or 2 Mbps) (Distaso, Lupi, & Manenti, 2006). AcFRUGLQJWRWKLVGH¿QLWLRQEURDGEDQGFRQQHFWLRQLV widely available via 3G (third-generation) mobile phones in Japan.
Wired (PC-Based) Internet ,QRI-DSDQ¶VZLUHGEURDGEDQG 3&EDVHG broadband) users, 62.3% had access via xDSL, 14.2% had cable Internet, 23.4% had FTTH ¿EHUWRWKHKRPH DQGKDG):$ )LJXUH Japanese people enjoy the highest-speed broadband services in the world. The International Telecommunications Union reports that, in 2004, Japanese consumers paid the smallest charge per 100 kilobits per second in the world (U.S. $0.06), compared to that paid by South Koreans (U.S. $0.77), Americans (U.S. $1.77), Germans (U.S. $2.77) and the Brit-
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Social Consequences of Broadband Access in Japan
Figure 1. Number of subscribers to wired broadband services in Japan (in millions) 0
10
20
30
2000
xDSL
2001 2002
Cable Internet
2003
FTTH
2004 2005 2006
Source: Ministry of Internal Affairs and Communications (2006)
ish (U.S. $6.18) (Ministry of Internal Affairs and Communications, 2005). For example, a 100 Mbps FTTH connection service with ISP, IP, phone, and VOD services for an apartment is provided for 7024 yen (U.S. $61; $1 = 115 yen) per month by Nippon Telegraph and Telephone Corporation (NTT). As a result of low charges, FTTH subscribers recently surpassed DSL subscribers. In March 2006, the number of FTTH subscribers reached 5.4 million. During the last three months of 2005, FTTH services saw a jump in subscriber numbers, which reached 658,000, almost three times as much as DSL (175,000).
Mobile Internet In addition to the cheapest wired broadband services, Japan also offers unique and advanced mobile Internet services. Figure 2 demonstrates that mobile phones outnumber PCs in the market for Internet-connected devices. Internet-enabled mobile phones accounted for 94% of all mobile phones, as of September 2004 (Ministry of Internal Affairs and Communications, 2005). Mobile Internet services, such as picture/video mail and picture characters in e-mail, are commonly used by the Japanese people (Ishii, 2004). In terms of business model, “i-mode”—the Web access proWRFRORQ177'R&R0R¶VWHUPLQDOV²LVRQHRIWKH
620
most successful mobile Internet business models in the worldwide telecom market. This service, launched in Japan in February 1999, attracted more than 33 million users three years after its launch (Ishii, 2004). In the mobile Internet market, mobile phone carriers play a dominant role because the carriers are the only ISPs in the mobile Internet. The carriers operate the major portal Web sites, FDOOHG³RI¿FLDOVLWHV´ZKLFKDUHOLQNHGGLUHFWO\WR WKH,63¶VPHQXSDJH&DUULHUVKDQGOHWKHELOOLQJ of those charges on behalf of the content provider (Ishii & Ogasahara, 2006). 3G mobile phones (cdma-one/W-CDMA/CDMA2000x) comprised 70% of all mobile phones in January 2007 (TCA, 2007). These mobile phone services enable connections that are as quick as the wired broadband services. Currently, NTT 'R&R0R -DSDQ¶V WRS PRELOH SKRQH RSHUDWRU provides a W-CDMA service with a maximum data transmission of 3.6 Mbps, while the second-largest operator, KDDI, provides cdma-one services with a maximum data transmission of 2.4 0ESV.'',¶V(=&KDQQHOVHUYLFHRIIHUVYLGHR content—including news, weather, movies, and music information—via push technology (Ishii & Ogasahara, 2006). In this chapter, the term “mobile Internet” will be used to refer to Internet access via mobile phones, and not to the wireless LAN access, because wireless LAN usually is connected to wired broadband only in a limited area.
Social Consequences of Broadband Access in Japan
Figure 2. Number of Internet-connected devices in Japan (in millions) 0 2001 2002
50 37
100
PC
24
49
150
25
2003
57
2004
62
2005
64
2006
66
Mobile phone
28 45 58 69
Others (game terminal, TV etc.)
Note: As of March, every year. Source: Ministry of Internal Affairs and Communications (2006).
Broadband Policy in Japan -DSDQ¶V LQIRUPDWLRQ FRPPXQLFDWLRQ WHFKQRORJ\ (ICT) policies underwent a fundamental change around the year 2000. Since the Japanese government declared in a white paper that the year 2002 was the “First Year of Broadband,” lower charges and faster access resulting from market-oriented ICT policies have caused the number of broadband users to grow rapidly. Following its e-Japan policy, the government proposed a new ICT plan called the U-Japan policy in 2004 (see Key Terms). In this new plan, “ubiquitous networks” are of particular importance. Ubiquitous networks enable easy person-to-person, person-to-goods, and goodsto-goods communication (Ministry of Internal Affairs and Communications, 2005). The e-Japan policy aims to develop a high-speed broadband infrastructure, while the U-Japan policy puts more emphasis on the usability of the system.
Local-Oriented Policy to Nation-Oriented Policy Previously, Japanese ICT policies aimed to create network infrastructure in local areas (Tsuda, 2005). These policies were expected to bridge the economic gap between urban and rural areas. For
example, in 1983, the Ministry of Post and Telecommunications (MPT) proposed the Teletopia Plan for the development of regional information and telecommunications infrastructure across the country using CATV and telecommunications PHGLD 037 )ROORZLQJWKH037¶VPRYH several government organizations proposed similar plans for local ICT projects. In the 1990s, these policies were implemented as economic stimulus measures, such as “public works spending,” which usually refer to the building of roads, dams, and railways. However, these projects were mostly unsuccessful. According to an MPT panel, many projects focused too much on hardware and too little on application software, and the projects were QRWVDWLVIDFWRU\LQVHUYLQJSHRSOH¶VQHHGVLQWKHLU respective regions. Tsuda (2005) pointed out that the local ICT projects did not necessarily lead to the revitalization of local communities. Another disadvantage of the locally-oriented projects was that they were not a part of a systematic nationwide strategy. Many regions worked separately on parallel versions of the local ICT projects, thus PDNLQJLWGLI¿FXOWWRDFKLHYHHFRQRPLHVRIVFDOH around the country (Ishii, 2003). For example, CATV operators were allowed to operate only in a ORFDOFRPPXQLW\EDVHGRQWKHJRYHUQPHQW¶VORFDO
621
Social Consequences of Broadband Access in Japan
community-oriented strategy; thus, most of the RSHUDWRUVZHUHVPDOODQG¿QDQFLDOO\ZHDN The government recently issued a critical advisory on local ICT projects, stating that they should WDUJHWPRUHVSHFL¿HGSXUSRVHVDQGWKDWGLIIHUHQW organizations should coordinate their projects. As a result of policy changes, the budget for local ,&7SROLFLHVGHFUHDVHGWRELOOLRQ\HQLQ¿VFDO year 2001 and to 15 billion yen in 2004, compared with 190 billion yen in 2000 (The Administrative Evaluation and Inspection Bureau, 2006).
Procompetitive Policy The introduction of local loop unbundling helped to enhance competition among telecommunications operators. The government ordered NTT, the dominant operator in Japan, to cut the price DWZKLFKLWRIIHUVLWVLQIUDVWUXFWXUHIRU¿EHURSWLFV to third parties in 2000. The unbundling policy prompted new operators to enter the market. For example, Yahoo BB entered the market with a vertical convergent business model that combines paid content with cheap connection. As of March 2005, the major operator, NTT, accounted for only 38% and 58% of the market for DSL and FTTH services, respectively (InfoCom Research, 2005). Despite its monopolistic status in the telecommunications market in the 1990s, NTT now holds a lower market share compared with telecommunications operators in other countries. The degree of intra and interplatform competiWLRQLVPHDVXUHGXVLQJWKH+HU¿QGKDO+LUVFKPDQ LQGH[ ++, ZKLFK LV GH¿QHG DV WKH VXP RI D ¿UP¶VVTXDUHGPDUNHWVL]H 'LVWDVRHWDO According to Distaso et al., the HHI index for the DSL market ranged from 0.51 to 0.82 (the median is 0.605) among 14 EU countries in 2004. In contrast, WKH++,LQGH[IRU-DSDQ¶VEURDGEDQGPDUNHWLVRQO\ 0.295 and 0.400 for DSL and FTTH, respectively.1 The HHI index for DSL service was as low as 0.272 in September of 2002. These low values indicate that the Japanese broadband market is much more competitive when compared with other countries, such as those in the EU.
622
LITERATURE REVIEW AND RESEARCH QUESTIONS Many studies examine the adoption of new media E\FRQVXPHUVSULPDULO\EDVHGRQ5RJHUV¶ diffusion of innovation paradigm (DiMaggio, Hargittai, Neuman, & Robinson, 2001; Hashimoto, Tsuji, Fukuda, Mori, & Yanagisawa, 1996; Jeffres & Atkin, 1996; Wei, 2006; Zhu & He, 2002). These VWXGLHVVHHNWR¿QGWKHFKDUDFWHULVWLFVRILQGLYLGXDOV that make them likely to adopt the Internet. Early adopters of the new information technologies are more likely to be male, wealthier, better educated, and younger than nonadopters (Ishii, 1996; Rogers, 2003; Wei, 2006; World Internet Project Japan, 2002). Differential access to and use of the Internet and new media in general, according to gender, income, race, and location has been cited as the digital divide (DiMaggio et al., 2001; Rice & Katz, 2003; van Dijk, 2006). Some argue that the digital divide remains substantial within some countries and is widening in some countries (Chen & Wellman, 2004; Hushing & Selhofer, 2004; Kubicek, 2003; Martin & Robinson, 2004; Parker, 2000), while others argue that the gap has been closing (U.S. Department of Commerce, 2002). It is widely believed that the broadband network will displace existing media content, such as Web casting over the Internet. The media substitution hypothesis has been tested in many previous studies. Many studies show that wired Internet use has something in common with the time-displacing technology of TV (Ishii, 2004; Robinson, Kestbaum, Neustadl, & Alvarez, 2000). Some studies reported a negative correlation between Internet use and television viewing (Pronovost, 2002; Robinson, Kestbaum, Neustadl, & Alvarez, 2002) while few reported a positive correlation (Rhee .LP 2WKHUVWXGLHVIRXQGQRVLJQL¿FDQW correlations between TV viewing and Internet use after demographic controls (DiMaggio et al., 2001; Gershuny, 2002; Haan & Huysmans, 2002). Many studies have examined the effects of the Internet on social capital. One concern is that the possibly solitary activity engendered by the Internet may displace time formerly spent on local social relations (Haythornthwaite & Wellman, 2002). A
Social Consequences of Broadband Access in Japan
study shows that greater use of the Internet was DVVRFLDWHGZLWKDGHFOLQHLQXVHUV¶FRPPXQLFDWLRQ with family members, and a decline in the size of their social circle (Kraut, Lundmark, Patterson, Kiesler, Mukopadhyay, & Scherlis, 1998). The more time that users spent on the Internet, the less time they shared at meals and watching TV with family members (Weng, 2000). However, the opposite ¿QGLQJV DOVR KDYH EHHQ UHSRUWHG +HDY\ HPDLO users have more social ties than do light e-mail users (Zhao, 2006). In Japan, many studies reported that mobile media users are more active in personal communications. Ishii (2004) documented that the PRELOH,QWHUQHWKDVDVLJQL¿FDQWO\SRVLWLYHHIIHFW on the amount of time spent with friends, whereas ZLUHG,QWHUQHWKDVDVLJQL¿FDQWO\SRVLWLYHHIIHFW on time spent with family. As compared with PCs, the mobile phone is considered to overcome the inequitable distribution of power that stems from differential access to ICT resources (Rice & Katz, 2003). However, only a few studies have examined the diffusion of mobile phones, especially Internet access over the mobile phone (Ishii, 2006; Katz & Aakhus, 2001). Based on the above brief review, the purpose of this chapter is to examine the social consequences of broadband access. Despite initial enthusiastic acceptance of the Internet, its social consequences KDYHQRWEHHQFODUL¿HG,QWKLVFKDSWHUZHLQWHQG to examine factors accounting for the adoption of broadband, and then to discuss the social consequences of the introduction of broadband services by comparing wired and mobile Internet. More VSHFL¿FDOO\WKLVFKDSWHUIRFXVHVRQWKHGLJLWDOGLvide, because many ICT policies (e.g., the e-Japan policy) aimed to close the digital divide. Additionally, this chapter discusses the future potential of the mobile Internet, in which Japan seeks to take a leading role. In sum, the research questions of this study are as follows:
METHOD
RQ1: To what extent are there gaps in wired and mobile Internet use in Japan? RQ2:What factors account for these gaps? RQ3: What are the social consequences of wired and mobile broadband services in Japan?
Digital Divides in Japan
This study uses data from the following questionnaire survey in its analyses. The World Internet Project Japan (JWIP) survey was conducted nationwide in Japan in 2000, 2001, 2002, 2003, and 2005, in cooperation with the World Internet Project (WIP), in which many countries are involved. Throughout this project, the author has participated in the research project (JWIP) headed by Prof. Shunji Mikami. This study will use two waves of this survey, years 2003 (World Internet Project Japan, 2003) and 2005 (World Internet Project Japan, 2005). The data collection method was a self-administered questionnaire. The survey was conducted under the auspices of the Communications Research Laboratory. The respondents were chosen from a national probability sample whose ages ranged from 12 to 74 for the JWIP2003 survey and 13 to 70 for the JWIP2005 survey. A questionnaire sheet was distributed by a research company and collected at a later date. The number of successful respondents was 1,520 (response rate: 69.1%) and 2,029 (response rate: 68.6%), respectively. The time diary method (time-budget survey) was adopted in 2003 and 2005 to measure the amount of time spent on a variety of information activities, such as viewing TV, reading newspapers, books, and magazines, and using the wired Internet and the mobile Internet. Respondents were requested to record basic activity codes, location, and information activities for every 15 minutes for 48 hours on two consecutive days. Details of the methods are documented in Suzuki, Hashimoto, and Ishii (1997) and World Internet Project Japan (2005).
MAIN FOCUS OF THE CHAPTER
Table 1 demonstrates that in the case of wired (PC) Internet, access gaps are found across various demographic factors, including gender, age, income, job, and region. Likewise, mobile Internet
623
Social Consequences of Broadband Access in Japan
Table 1. Internet penetration rates by demographic factors Variable
Gender
Values
N
Male
925
Female
1074
Chi-square statistic Age
47.8
37.4
34.1
49.1
38.7(***)
27.5(***)
50.5
69.7
20–29
226
55.3
72.6
30–39
399
53.9
64.9
40–49
362
50.6
40.3
50–59
436
28.9
24.1
60–74
368
14.7
14.7
200.3(***)
402.7(***)
Junior high school
230
20.4
20.0
High school
1028
30.5
42.2
College
321
48.0
60.4
University
414
70.8
47.3
245.5(***)
92.2(***)
Less than 2 million yen
154
21.4
29.2
annual
2-4 million
481
26.6
38.5
income
4-6 million
499
39.3
44.7
6-8 million
352
51.1
45.2
8-10 million
188
53.7
49.5
10-12 million
85
56.5
48.2
More than 12 million
80
62.5
53.8
117.6(***)
25.2(***)
Chi-square statistic Job
Full-time
925
50.2
40.9
Part-time
355
30.1
45.6
Housewife
321
24.9
41.7
Student
228
54.8
72.8
Unemployed
170
18.8
19.4
Chi-square statistic Region
136.6(***)
123.4(***)
1 million or more
474
45.5
44.7
100 thousand or more
784
40.3
46.7
Less than 100 thousand
376
38.8
39.6
County
365
35.9
40.0
8.3(*)
7.6(n.s)
Chi-square statistic
1RWH S
S
S QVS! 6RXUFH -:,3
624
rate (%)
208
Chi-square statistic Household
Mobile Internet
(%)
12–19
Chi-square statistic Education
PC Internet rate
Social Consequences of Broadband Access in Japan
7DEOH*LQLFRHI¿FLHQWVIRU,QWHUQHWSHQHWUDWLRQUDWHVE\\HDU Factor/year
2000
2001
2002
2003
2005
Age
0.308
0.266
0.242
0.241
0.204
Income
0.189
0.155
0.153
0.158
0.125
Education
0.291
0.255
0.284
0.198
0.172
Region*
0.090
0.030
0.072
0.064
0.038
* Region is categorized as (1) cities with over 1 million population, (2) cities with 100 thousand through PLOOLRQ RWKHUFLWLHVDQG WRZQVDQGYLOODJHV6RXUFH ,VKLL Table 3. Regression models predicting PC Internet use and amount of time spent on PC Internet PC Internet use
Time spent on PC Internet a
(logistic regression;
(regression; n = 707)
n = 1732) Standardized
Wald chi-
Standardized
parameter
square
parameter
t value
Gender (M = 1, F = 2)
-0.10
8.1**
-0.15
-3.5***
Age
-0.39
132.4***
-0.19
-4.7***
Education
0.37
116.1***
0.06
1.6
Job status
0.17
19.5***
-0.03
-0.7
0.20
40.2***
-0.01
-0.3
0.08
7.2**
0.04
1.0
(full-time = 1, else = 0) Income (10 thousand yen) Region b
Note: (a) Only PC Internet users are considered; (b) City with more than 1 million population = 1, others = 0.
S
S
S 6RXUFH -:,3
has similar access gaps across these demographic factors.
Closing Gaps in Internet Penetration The Gini index has been used widely to measure the extent of gaps across the social strata. The *LQLFRHI¿FLHQWUDQJHVIURPWKURXJKZKHUH 0 represents complete equality (no gap) regarding the target variable, while 1 represents complete inequality (the largest gap). To measure gaps in ,QWHUQHW SHQHWUDWLRQ ,QWHUQHW DFFHVV LV GH¿QHG as either wired or mobile Internet access. In Ja-
pan, gaps among age and education groups are UHODWLYHO\ODUJHDVUHÀHFWHGE\-DSDQ¶VFXOWXUDOO\ and economically homogeneous society. Table 2 demonstrates that almost all gaps in Internet access have narrowed over time. These results demonstrate that Internet access gaps based on age, income, region, and education have narrowed in the past 5 years. The regional gap is especially small as compared with other demographic factors. As discussed previously, the Japanese ICT policy in the 1990s targeted reduction of regional gaps. Although the present ICT policy still supports ICT development in rural areas, the mainstream policy
625
Social Consequences of Broadband Access in Japan
Table 4. Regression models predicting mobile Internet use and amount of time spent on mobile Internet Mobile Internet use
Time spent on mobile Internet a (regression;
(logistic regression;
n = 701)
n = 1732) Standardized
Wald chi-
Standardized
parameter
square
parameter
t value
Gender (M = 1, F = 2)
0.25
49.8***
-0.09
-2.1*
Age
-0.51
230.1***
-0.23
-5.8**
Education
0.21
38.7***
0.02
0.49
Job status
0.06
2.7
-0.06
-1.5
0.03
0.7
0.03
0.9
-0.02
0.4
0.05
1.4
(full-time = 1, else = 0) Income (10 thousand yen) Region b
Note: (a) Only PC Internet users are considered; (b) City with more than 1 million population = 1, others = 0.
S
S
S 6RXUFH -:,3
has shifted direction; where it was locally-oriented, it is now competition-induced (Ishii, 2003). These UHVXOWVVKRZWKDWWKH¿UVWW\SHRIJDS WKHDFFHVV gap) has tended to close in recent years, despite WKHLQWURGXFWLRQRIWKHJRYHUQPHQW¶VFRPSHWLWLYH ICT policies.
The Second-Level Digital Gap Many studies have looked at the digital divide only in terms of Internet access and have ignored Internet use (Peter & Valkenburg, 2006). Yet, the closing of access gaps does not mean that everyone EHQH¿WVHTXDOO\IURPWKHQHZWHFKQRORJLHV:H need to consider the second-level digital gap—the usage gap—apart from the access gap. In order to measure how these two gaps are associated with social status, the use of wired Internet and the amount of time spent on PC Internet are regressed on demographic factors. Table 3 demonstrates the estimated parameters of the regression models predicting access to the Internet, and the amount of time spent using the Internet; in the second
626
model, only users of the Internet are considered. The results of the logistic regression model show that all of the six variables—gender, age, education, ZRUNVWDWXVLQFRPHDQGUHJLRQ²DUHVLJQL¿FDQWO\ correlated with the Internet access, while the second regression model shows that second-level gaps H[LVWHYHQDPRQJXVHUV0RUHVSHFL¿FDOO\PDOHV and younger people are more likely to access the Internet and use it for longer durations. In other ZRUGV JHQGHU DQG DJH DIIHFW ERWK WKH ¿UVW DQG second levels of the digital divide in similar ways. The second digital divide suggests that new techQRORJLHVSURPRWHGE\WKHJRYHUQPHQW¶VSROLFLHV do not necessarily lead to effective use. Finance support from the government may boost adoption of the Internet, but it will not necessarily result in effective use of the Internet. The same regression models are applied to the mobile Internet (Table 4). Unlike the wired Internet, females are more likely to adopt the mobile Internet. Job status, income level, and region are QRWVLJQL¿FDQWO\FRUUHODWHGZLWKPRELOH,QWHUQHW use. In this sense, the digital divide is smaller in
Social Consequences of Broadband Access in Japan
7DEOH5HJUHVVLRQPRGHOVSUHGLFWLQJDPRXQWRIWLPHVSHQWRQYLHZLQJ79UHDGLQJQHZVSDSHUVDQG magazines TV
Newspapers
Magazines
Parameters
t value
Parameters
t value
Parameters
t value
Intercept
215.87
10.2
-28.38
-5.6
5.45
2.1
Time on Internet
-0.01
-0.2
0.03
2.0*
0.00
0.4
2.04
10.2***
0.99
20.6***
-0.05
-2.1*
-5.84
-3.9***
0.89
2.5*
0.21
1.2
PC a Age Education Income
b
Job status c
-0.03
-2.9**
0.00
1.5
0.00
-1.1
-67.55
-10.8***
-4.39
-2.9**
-2.38
-3.2**
Note: (a) minute; (b) yearly household income (million yen); (c) full time job= 1, else = 0.
S
S
S 6RXUFH -:,3
the mobile Internet than in the wired Internet. This difference between wired and mobile Internets suggests that the mobile Internet may bridge the digital divide in the wired Internet. The implications of these results will be discussed later in this chapter.
Social Consequences of Internet Access Closing the digital divide is not the ultimate goal RI WKH JRYHUQPHQW¶V ,&7 SROLF\ 7KH XOWLPDWH JRDOLVWRLQFUHDVHVRFLDOEHQH¿WVWKURXJKXVHRI WKH,QWHUQHW+HQFHWKHLQÀXHQFHRI,QWHUQHWXVH should be measured appropriately. To this end, this chapter will make two comparisons—between users and non-users and between broadband and narrowband users.
Users and Nonusers: Media Substitution Hypothesis Technological enthusiasts might simply expect that WKH,QWHUQHWPRUHVSHFL¿FDOO\WKHZLUHG,QWHUQHW will displace other media (media substitution hypothesis), such as TV. We test this hypothesis by comparing media usage patterns between wired
Internet users and nonusers, controlling for main demographic factors. During the two-day survey period, 30.5% of the respondents used the Internet at least once via PCs, with an average of 22.2 minutes of wired Internet usage per day. To assess the effects of wired Internet usage on the use of television, newspapers, and magazines, the regression model was employed. Table 5 demonstrates that ZLUHG,QWHUQHWXVHKDVQRVLJQL¿FDQWHIIHFWRQWKH amount of time spent on these media. Regression FRHI¿FLHQWVLQGLFDWHWKDWWKHDPRXQWRIWLPHVSHQW RQ ZLUHG ,QWHUQHW LV QRW VLJQL¿FDQWO\ FRUUHODWHG with the amount of time spent on other media, after controlling for age, education, income, and job. Contrary to the hypothesis, the wired Internet users spent more time reading newspapers than nonusers. The regression parameter suggests that WKH XVH RI ZLUHG ,QWHUQHW OHDGV WR VLJQL¿FDQWO\ increased newspaper reading. These results show that wired Internet use does not conform to the media substitution hypothesis.
Effects of the Internet on Social Capital There are utopian and dystopian views on the impacts of the Internet on social capital. To estimate the effects of the Internet on social capital, the
627
Social Consequences of Broadband Access in Japan
Table 6. Regression models predicting amount of time with family and friends Amount of time
Amount of time
spent with family
spent with friends
Standardized
t value
parameter Gender (M = 1, F = 2)
Standardized
t value
parameter
0.04
1.18
-0.07
-2.1*
Age
0.07
1.98*
-0.05
-1.35
Education (years)
-0.01
-0.20
-0.07
-2.52*
Job status (full-time = 1, else = 0)
-0.20
-5.76***
-0.04
-1.08
Student (student=1, else=0)
-0.10
-2.82**
0.44
11.6***
Family income (one million yen)
-0.07
-2.38*
-0.45
-1.63
Wired Internet use
0.01
0.29
-0.03
-1.19
Mobile Internet use
-0.01
-0.39
0.12
4.39***
R-Square
0.074
0.08
N
1183
969
S
S 6RXUFH -:,3 Table 7. Logistic regression model predicting broadband subscription vs. narrowband subscription Standardized parameter
Wald chi-square
Gender (M = 1, F = 2)
-0.121
3.85*
Age
-0.045
0.65
Education
0.080
1.95
Work (full-time = 1, else = 0)
-0.140
4.87*
Income (10 thousand yen)
0.023
0.16
0.010
0.03
Region
a
1RWH 1 $SRVLWLYHSDUDPHWHUPHDQVDSRVLWLYHHIIHFWRQEURDGEDQGVXEVFULSWLRQ D &LW\ZLWK more than 1 million population = 1, others = 0.
S 6RXUFH -:,3
amounts of time spent with family and friends are regressed on wired Internet use and mobile Internet use, controlling for demographic factors. Table 6 LQGLFDWHVWKDWRQO\PRELOH,QWHUQHWLVVLJQL¿FDQWO\ associated with the amount of time spent with friends. This suggests that the mobile Internet use facilitates socializing with friends, although one cannot conclude a causal relationship.
628
Broadband Users vs. Narrowband Users Japanese ICT policies have aimed to shift users from narrowband to broadband access. Hence, we compare usage patterns of broadband and narrowband users based on survey results. In the following DQDO\VLVEURDGEDQGXVHUVDUHGH¿QHGDVXVHUVRI wired Internet through FTTH, DSL, or CATV at KRPHZKLOHQDUURZEDQGXVHUVDUHGH¿QHGDVXVHUV of wired Internet through dial-up or ISDN at home.
Social Consequences of Broadband Access in Japan
Table 8. Comparison of penetration rate (%) of Internet services between narrowband and broadband users Narrowband users
Broadband users
Chi-square
6LJQL¿FDQFHOHYHO
Viewing Web sites for entertainment
73.0
81.9
5.4
*
Getting information or making travel
65.2
69.1
0.8
Using health/medical information
56.5
60.2
0.7
Studying for school
28.3
33.3
1.4
Working at home
42.2
43.8
0.1
Getting news
54.8
68.7
9.8
Viewing Web sites about religion
1.7
2.0
0.0
Playing online game
18.7
25.3
3.0
Activities via the Internet
registrations
Using online banking
6.1
17.7
15.1
Searching job information
12.6
15.7
0.9
Chatting
6.1
13.7
7.6
**
***
**
Joining BBS/news group
11.3
26.1
17.0
***
Using online shopping
35.2
52.2
14.0
***
Using Internet auction
20.4
26.5
2.4
Downloading documents or data
53.5
66.3
8.2
**
Downloading music or ring-tone
21.7
29.7
4.0
*
Playing videos
24.8
44.6
20.6
***
Streaming video
7.8
26.1
27.9
***
Using matching Web sites
1.3
0.4
1.2
Using adult Web sites
3.9
11.2
9.0
N
230
249
**
1RWH S
S
S 6RXUFH -:,3
To assess the effects of factors that promote the shift from narrowband to broadband, the logistic regression model is employed. Table 7 denotes that males are more likely to shift services from narrowband to broadband. With regard to regional JDSVQRVWDWLVWLFDOO\VLJQL¿FDQWGLIIHUHQFHLVIRXQG between rural and urban residents. An unexpected ¿QGLQJLVWKDWRWKHUWKLQJVEHLQJHTXDOIXOOWLPH workers are less likely to change from narrowband service to broadband. This suggests that only few people access the Internet at home through broadband connection for business use. As expected, broadband users are more active online than narrowband users. Broadband users spend 29.8 minutes per day on wired Internet;
narrowband users spend only 11.9 minutes on wired Internet (t = 3.24, DF = 447, p < .01). In addition, broadband users employ the Internet rather differently than narrowband users. For example, 52% of broadband users engage in online shopping, compared with 36% of narrowband users (chi-square = 13.4, DF = 1, p < .001). Broadband users spend 23.8 thousand yen per year on online shopping, while narrowband users spend 12.1 thousand yen (t = -1.98, DF = 475, p < .1). 45% of broadband users view videos via the Internet, while only 25% of narrowband users view videos (chi-square = 20.6, DF = 1, p < .001). Table 7 also indicates that broadband users generally are more active online than narrowband users when it comes
629
Social Consequences of Broadband Access in Japan
Table 9. Regression models predicting the perceived importance of the Internet Perceived importance as an information
Perceived importance as an entertainment
source
medium
Standardized
t value
parameter
Standardized
t value
parameter
Connection type a
0.08
1.56
0.10
1.98*
Gender (M = 1, F = 2)
-0.00
-0.00
0.08
1.44
Age
-0.14
-2.97**
-0.16
-3.17**
Education
0.17
3.30**
0.01
0.23
Job status
-0.01
-0.1
0.02
0.29
-0.07
-1.4
-0.05
-0.96
(full-time = 1, else = 0) Income (10 thousand yen)
Note: (a) broadband = 1, narrowband = 0.
S
S 6RXUFH -:,3 Table 10. Access rates of information via PC/mobile internet (%) Via PC
Via mobile phone
Economic/business news
24.83
1.33
Weather forecast
35.68
16.30
Tour information
46.14
4.19
Shopping/products information
45.90
5.76
Health/hospital information
27.96
1.12
TV program information
20.31
1.77
Music information
35.41
7.80
Restaurant information
33.26
4.21
Note: Only users of both mobile and PC Internet are considered. 6RXUFH -:,3
to viewing Web sites for entertainment, getting news, using online banking, using BBS (bulletin board system), chatting, downloading documents and music, playing and streaming videos. These results suggest that the broadband access promotes some activities that require high connection speed (e.g., downloading and streaming). At the same time, we should note that broadband connection does not promote all online activities, especially those that are more practical (e.g., searching job information and working at home).
630
To compare the purposes of using the Internet, the perceived usefulness of the Internet2 is regressed RQWKH,QWHUQHWFRQQHFWLRQZKLFKLVGH¿QHGDVD dichotomous variable (narrowband = 0, broadband FRQWUROOLQJ IRU UHVSRQGHQW¶V GHPRJUDSKLF factors. Table 9 demonstrates the estimated parameters. Younger people perceive the Internet to be more useful, either as an information medium or as an entertainment medium. As compared to narrowband users, broadband users perceive the Internet to be more important as an entertainment
Social Consequences of Broadband Access in Japan
Table 11. Regression model predicting the information-handling competency index Standardized parameter
t value
Gender(M = 1, F = 2)
-0.030
-1.190
Age
0.009
0.340
Education (years)
0.088
3.540***
Job status
0.013
0.510
0.045
1.970*
(full-time = 1, else = 0) Income Region
a
Key-in ability PC Internet use b Mobile Internet use
b
R-square
-0.046
-2.060*
-0.314
-11.470***
0.069
2.890**
0.037
1.560 0.166
Note: (a) City with more than 1 million population = 1, others = 0. b. minute.
S
S
S 6RXUFH -:,3
medium (t = 1.98, DF = 1, p < .05), whereas no VWDWLVWLFDOO\VLJQL¿FDQWGLIIHUHQFHLVIRXQGLQWKH perceived importance of the Internet as a source of information. These results suggest that the purpose of broadband access is entertainment rather than information-gathering.
Mobile Phones as Potential Alternative to PCs Unlike the youth in other countries, Japanese youth are not very interested in using the Internet via PC (Ishii & Wu, 2006). Table 1 indicates that younger people (14-39 years) more often use mobile phones to access the Internet, while older people (40-59 years) more often use PCs. Despite its similarity to wireless LAN (WLAN), Japanese mobile Internet users differ greatly from WLAN users, who are younger, better educated, and more often male than non-users (Wei, 2006). Table 4 also shows that access gaps in mobile Internet are smaller than in wired Internet. Such a unique usage pattern suggests that Japanese ICT development may take a different route from that in other countries. Will the mobile phone displace PCs? To answer this question, we should bear in mind that the two technologies are used quite differently. Table 10
demonstrates that fewer users engage in information-gathering via mobile phones, compared with that through PCs. Despite a high usage rate (94%) of Internet-enabled mobile phones in Japan, few take advantage of the advanced Internet services, except for such personal communications as text messaging and photo e-mail. Mobile phones are used mostly for communication with close friends to maintain existing ties, rather than to create new ones (Ishii, 2006). The “information handling competency index”3 (IHC index) is a scale to measure the competency level of information-gathering activities in everyday life. To examine the effects of mobile Internet and wired Internet use on the competency level of information-gathering activities, the score of this index is regressed on the amount of time spent on wired and mobile Internets, controlling for demographic factors. Table 11 indicates that mobile ,QWHUQHWXVHLVQRWVLJQL¿FDQWO\DVVRFLDWHGZLWKD higher level of information handling competency. In FRQWUDVWZLUHG,QWHUQHWXVHLVVLJQL¿FDQWO\DVVRFLated with a higher level of information-gathering activities (t = 2.89, p < .01). These results demonstrate that, compared with wired Internet, mobile Internet use is not associated with a higher level of information-gathering competency.
631
Social Consequences of Broadband Access in Japan
FUTURE TRENDS ,WLVGLI¿FXOWWRSUHGLFWWKHIXWXUHEDVHGRQVXUYH\ results. Our survey suggests that wired Internet will become increasingly advanced, at least as an entertainment medium. For example, Web casting will become more popular for entertainment use services (Ha & Ganahl, 2006). In this sense, wired Internet will displace some of the functions RI H[LVWLQJ PHGLD HJ YLGHR YLHZLQJ -DSDQ¶V new ICT policy (U-Japan policy) and the present unique Internet usage patterns suggest that Japan will move toward the development of mobile Internet. Lessons from ICT policy history show that the key factor in making a successful system is how mobile Internet will be designed for users in the next generation.
CONCLUSION Findings indicate that gaps in wired Internet and mobile Internet are very different. The distinction between users and nonusers of wired Internet are associated with socioeconomic characteristics such as gender, age, education, job status, income, and region. This suggests that the traditional digital divide persists in wired Internet despite the narrowing trend of the divide. In contrast, the distinction between users and non-users of mobile Internet is associated only with gender, age, and educational levels. Gaps in the mobile Internet are smaller than adoption gaps in the wired Internet, except age gaps. Unlike early adopters of wired Internet and other new media, young women are the heaviest users of mobile Internet. This difference is attributable to the fact that mobile Internet is used for entertainment and communication rather than information-gathering, as mobile Internet supports a unique media culture among young people in Japan (Ishii & Wu, 2006). The results indicate that the social consequences of the Internet cannot be overstated. The impact of the Internet is more limited than either the utopian or dystopian visions suggest (DiMaggio et al., 2001). The media substitution is not the case and the Internet has not yet substituted for TVs. The
632
ZLUHG,QWHUQHWLVQRWVLJQL¿FDQWO\FRUUHODWHGZLWK either socializing with family or friends. In sum, despite the plethora of technological possibilities, the broadband Internet has not yet changed our everyday life. From a policy perspective, the adoption of EURDGEDQGDFFHVVFDQQRWEHWKH¿QDOVROXWLRQIRU bridging the second-level digital divide (the usage gap). Presently, in Japan, it is more important to provide a user-oriented system than to introduce broadband network infrastructure. The governPHQW¶VSURFRPSHWLWLYHSROLFLHV HJWKHH-DSDQ policy) have effectively promoted a fast and cheap broadband network, and should be continued for several years. However, the technology cannot be an end in itself. It is time to consider how the Internet should be promoted to make social activities more HI¿FLHQWDQGFRPIRUWDEOH,QRWKHUZRUGVVRFLDO EHQH¿WVRIWKH,QWHUQHWVKRXOGEHHPSKDVL]HGPRUH in future policies. One lesson from the history of -DSDQ¶V,&7SROLF\LVWKDWLWVKRXOGEHXVHURULHQWHG -DSDQ¶VORFDO,&7SURMHFWVLOOXVWUDWHWKHIDLOXUHRI a government-led ICT policy. In contrast, NTT 'R&R0R¶V³LPRGH´GHPRQVWUDWHVWKHVXFFHVVRI a customer-oriented technology; “i-mode” was designed as a user-friendly system based on relatively old technologies (Ishii, 2004). The history of the mobile Internet in Japan shows that it has been promoted by user needs, not by technology or policy (Ishii, 2004). Thus, if an ICT policy is guided only by the government or engineers, it cannot achieve the desired result. In this sense, there is a risk that, without appropriate customer-oriented systems, the U-Japan policy will end in failure. The future of the mobile Internet is still unknown, but the survey results indicate that, unlike the wired Internet, mobile Internet use is not associated with a higher level of information-gathering activities. Kimura (2004) even pointed out that mobile Internet use inhibits the information-gathering abilities of young people. Likewise, a survey of junior high school students showed that mobile phone users got worse grades than nonusers (Fukano, 2002). This evidence suggests that the mobile phone is a tool for the socially lower class, while previous studies suggest that the PC is a tool for the
Social Consequences of Broadband Access in Japan
socially higher class in Japan (Ishii, 2003, 2006). The present study did not support such negative effects of the mobile phone, but results demonstrate that, unlike wired Internet, mobile Internet is not positively associated with information-gathering activities. If the U-Japan policy overemphasizes the technological aspects and overlooks the social and psychological aspects, it may fail to enrich the everyday life of users.
Hashimoto,Y., Tsuji, D., Fukuda, M., Mori, Y., &
REFERENCES
+XVKLQJ 7 6HOKRIHU + ',',; $ digital divide index for measuring inequality in IT diffucion. IT & Society, 1(7), 21-38.
The Administrative Evaluation and Inspection Bureau, Ministry of Internal Affairs and Communications. (2006, February). IT ka no kouritutekina suisin no tameni [Toward an effective promotion of IT policies]. Gyosei Hyoka Geppo, 3-39.
Haythornthwaite, C., & Wellman, B. (2002). The Internet everyday life: An introduction. In C. Haythornthwaite & B. Wellman (Eds.), The Internet in everyday life (pp. 3-41). Malden, MA: Blackwell Publishers Ltd.
InfoCom Research. (2005). Joho tsusin handobuku 2006 [Telecom Handbook 2006]. Tokyo: InfoCom Research.
Chen, W., & Wellman, B. (2004). The global digital divide within and between countries. IT & Society, 1(7), 39-45.
Ishii, K. (2003). Diffusion, policy and use of broadband in Japan. Trends in Communication, 11(1), 45-61.
DiMaggio, P., Hargittai, E., Neuman, W.R., & Robinson, J.P. (2001). Social implications of the Internet. Annual Review of Sociology, 27, 307-336.
Ishii, K. (2004). Internet use via mobile phone in Japan. Telecommunications Policy, 28(1), 43-58.
Distaso, W., Lupi, P., & Manenti, F.M. (2006). Platform competition and broadband uptake: Theory and empirical evidence from the European Union. Information Economics and Policy, 18, 87-106. Fukano, Y. (2002). Chugakusei to keitaidenwa [Junior high school students and mobile phones]. Retrieved July 21, 2007, from http://www.crn. or.jp/LIBRARY/CYUU/VOL710/index.html Gershuny, J. (2002). Mass media, leisure and home IT: A panel time-diary approach. IT & Society, 1(1), 53-66. Ha, L.S., & Ganahl, R.J. (Eds.). (2006). Webcasting worldwide: Business models of an emerging global media. Mahwah, NJ: Lawrence Erlbaum Associates. Haan, J.D., & Huysmans, F. (2002). Differences in the use between Internet users and nonusers in the Netherlands. IT & Society, 1(2), 67-85.
Ishii, K. (2005). Internet usage and time spent on the Internet based on a questionnaire survey. In World Internet Project Japan (Ed.), Internet usage WUHQGV. Ishii, K. (2006). Implications of mobility: The uses of personal communication media in everyday life. -RXUQDORI&RPPXQLFDWLRQ(2), 346-365. Ishii, K., & Ogasahara, M. (2006). Japan- the world third largest user in the world. Mobile and broadband: Two emerging business models in Japan. In L.S. Ha & R.J. Ganahl (Eds.), Webcasting worldwide: Business models of an emerging global media (pp. 335-356). Mahwah, NJ: Lawrence Erlbaum Associates. Ishii, K., & Wu, C. (2006). A comparative study of media cultures among Taiwanese and Japanese youth. Telematics and Informatics, 23, 95-116. ,73ROLF\2I¿FH&DELQHW6HFUHWDULDW ,7 strategies headquarters. Retrieved July 21, 2007, from http://www.kantei.go.jp/jp/singi/it2/index. html
633
Social Consequences of Broadband Access in Japan
Jeffres, L., & Atkin, D. (1996). Predicting use of technologies for communication and consumer needs. Journal of Broadcasting and Electronic Media, 40, 318-330. Katz, J.E., & Aakhus, M. (Eds.). (2001). Perpetual contact. New York, Cambridge: Cambridge University Press. Kimura, T. (2004). Netto waku riariti (Network reality.) Tokyo: Iwanami Shoten. Kraut, R., Lundmark, V., Patterson, M., Kiesler, S., Mukopadhyay, T., & Scherlis, W. (1998). Internet paradox: A social technology that reduces social involvement and psychological well-being? American Psychologist,(9), 1017-1031. Kubicek, H. (2003). Fighting a moving target: Hard OHVVRQVIURP*HUPDQ\¶VGLJLWDOGLYLGHSURJUDPV IT & Society, 1(6), 1-19. Martin, S.P., & Robinson, J.P. (2004). The income digital divide: An International perspective. IT & Society, 1(7), 1-20. Ministry of Internal Affairs and Communications. (2005). Joho tsusin hakusyo [White Paper Information and Communications in Japan]. Tokyo: Gyosei. Ministry of Internal Affairs and Communications. (2006). Joho tsusin hakusyo [White Paper Information and Communications in Japan]. Tokyo: Gyosei. Ministry of Post and Telecommunications. (1997). White paper 1997: Information and communications in Japan. Tokyo: Gyosei. Parker, E.B. (2000). Closing the digital divide in rural America. Telecommunications Policy, 24, 281-290. 3HWHU- 9DONHQEXUJ30 $GROHVFHQWV¶ internet use: Testing the disappearing digital divide versus the emerging digital differentiation approach. Poetics, 34, 293-305. Pronovost, G. (2002). The Internet and time displacement: A Canadian perspective. IT & Society, 1(2), 44-52.
634
Rhee, K.Y., & Kim, W-B. (2004) The adoption and use of the Internet in South Korea. Journal of Computer Mediated Communication, 9(4). Retrieved July 21, 2007, from http://jcmc.indiana. edu/vol9/issue4/rhee.html Rice, R.E., & Katz, J.E. (2003). Comparing internet and mobile phone usage: Digital divides of usage, adoption, and dropouts. Telecommunications Policy, 27, 597-623. Robinson, J.P., Kestnbaum, M., Neustadl, A., & Alvarez, A. (2000). Mass media use and social life among Internet users. Social Science Computer Review, 18(4), 490-501. Robinson, J.P., Kestnbaum, M., Neustadl, A., & Alvarez, A. (2002). Information technology and functional time displacement. IT & Society, 1(2), 21-36. Rogers, E.M. (2003). Diffusion of innovations (5th ed.). New York: Free Press. Suzuki, H., Hashimoto, Y., & Ishii, K. (1997). Measuring information behavior: A time budget survey in Japan. Social Indicators Research, 42(2), 151-169. Telecommunication Carriers Association (TCA). (2007). Subscriber statistics. Retrieved July 21, 2007, from http://www.tca.or.jp/index.html Tsuda, S. (2005). Chiiki niokeru johoka no sinten to simin ishiki no henyo [Progress of regional informatization and transformation of civic concsiousness]. Keio Media and Communication Research, 33-48. U.S. Department of Commerce. (2002). A nation online: How Americans are expanding their use of the Internet. Retrieved July 21, 2007, from http://www.ntia.doc.gov/ntiahome/dn/html/anationonline2.htm van Dijk, J.A.G.M. (2006, August-October). Digital divide research, achievements and shortcomings. Poetics, 34(4-5), 221-235. Wei, R. (2006). Wi-Fi powered WLAN: When built, who will use it? Exploring predictors of wire-
Social Consequences of Broadband Access in Japan
less Internet adoption in the workplace. Journal of Computer-Mediated Communication, 12(1). Retrieved July 21, 2007, from http://jcmc.indiana. edu/vol12/issue1/wei.html Weng, H. (2002). Internet adoption in Macao. Journal of Computer Mediated Communication, 7(2). Retrieved July 21, 2007, from http://jcmc. indiana.edu/vol7/issue2/macao.html World Internet Project Japan. (2002). Internet usage trends 2001. Tokyo: Communication Laboratory. World Internet Project Japan. (2004). Internet usage trends 2003. Tokyo: Communication Laboratory. World Internet Project Japan. (2005). Internet XVDJHWUHQGV. Tokyo: Communication Laboratory. Zhao, S. (2006). Do Internet users have more social ties? A call for differentiated analyses of Internet use. Journal of Computer-Mediated Communication (Article 8), 11(3). Retrieved July 21, 2007, from http://jcmc.indiana.edu/vol11/issue3/zhao.html Zhu, J.H., & Zhou, H. (2002). Diffusion, use and impact of the Internet in Hong Kong: A chain process model. Journal of Computer-Mediated Communication, 7(2). Retrieved July 21, 2007, from http://jcmc.indiana.edu/vol7/issue2/hongkong.html
KEY TERMS E-Japan Policy: In March of 2001, the IT Policy 2I¿FH&DELQHW6HFUHWDULDWSXEOLVKHGWKHH-DSDQ VWUDWHJ\ ,7 3ROLF\ 2I¿FH &DELQHW 6HFUHWDULDW 2001). It aimed to create a “knowledge-emergent society” in which everyone can actively utilize IT DQGIXOO\HQMR\LWVEHQH¿WV7KLVVWUDWHJ\DLPHG to establish an environment in which the private sector, based on market forces, can exert its full SRWHQWLDODQGPDNH-DSDQWKHZRUOG¶VPRVWDGYDQFHG IT nation within 5 years.
PC Internet: Internet access through a PC that is connected to the Internet. Second Digital Gap: Access gaps in the Internet among existing Internet users. U-Japan Policy: In December 2004, the MIC announced a U-Japan policy aimed at achieving a “ubiquitous network society” (U-Japan) in which “anything and anyone” can easily access networks and freely transmit information “from anywhere at any time” by 2010 (Ministry of Internal Affairs and Communications, 2006).
ENDNOTES 1
2
3
The author computed the index based on the data published in InfoCom Research (2005). Perception of importance is scored as follows: “not important at all” = 1, “not important” = 2, “neither important nor unimportant” = 3, “important” = 4 and “very important” = 5. The information handling competency index FRQVLVWVRI¿YHVWDWHPHQWVUHJDUGLQJLQIRUmation-gathering activities in everyday life (Kimura, 2004; World Internet Project Japan, 2005). They include “I have my own method WRJDWKHULQIRUPDWLRQ ´³,FDQ¿QGQHFHVVDU\ information from a lot of information;” “I like to search a variety of information about interesting topics;” “I can communicate when I need to convey something;” and “I like to cooperate to produce new ideas through discussing with other people.” These items are rated on a 1-4 likert-type scale. The reliability FRHI¿FLHQW &URQEDFK¶V$OSKD LV
Mobile Internet: Internet access through an Internet-enabled mobile phone.
635
Division IV
Communication
637
Chapter XL
Internet-Based Changes in Organizational Communication Erik Lundmark Linköping University, Sweden Alf Westelius Linköping University, Sweden
ABSTRACT This chapter presents a descriptive study of the use of information and communication technology (ICT) and the change in communication patterns in Swedish sport associations over the period 1994 to 2003. The change is discussed in light of Internet and broadband diffusion. Results show that new channels for communication have been adopted, primarily Web sites and e-mail, but few established channels have been dropped. While there are associations that save time and money and increase the spirit of community using ICT, many organizations experience the increased number of communication channels DVDEXUGHQVLQFHPDLQWDLQLQJWKHPWDNHVH[WUDUHVRXUFHVEXWWKHEHQH¿WVDUHQRWDOZD\VHDV\WRGHWHFW RUPHDVXUH&HUWDLQFKDUDFWHULVWLFVFRPPRQDPRQJQRQSUR¿WRUJDQL]DWLRQV 132V DVZHOODV,QWHUQHW DQGEURDGEDQGDFFHVVKDYHLQÀXHQFHGWKHGHYHORSPHQWRI,&7XVH
INTRODUCTION The diffusion of Internet and broadband has changed the way we communicate. At the beginning of the 1990s, the Internet was not used at all by the vast majority of Swedes. Since then, the number of people having access to Internet at home has risen dramatically, from 31% in 1998 to 73% in 2003 (Statistics Sweden, 2004a). Although this number has since seemed to reach a plateau at around 80% of the population, the proportion having broadband connections continues to increase. By 2006, more than half of the homes had a broadband connection.
(Statistics Sweden, 2006). About 70% of the adult population use e-mail (Statistics Sweden, 2005a) and 85 % of the companies with more than 10 employees have a Web site (Statistics Sweden, 2005b). The use of modern information and communication technology (ICT) has drawn academic interest, both regarding individuals (e.g., Ritter, Powell, & Middleton, 2004; Nilsson, 2006) and companies (e.g., Cascio, 2000; Davenport & Pearlson, 1998; Westelius & Mårtensson, 2004). The basis for this study is the changes in communication patterns that occurred in associations connected to the Swedish Sports Confederation
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Internet-Based Changes in Organizational Communication
66& IURP WR7KH66&LVDQRQSUR¿W organization (NPO) that represents and supports the Swedish sports movement. There are many reasons for studying NPOs. One is that they concern a large part of the population (Vogel, Amnå, Munck, & Häll, 2003). Another UHDVRQLVWKH132V¶LPSRUWDQFHWRWKHGHYHORSPHQW of social capital, which in turn is important for both the political and the economical functioning of society (Putnam, 1993). Furthermore, there are, to a certain extent, parallels to developments within organizations in both the private and public sectors, but there are also some highly pronounced characteristics of the NPO which set it apart; both are reasons for studying NPOs (Westelius, 2006). In addition, studying Swedish organizations is interesting from a broadband diffusion perspective VLQFH LQ DQ LQWHUQDWLRQDO FRPSDULVRQ 6ZHGHQ¶V Internet and broadband penetration is among the highest in the world (Organization for Economic Co-operation and Development, 2006; Statistics Sweden, 2005a). At the turn of the millennium, the SSC embarked on a journey of technological change. Riding the wave of enthusiasm for ICT, the SSC aimed at creating a system with the capacity of including more than 20,000 sports associations and their WKUHHPLOOLRQPHPEHUV$VD¿UVWVWHSLQDVVHVVLQJ the effects of this project, we studied the extent of spontaneous use of ICT among the associations connected to the SSC to put the change in a FRQWH[W7KDW¿UVWVWHSSURYLGHGDPDWHULDOKLJKO\ interesting in its own right (Lundmark & Westelius, 2004). Some of that material is presented here, in (QJOLVKIRUWKH¿UVWWLPH Although the large-scale, central initiative in the SSC to implement an organization-wide application provided the impetus for our study, this chapter primarily deals with the spontaneous use of ICT, mainly e-mail and Web sites, which form the context in which that initiative took place. Spontaneous use in this context means that it is not part of, or initiated by, the central SSC project. The effects of the central initiative on the technology use by associations were generally limited up to 2003, but considerable change in communication patterns took place during the period 1994-2003.
638
Thus, from a central change initiative perspective, the sport associations have mainly started using modern ICT spontaneously. This study investigates the use of different communication channels by Swedish sport associations in communicating with members, coaches/leaders, and central administration. The study presents how communication took place in 1994 and how it took place 2003, including the whole range of communication alternatives, that is, from face-toface contact to broadband and narrowband alternatives. The changes that occurred in the period 1994-2003 are discussed in the light of Internet and, particularly, broadband diffusion. We analyze the patterns of communication and the choices of information channels through the lens of information richness theory (IRT) (Daft & Lengel, 1986) and by using the metaphor of intertwining suggested by Robey, Schwaig, and Jin (2003).
BACKGROUND AND FRAME OF REFERENCE In this section, we present the SSC and certain characteristics that make the organization and LWV DI¿OLDWHG DVVRFLDWLRQV DQ LQWHUHVWLQJ REMHFW of study. Thereafter we present the changes in the Swedish society with regard to ICT and ICT infrastructure during the time from 1994 to 2006. Finally we present the theory we use to analyze our empirical material.
The Swedish Sports Confederation: A 1RQ3UR¿W2UJDQL]DWLRQ The SSC is an umbrella organization for the Swedish sports movement. It has 68 (in 2003 it was 67) member organizations, so called Special Sports Federations (SSF), each representing a sport or group of sports. Each SSF has member associations, which in turn have individual people as members. Thus, the SSC is an organization of organizations of organizations of people which in total includes one third of the Swedish population.
Internet-Based Changes in Organizational Communication
The associations at the bottom of the pyramid, sport associations, vary in size and activity and can be members of one or more SSFs. For most associations connected to the SSC, a substantial part of the work is done by volunteers. About 460,000 people had a commission of trust within a sport association in Sweden in the year 2000 (Vogel et al., 2003). On average these people spent 14 hours a month working for the association (Ibid.). There are characteristics of NPOs that pose additional challenges to the would-be leader of change FI)LRO 2¶&RQQRU 2QHVXFKWUDLWLVWKDW membership of an NPO is more voluntary-based compared with that of most other organizations. In terms of I/E-ratios (I = people involved, E = people actually employed) (Hedberg, Dahlgren, Hansson, & Olve, 1994), most NPOs would probably rate as extreme imaginary organizations. Very few people are actually employed by the organization and in order to function, it relies heavily on a network of individuals and related or partner organizations. You work for the organization because of idealism or because you feel that it is rewarding in some way. In an NPO, relying on unpaid work, the usefulness of an information system is therefore even more important for its use than in an organization where managers, based on the employment contract, can exert some degree of pressure on the users to conform to expectations (Simon, 1947/1997). An NPO like the SSC is thus an extreme case, showing more clearly some aspects of adoption factors. However, the setting is also in some way similar to company/customer relationships – but not quite; in many customer relationships there is no sense of identity between customer and company, whereas in most NPOs, the spirit of community is an important aspect.
Internet and Broadband Use in Society Firth and Mellor (2005) argue that, despite calls for rapid spread of broadband and great optimism about the effects on society and business, there is OLWWOHVFLHQWL¿FVXSSRUWIRUWKHVRFLHWDOEHQH¿WVRI a rapid spread of broadband being greater than the problems. They argue that there is little support for
H[WUDEHQH¿WVGHULYHGIURPEURDGEDQGFRPSDUHG to what can be attained through narrowband on a societal level. Our study does not allow us to draw conclusions about the summarized effects of broadband on society. However, we draw some conclusions about the effects of the spread of broadband on a large organization encompassing one third of the Swedish population and identify aspects that deserve attention in future research. As mentioned in the introduction, the access to Internet has risen dramatically in the 10 years following 1994 but seems to have reached a plateau at about 80%. In the last few years, broadband access at home has risen considerably. Between 2004 and 2006 the access to broadband at home increased with about 87%, from about 30% to 56% of the population. Consequently, two thirds of all the Internet connections in Swedish homes are broadband connections (Statistics Sweden, 2004a, 2004b, 2005a, 2006). People with broadband access tend to spend more time online then do narrowband users. Not only bandwidth but also experience seem to LQÀXHQFH WKH WLPH VSHQW RQOLQH )LQGDKO Statistics Sweden 2004a, 2004b, 2005a). There is a difference between broadband and modem users even when beginners are studied; broadband users spend 26% more time online than do modem users (Findahl, 2004). The difference is even stronger for more experienced users; when users have 4 to 5 years of experience, the broadband users spend more than three times as much time online as modem users do (5 compared to 15,5 h/week) (Findahl, 2004). However, these kinds of statistics have a tendency to oversimplify Internet use patterns reducing it to a model where a user is only becoming more and more advanced, spending more and more time online (Ritter et al., 2004). This has implications for the so called digital divide (Vehovar, Sicherl, Hüsing, & Dolnicar, 2006). The digital divide is often measured in the number of people without access to the Internet (Ibid). However, access is not the same as use (Nilsson, 2006). Furthermore, users are not a homogenous group and a richer description is needed (Ritter et al., 2004). Ritter
639
Internet-Based Changes in Organizational Communication
et al. categorize experienced users into four categories: dropouts, suspicious, instrumental, and H[WUHPHO\FRQQHFWHG$PRQJXVHUVZH¿QGSHRSOH who have been excessive users but reduced their use because it affected their social life negatively or users who have to use the Internet because Internet is the medium through which the social group is communicating. Another problem in the discussion about broadband diffusion is that there is no consensus of what is considered broadband. Swedish authorities have QRWVWDWHGDQ\RI¿FLDOGH¿QLWLRQRIEURDGEDQGVLQFH what is considered broadband is likely to change over time. The Organization for Economic Co-operation and Development (OECD) has previously GH¿QHG EURDGEDQG DV NELWV LQ DW OHDVW RQH GLUHFWLRQ7KLVGH¿QLWLRQVHWVWKHOLPLWDWDERXW¿YH times the speed of a dial-up connection. Within the broadband category there are connections more than 100 times faster than the lower limit. However, data exchange over the Internet requires at least two parties and thus the bottleneck could DOZD\VEHWKHRWKHUSDUW\,QWKLVVWXG\ZHGH¿QH broadband as a connection with a speed of at least 256 kbit/s in at least one direction. Among modem users, 40% considered that using the Internet was to “wait and wait.” The corresponding number for broadband users was still 16%, indicating that broadband does not solve all problems. One reason for this is, as mentioned above, other bottlenecks. Another might be that the need for high bandwidth varies with the task. 'RZQORDGLQJPXVLFRU¿OPVUHTXLUHVPRUHEDQGwidth than searching for information. It is also in the realm of multimedia that we see the greatest differences between broadband and nonbroadband users (Findahl, 2004). What is considered broadband varies with the task and the societal context. The sizes of many Web pages have changed over the years. For example the front page of aftonbladet.se (http://www. DIWRQEODGHWVH RQHRIWKH¿UVW6ZHGLVKQHZVSDSHUV to go online, has changed from 55KB in 1996 to 1,000 KB 2006. Thus, a person with the same connection would have had to wait almost 20 times as long for the download today. However, the general trend might not be as strong; according to
640
the Swedish Royal Library that saves snapshots of Swedish Web sites, the average size of a Web page has roughly doubled from 1997 to 2003 (Swedish Royal Library, 2005).
A Framework for Assessing and Analyzing the Effects of ICT Use ICT is commonly seen as an enabler for making members of an organization work together even when not located in the same place. This has attracted considerable academic attention (e.g., Cascio, 2000; Davenport & Pearlson, 1998; Schultze & Orlikowski, 2001; Westelius & Mårtensson, 2004). Among positive effects are reduced costs, OHVVHQYLURQPHQWDOLPSDFWDQGJUHDWHUÀH[LELOLW\ for individuals and organizations (e.g., Cascio 2000; Davenport & Pearlson, 1998; Westelius, 2006, Westelius & Valiente, 2006). Negative effects are usually connected to social phenomena such as fewer face-to-face meetings or high costs for technology and implementation (Ibid). Since the effects of ICT use in Swedish sport associations have not been studied before, we want to asses the general picture. That is, we want to know the proportion of associations that consider ,&7 KDYLQJ DIIHFWHG WKH HI¿FLHQF\ RU WKH VRFLDO situation in the associations. In view of the fact that the associations have a substantial volunteer workforce and thus time is not easily measurable in money, we asses the effects in both monetary terms and a time aspect. Accordingly, in this chapter we focus on the areas of costs, time spent on administration, and the spirit of community. The latter is a translation of the Swedish word gemenskap or German Gemeinschaft which is best described as a feeling of belonging to the association and group/team spirit, that is, spirit of community. The word is common in Swedish and the pilot study showed that the meaning was clear to the interviewees. In analyzing effects on the spirit of community, we touch upon the concept organizational identity. 7KHUHLVDODFNRIFOHDUGH¿QLWLRQVRIRUJDQL]DWLRQDO identity both among practitioners and academics (Melewar & Karaosmanoglu, 2006). In this study RUJDQL]DWLRQDOLGHQWLW\LVGH¿QHGDVDFRPPRQO\
Internet-Based Changes in Organizational Communication
held representation of the organization. “Commonly held” in this context is meant to exclude individual perceptions if they are not shared within a group. Organizations can have multiple identities depending on both differences in how different groups perceive the organization and ambiguities in the organizational identity (Pratt & Foreman, 2000). Organizational identity has a reciprocal relationship with individual behavior; organizational identity can affect individual behavior and individual behavior can affect organizational identity (Pratt & Foreman, 2000). The way virtual work is perceived, especially ZKHQSHUVRQDOH[SHULHQFHLVOLPLWHGLVLQÀXHQFHG by the metaphors used to describe it (Schultze & Orlikowski, 2001). Metaphors can be useful in academic work both in analysis and as a pedagogical tool (Huang, Newell, Galliers, & Pan, 2003). In fact, metaphors are present in most scholarly work. As soon as we label a category or an instance of a phenomenon, the label becomes a metaphor. However, any metaphor bears the risk of leaving out certain aspects of the phenomenon it describes. Furthermore, metaphors can mislead people into seeing aspects that do not exist in realLW\6LQFHPHWDSKRUVKDYHWKHSRZHUWRLQÀXHQFH and metaphors can lead as well as mislead, it is important to be mindful about which metaphors are being used to describe phenomena (Schultze & Orlikowski, 2001). Robey et al. (2003) use the metaphor of intertwining material and virtual work. They suggest four aspects of intertwining: reinforcement (e.g., two channels are more effective than one), complementarity (e.g., using different channels for reaching different groups), synergy (using a mix of channels with a result greater than each individual FKDQQHO¶VFRQWULEXWLRQ DQGUHFLSURFLW\ FKDQQHOV are mutually dependent). We use this metaphor in DQDO\VLQJVSRUWDVVRFLDWLRQV¶XVHRI,&78VLQJWKH metaphor for analysis serves two purposes. First, it gives us a framework for analysis. Second it allows for an evaluation of the usefulness of the metaphor in this type of analysis. Choices of information channels are also analyzed through the lens of IRT. According to IRT, there are two reasons for communication: reducing
uncertainty and reducing equivocality. Different media have different richness. Face-to-face meetings are considered having the highest richness followed by telephone, personal documents such as letters or memos, impersonal written documents, and numeric documents (Daft & Lengel, 1986). The richness of a medium depends, among other WKLQJVRQWKHPHGLXP¶VFDSDFLW\IRULPPHGLDWH feedback and the number of social cues that are transferred (Ibid). Generally, reducing uncertainty is possible with channels having lower richness than what is needed to reduce equivocality. However, IRT has been the subject of critique for the lack of ability to handle context and meaning (Ngwenyama & Lee, 1997).
METHODOLOGY The study includes the use of several different instruments. This chapter will focus on three of them. First, we compare the results from two postal surveys conducted 1994 and 2003 within the SSC. 7KHVXUYH\VVKRZHGDVLJQL¿FDQWFKDQJHLQWKH way sport associations informed their members. The change was strongly connected to the use of Internet and the World Wide Web (www). A follow-up study was made to investigate what effects these changes brought. The follow-up study consisted of semistructured telephone interviews with 110 sport associations.
Postal Questionnaire In 1994, a large survey targeting sport associations within the SSC was carried out. The postal questionnaire used had a broad spectrum of questions about the association and its activities. Among these questions one focused on channels for communication. The respondent was asked to tic the channels he or she considered being the main channels for informing the members. In 2003, a follow-up study was made and the same question was asked, also as one of many questions in a broad spectrum. 7KH¿UVWVXUYH\ZDVVHQWWRDVVRFLDWLRQV and had a response rate of 58%. The follow-up survey was sent to 1,196 associations. However,
641
Internet-Based Changes in Organizational Communication
261 were returned to sender due to unknown addresses. Thus 935 questionnaires were delivered DQGZHUH¿OOHGLQDQGUHWXUQHGJLYLQJDUHsponse rate of 56%.
The Main Study: Semistructured Interviews
Each association was represented by the chairman, another member of the board, a hired administrator, the Web master or another person with good NQRZOHGJHRIWKHDVVRFLDWLRQ¶VDGPLQLVWUDWLRQDQG ICT use. When necessary; two or more representatives per association were interviewed.
Research Instrument Population and Sample We have investigated sport associations that are members of one or more of the 67 SSFs that in turn are members of the SSC. First, a pilot study was conducted. About 20 sport associations participated. Those associations were excluded from participation in the main study. The purpose of the pilot study was to validate the research instrument. In the main study, we randomly selected 150 of the about 20,000 sport associations constituting our population. Out of these 150 associations, we got in contact with 119. Twelve stated that they did not have any activity at present, another 11 stated they were closed and 1 stated they did not want to participate. Because of the general low adoption of the centrally developed system, another 15 associations were selected for their high use of that system. These 15 associations have not been included in statistical analysis unless it is explicitly stated.
The telephone interviews were conducted acFRUGLQJ WR WZR TXHVWLRQQDLUHV 7KH ¿UVW TXHVtionnaire addressed general issues regarding the associations and their communication channels. The second questionnaire addressed the use of ICT and experienced effects. The interviews were semistructured.
Statistical and Interpretative Analysis The quantitative data was coded in SPSS for statistical analysis. The scales employed are ordinal, and not strictly of an interval scale character. 7KHUHIRUH0DQQ:KLWQH\DQG6SHDUPDQQ¶VUKR rank correlation analyzes have been used and no multivariate analyzes have been conducted. In comparing groups, we test the null-hypothesis of no difference between the groups. At a 5% level of VLJQL¿FDQFHZHUHMHFWWKHQXOOK\SRWKHVLV
Table 1. Information channels used by the sport associations to inform members Communication channels
642
2003
1994
Change 1994-2003
(%)
(%)
(percent units)
Meetings
41
47
-6
Postal letters
35
43
-8
Web site
48
0
+48
Flyers
10
11
-1 +23
E-mail
23
0
Bulletin board
36
38
-2
Advertisements
13
13
+/-0
Association magazine
16
20
-4
Internet-Based Changes in Organizational Communication
Table 2. Information channels used by the sport associations in communication with members, leaders/coaches, and central federations (2003) Communication Channels
Members
Leaders
Central
(%)
/coaches
federations
(%)
(%)
Personal meetings/practice
38
67
12
Postal letters
41
10
48
Web site
38
7
10
Flyers
21
7
0
E-mail
28
33
51
Bulletin board
25
0
0
Telephone
20
51
46
Advertisement
10
0
0
Association magazine
13
0
0
SMS
2
0
0
Fax
0
0
3
RESULTS Communication Before and After the Growth of the WWW In the postal questionnaire, a representative for each association was asked which channels the association mainly used for informing the members. The respondent could tic any number of suggested channels but also write down extra channels not suggested in the question. Web site and e-mail were not among the suggested channels 1994, and no association reported using these channels. In the question in the survey 2003, e-mail and Web site were among the suggested channels. There was no limit on the number of answers the respondents could give. Therefore, the sum exceeds 100%. On average a mean of 2.2 channels were mentioned in the most recent postal survey and 1.7 in the earlier one. The proportion using Web site and e-mail has LQFUHDVHGVLJQL¿FDQWO\ 0DQQ:KLWQH\VLJ The use of postal letters has decreased by 19% (eight percent units) and that change is statistically VLJQL¿FDQW 0DQQ:KLWQH\ VLJ 1R RWKHU
FKDQJHV ZHUH VWDWLVWLFDOO\ VLJQL¿FDQW +RZHYHU the decrease in the use of association magazine and meetings (20% and 13% respectively) could be WHQGHQFLHVWKDWFRXOGEHVLJQL¿FDQWLQWKHIXWXUH
Communication with Different Hierarchical Levels In the follow-up study we focused on the use of ICT and Internet as well as communication in more traditional ways. The scope of communication was EURDGHUWKDQLQWKH¿UVWWZRVWXGLHV:HLQYHVWLgated communication not only with members but also with coaches and other leaders as well as with central federations. When comparing the results from the interviews with the results in the postal survey, one has to bear in mind that in the interview the respondent answered freely whereas they got a list to choose from in the postal survey. Furthermore, there are other slight differences: categories do not correspond fully. In the interview, all physical meetings, practice/training and more formal meetings, would end up in the same category whereas in the postal survey there was one suggested channel named
643
Internet-Based Changes in Organizational Communication
sig 0,007). This indicates that more channels is not necessarily better and/or only the associations that can reach their members via new channels such as e-mail and Web site give up their old channels.
meetings (in Swedish Mötesverksamhet) which could be interpreted as more formal meetings, excluding training/practice from consideration. Thus, we recommend caution when comparing the interview with the postal survey. There was no limit on the number of answers the respondents could give. Therefore, the sum exceeds 100%. Respondents mentioned on average 2.3 channels to members and 1.7 to leaders and central federations.
Costs Among frequently mentioned reduced costs are cost for sending letters and making telephone calls. In a few cases, the cost reduction was considered large. In those cases, communication costs and time spent for hired staff on administration were mentioned. Among increased costs, costs for Web hotels and domain names were mentioned. No association considered the increased costs to be large.
The Effects of ICT The effects of ICT on associations categorized by use of ICT are presented in Tables 3-5. The results show that the associations experiencing the strongest positive effects of ICT use IHZHUQRQ,&7FKDQQHOV 6SHDUPDQ¶VUKR
Table 3. Low use of ICT Effect
Costs (%)
Time spent on
Spirit of community (%)
administration(%) Decreased
13
13
1
Unchanged
80
60
93
Increased
7
27
7
Table 4. Moderate use of ICT Effect
Costs (%)
Time spent on
Spirit of community (%)
administration(%) Decreased
13
15
3
Unchanged
80
39
49
Increased
8
46
49
7DEOH+LJKXVHRI,&7 Effect
Costs (%)
Time spent on
Spirit of community (%)
administration(%) Decreased
53
78
1
Unchanged
47
11
15
11
85
Increased
644
Internet-Based Changes in Organizational Communication
Time Spent on Administration Mentioned timesavers include: less time spent on informing, easier administration regarding contests/races/matches and other arrangements, and less time spent on answering phone calls. Time consumers mentioned are primarily updating and working with content for the Web site. Some of the interviewees emphasized that ICT increases the work load for some association administrators and decreases it for others.
Spirit of Community Contrary to concerns about ICT having a negative effect on spirit of community (Westelius, 2006), RXU¿QGLQJVLQGLFDWHWKDW:HEVLWHVFDQEHDJRRG tool for increasing it. The Web site is considered to contribute to awareness about what different subgroups, such as teams, are doing. Commonly the interviewee mentioned that pictures from games/matches, practice, or journeys adds to the interest of those not present (as well as those present) at the event. The Web page helps keeping the collective memory up to date; in doing so it affects the way the organization is perceived. Another factor contributing to the spirit of community is forums. But there are also some problems with forums. Sometimes there can be heavy insults against members or people from competing associations, necessitating monitoring and in some cases closing of the forum. Primarily, effects on spirit of community are noticeable within the associations, but some interviewees said that it has also improved the relation to other associations.
Reasons for Use of ICT The reasons for starting to use ICT in the associations were often unclear. A frequently mentioned reason was that an individual with special interests in computers or ICT initiated the use. Some respondents said that they felt they had to have a Web site since other associations did or that information technology (IT) was in fashion. Mentioned limiting
factors for the use of ICT were lack of knowledge among leaders or lack of knowledge about, and access to, ICT among members. Many organizations also said there were differences between different groups, for example, older and younger members, in attitudes towards and use of ICT.
Broadband Communication Channels Any of the reported mediated communication types would have been possible with a narrowband connection. In general, the advantage of a broadband connection would lie in reducing the time spent waiting and also removing the marginal cost for time spent online. However, there were a few instances where a narrowband connection would have been a substantial obstacle, for example, ODUJH¿OHVVXFKDVYLGHRGRFXPHQWVDYDLODEOHIRU GRZQORDGRQWKHDVVRFLDWLRQV¶:HEVLWHV7KHUH ZHUHDOVRH[DPSOHVRIODUJH¿OHVLQSLFWXUHIRUPDWV that would have been excessively time consuming to download with a narrowband connection.
Internet Habits Comparing with the Swedish average at the time (2003), the respondents were well above average in Internet use. In 2003, about 40% of the population used Internet on a weekly basis (Statistics Sweden, 2004a), compared with 82% in our sample. In general, the demographic use patterns were the same as the national standard, with somewhat less use in the countryside compared with the cities and less use among older people compared with younger (Statistics Sweden, 2004a). However, our research instrument did not include certain commonly investigated aspects, for example education, that could be underlying parameters affecting demographic variation, (Vehovar et al., 2006).
DISCUSSION Uncertainty and Equivocality Earl (2003) suggests three types of uncertainties UHJDUGLQJEHQH¿WVIURP,7XVHZKHQHQYLVDJHGLQ
645
Internet-Based Changes in Organizational Communication
advance: enabling uncertainty, what technology can do; commissioning uncertainty, whether the technology can be built; and impact uncertainty, whether the application will be used as planned. Looking at the rather conventional use of ICT LQPRVWVSRUWDVVRFLDWLRQVWKH¿UVWWZRXQFHUWDLQties (enabling uncertainty and commissioning uncertainty) are less common than impact uncertainty, although not completely absent. Impact uncertainty, however, is common and remains so for many associations even when the technology is implemented; for example there is much uncertainty about whether people are reached by the information published on the Web page, how many of the members have access to Internet/an e-mail address, or how often people can be expected to read their e-mail. In the context of uncertainty, there is a difference between one- and two-way communication. In two-way communication, quick feedback reduces uncertainty about whether the message has come across and how it is interpreted. In IRT, feedback potential is one of the factors determining the richness of a medium (Daft & Lengel, 1986). In general, the communication patterns in our study are in line with what could be predicted from IRT. Rich two-way channels (e.g., meetings or telephone) are used in communication with leaders/coaches to reduce equivocality and one way channels of lower richness (e.g., postal letters) are used towards members to reduce uncertainty. The way to get feedback from members was primarily exchanges during meetings/practices. In a few cases, the forum on the Web site was mentioned. The reason for communication with central federations is mixed, both to reduce uncertainty and to reduce equivocality, and so is the richness of the channels used.
Coping with Uncertainty Uncertainty about the reach of the information channels almost exclusively focused on members. The associations used different strategies to cope with the uncertainty about the reach of their information channels. Some associations used extra channels in the hope that in using many channels
646
at least one would get through to each member, for example, sending both conventional postal mail and e-mail. This can also be seen in Table 2; information channels used towards members are much more diverse and the number of channels used is higher compared with communication with leaders/coaches or central federations. Other associations tried to appreciate the reach of their information channels, for example, using Web statistics tools for their Web site. However, WKHDVVRFLDWLRQVPRVWFRQ¿GHQWDERXWWKHUHDFKRI their Web site were not primarily the associations that used Web statistics tools but the associations which had a clear pull perspective on information :HVWHOLXV 7KDWLVWKH\ZHUHFRQ¿GHQWWKDW their members had a demand for information and thus making it available was enough. For associations with a push perspective, the problems were connected not only with reaching the members with information but also with reminding members about what was on the agenda. Generally, this kind of culture required more work from leaders/coaches and sometimes members.
The Dual Role of the Web Site In analyzing the effects of Web sites among sport DVVRFLDWLRQVZH¿QGWKDWWKH:HEVLWHSOD\VPDLQO\ two roles. ,QWKH¿UVWUROHLWLVDQLQIRUPDWLRQFKDQQHOIRU reducing uncertainty in the terms of IRT (Daft & Lengel, 1986): for example, providing a schedule for practice sessions, results from games/matches or contact information to coaches or other key people. For this, a medium with relatively limited ULFKQHVVZRXOGEHVXI¿FLHQWZKLFKLVDOVRWKHFDVH the information is mainly transferred via text and tables through a one-way channel. The second role we would describe as a catalyst for development of organizational identity. See the section entitled Spirit of Community. Such a catalyst would, in terms of IRT, require a rich medium to reduce equivocality. In this sense, a Web page VKRXOGQRWEHVHHQDVDPHGLXPZLWKD¿[HGULFKness even within the original IRT. A Web page can contain many different media, for example text, pictures, sound, and video. Usually a Web page is
Internet-Based Changes in Organizational Communication
a one-way medium. However, forums and bloggs can make it a two- or rather multiway medium. We argue that pictures and forums can make a Web page a very rich medium, even within IRT. However, especially in the identity development UROHRIDVVRFLDWLRQV¶:HEVLWHVPHDQLQJDQGLQWHUpretation are crucial; organizational identity would QRWH[LVWZLWKRXWLW,QWKLVFRQWH[WWKHSDUWLFLSDQWV¶ beliefs about the media and the sender are important in analyzing communication. For example, people tend to react very differently to advertisements and newspaper articles. This aspect is not captured in IRT. Thus, we agree with Ngwenyama and Lee LQWKDWWKH,57LVLQQHHGRIPRGL¿FDWLRQ shifting the focus from the interaction between a sender and a receiver to the creation of meaning, and reassessment of communicated cues in the light of the wider organizational context in which the communicators operate.
Intertwining Not one of the associations in our sample used ICT for all communication. That is, all associations using ICT combined it to some degree with non ICT channels. However, the general trend seems to be that there are too many channels in use rather than too few. Many association representatives expressed a wish to limit the number of communication channels. They believed that if they could stop using some traditional ways of communicating they could save both time and money. The reason for not doing so was most often internal resistance or uncertainty about the reach of the new channels. As mentioned, the reasons for using ICT in sport associations are often unclear. Often ICT channels have just been extra channels on top of the others. In the terms of Robey et al. (2003), most associations use reinforcement. However, some associations use combination, for example, they have an e-mail list for members preferring e-mail and then they send postal letters to the rest of the members. This way they save costs for stamps and letters and they still reach all members and in a convenient way for the members. We have seen examples of synergy and reciprocity but they are rare. One example was an
association recruiting members by promoting the :HESDJHYLDÀ\HUVZKLFK¿WVERWKDVSHFWV WKH result is greater than the sum of the parts and they are mutually dependent. Many sport associations that have a Web page still use the same information channels that they used before, for example, sending letters to all members on a regular basis. This could be seen as an instance of reinforcement (Robey et al., 2003); LISHRSOHORVHWKHDJHQGDVHQWWRWKHPWKH\FDQ¿QG it on the Web site. Viewed this way, reinforcement LVEHQH¿FLDO+RZHYHUYLHZHGIURPDQRWKHUDQJOH we could see that reinforcement makes the association bear the costs of both channels but only get the effect of one. To draw on the metaphors suggested by Robey et al., this could be seen as an unnecessary overlap creating a more expensive and thick material but the extra thickness is not needed, thus being a negative aspect of intertwining. Confusion about which channel is the appropriate for what type of information is another negative aspect of intertwining that was reported. Using different channels for different groups can lead to different groups having different information. For example, when a misprint was found in a letter, the Web master made a correction on the Web site and then reported to the chairman that information had gone out to the members. However, since a large part of the members did not use the Web site, there was confusion about which information was correct. This could be seen as a negative aspect RIWKHPHWDSKRU LQWHUWZLQHGFRXOGPHDQGLI¿FXOW to keep apart or mixed up. We would label that aspect confusion. According to Robey et al. (2003), failure in one or more of their suggested four aspects of intertwining of virtual and material work could lead to ironic and confused outcomes and results. This is a rather weak statement, since many instances of intertwining would meet only one or two of these aspects and still create neither ironic nor confused outcomes. We see that many sport associations gain from intertwining old and new information channels. However, there are also many who fail to intertwine in a constructive way. Mindful intertwining (Westelius & Valiente, 2006) is about trying to catch the positive aspects of intertwin-
647
Internet-Based Changes in Organizational Communication
ing and minimize the negative aspects; both are equally important. Robey et al. (2003) also give examples of unsuccessful intertwining, but the PHWDSKRUIRFXVHVRQWKHSRVLWLYHDVSHFWV:H¿QG the metaphor a useful analytical tool but we suggest that the metaphor is extended to also include negative aspects of intertwining. This study gives a few examples (unnecessary overlap, confusion) but it is not exhaustive.
Differences Between Subgroups In our sample, there is an average age difference of 14 years between leaders who are involved in the association Web site compared with the ones not involved (not involved, average 55.4; involved, average 41.8, Mann-Whitney sig 0,000). In society as a whole, there seems to be a greater demand for Internet among younger people compared with older. About 20% of people in the age span 44-74 do not want Internet at home. The corresponding number for people in the age span 16-44 is 3%. There are also more young people using broadband connections. Among people 16-24, 60% use a broadband connection whereas only 26% of the people 55-74 do so (Statistics Sweden, 2005a). The average number of leaders/managers/ coaches per association within the SSC has decreased with about 11% over the last decade, leading to an average increase in workload with about 8% per leader/manager/coach (Riksidrottsförbundet, 2005). Within the SSC, some concerns are expressed about this trend, since the most common reason for quitting is “lack of time.” Particularly, there is concern for the ability to continue attracting and keeping young leaders. Many respondents mentioned that young members and coaches used Internet and the association Web site more than older ones. Differences in use could lead to non users being cut off from important information channels as well as frustration about ODFNRIHI¿FLHQF\LQ,&7FKDQQHOVDPRQJXVHUV This frustration could have stronger effects in the group most likely to experience it, namely younger volunteers. Research indicates that there is a difference in what makes young and old volunteers quit. Younger volunteers have a stronger tendency
648
to quit when they do not see the results they expect from their work than do older volunteers (Cadwell, Farmer, & Fedor, 2004). In the short term, changing culture could be GLI¿FXOW &RRSHU +XDQJHWDO &HUWDLQ characteristics of associations within the SSC, such as democratic structure and voluntary workforce, might even make such attempts more complicated. Furthermore, often there are limited means to train members, let alone provide them with technology. Thus sport associations are to a large extent dependent on the development in society as a whole. However, many associations have members who do not mirror the Swedish society. Even so, most of WKHPKDYHVXEJURXSVZLWKVLJQL¿FDQWGLIIHUHQFHV in attitudes towards, and use of, ICT. The limited possibilities to force change in NPOs might both facilitate and restrain the use of ICT. On the one hand, it makes subgroup iniWLDWLYHVGLI¿FXOWWRKROGEDFNHQDEOLQJSHRSOHWR start using ICT. On the other hand, it also makes it GLI¿FXOWWRPDNHHYHU\RQHXVHWKHVDPHFKDQQHOV making mindful intertwining necessary but also more of a challenge. $OWKRXJK LW LV GLI¿FXOW WR IRUFH FKDQJH LQÀXHQFLQJ PHPEHUV LV VWLOO SRVVLEOH ,&7 XVH LQ DVVRFLDWLRQV FDQ LQÀXHQFH PHPEHUV WR VWDUW XVing ICT or to get an Internet connection at home. Especially for people who do not get in contact with ICT in other areas of their life, for example, UHWLUHGSHRSOHWKHVSRUWDVVRFLDWLRQ¶VXVHRI,&7 might be an incentive to start. The differences in use of ICT between primarily young and old members could create or amplify existing cultural differences. If, for example, the Web site is created and visited primarily by one subgroup of the association (e.g., the younger members) and the Web page has the role of catalyst for the development of organizational identity, then one can expect the identity to diverge and different representations to develop in the different groups. Thus, the successful use of ICT in sport associations could lead to multiple organizational identities rather than promote a more homogenous organizational identity.
Internet-Based Changes in Organizational Communication
Effects of Broadband Diffusion Most of the described changes in this study are spurred by the rapid and extensive diffusion of Internet connection in Swedish homes and workplaces. During the whole period studied, modem has been the most common Internet connection (Statistics Sweden, 2004a). Although the change in communication patterns has occurred in a time period when modem connection has been the most common connection, this does not exclude broadband playing a role. As we have seen, the initiative to start using ICT was often from individuals with an interest in computers and ICT. These individuals more often have access to broadband connections. Modem connections in general were seen as inhibitors to working with administrative tasks online for the association. According to several respondents, access through a connection with a positive marginal cost in particular was seen as an inhibitor to Internet use associated with administrative work for the association. This indicates that there are characteristics of the Internet access that can facilitate or inhibit the use. Bandwidth determines to some degree what services can be easily used, but marginal cost for extra time spent online must also be considered. ,WLVGLI¿FXOWWRVD\LILWLVWKHEDQGZLGWKSHUVH or the lower marginal cost that is the most important factor of the two in explaining the use patterns. Not to forget, one important factor explaining broadband users spending more time online is self-selection into the broadband user group; people who value Internet services more tend to acquire a broadband connection more often. In making the choice of connection, bandwidth is most certainly an important factor. But given a connection, the marginal cost could be a more important inhibitor than low bandwidth for many services.
grow as the average bandwidth increases. Thus there is a risk that narrowband users will have to spend even more time waiting for downloads (Findahl, 2004). This poses a very concrete question to Web site administrators: How large should Web pages be? To some extent, size of a page is determined by the type of content, for example, pictures usuDOO\UHTXLUHODUJHU¿OHVWKDQGRHVWH[W%XWWKHUH is also a question of how much graphics should be included for aesthetic reasons. The reason SwedLVKQHZVSDSHUV¶:HESDJHVKDYHJURZQODUJHULV not that there is more news now than in the early nineties, but that more and larger pictures are used for aesthetic reasons. The spread of broadband connections has made this transition possible. Had modem connections still dominated the Swedish home scene, commercial sites trying to reach as large an audience as possible could not have indulged in ever larger pages, more multimedia, and so forth. 7KLVFRQÀLFWEHWZHHQ³PRGHUQ´JUDSKLFDQG multimedia pages and time to access the pages will be current for any Web master, but considering intraorganizational Web sites, a site constructed for intranet use has a much more homogenous target group in terms of connections, equipment, knowledge, incentives, etc, than do most sport DVVRFLDWLRQV¶VLWHV If the producer of a Web page has considerably larger bandwidth than the common visitor, and the producer is not mindful about the discrepancy, there is a risk that the pages may not be user-friendly due to size contra bandwidth. Since ICT use is often initiated and promoted by individuals or subgroups that are more active on the Internet and more likely to have broadband, there is additional cause to be mindful of dispersion of bandwidth and its implications for Web page size.
Bandwidth vs. Page Size
Broadband and the Dual Role of Web Sites
Much of the Internet-based communication that is used in Swedish sport associations does not require a broadband connection. However, as demonstrated by aftonbladet.se, Web sites have a tendency to
In the two roles of the association Web page (information channel and catalyst of organizational identity), we see that broadband has a different function for each role. For the role of an informa-
649
Internet-Based Changes in Organizational Communication
tion channel, marginal cost is a more important inhibitor than bandwidth since the typical data downloaded is in text format, which does not require wide bandwidth. For the role of catalyst of organizational identity, bandwidth is more important since much of the data involved in this aspect is facilitated by broader bandwidth (e.g., pictures and video). A page on a site is rarely playing just one of the two roles. However, they can have a stronger emphasis on one of the two roles. When the two roles are kept less intertwined and more separated, we can expect to get a higher overall use but also a more diverse use between different user groups. For example, narrowband users can access material that does not require broadband, such as schedules for practice sessions, results from games/matches or contact information to coaches or other members, without having to wait for pictures of the latest match to be downloaded.
CONCLUSION AND FUTURE TRENDS Our results show that the associations within the SSC have changed their ways of communicating over the last decade. New channels for communication, primarily e-mail and Web sites, have been adopted but few established channels have been dropped. While there are associations that save time, money and increase the spirit of community using ICT, many organizations experience the increased number of communication channels as a burden, since maintaining them takes extra UHVRXUFHVEXWWKHEHQH¿WVDUHQRWDOZD\VHDV\WR identify or measure. Even if the general pattern in this study seems to be the fewer non ICT channels, the better, we would not say that only using ICT would be preferable. One-sided use of virtual work can lead to neglect of important aspects of social relationships like face-to-face meetings, which are important to create relationships and to build trust (Robey et al., 2003; Westelius & Mårtensson, 2004). Rather, what we see in the sport associations is a movement
650
from the other extreme; that is, from only using traditional channels, to some use of modern ICT. We could identify instances of all of the aspects of intertwining suggested by Robey et al. (2003). +RZHYHUZH¿QGWKDWWKHDVSHFWVIRFXVWRRRQH sidedly on the positive aspects of the metaphor. We found two negative aspects of the metaphor: unnecessary overlap and confusion. However, we do not propose that this is an exhaustive list of negative aspects. Thus we welcome further research on the subject. ,QJHQHUDOZH¿QGWKDWWKHFRPPXQLFDWLRQSDWterns in Swedish sport associations follow what could be predicted by IRT. Nevertheless, we agree with Ngwenyama and Lee (1997) in that the IRT is LQQHHGRIPRGL¿FDWLRQWRLQFOXGHWKHFUHDWLRQRI meaning and reassessment of communicated cues in the light of the wider organizational context in which the communicators operate. Turning the attention to bandwidth, more people having broadband connection will increase Internet WUDI¿FLQJHQHUDODQGDOVRRQVSRUWDVVRFLDWLRQ:HE sites. Already today, the Web site has become one of the most important information channels for Swedish sport associations. This trend will most certainly continue with the diffusion of broadband. Furthermore, associations will have more people who can contribute to the Web site and other Internet-based tasks, since inhibitors such as marginal cost and time-spent-waiting will decrease. However, we will probably also have more people reacting against ubiquitous Internet (Ritter et al., 2004). Acknowledging cultural differences and trying to make the best of them is important (Huang et al., 2003). It will be a challenge for many sport associations to form a communication strategy in this time of transition from modem to broadband connections. In doing so, mindful intertwining of channels (Robey et al., 2003; Westelius & Valiente, 2006) is an important part. With more people acquiring broadband, the concentration of the most eager users in this category is likely to decrease. A further push towards broadband use is given by the probability that the time spent waiting will grow longer for the modem users due to Web pages growing with average bandwidth (although not proportionally).
Internet-Based Changes in Organizational Communication
Firth and Mellor (2005) claimed that there is little evidence that broadband connections are EHQH¿FLDOWRVRFLHW\FRPSDUHGZLWKQDUURZEDQG connections. We have not attempted to assess the societal consequences of broadband diffusion, but in our investigation of ICT use in a very large 6ZHGLVKRUJDQL]DWLRQZH¿QGEHQH¿WVRIEURDGband access, even if broadband access was not a prerequisite for ICT use. In principle, none of the communication types used by Swedish sport associations required broadEDQGDFFHVV7KHEHQH¿WVLGHQWL¿HGDUHPDLQO\WKDW broadband access facilitates access and interaction. This is due to both lack of marginal cost for use and for decreasing time spent waiting using Internet. Both of these aspects are particularly important to people administrating the Web sites or other Internet-based processes. Furthermore, we do not exclude communication types requiring broadband once critical mass of broadband connections has been established. On the other hand, the span of bandwidth can pose problems for organizational communication. Successfully dealing with a range from slow dial-up to very high bandwidth broadband connections poses a substantial challenge. The 1994 broad study of communication was vital in making possible our assessment of the changes in communication patterns brought about by the rise of the World Wide Web. In the same way, we believe our 2003 investigation can form a good basis for future studies of the effects on communication that the spread of broadband will bring.
Implications for Sport Associations 0DQ\ VSRUW DVVRFLDWLRQV ZRXOG EHQH¿W IURP D communication strategy. That is not implying that all sport associations should necessarily increase their ICT use; rather, sport associations should make an informed choice. For some associations this would lead to less use of ICT. In general, we found that e-mail was a good substitute for postal mail. The Web site was the channel that consumed extra time for content providers and administrators, and this could be the
source of bad conscience. We found that it was the discrepancy between the ambition with the Web site and the activity on it that caused dissatisfaction. :HLGHQWL¿HGWKUHHGLIIHUHQWW\SHVRI:HEVLWHXVH recommendable to sport associations. Identifying the appropriate level and setting the ambition accordingly would help many associations.
Static Web Site If interest for the Web site is low among members, RULILWLVKDUGWR¿QGSHRSOHZLOOLQJWRVSHQGWLPH updating the Web site, then a static Web site can be a good alternative. By static, we mean a Web site that does not require regular updates, only presenting information that does not change frequently, like contact information and a short description of the association. The main target group would be nonmembers looking for information about the sport or the association, or members looking for basic contact information, such as physical and electronic addresses and telephone numbers. This kind of Web site would not replace any other information channel, but neither would it require much extra resources.
Web Site with a Clear Focus on Spirit of Community If there is an interest in the Web site, but the members can not be expected to frequently visit the Web site, then critical information must be spread using other channels as well. However, the Web page can still serve as a means of increasing the spirit of community. Important features of such a Web site ZRXOGEHSLFWXUHVIURPWKHDVVRFLDWLRQ¶VDFWLYLWLHV and articles or forums where the members should be encouraged to contribute with material. A Web site with this focus could work as a substitute for an association magazine.
Web Site as the Main Source of Information If all the members have access to the Internet and can be expected to visit the Web site regularly, then the Web site could serve as an information
651
Internet-Based Changes in Organizational Communication
channel that could replace, or reduce the use of, other channels. It is important to notice that this requires a clear pull perspective on information. If the members do not actively look for news and updates, the Web site will not be a good substitute for push media, such as direct mail and telephone calls. In addition, although broadband is spreading rapidly, a considerable minority still have narrowband connections. It is therefore wise to make sure that critical information can be obtained conveniently even with narrowband. We would also recommend Web masters keep an eye on the capacity of the site. With broadband connections becoming increasingly common, large ¿OHVWKDWSUHYLRXVO\ZHUHGRZQORDGHGVHOGRPFDQ be requested more frequently, thus increasing the WUDI¿F FRQVLGHUDEO\ 7KH VSUHDG RI EURDGEDQG connections can change the bottleneck from connection speed at the user end to connection speed at the Web site server end. Connection speed is also not the only obstacle; server capacity can also constitute a barrier. The more mission critical and frequently updated the messages on the Web site become, the more important the server capacity and server connection speed will be.
REFERENCES Cadwell, S.D., Farmer, S.M., & Fedor, D.B. (2004, August). $JHDVDPRGHUDWRURIWHDPLQÀXHQFHVRQ YROXQWHHUFRQWULEXWLRQVLQDQRQSUR¿WRUJDQL]Dtion. Paper presented at the Annual Meeting of The National Academy of Management Conference, New Orleans. Cascio, W.F. (2000). Managing a virtual workplace. Academy of Management Executive, 14(3), 81-90.
Davenport, T.H., & Pearlson, K. (1998). Two cheers IRU WKH YLUWXDO RI¿FH MIT Sloan Management Review, 39(4), 51-65. Earl, M. (2003). IT: An ambiguous technology? In Sundgren, B., Mårtensson, P., Mähring, M., & Nilsson, K. (eds.), Patterns in information management (pp. 39–48). Stockholm: EFI. Findahl, O. (2004). Svenskarna och Internet 2003 [The Swedes and Internet 2003]. Gävle: World Internet Institute. )LRO0 2¶&RQQRU(- :KHQKRWDQG cold collide in radical change processes: Lessons from community development. Organization Science, 13(5), 532-546. )LUWK/ 0HOORU' %URDGEDQG %HQH¿WV and problems. Telecommunications Policy, 29, 223-236. Hedberg, B., Dahlgren, G., Hansson, J., & Olve, N.G. (1994). Imaginära organizationer [Imaginary organizations]. Lund, Sweden: Liber-Hermods. Huang, J.C., Newell, S., Galliers, R.D., & Pan, S.L. (February, 2003). Dangerous liaisons? Component-based development and organizational subcultures. IEEE Transactions on Engineering Management, (1). Lundmark, E., & Westelius, A. (2004). IT-användning inom idrotten [IT-use within sports] (FoU-rapport 2004:4). Stockholm: Riksidrottsförbundet. Melewar, T.C., & Karaosmanoglu, E. (2006). Seven dimensions of corporate identity: A categorization IURP WKH SUDFWLWLRQHUV¶ SHUVSHFWLYHV European Journal of Marketing, 40(7/8), 846-886.
Cooper, R. (1994). The inertial impact of culture on IT implementation. Inform. Manage, 27, 17-31.
Ngwenyama, O.K., & Lee, A.S. (1997). Communication richness in electronic mail: Critical social theory and the contextuality of meaning. MIS Quarterly, 21(2), 145-167.
Daft, R.L., & Lengel, R.H. (1986). Organizational information requirements, media richness, and structural design. Management Science, 32(5), 554.
Nilsson, O. (2006). To judge access – a user centered approach. International Journal of Public Information Systems, 2. Retrieved July 22, 2007, from www.ijpis.net
652
Internet-Based Changes in Organizational Communication
Organization for Economic Co-operation and Development. (2006). OECD Broadband Statistics to June 2006. Retrieved July 22, 2007, from http://www.oecd.org/sti/ict/broadband Pratt, M.G., & Foreman, P.O. (2000). Classifying managerial responses to multiple organizational identities. The Academy of Management Review, (1), 18-42. Putnam, R.D. (1993). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press. Riksidrottsförbundet. (2005). Idrottens föreningar – en studie om idrottsföreningarnas situation [The VSRUW¶VDVVRFLDWLRQV±DVWXG\RIWKHVSRUWDVVRFLDWLRQV¶VLWXDWLRQ@ )R8UDSSRUW 6WRFNKROP Riksidrottsförbundet. Ritter, T., Powell, A., & Middleton, C. (2004, September). Should Internet be everywhere? Perspectives on ubiquitous Internet access. Paper presented to the Association of Internet Researchs Conference (AOIR 5.0). UK: University of Sussex. Robey, D., Schwaig, K.S., & Jin, L. (2003). Intertwining material and virtual work. Information and Organization, 13, 111-129. Schultze, U., & Orlikowski, W.J. (2001). Metaphors of virtuality: Shaping an emergent reality. Information and Organization, 11(1), 45-77. Simon, H.A. (1997). Administrative behavior. New York: The Free Press. (Original work published in 1947)
Statistics Sweden. (2005b). Use of ICT in Swedish HQWHUSULVHV Stockholm: SCB, Enheten för investeringar, FoU och IT. Statistics Sweden. (2006). Use of computers and the internet by private persons in 2006. SCB, Enheten för investeringar, FoU och IT (Press release 2006-12-18 09:30 Nr 2006:343). Retrieved July 22, 2007, from http://www.scb.se/templates/pressinfo____187176.asp Swedish Royal Library. (2005). Kulturarw.3 Retrieved July 22, 2007, from http://www.kb.se/ kw3/Statistik.htm Vehovar, V., Sicherl, P., Hüsing, T., & Dolnicar, V.(2006). Methodological challenges of digital divide. The Information Society, 22, 279-290. Vogel, J., Amnå, E., Munck, I., & Häll, L. (2003). Föreningslivet i sverige - välfärd, socialt kapital, demokratiskola [Associational life in Sweden: General welfare production, social capital, training in Democracy]. Rapport 98. Stockholm: Statistics Sweden. Westelius, A. (2006). Images and imaginators in virtual organising – the NPO Friluftsfrämjandet and www.frilufts.se. International Journal of Public Information Systems, 2(1). Westelius, A., & Mårtensson, P. (2004). The Midas touch in knowledge management projects – beware, your wish could come true. The Electronic Journal of Knowledge Management, 2(2), 35-44.
Statistics Sweden. (2004a). Use of computers and the internet by private persons in 2003. Stockholm: SCB, Enheten för investeringar, FoU och IT.
Westelius, A., & Valiente, P. (2006). Bringing the enterprise system to the frontline - intertwining computerized and conventional communication at BT Europe. In S.J. Barnes & E. Scornavacca (Eds.), Unwired business: Cases in mobile business. Hershey, PA: IRM Press.
Statistics Sweden. (2004b). Use of computers and the internet by private persons in 2004. Stockholm: SCB, Enheten för investeringar, FoU och IT.
:LNLSHGLD 1RQSUR¿WRUJDQL]DWLRQ5Htrieved July 22, 2007, from http://en.wikipedia. RUJZLNL1RQSUR¿WBRUJDQL]DWLRQ
Statistics Sweden. (2005a). Use of computers and WKHLQWHUQHWE\SULYDWHSHUVRQVLQ Stockholm: SCB, Enheten för investeringar, FoU och IT.
653
Internet-Based Changes in Organizational Communication
KEY TERMS Broadband:,QWKLVVWXG\ZHGH¿QHEURDGEDQG as a connection with a speed of at least 256 kbit/s in at least one direction. Information and Communication Technology (ICT): ICT could refer to a wide range of communication tools. In this study we use a narrow GH¿QLWLRQ RI ,&7 :H LQFOXGH RQO\ PRGHUQ ,&7 such as Internet and computerised communication channels and SMS. Of the channels named by the sport associations in this study, only e-mail, Web sites, and SMS are included in ICT. 1RQSUR¿WRUJDQL]DWLRQ 132 $QRQSUR¿W organization is an organization whose primary objective is to support some issue or matter of private interest or public concern for non-commercial SXUSRVHV1RQSUR¿WVPD\EHLQYROYHGLQDQLQQXmerable range of areas relating to the arts, charities, education, politics, religion, research, sports, or some other endeavor (Wikipedia, 2006). Organizational Identity: Organizational identity is a commonly held representation of the organization. “Commonly held” in this context is
654
meant to exclude individual perceptions if they are not shared within a group. Organizations can have multiple identities depending on both differences in how different groups perceive the organization and ambiguities in the organizational identity (Pratt & Foreman, 2000). Organizational identity has a reciprocal relationship with individual behavior; organizational identity can affect individual behavior and individual behavior can affect organizational identity (Pratt & Foreman, 2000). Pull Perspective: Pull perspective is when the focus is on the receiver of information actively seeking information, for example, via Web page or a bulletin board (Westelius, 2006). Push Perspective: Push perspective is when the focus is on the receiver of information being alerted to information, for example, via letters or SMS (Westelius, 2006). Spirit of Community: Spirit of community is a translation of the Swedish word gemenskap or German Gemeinschaft which is best described as a feeling of belonging to the association and group/team spirit. The word is commonly used in Swedish.
655
Chapter XLI
Ubiquitous Communication via Residential Gateways Alex De Smedt Thomson, Belgium
ABSTRACT This chapter focuses on the equipment in the home, providing broadband access to the users into the home, called residential gateway. The chapter shows how such a device evolves from a simple modem to an advanced gateway system that contributes to: • • • •
The access inside the home of any data on any compatible device The ubiquity of access points for the broadband network The ubiquity of media data in the home from remote devices 7KHFRPPXQLFDWLRQRIPRELOHKDQGVHWVYLDWKH¿[HGQHWZRUN
The objective of this chapter is to give the reader a technical insight into the enabling mechanisms and technologies dealing with such functionalities. The text indicates particular technical solutions, but explanations are kept high-level in order to allow nontechnical readers to understand the basics and concepts of the solutions. A number of references show that the technical exposé is becoming a reality.
INTRODUCTION In modern countries, the evolution of a simple xDSL modem to a residential gateway system (RGW) for broadband access offers new features which probably will speed up introduction of broadband services. A multiservice residential gateway is a key element in the home towards integration of media (documents, mail, telephony, radio, video, TV, and so forth), which are carried over a single access
line to the RGW. All home devices are connected to the rest of the world via this single access box in the home. Some home devices are storing family data (photos, video, recordings, and so forth). An important step is that devices may communicate wirelessly to the RGW, which allows access to the outside world from anywhere in the home. A next step is that the RGW allows access to home devices from outside so that the family can show media (photos, etc.) to friends when they are visiting them. A further step is the integration of RGW with mobile devices and portable equipment, so
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Ubiquitous Communication via Residential Gateways
that it is possible to use broadband network access anywhere where there is a hotspot facility or guest access supported by a RGW. The following sections start with a background and then show a technical evolution, starting from a simple home modem to a complex residential gateway, adding respectively multiple LAN interfaces (wireless, Powerline), mobility in the home, integration with public mobility networks, roaming and public access via hotspot in the homes and guest access, and remote access capabilities for home automation and remote metering purposes. This chapter is focusing on technology, but references show that this is an evolution becoming reality.
the home for a single personal computer (PC). The users need or will need more extensive capabilities, which can be offered by a residential gateway, and which is still reasonably cheep in relation to the monthly fee to be paid for the broadband access. The home user, with regard to the aspect ubiquity, will gradually require more and more the following capabilities (De Smedt, 2004; HGI, 2005; IST European Project, 2005):
BACKGROUND
•
7KHEURDGEDQG,3EDVHG¿[HGQHWZRUNVDUHWDNing their place in the communication networks landscape (Home Gateway Initiative, 2005). The success of the broadband network are services and in particular the data services like Web, mail, and so forth (De Smedt, 2004). These services are now also being offered via the mobile networks. 7KHUHLVDFRQYHUJHQFHRIERWKZRUOGV7KH¿[HG networks are providing functionality to allow more mobility to the user. The mobile communication is DOVRXVLQJWKH¿[HGDFFHVVOLQHVLQRUGHUWRSUR¿W of their bigger bandwidths. The real-time services (telephony, video, TV, etc.) are getting more and more important in this communication area. Also, wireless communication is of greater interest to users. There is a growing communication need to access some home stored media data by remote users, in particular the traveling user who wants to access his home data. Residential gateways may play twice a role in such communication. At home, the user needs to get access to the broadband network and for the traveler the residential gateway at home needs to give access to his home data.
USER NEEDS AND USE CASES The original aim of a residential gateway is to provide access to the Internet broadband services in
656
• •
•
•
•
•
Use of more PCs in the home which are communicating with each other Use of home laptops with wireless access, and telephony and other devices with wireless access, or at least solutions not providing additional cabling The ability to use the mobile phones for FRPPXQLFDWLRQ RYHU WKH ¿[HG QHWZRUN RU for communication in the home at reduced calling prices The ability use the wireless (and wireline) devices at places outside the home, for example, in station, airports, pubs, or even anywhere on a parking place The ability to have broadband access in homes of family or friends 7KH DELOLW\ WR DFFHVV SKRWRV PRYLHV ¿OHV which are on storage devices in the home, from remote locations (e.g., from public places or home of family or friends) The ability to check security in the home (doors closed? oven off? who is ringing the doorbell? motion detecting in the home, etc.) The ability to give utility companies access for remote metering.
$OO WKHVH QHHGV ZLOO KDYH WR EH IXO¿OOHG E\ technical solutions that are technically, at highlevel, described in the following sections.
THE EVOLUTION FROM A SIMPLE MODEM TO A RESIDENTIAL GATEWAY A few years ago, it was only necessary to connect one single personal computer to the Internet.
Ubiquitous Communication via Residential Gateways
Dial-up modems were used, requiring dialing a telephone number for access to the broadband network, but the quality was not very good, and digital subscriber line (xDSL) modems, offering a much better bandwidth, were successfully introduced into the market. The equipment consisted of DVSOLWWHUVSOLWWLQJWKHWHOHSKRQHWUDI¿FIURPWKH ['6/ ,QWHUQHW WUDI¿FDQGDQ['6/PRGHP7KH modem converts the analogue signal to ethernet frames readable by the personal computer (PC). The PC is getting the information via a USB or ethernet interface. This small modem box has evolved to a bigger box (De Smedt, Balemans, Onnegren, & '¶+DHVHOHHU VWLOOZLWKWKH['6/LQWHUIDFH on the network side, but with a number of different interfaces to the user side. These interfaces are in general ethernet ports, each of which may be connected via an ethernet cabling to a number of devices. This way of connecting is called a
local access network (LAN). The ethernet LAN port is a general interface to which a number of ethernet/IP(internet protocol) enabled devices may be attached, for example, PCs, voice over IP equipment, Powerline interface equipment, and so forth. The RGW will have a number of ethernet interfaces and to one interface multiple devices may EHFRQQHFWHGLQDSRLQWWRPXOWLSRLQWFRQ¿JXUDWLRQ (Wilson, 2006). The modem box had some switching capability, in the broadband world called either bridging or routing. Having such capability the box is no longer called an xDSL modem, but rather a residential gateway. In addition, some of the boxes may (also) have an antenna for wireless connections and, with this, ubiquity starts coming into the picture. This way of connecting is called a wireless local access network (WLAN). The wireless interface technology is called Wi-Fi, and is according to standard IEEE 802.11 of which there are a number of types
Figure 1. xDSL modem and PC Telephone/ xDSL line
Ethernet xDSL modem
Broadband network
Splitter
Internet
Figure 2. RGW supporting local LAN, WLAN, and Powerline networks Wi-Fi wireless network
Internet Ethernet/IP
Wireline network
Broadband network Residential gateway Powerline adapter Other services
Powerline network
HOME
657
Ubiquitous Communication via Residential Gateways
(a, b, g, and, in 2008, n), which mainly differ in the bandwidth supported. This type of interface gives the possibility of attaching a number of wireless communication equipments, including laptops, PCs, wireless phones, PDAs, and mobile dual handsets. Whether the device is either connected wirelessly (over WLAN) or via wireline (LAN), the basic protocol stack for transferring data is a combination of ethernet protocol and IP protocol, and all devices directly connected to the LAN or WLAN are ethernet/IP enabled. A number of ethernet/IP devices can be connected to the residential gateway, in order to get access to the broadband services offered by the Internet service provider or other providers (telephone service, television broadcast, etc.). There are also other ways of offering some mobility in the home apart from Wi-Fi. One such technology is called Powerline (Dhir & Mousavi, 2001; Rao & Horvath, 2006). If the residential gateway either contains internally or externally a Powerline adapter, data to be communicated to other home devices is transferred via the power cord of the residential gateway to the electricity network inside the home and via their power cord to other devices. This way of acting supports portability (from one power plug to the other) rather than mobility, although via an extension cord, some ÀH[LELOLW\LVSRVVLEOH0DQ\VD\WKDW3RZHUOLQHLV
too expensive (and offers less bandwidth) to push away the Wi-Fi, but the future will show. Today both solutions are sold, although the market share of Wi-Fi is higher. More functionality can be put in the same box (Ishihara, Matsuoka, & Takeyoshi, 2006), in order to adapt the ethernet interface to present legacy interfaces for voice and video (TV). The box may contain a terminal adaptor for telephony called an analogue terminal adapter (ATA), which ofIHUVDFRQQHFWRUFDOOHG);6WRZKLFKDVWDQGDUG telephone may be connected. There are also other ways for telephony connectivity using the residential gateway. On the market there exist devices for telephony which may be connected to the residential gateway via the ethernet/IP protocols, either via LAN (using an ethernet connector) or via WLAN (using the Wi-Fi wireless technology). This method is called voice over IP (VOIP) (Atkinson, 2005). The set may look like a normal telephone, but the interface will be different. The residential gateway may also have an internal terminal adapter (called a set top box or STB) for TV, which offers a connector called SCART, to which an analogue TV set can be connected. Of course, both the ATA and the STB can also be located outside the residential gateway. With all this, the multiservice RGW (De Smedt et al., 2006) is a key element in the delivery of triple play (Internet data services, telephony, and multi-
Figure 3. In-house and external communications: Examples
DECT
Wi-Fi
Broadband network Residential gateway Powerline adapter Electric wires
658
Ubiquitous Communication via Residential Gateways
cast TV) services. It allows making connections between user devices in the home or between user devices and terminals of other users somewhere attached to the public broadband network. It is possible to transfer a connection from one device to another device. Today research is still needed on how to realize end-to-end connectivity with guaranteed QoS.
ADDING MORE MOBILITY IN THE HOME It was shown that there is some mobility in the neighborhood of the home for ethernet/IP enabled devices via Wi-Fi. Now we could add mobility for telephony by adding a DECT base station. DECT is a standard (DECT Forum, 1997) supporting a local wireless telephony connection from a portable DECT phone to a DECT base station, which will be attached to (or integrated in) the residential gateway. There is no particular local wireless standard for a portable TV (such as DECT for voice), since a traditional TV is rather heavy to carry, but nowadays a number of wireless portable video devices show up (PDAs with a color screen of about 10 cm diameter or more) which have a direct Wi-Fi interface and, as such, can receive video information from the Internet, from a TV server, or from a personal video recorder in the home network. All these communications pass via the residential gateway. Big TV screens are not yet supported over wireless (although codings offer better and better compression). The reason is that the present WiFi standards do not offer higher bandwidth that is needed for good quality video on big screens. HowHYHUIRUVPDOOHUVFUHHQV RUZLQGRZV VXI¿FLHQW bandwidth is supported. As wireless technology evolves, it will be possible to have a good QoS also for bigger screens. The wireless bandwidth nowadays is acceptable for all small screen devices, speech devices, and data devices.
INTEGRATION WITH MOBILITY NETWORKS: FIXED MOBILE CONVERGENCE There are other ways for wireless communication, but on a longer distance scale. This is the mobile communication, where the user has a mobile handset and connects over a mobile network. The wireless connection of the mobile handset is with a so-called mobile base station. This network is undergoing an evolution where more and more digital communication is added. The network nowadays is the GSM network, also called 2G (second generation), but there is an evolution to a 3G WKLUGJHQHUDWLRQFDOOHG8076 ZKLFKLVGH¿QHG and being introduced into the market. Meanwhile a further evolution (Valentzas & Dagiuklas, 2005) to a 4G (totally integrated solution with IP networks) LVEHLQJGH¿QHGLQVWDQGDUGL]DWLRQERGLHV7KLVLV an evolution starting from an analogue voice communication and going to full digital voice, message, picture, and video communications. The mobile handset can be adapted so that it is also able to set up a local wireless connection using Wi-Fi. By doing so, connections can be set up to the mobile network, but instead of using the mobile wireless transmission to a base station, there is a Wi-Fi transmission to the gateway and from the gateway the data is transferred over a wireline connection via the broadband network to the mobile network (MobileIGNITE, 2006). This way of connection may have advantages, for example, the price of a connection may be less. Since the broadband bandwidth is underused in most homes, this may be a cheap way of communication for mobile phone. Such a mobile handset, which in addition to the mobile radio transmission is also enabled for communication over Wi-Fi, is called a dual handset because it supports two transmission schemes (although only one at a particular time). One transmission scheme will handle the traditional mobile communication and the signal is picked up by mobile base stations. For the Wi-Fi transmission scheme, the signal is picked up by the residential gateway, and via wireline (xDSL,
659
Ubiquitous Communication via Residential Gateways
Fiber) transmitted over the broadband network to the mobile network. The dual handset evolution (Stevenson, 2006) has just started, and it is really an evolution of integration of the mobile communication world with the broadband communication networks. 7KLVHYROXWLRQLVFDOOHG¿[HGPRELOHFRQYHUJHQFH (FMC) (Wikipedia, 2006). Michelson et al. (2006) shows a number of FMC use cases.
FIXED MOBILE CONVERGENCE VIA HOTSPOTS This evolution to wireless is interesting inside the home, since cables and wires are eliminated. This has also consequences for people traveling with their wireless device(s) assuming that there are public access points, called hotspots. These may be in public places like hotels, stations, airports,
and so forth, where people want to connect to the ,QWHUQHW RU YLD WKH ,QWHUQHW WR WKHLU FRPSDQ\¶V intranets, or where people want to perform other IP-based services. The wireless devices, when started up, will see the local accessible hotspot(s) on the physical level. The user is able to connect to a particular access point, which will present some login screen to the user. This will be related to some account, either via subscription, or via a pay card allowing some wireless access time. The login VFUHHQLVIRUXVHULGHQWL¿FDWLRQDQGDXWKHQWLFDWLRQ (e.g., via password) (Reddy, Krishman, Zhang, & Das, 2002). In order to make it easy to people, even when the visitors of a hotspot do not have a subscription or pay card, there will be a facility to buy time (minutes, hours, days) via the Internet using secure pay-over-the-Internet techniques, for example, entering the credit card number. Since there are quite some different providers of hotspots, an attractive feature is the possibility of
Figure 4. Dual access to IP broadband services
xDSL line
Wi-Fi
Broadband network
Internet or other services
RGW
Dual handset
Mobile network Base station
Mobile
)LJXUH+RWVSRWRSHUDWLRQ
Internet or other services
SSID x
xDSL line RGW supporting hotspot
660
Hotspot communication interception
Broadband network
Hotspot server
Ubiquitous Communication via Residential Gateways
using one single account for a number of different hotspot providers. In that case the providers need to have service level agreements (SLA) supporting this feature, which is called roaming. Roaming is very common in the mobile network. It means that when somebody using a mobile phone exits the area in which the mobile operator is active, he is switched over to a second mobile RSHUDWRUZLWKZKLFKWKH¿UVWRSHUDWRUKDVDQ6/$ However, all the subscriber data, for example, his account data and password, remain technically LQWKHQHWZRUNRIWKH¿UVWRSHUDWRU6RZKHQWKH mobile subscriber enters the second network (called visited network) and has to authenticate (Reddy et al., 2002) himself via his password, this is sent WRWKHQHWZRUNRIWKH¿UVWRSHUDWRU FDOOHGKRPH network) to check and authorize. Exactly the same principle can be used for hotspot providers. It should be noted, however, that a hotspot provider can have a very high number of hotspots, for example, throughout a country, so that if a subscriber remains in a country, he always ¿QGV D KRWVSRW RI KLV ³KRPH´ KRWVSRW SURYLGHU somewhere in the neighborhood. In that case the roaming could be avoided. Also note that where the hotspot provider basically provides a wireless access, he could also have some dedicated room in the public place where he could connect over wireline via an ethernet cord and still using his credentials in the same way. This has the advantage that he
can communicate at much higher bandwidths than in the wireless case, which is needed, for example, ZKHQKHZDQWVWRWUDQVIHUELJ¿OHV
FIXED MOBILE CONVERGENCE VIA HOTSPOT ACCESS IN THE HOME It has already been said that the broadband bandwidth on the access line in general is not fully used in the homes. For instance during the day, the family members are at school or at work. Also in general the available xDSL bandwidth is not fully taken, IRUH[DPSOHZKHQVXU¿QJ7KLVRSHQVVRPHSHUVSHFWLYHVRIWKH¿[HGEURDGEDQGQHWZRUNSURYLGHU becoming also a hotspot provider, or opening the bandwidth for a hotspot provider with whom he needs to have an SLA. The pure hotspot providers have a network of hotspots which may be at a distance of a few hundred meters in a town, up to a few thousand meters in the country. Note however that Wi-Fi access works well in reality situations up to about 50 meters, so there is still space where there is no reception of hotspot Wi-Fi signal. 7KH¿[HGEURDGEDQGSURYLGHUHVSHFLDOO\WKH incumbent one, having in general a big market share, has a device in the home that may serve as a hotspot: the residential gateway. Of course this needs to be according to a business case, where WKHUHLVDQDJUHHPHQW FRQWUDFW EHWZHHQWKH¿[HG
Figure 6. Using multiple hotspots with one (PC) terminal Customer of Provider A
HotspotA1
Provider A Hotspot A Server
Moving
Broadband network
HotspotA2
Moving HotspotB1
Provider B Hotspot B Server
661
Ubiquitous Communication via Residential Gateways
broadband operator and the home customer that part of the bandwidth could be used for hotspot activity. The customer needs to have some advantage in this business case too, for example, he gets a lower price or free access at certain times, or he may download more as usual, and so on. But of course the provider also has an advantage, since ZLWKWKHKRWVSRWIXQFWLRQKHDWWUDFWVPRUHWUDI¿F and consequently more income. This type of access via residential hotspot opens a totally new perspective of ubiquity. With such a setup, there may be found a hotspot at small distances. This is at a totally new scale that wireless access can be offered to mobile users, compared WRWKHGHQVLW\RISXEOLFKRWVSRWV7KH¿[HGEURDGband operator now gets a coverage by wireless access which is almost equivalent to that of the mobile operator. There is the fact that at present the main services differ: mobile operators focus on voice and messaging while the main service of WKH¿[HGEURDGEDQGRSHUDWRULVVWLOOWKH,QWHUQHW However the common service area in which the mobile operators work becomes broader. Mobile operation is focusing more and more on data, and via PCs and laptops, voice over IP is also becoming very popular. In general it is good for the mobile user that he is able to choose how and with which provider he wants to make his communication services.
However, providing hotspot functionality in the home is not free. Some software integration in the residential gateway is necessary in order to make it virtually two gateways where security (Geier, 2006) between the hotspot network and the home network are fully guaranteed. With regard to bandwidth assignment some precautions need to be taken according to the contract with the customer. This could stipulate that the home customers have always priority when aiming for service bandwidth, some bandwidth up to a limit could be dedicated to hotspot users on the residential gateway, or a service like telephony over IP could always have priority (however still more priority is dedicated to the home user). Presently, the technologies for separating wireless networks and for securing communication inside the residential gateway are available. One residential gateway can provide multiple wireless networks that all have a name. Such a network name LVFDOOHGDVHUYLFHVHWLGHQWL¿HU 66,' :KHQD powered-on computer with Wi-Fi ability is in the neighborhood of a residential gateway supporting Wi-Fi (this means that the RGW is acting as a Wi-Fi access point [AP]), it may “sense” the different wireless networks which are present. Their names are offered as a list to the user. A simple list could be: home_network, hotspot_access1, hotpot_access2. The user may choose one of them to connect to. However, the hotspot user should not
Figure 7. Home RGW providing home network, hotspot, and guest access Traveler 2 Traveler 1
Traveler 3
Internet or other services
Traveler 4
Hotspot
Home network
Hotspot server
SSID x Wi-Fi
xDSL line Ethernet/IP SSID y SSID z
Guest
662
RGW supporting hotspot
Broadband network Home and guest communication
Hotspot communication interception
Ubiquitous Communication via Residential Gateways
see the home network so there is an option that the home_network is not advertised, and, as such, is not in the list. Since the home user is aware of the home_network name, he is still able to connect without seeing the name in the list. Once locally connected to a wireless network, one should take care that even if a wireless network name is not visible, access to the radio signal is still possible and more security is necessary. The methods used are authentication (Reddy et al., 2002) and encryption. Authentication is done by the user entering his credentials, and this could be on a subscriber identity module (SIM) card, to get access authority. Encryption is done by keying the transmitted data by some encryption coding (and there are lots of these). It is not a coincidence that for mobile networks the same methods are used, although the encryption is generally called “ciphering” in the mobile world. The hotspot access is provided to travelers passing by the home. It should be indicated in the neighborhood via a sign that a hotspot is available or on the Internet the hotspot access points are mapped. The passengers-by stop in the neighborhood of the home and connect via the Wi-Fi communication and the residential gateway to the broadband network in order to use any kind of service (data, voice, video), but the Internet access is, RIFRXUVHDYHU\SRSXODURQH$OOWUDI¿FPDGHE\ hotspot users is passed transparently and secured over the residential gateway. A method for doing this at the IP layer is encrypting the transmitted IP messages and sending them through an IP tunnel. A security protocols like IPSec is supporting this functionality. Some new operators (e.g., Malik, 2005) base their business cases on implementing hotspot functionality in RGWs and on operating the hotspot system. The operator needs to have a particular agreement with the RGW customer. The deal made is both advantageous for the RGW customer ¿QDQFLDO IUHH EDQGZLGWK HWF DQG WKH KRWVSRW operator (hotspot revenue). When the residential gateway supports a home network and a hotspot network, still something is missing for the home customer in the case that he receives guests, family, or friends whom he wants
to give access to the broadband network for Internet or other services. He could direct them to the hotspot supported by his residential gateway but then they have to pay. The home customer does not want that, but he also does not want that the guests have access to his home network data. To this purpose, a Wi-Fi network with a SSID called something like guest_network could be created. 7KHWUDI¿FRYHUWKLVZLUHOHVVQHWZRUNQHHGVWREH secured in the residential gateway also. Again this may be done by creating kind of a transfer pipe in the residential gateway so that access to home devices is impossible. The difference with the hotspot is that this functionality is totally under control of the home customer. Guest access can be enabled by the home customer via the residential gateway to regular or occasional visitors entering the home for connection of their laptop to the Internet, for voice communication with their DECT or Wi-Fi phone, or for communication with their PDA or other small devices, and any other service. This type of communication may or may not be on the account of the residential customer. For example, if the guest wants to place a telephone call, the access to the broadband network is free, since he uses the guest access. But for the telephone service, the guest could either use his own account or the account of the home customer. This depends upon the credentials used. In any case the communication should be secured with regard to the home network communications and the private data should not seen by guests. The quality of service is an important issue when dividing the bandwidth of the access line of a particular customer over a number of users. One should take care that video and voice communications remain good, and that the customer family members have priority for similar services against the foreign visitors or travelers. A connection admission control mechanism must be implemented. As a summary, a residential gateway may offer three scenarios for which the communications are secured against each other:
663
Ubiquitous Communication via Residential Gateways
• •
•
Communication within the home network Communication by a guest using the broadband access for free in the home and under control of the home user Paid communication by travelers using the hotspot in the environment of the home.
This setup, in combination with the communication to the mobile network, either through the air or via the residential gateway, is very powerful for a user moving around. A home user is able to access services he subscribed to, inside his home; he is able WRDFFHVVWKHVHVHUYLFHVLQVRPHRQHHOVH¶VKRPH KH is able to access his services via hotspots, present in almost every street (of course depends upon penetration); he may access services anywhere via the mobile network, and even access the mobile QHWZRUNYLD¿[HGOLQH2QHLPSRUWDQWFDSDELOLW\ is still missing: access to his data (photos, movies, etc.) at home when he is away from home.
REMOTE ACCESS BY THE REMOTE USER TO HIS HOME NETWORK An additional approach is when the customer himself is in another home (in his circle of acTXDLQWDQFHVIRUH[DPSOHLQDIULHQG¶VKRPH 7KH customer there (e.g., his friend) could give him a guest access so he is able to access the Internet and,
via login in to his home network, he will be able to access his own home information environment. In his home information environment, he has a number of devices with interesting information to show to his friend. There is a PC with digital photos of his latest trip to Venice. There is a hard disk, connected to the residential gateway, containing a number of electronic books he wants to indicate to his friend, and there is a personal video recorder (PVR) containing a fun movie that he recorded and wants to enjoy together with his friend. The home data needs to be protected against access from the outside. To this purpose a number of precautions are made in the home network. A residential gateway in general gets a new IP address every time it is powered up. Typically a residential gateway is powered on in the course of a day and powered off in the evening so the address changes, but this is not really a protection, rather DQRWKHUGLI¿FXOW\ The most used method for securing a home HQYLURQPHQWLVD¿UHZDOO$¿UHZDOOPDLQO\OLPLWVLQFRPLQJDQGRXWJRLQJWUDI¿FGHSHQGLQJRQ DQXPEHURISDUDPHWHUV,QSDUWLFXODUD¿UHZDOO controls the programs that are actually sending information to the outside world. For instance, a mailing program will need permission from the ¿UHZDOO WR VHQG GDWD ,Q WHFKQLFDO WHUPV WKLV LV called the mailing program is acting as a server. 7KH¿UHZDOOZLOOUHDOO\DFWDJDLQVWSURJUDPVDFWLQJ as a server, since hackers can put some nasty server
Figure 8. Remote access concept Home server
Address = https://myname.dyndns.org = IP address
Firewall
Wi-Fi
Ethernet/IP
Residential gateway PVR
HOME
664
Broadband network
AT FAMILY/FRIEND’S HOUSE
Ubiquitous Communication via Residential Gateways
programs (called “Trojan”) on home devices (e.g., PC) in an unknown way (via mail or downloads). Via this server, the malicious user could get access to private data on the PC and have the Trojan send information to the malicious user. This is all important because if a user wants to access his data from a remote side, he needs to have a “personal server” in his home environment, allowing him to access his private data. A residential gateway is the ideal location to put such a server on since from there all home devices connected to it are accessible. There are basically two problems rising with remote access. First the user should be able to access his network, taking into account that the IP address might change. Second, there should be enough security build in so that the user is able to access his private data, but others are absolutely not able to access this data so precautions need to be taken in order to securely identify the user. /HWXVVWDUWZLWKWKH¿UVWSUREOHP7KH,3DGdress of his home is not always the same. However, in general, people do not type in IP addresses in browsers. They type in a name, called the uniform resource locator (URL), for example, http://www. someprovider.com/myhome.html. The translation between a URL and an IP address is done via another server in the network called domain name server (DNS). However this is static DNS, where D¿[HG85/LVDOZD\VGHGLFDWHGWRWKHVDPH,3 address. For an IP address that changes (which is dynamic), a new service has been invented called '\QDPLF '16 RU '\Q'16 7KLV LV D VSHFL¿F DNS server which gives residential equipment an address similar to http://myname.dyndns.org, which allows you to address your home network always in the same way. However, in the residential gateway there is a little program running at each new power-up of the gateway, that, after getting a new IP address, and able to access the Internet, informs the DynDNS server about this change. In this way, the remote user, working on a computer of his family or friends, in order to get connected to his own PC in his home network, is just typing his http://myname.dyndns.org address in the browser and is connected to the right IP address, although it did change since last time.
When the remote user is accessing his home network, the second problem needs a solution. The ¿UHZDOO RUWKHKRPHVHUYHULWVHOI PXVWDVVXUHWKDW it is the home user who is behind the request to enter the home network and not some malicious person trying to access his data. This may be done via a password, but this is not advisable since passwords are in general small in length and not that secure. It is much better to work with secure mechanisms like exchanging (long) public and private keys. Such types of authentication of methods are a technology in its own right. When the access is achieved via Web browser, an indication whether the access is secured is the https indication (instead of the http), so the address of accessing the home network would be https://myname.dyndns.org. The remote user is now able to show the inIRUPDWLRQRQDSDUWLFXODUGHYLFHLQKLVIULHQG¶VRU IDPLO\¶VKRPHZKLFKKHFDQXVHZLWKSHUPLVVLRQ of his friend or family, so that he can show the photos, e-book, and movies from his home to the ÀDWVFUHHQLQWKHOLYLQJURRPRIKLVIULHQG$OOWKH mentioned capabilities allow the customer to use the broadband network as a ubiquitous access from anywhere without having to make hard copy photos or drag laptops and video tapes around.
ACCESS OF HOME DEVICES AND DATA FOR NONENTERTAINMENT PURPOSES Access to services or devices in the home for remote access opens new perspectives for other applications too. Examples are utility companies and home automation. Utility companies today are searching ways to facilitate meter readout. When meters get digitized and getting a digital communication interface, the way is open for remote metering and presenting quite detailed statistics of the metering to the user, together with an advice for savings. The meter is connected via the home network to the residential gateway. The meter is generating pulses which are handled by a local program in the meter. This program counts the pulses and stores time and amount information. When the residential gateway
665
Ubiquitous Communication via Residential Gateways
Figure 9. Remote access for metering and home automation Address = https://myname.dyndns.org = IP address
Home server
Utility company
Firewall
Broadband network Home automation system
Utility server
At work or in hotspot Residential gateway HOME
is powered on, the program senses this and may send the metering information to a dedicated utility server in the network. The utility server may try, at dedicated times, to get access to the home utility meter in the same way as detailed, where the utility server should be authenticated. This access should be secured against the utility server so that it is not able to access the home data. The security protocol IPSec could be used in this case. The assembled metering data may be put on timelines and in charts and may be presented to the user via a Web server of the utility company, again via a personal login so that not everybody FDQWDNHDORRNDWWKHXVHU¶VFRQVXPSWLRQ7KHGDWD may be analyzed so that the user may get some advice for approving his consumption. The remote access facility in combination with a home automation system will provide the customer with the ability to check/control his home devices: Are the doors closed? Is the heating switched off? And even, who is ringing the GRRUEHOO"DQGQRWL¿FDWLRQRIXQH[SHFWHGYLVLWRUV via the security and alarm system. Again there is a two-way access possible (providing the necessary security measures as described). First there is the user who is just at work and wants to check whether he, for example, switched off the lights. Second, there may be some alarms like unexpected visits that are announced to the remote user.
666
REALITY AND FUTURE TRENDS The residential gateway supporting audio, video, and data communication is available on the market. Telephony over the broadband network is already offered by new providers on the market, offering a RGW supporting this to their customers. Also digital television is offered via the access line, but for this still dedicated adaptor boxes are needed. The access line to the broadband network is still xDSL but could in the future evolve to higher EDQGZLGWKVIRUH[DPSOHRYHU¿EHU The wireless technology is being more and more introduced into the homes. At this moment the IEEE 803.11g is very common, but within 3 years from now the IEEE 803.11n, offering much more bandwidth, could take over. The mobile networks offering IP based data communication is at this moment evolving from the second generation (GSM/GPRS) to the third generation (UMTS), while the fourth generation is still under development. An alternative method FDOOHG:,0$;LVGH¿QHGEXWQRWYHU\PXFKLPSOHmented. There are already dual mobile handsets on the market supporting Wi-Fi/2G and Wi-Fi/3G, but at this time they are very scarce. Also scarce are residential gateways supporting GSM or UMTS handsets as an access point. Remote access to residential gateways is quite common in the products, but not often used. One of the problems here is the general inability of the user to program his residential gateway, in addition to the DynDNS problem. There is also
Ubiquitous Communication via Residential Gateways
a lack of providers who offer this as a feature to the user, while they assure they can program the feature remotely in the residential gateway for the user. It should, however, be noted that remote management by providers is becoming more and more utilized.
CONCLUSION The residential gateway is not the only, but is certainly a crucial, element in providing wireless DFFHVV DQG ¿[HGPRELOH FRQYHUJHQFH IHDWXUHV It provides multimedia access to the broadband network and has the possibility of allowing remote users to access local data and media. It offers wireless communication in the home using Wi-Fi, but in the near future will also be able to connect mobile dual phones. It needs enablers to provide good quality of service and security. Taking all these capabilities of the residential gateway into account, it has been shown that it is contributing a lot to the ubiquitous communication.
ACKNOWLEDGMENT This research is performed within the MUSE project, which is partially funded by the European Commission as part of the European IST 6th Framework Program.
REFERENCES Atkinson, R.D. (March 2005). Internet telephone service: A new era of competition in telecommunications. Washington, D.C.: Progressive Policy Institute. DECT Forum. (1997, February). DECT, the standard explained. Retrieved July 23, 2007, from http:// www.dect.ch/pdf/TechnicalDocument.PDF De Smedt, A. (2004, December). Home Gateway: Key to convergence of BB and consumer industry. Paper presented at BBEurope, Brugge, Belgium. Retrieved July 23, 2007, from http://www.BBeurope.org/
De Smedt, A., Balemans, H., Onnegren, J., & '¶+DHVHOHHU6 'HFHPEHU The multi-play service enabled Residential Gateway. Paper presented at BBEurope, Geneva, Switzerland. Dhir, A., & Mousavi, S. (2001). Home networking using “no new wires” phoneline and powerline LQWHUFRQQHFWLRQ WHFKQRORJLHV 3XEOLVKHU ;LOLQ[ Retrieved July 23, 2007, from http://direct.xilinx. com/bvdocs/whitepapers/wp133.pdf Geier, E. (2006, August). Wi-Fi hotspot security: Solutions for users. Retrieved July 23, 2007, IURPKWWS ZZZZL¿SODQHWFRPWXWRULDOVDUWLFOH php/3625016 Home Gateway Initiative Consortium (HGI). (2005). Vision and white paper. Retrieved July 23, 2007, from http://www.homegatewayinitiative. org/public/docs/HGI_white_paper.pdf Ishihara, T., Matsuoka, N., & Takeyoshi, H. (2006). Home appliance networking technology creating new service beyond triple play. Paper presented DWWKH3DFL¿F7HOHFRPPXQLFDWLRQV&RXQVHO Conference. IST European Project MultiService Everywhere (MUSE). (2005). D TF3.2 Detailed description of Residential Gateway and advanced features. Retrieved July 23, 2007, from https://www.ist-muse. org/Abstracts/abstract_DTF3.2.htm Malik, O. (2005, December). FONing a WiFi revolution. Retrieved July 23, 2007, from http://gigaom. FRPIRQLQJDZL¿UHYROXWLRQ Mickelsson, H. et al. (2006, December). Nomadism/ FMC use cases and AAA impact. Paper presented at BBEurope, Geneva, Switzerland. MobileIGNITE. (2006). An FMC context for voice call continuity. Retrieved July 23, 2007, from http://www.mobileignite.org/documents/MobileIGNITE_Handover_FS_v1.0.doc Rao, S., & Horvath, S. (2006, December). Broadband for all over powerline to bridge the ICT divide in Europe. Paper presented at BBEurope, Geneva, Switzerland.
667
Ubiquitous Communication via Residential Gateways
Reddy, P., Krishnan, V., Zhang, K. & Das, D. (2002, August 2). Authentication and authorization of mobile clients in public data networks (HP Labs 2002 Technical Reports HPL-2002-213). Stevenson, T. (2006, July). F/MC watch: Dualmode handset growth to explode. Retrieved July IURPKWWS ZZZZL¿SODQHWFRPQHZV article.php/3618556 Valentzas, S., & Dagiuklas, T. (2005, July). 4G cellular/WLAN interworking. Paper performed in the framework of the IST-2001-32449 project EVOLUTE and presented at HET-NETs 2005 Conference, West Yorkshire, UK. Wikipedia. (2006). Technology convergence. Retrieved July 23, 2007, from http://en.wikipedia. org/wiki/Technological_convergence Wilson, T.V. (2006). How home networking works. Retrieved July 23, 2007, from http://www.howstuffworks.com/home-network.htm
KEY TERMS Residential Gateway: A device in the home offering broadband network access to the home user, supporting home LAN and WLAN networks and interfacing to legacy devices.
668
Broadband Network: IP-based network which gives access to the Internet and other IP based services. Dual Phone: Phone able to access both the mobile network and IP-based services directly or via the residential gateway and the Broadband network. Wi-Fi Network: Local home network based on Wi-Fi (standard IEEE 802.11) wireless communication Remote Access: Access via the broadband network to the home network, by an authorized user; this user will have to authenticate before being able to enter the home network. Hotspot: A location where a passenger-by may get access to the broadband network, based on credentials related to a hotspot account. Guest Access: Access to the broadband network via the home network and residential gateway of family or friends. Server: An addressable entity attached to the broadband network which provides some services; there are several types of servers, for example, for address translation, authentication, home access, and so forth. Authentication: A method to verify and ensure that users are who they say they are – that the user who attempts to get access to networks or functions is in fact the user who is authorized to do so.
Ubiquitous Communication via Residential Gateways
APPENDIX Abbreviations ATA DECT DNS FMC FXS GPRS GSM IEEE IPSec IST LAN MUSE PC PDA RGW SCART SLA SSID STB UMTS URI URL USB VOIP WIMAX WLAN xDSL
Analogue terminal adapter Digital enhanced cordless telecommunications Domain name server Fixed mobile convergence Foreign exchange subscriber General packet radio service Global system for Mobile Communications Institute of Electrical & Electronics Engineers Secure Internet Protocol Information Society Technologies Local Access Network Multiservice Access Everywhere project Personal Computer Personal Digital Assistant Residential Gateway 6\QGLFDWGHV&RQVWUXFWHXUVG¶$SSDUHLOV5DGLRUpFHSWHXUVHW7pOpYLVHXUV 79FRQQHF tor) Service Level Agreements 6HUYLFH6HW,GHQWL¿HU Set Top Box Universal Mobile Telecommunications System 8QLIRUP5HVRXUFH,GHQWL¿HU Uniform Resource Locator Universal Serial Bus Voice Over IP Worldwide Interoperability for Microwave Access, Inc. (IEEE802.16 standard) Wireless Local Access Network A form of Digital Subscriber Line
669
Section V
Cross-Country Analysis
671
Chapter XLII
Adoption of Broadband Services:
The Role of National Policies Morten Falch Technical University of Denmark, Denmark
ABSTRACT Broadband is seen as a key infrastructure for developing the information society. For this reason many Governments are actively engaged in stimulating investments in broadband infrastructures and use of broadband services. This chapter compares a wide range of broadband strategies in the most successful markets for broadband. This is done through analysis of national policies in three European countries—Denmark, Sweden, and Germany—and the U.S., Japan, and South Korea. We concluded that successful implementation of broadband depends on the kind of policy measures to be taken at the national level. Many countries have provided active support for stimulating diffusion of broadband and national variants of this type of policies in different countries are important for an explanation of national differences in adoption of broadband.
INTRODUCTION Broadband is radically transforming the telecom sector. First broadband offers opportunities for provision of a host of new services and stimulates demand for broadband connections. Second, provision of triple play enables a bundling of voice, data, and broadcasting services. And third, broadband has created a window of oportunity for new actors including cable operators and public utility companies to enter the telecom market. Since the year 2000, there has been a tremendous growth in the demand for broadband, and within a few years broadband will become the
most common mode of Internet access in many countries. In most countries promotion of broadband access constitutes an important element in their information society strategy. In the year 2004, American presidential, national performance in broadband was seen to be almost as important as GDP (Fransman, 2006). A comparison of broadband development in the OECD countries reveals that national policies are important for the penetration of broadband. Successful policies include direct intervention LQWKHIRUPRI¿QDQFLDOVXSSRUWWRLQIUDVWUXFWXUH development in disadvantaged areas, regulation ensuring facility based competition, and facility
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Adoption of Broadband Services
measures such demand stimulation and providing common visions for the information society. It seems that the goal of achieving universal access to broadband services has legitimized a revitalization of traditional industrial policy tools. Broadband policies do not only include improvement of framework conditions and demand stimulation, but also direct or indirect support to the supply side, and it is being discussed to include broadband in the universal service obligation. Can such types of policies be effective in a globalized liberal economy? Will it be possible WR LPSURYH D QDWLRQ¶V VWDWXV DV DQ LQIRUPDWLRQ society and its overall competitiveness through the implementation of a dedicated broadband policy? And what types of initiatives should such a policy include? First the chapter outlines a number of factors considered to be decisive for penetration of broadband. Second, these factors are used to explain why development has differed in a number of countries both within the EU (Denmark, Sweden, and Germany) and outside the EU (U.S., South Korea, and Japan). Denmark has the highest penetration of broadband and Sweden is among the leading countries with regard to provision of fiber to the home (FTTH). Germany has been included, as it is the largest country within the EU and represents an example of how lack of competition has hampered development. The United States, South Koreas, and Japan are the most advanced key PDUNHWVRXWVLGH(XURSH7KHFKDSWHUZLOO¿UVWRI all focus on the political initiatives taken in different countries. We will here distinguish between facilitation, regulation, and direct intervention. Facilitation includes improvement of general framework conditions, regulation includes measures stimulating competition and improving market transparency, while direct intervention include direct support to infrastructure development and content production. The chapter draws partly on results from the EU funded project Broadband in Europe for All: A Multi Disciplinary Approach (BREAD), which has collected information on national experiences within the EU countries and in all the EU member
672
states and in Canada, the U.S., Japan, Korea, and Iceland (Daele, 2005).
OVERVIEW OF THE BROADBAND MARKET The penetration of broadband connections has on a global scale increased dramatically during the past few years. The penetration rates within OECD have grown from 2.9 per 100 inhabitants in 2001 to 13.6 in 2005. Within EU15 the growth has been even higher (from 1.6 to 14.2). However, the growth has been very unevenly distributed among countries (Figure 1). Penetration is highest in the Scandinavian countries and in South Korea, the Netherlands, and Switzerland. Greece and Ireland are the countries within EU15 with the lowest penetration rates. As indicated by the OECD data, there is a positive relationship between income and broadband penetration. But this relationship is not without exceptions. In spite of its high income and its position as the country with the longest tradition for using the Internet, the U.S. is positioned in the middle of the OECD countries. Outside the OECD area, penetration of broadband is more modest. It may be argued that broadband is not relevant in developing countries, where universal access to basic telephone services still is an issue. However also in developing countries broadband is becoming more important as a way to ensure ICT access. By April 2006, broadband was available in 166 countries—twice as many as 4 years earlier (ITU, 2006). This includes low income countries like Ghana and Rwanda. One reason for this rapid expansion is rapidly falling prices. Broadband tariffs have been reduced by 40% during the past two years. A continuation of this development could make broadband affordable for major parts of the population also in developing countries (ITU, 2006). In Beijing, 50% of the populations is connected to the Internet and 64% of those uses a broadband connection (Fransman, 2006, p. 63). In absolute numbers China is today the second largest broadband market. Although the penetration rate for the entire country is comparatively low, 1.3 per 100 inh. in 2004 (UNCTAD,
Adoption of Broadband Services
Figure 1. OECD broadband penetration and GDP per capita Broadband penetration
GDP per capita
30
70.000 Broadband penetration (subscribers per 100 inhabitants, June 2006) GDP per capita (USD PPP, 2004)
25
60.000
Simple correlation = 0.629 50.000 20 40.000 15 30.000 10 20.000 5
10.000 0
N
D en et ma he r rl a k nd Ic s el an d Sw Ko i tz rea er la n Fi d nl an N or d w Sw ay ed e U ni Ca n te d nad Ki a ng do m B U ni el g i te d um St at es Lu J a x e pa n m bo ur Au g st r Fr ia an Au ce st r G ali a er m an y Sp ai n Ita N Por l y ew tu g C ze Zea a l ch la R nd ep ub l Ire i c la H nd un ga r Po y la Sl nd ov T a k urk e R ep y ub M lic ex ic G o re ec e
0
Source : OECD
2005), 27% of the Internet subscribers are using broadband to access the Internet (ITU, 2006). Still there is a considerable digital divide between high and low income countries. Apart from a few Asian nations, penetration rates in most developing countries are lagging behind. In spite of impressive growth rates in Africa and Latin America, it will take years before they reach the levels achieved in the U.S., Europe, and the most successful nations in Asia. This digital divide may have serious implications for local enterprises. It has been demonstrated WKDW GLDOXS DFFHVV LV VXI¿FLHQW IRU FRPSDQLHV starting to use basic communication services such as e-mail and posting of static information on a Web site (UNCTAD, 2005). More advanced applications such online ordering demand always on functionality. Broadband access offered at affordable prices is also important for ICT-based community centers offering telelearning, telemedicine, or similar services. In particular it will greatly improve WKHLU ¿QDQFLDO VXVWDLQDELOLW\ LI WKH\ DUH DOORZHG to use their broadband connection to provide low cost telephony services by use of VoIP (Proenza, 2005).
DRIVERS AND INHIBITORS TOWARDS BROADBAND A number of econometric studies have been carried out in order to identify the most important drivers and barriers towards adoption of broadband (Bauer, 2004; Cava-Ferreruela, 2006; Distaso, 2006; Prieger, 2003). However, these studies come out with very different and sometimes contradictory results. One reason for this is that penetration of broadband services depends on a wide range of different factors affecting supply and demand conditions. Broadband access is an intermediate product in the sense that the value depends on the kinds of services offered. Another reason is that factors such as cultural preferences and educational policies—factors ZKLFK DUH GLI¿FXOW WR TXDQWLI\²SOD\ D GHFLVLYH role with regard to the demand for broadband services. The decisive factors affecting broadband development can be categorized according to three different dimensions: •
7KH ¿UVW GLPHQVLRQ GLVWLQJXLVKHV EHWZHHQ factors affecting supply and factors affect-
673
Adoption of Broadband Services
•
•
ing the demand. These factors are, of course, interrelated. The demand depends on how and on what conditions broadband services are supplied. High quality services offered at low cost generates more demand than poor services offered at high costs. On the other hand, a certain level of demand is necessary to stimulate investments enabling supply of broadband services. The second line of division goes between content and infrastructure. Broadband networks are essential for any economy and information society. But broadband networks are only useful if content services are developed and made available to the subscribers. Development of content and infrastructure stimulate each other. But it is important to distinguish between the factors stimulating content development and those stimulating the infrastructure development. The third line of division distinguishes between technological, economic, as well as political/cultural factors affecting both supply and demand conditions for both content and infrastructure development. Technological aspects include development of new transmission technologies and development of new types of services that can be transmitted via a broadband infrastructure. Economic factors include market conditions such as the overall market size and the level of competition. Cultural/political factors include regulation and other types of policy intervention as well as differences in lifestyles.
These factors are highly interrelated. Supply of an adequate infrastructure will stimulate generation of broadband services. Availability of services will stimulate demand for both services and infrastructure. Technology is an important factor in broadband development, but technology is also—with certain limitations—global in the sense that the market actors in various countries have access to the same technologies. Policy makers may be able to direct national choices of technology, EXWWKHLULQÀXHQFHRQWKHGLUHFWLRQVLQQRYDWLRQV in new technology may take is limited.
674
With regard to nontechnical factors, it is possible to distinguish between environmental factors such as demography, economy, and so forth, and factors which are a result of a deliberate policy towards stimulation of growth in supply and demand for an advanced broadband infrastructure and for a high usage of broadband services. In this chapter the latter type of factors are of particular interest. We will distinguish between facilitation, regulation, and direct intervention. Facilitation includes improvement of general framework conditions, regulation includes measures stimulating competition and improving market transparency, while direct intervention include direct support to infrastructure development and content production (Falch, Saugstrup et al., 2006).
Supply of Broadband Services Broadband can be provided by use of a wide range of different technologies. The costs depend on the architecture and availability of existing infrastructures, density of customers, and various environmental characteristics. Table 1 compares the annual cost per line for three different access technologies. DSL and cable are the cheapest and also most widespread access technologies, if access to a telecom or a cable network is installed already. Cable modem is the cheapest solution for every line density. However others have calculated that DSL is cheaper than cable modem calculations (Falch & Sigurdsson, 2006). For both technologies, cost increases rapidly ZLWKORZHUGHQVLW\DQGLQORZGHQVLW\DUHDV¿[HG wireless becomes a more cost effective solution. The most successful access technologies for broadband have so far been ADSL and cable modems. Cable modems offer generally higher bandwidth, but cable networks are not as widespread as telecom networks. Cable modems are only offered to a certain segment of users mainly in urban areas. Wireless connections such as 3G and FWA have so far had a limited penetration. In Japan and South Korea, 3G services have some success and are expected to take off in other countries as well. However the bandwidth offered is limited and can
Adoption of Broadband Services
Table 1. Annual cost per line for broadband over three platforms ($/line/year) Density (lines/sqare mile)
Fixed wireless
DSL
Cable modem
0-5
250-336
707
646
5-100
248-308
364
292
100-200
230-304
274
189
200-650
233-287
228
136
650-850
227-302
212
121
850-2.250
217-270
202
113
2.250-5.000
212-259
195
109
5.000-10.000
207-258
199
114
>10.000
214-241
181
110
Average
225-286
236
151
Note: Spectrum costs are excluded 6RXUFH &UDQGDOO
hardly be a full substitute for a wired broadband connection. FWA is more expensive than DSL and cable modem in high density areas, especially if the costs of licenses are included. FWA is therefore mainly used by businesses. But there is a huge potential for use of FWA in rural areas not served by ADSL or cable. WLAN technologies may be used for providing broadband access in public spaces or in neighborhoods. Fiber to the home (FTTH) is the technology which is to offer the highest bandwidth. However, this solution is also by far the most expensive due to the need for new cabling. However in areas where new cabling is needed anyway or if combined with cabling for provision of electricity, FTTH may offer DFRVWHI¿FLHQWDOWHUQDWLYH Wireless technologies and FTTH can be especially attractive for new entrants such as municipalities or power companies. Local community associations may also be a driving force in provision of broadband, for example through existing cable networks or by use of WLAN. Some of the largest of these networks are established in cooperation with the municipalities. This has been the case in, for example, Germany and in Denmark. Unlike network investments made by telecom or cable operators, these initiatives are customer driven. Pooling of demand from a group of customers
can be a very effective tool to decrease prices by EHQH¿WLQJIURPHFRQRPLHVRIVFDOH5DWKHUWKDQ approaching the network operator individually, communities considerably strengthen the bargaining power of the customers. Incidentally, such initiatives also stimulate demand. The extent to which these technologies can be applied by use of existing network structures affect the total cost of investments needed for supply of broadband services. The development of broadband facilities depends on the total cost of investment and on the revenues that these facilities can be expected to generate. Expected revenues depend on demand: the number of customers and the prices that can be charged for delivery of broadband services. Bauer ¿QGVWKDWFRVWFRQGLWLRQVRIQHWZRUNGHSOR\ment is the most constant supply side factor. This is in line with other studies, where the low level of urbanization (implying high costs) is found to be an important barrier for supply of broadband (Cava-Ferreruela, 2006; Prieger 2003). ,QDGGLWLRQWRWKLVWKHPDUNHWVWUXFWXUH²¿UVW of all, the level of competition—plays a key role in driving supply. Facility based competition seems to be very important in this context, while competition within the DSL segment does not seem to have played a role (Cava-Ferreruela, 2006; Prieger,
675
Adoption of Broadband Services
2003). However, in Bauer (2004), competition did not show any statistically relation. Infrastructure supply is also helped by standardization. Increased industry-wide commodization of network elements and modularization of infrastructure allow easy third party interoperability. In other words, the more industry standards emerge, the more third parties investments can complement the investments by the network operator itself. This is of particular relevance in relation to the high level of debts of many network operators.
Demand for Broadband Services The most obvious economic factor affecting the demand for broadband is national income. However, the expected positive relationship between income and penetration of broadband is not as consistent as could be expected. A positive relation is indicated in Figure 1 provided by OECD. But the picture is muddy, as there are important outliers such as Korea and U.S. Miralles (2006) concludes on this background that “broadband penetration is not correlated to the GDP per capita or other WHFKQRORJLFDOLQGH[HVRIDFRXQWU\´%DXHU¿QGV even a negative correlation between income and broadband diffusion, and he suggests that this can be explained by path dependency and leap frogging (Bauer, 2004). This conclusion is however disputed by other studies including (Cava-Ferreruela, 2006) where income, local content, and education all are VKRZQWREHVLJQL¿FDQWSDUDPHWHUVZLWKDSRVLWLYH impact on demand. A possible explanation on this dispute could be that income is important only up to a certain threshold: for countries with a GDP above US$2025,000, income does not play a decisive role, while importance increases for lower levels of income. Figure 1 could indicate such a relationship. Another factor is price compared to income. It may be argued that this factor is more important than income alone. But even if prices are taken into account, there are substantial differences in the percentage of disposal income used for broadband and other communication services. One reason for this is the availability of relevant content demanding a broadband connection to be
676
accessed. Although the market for content to a certain extent is international, there are differences in national preferences, and the markets for many services, for example, e-government services, are still national. Another sociocultural factor is what can be termed as the e-readiness of the society: a combination of skills, availability of terminal equipments, and openness to new technologies. Although households dominate demand in the most advanced countries, the use of broadband in businesses is important in the less advanced economies. Business applications may also stimulate demand from households; therefore, the structure of the economy is important for the overall demand. An economy dominated by informational activities must be expected to generate more demand for broadband services than an economy based on agricultural production.
POLICIES STIMULATING BROADBAND DEVELOPMENT WITHIN THE EU Although the EU member states adhere to the same regulatory framework, there are substantial differences in how the national markets work. EU includes both some of the countries with highest penetration of broadband (Netherlands and the Scandinavian countries), as well as countries with penetration rates far lower than other countries with similar levels of income (e.g., Ireland). The current regulatory framework is designed to stimulate competition in the market for broadband services as well as in other telecom services. In particular, a demand for unbundling of the local loop has stimulated competition in DSL services. However, there is no demand for separation of different kinds of networks and in several member states the incumbent operator owns telecom as well as cable networks, which limits facility-based competition in those countries. In addition to the telecom regulation, the EU commission has initiated a number of measures stimulating broadband. Most of these programs are included in the eEurope program. The eEurope program was initiated in 2000 as a commission
Adoption of Broadband Services
initiative for the special European Council of Lisbon (CEC, 2000). The program has later been followed with the eEurope 2005 program (CEC, 2002) and the successor of this program—i2010 —has been announced (CEC, 2005b). The eEurope plan 2000-2002 had three overall objectives: cheaper, faster, and more secure Internet, investment in people and skills, and greater use of the Internet. In the eEurope 2005 plan, these aims were replaced with: modern online public services, dynamic e-business environment, widespread availability of broadband access, and secure information infrastructure. In i2010 the objectives are: a single European information space, innovation and investment in research, and inclusion, better public services, and quality of life. All plans are building on similar strategies: (1) improvement of available infrastructure and content, (2) strengthening of the European ICT industry primarily through R&D programs, and ¿QDOO\ VWLPXODWLRQRIXVDJHRI,&7VHUYLFHV in order to improve quality of life and to improve European competitiveness. The mutual dependence between network development and content development is most clearly formulated in the 2005 plan: funding more advanced multimedia services depends on the availability of broadband for these services to run on, while funding broadband infrastructure depends on the availability of new services to use it. Action is, therefore, needed to stimulate services and infrastructure to create the dynamic where one side develops from the growth of the other(CEC, 2003). This is the typical market failure argument used for implementing industrial policy: In order to achieve the most favorable development, it is necessary to help the market to take the right decisions. All plans include a large number of actions points, but none or very few of these points include initiatives that would be termed as industrial policy in the supply oriented traditional sense. However, they include the development of a wide range of public sector applications such as e-learning and e-health and the provision of broadband access to public institutions. In this way, the public sector can stimulate broadband development by increasing the demand for network access directly as well as
by the provision of content stimulating demand for broadband access in the private sector. The plans include also a large number of initiatives involving the private sector. These initiatives include support to R&D in telecom networks as well as applications like e-business. The objective of many of these initiatives is to develop pilots, which can inspire other parts of the private sector. Finally, the plans include a continuation of the telecom reform process and measures for improvement of legislation in order to facilitate e-business. Although the 2005 plan states the importance of the supply side, the only really new initiative, in addition to a continuation of the telecom reform SURFHVVLVDVXJJHVWLRQWR¿QDQFHQHWZRUNGHYHOopment in less favored regions, for example, by the use of structural funds. This line of thinking has been strengthened in a communication from the commission issued under the i2010 initiative (CEC, 2006). Here the commission calls for a more active participation by national and local authorities in stimulating uptake of broadband services in particular in less developed areas. It is among others proposed to use public intervention in form of loans and grants and public private partnerships as instruments in this process. Another initiative, which is not mentioned in this context, but is proposed in the i2010 program, is the review of the Universal Service Directive. Inclusion of broadband access within this frameZRUNFRXOGEHDZD\WRHQDEOH¿QDQFLQJRIQHWZRUN development in periphery regions. In addition to initiatives at the EU level, most RIWKHPHPEHUFRXQWULHVKDYHGH¿QHGWKHLURZQ national broadband strategies, as a part of eEurope. National programs within the ICT area are far from a new phenomenon. In the early 1990s, it became a fashion for each country to develop their own plan for how they could become global champions in ICT (Henten, 1995). But new plans with similar aspirations are being published regularly in most countries. 7KHQDWLRQDOSURJUDPV¿WYHU\ZHOOZLWKWKH EU programs, as they basically use the same type of remedies: Support for R&D and pilot projects, e-government initiatives, and demand stimulation through demand from public sector institutions.
677
Adoption of Broadband Services
Very few countries have discussed direct subsidies for network development. Sweden has had programs for public funding of broadband networks and in Denmark broadband access receives substantial taxation subsidies, as companies are allowed to offer broadband connections to their HPSOR\HHVDVDWD[IUHHIULQJHEHQH¿W During the process of creation of the EU internal market it has been important to remove all barriers for international trade between member states. This includes national subsidies as well as other measures favouring domestic industries. On the other hand, it has been important to ensure development of the European industry in competition with the U.S. and Japan. This has led to a strategy with policy goals, which are not very different from those set by traditional national industrial policy, but at the European level. However, a strategy where many of the traditional remedies are replaced by a new set of initiatives with focus on stimulation of ICT-based innovations and applications.
Denmark Denmark ranks as number one with regard to penetration of broadband services. This is not surprising as Denmark is leading with regard to penetration of most ICT services and for this reason is placed in many international comparisons of epreparedness (Unit, 2006). This has been explained by the fact that the major share of the population is able to afford these services. In addition, Danish consumers and enterprises are among the fastest to take up new technology (Falch & Henten, 2007). However, Denmark is not equally successful in all aspects of broadband development. Denmark has since the mid 1990s followed a deliberate policy stimulating competition on the telecom market, and the Danish market is considered to be one of WKH PRVW FRPSHWLWLYH ZLWKLQ WKH (8 ZLWK ¿HUFH competition in particular on the market for mobile VHUYLFHV'HQPDUNZDVDOVRDPRQJWKH¿UVWFRXQtries to demand unbundling of the local loop. In contrast to most other EU countries, unbundling is also demanded for optical networks. However, facility-based competition has been limited by common ownership across technologies. The largest cable
678
operator is owned by the incumbent operator TDC and this has delayed introduction of cable modem services. Moreover, the capacities offered are more limited than in other leading countries. In spite of low interconnection rates, prices for broadband subscription are among the highest in Europe. This is partly offset by a special taxation scheme, which enables employers to offer PCs as well as broadband connections to their employees DVDWD[IUHHEHQH¿W&RQVLGHULQJWKHKLJKOHYHOV of income taxes in Denmark, this implies that tax reductions in reality pay more than 50% of the costs. This scheme has become very popular and many companies provide this opportunity to all of their employees as part of their salary. This scheme has been an important contribution to the growth in demand for broadband. A special characteristic of the Danish market is the emergence of a number of alternative providers bypassing the local telecom network facilities in their provision of broadband services (Tadayoni & Sigurdsson, 2007). These alternative providers include public-private partnerships and QRQSUR¿WFRPPXQLW\EDVHGRUJDQL]DWLRQVDVZHOO as public utility companies. So far their market share is around 5%, but in particular public utilLW\FRPSDQLHVDUHLQYHVWLQJKHDYLO\LQ¿EHUEDVH infrastructures and will in a few years provide a real alternative to the existing telecom network in many areas.
Sweden Sweden has been an early innovator in broadband communication. Broadband to households took off in 1999 (Lindmark & Björstedt, 2006). In terms of penetration, the country ranks as number eight within OECD. Sweden offers the best overall for value for broadband services within OECD and the average bandwidths are higher than anywhere else in Europe and America (Williamson, 2006).1 The bandwidths offered are higher in Sweden than in any other European country. Bredbandsbolaget has played a key role in the Swedish success. BredEDQGVERODJHWDQG7HOLDKDYHFRPSHWHG¿HUFHO\LQ offering broadband access to households since 1999. Bredbandsbolaget provides 10Mbps using
Adoption of Broadband Services
ethernet LAN in apartment buildings at relatively low prices, and this has forced Telia to offer high speed at low prices also. Due to this Sweden has DKLJKHUSHQHWUDWLRQRI¿EHUDFFHVVWKDQDQ\RWKHU European country. The public sector has been very active in developing the Swedish broadband market. First of all, a number of key players on the broadband market including Telia are state owned. Second, the government has taken a number of initiatives VXSSRUWLQJ¿QDQFLQJRIEURDGEDQGLQIUDVWUXFWXUH development in sparsely populated areas. The government has also obliged state owned companies, for example, electricity companies, to roll out high-speed backbone infrastructures to which municipalities can connect. In addition demand is stimulated by tax reduction schemes both for PCs and broadband access similar to those applied in Denmark.
Germany Germany has had a late start in developing its market for broadband services. This has been explained by a high penetration of ISDN and lack of competition. Today broadband prices are, however, among the lowest in Europe (FredebeulKrein, 2006) and the growth rate in penetration is well above OECD average (but below the growth in the Scandinavian countries). This has resulted in a penetration close to both the OECD and the EU15 average—a position which corresponds to income level in Germany (OECD, 2006). A striking feature of the German market is the lack of inter- and intramodal competition. Germany has the second largest cable-TV network after the U.S., connecting more than 21 million households (57%) in 2003 (Büllingen, 2006). In spite of a high coverage of cable TV, penetration of cable access is as low as 0.3 per 100 inhabitant, which is among the lowest in Europe. One reason is that the German cable network until recently has been controlled by Deutsche Telekom. The EU Commission requested a divestiture back in 1996. Therefore German Telecom did not have an interest in modernizing the outdated cable network, and
in this way created a new competitor to their own service offerings at the DSL market. Deutsche Telekom dominates the DSL market with a market share of 67% in 2005. Due to the dominating position of DSL at the German market, this translates into an overall market share of 65% at the entire broadband market. This is well above the EU25 average market share of 50% for incumbent operators. However, this is a dramatic improvement in competition compared to 2004, where the market share of Deutsche Telekom was as high as 88 % (CEC, 2005a). Most competition is based on unbundled lines, and Germany is by far the country with most DSL connections provided in this way. The German regulator has not mandated Deutsche Telekom to offer bitstream access (Fredebeul-Krein, 2006).
NATIONAL POLICIES STIMULATING BROADBAND DEVELOPMENT OUTSIDE THE EU South Korea South Korea has been an early mover with regard to introduction of broadband. It has for a number of years by far been the leading country with regard to penetration of broadband. It seems, however, the market now, at least temporarily, has reached its saturation level—and a few other countries DUHFDWFKLQJXSZLWK6RXWK.RUHD7KHFRXQWU\¶V performance is however still exceptional, if the income level is taken into account, and the average bandwidth offered (4Mb in 2004) is much higher than in Europe. The reasons for this success have been debated in a vast number of reports (DTI, .HOO\*UD\ 0LQJHV /HH2¶.HHIH & Yun, 2003; Lee 2004) The DTI report summarizes the success factors as: • • • •
Geography and demographics Facility based competition Price Internet cafés (PC baangs)
679
Adoption of Broadband Services
• •
Available content Government policy
A favorable geography and demography has lowered the infrastructure costs. More than 50% of the households live in apartment complexes and 93% live within four km of the central of¿FHV &KXQJ 7KHFRVWRISURYLGLQJDFFHVV is therefore as low as 14% of the costs that apply in other countries. This has also favored facility based competition. But even more important in this respect is the fact that the block wiring in the apartment complexes are owned by the landlords. Internet cafés, the so-called PC-baangs, have stimulated the demand for broadband services, in particular the demand for online games. Here, people can try out and learn how to use broadband services without paying a monthly subscription fee. The PC baangs also provide a critical mass for content providers. Prices for broadband services measured per Mb are among the lowest, but the Koreans also seems willing to pay a larger share of their income in order to have a high bandwidth available. There is a vast amount of services available for users, not only entertainment services like online games (which is seen as a killer application), but also a wide range of educational services. Government policy has had an impact on most of the abovementioned factors. First of all, the government has set out clear visions for broadband deployment. These visions are expressed in the IT839 and the u-IT839 plans (eight services, three infrastructures, and nine growth engines).2 The visions have been followed by a wide range of initiatives stimulating both supply and demand. In contrast to Europe, the income generated from 3G licences fees has been recycled into the telecom sector providing funds for research and network investments The government started already back in the 1980s to build a nationwide backbone for broadband services. This has been followed by a number of infrastructure projects serving the public sector and universities and providing the basic infrastructure for connecting the private sector (e.g., the PC bangs) as well as private homes. Support
680
is also given to the provision of broadband in less favored regions. Government policy has also stimulated investments in the local loop and facility based competiWLRQWKURXJKWKHF\EHUEXLOGLQJFHUWL¿FDWHV\VWHP introduced in 1997. Through this system, buildings are ranked according to their capacity to handle high-speed Internet and, since 2001, the Ministry of Construction and Transportation has demanded that those information and communication networks are installed in all new large apartment complexes (Lee, 2004). The government interfered indirectly in the price setting by stating that prices should be not more than $30 per month to become affordable (2Mb connections cost $25 per month). Demand has been stimulated through extensive educational programs in IT involving all tiers of the society including housewives and residents in local communities. In total 14 million out of a population of 50 million have received training during 3 years. The government has also been instrumental in developing content, as they supply high-quality educational content on the Web. It follows that the initiatives cover a wide range of policy areas, most of them considered to be outside the scope of industrial policy. However, they are formulated into a context, where the aim seems to be the strengthening of the Korean sociHW\ZLWKLQDVSHFL¿FDUHD ,&7 7KHREMHFWLYHLV here not only to strengthen the ICT sector itself, but also to strengthen the national capabilities in knowledge based industries in general.
Japan Japan is not among the leading countries with regard to penetration. However, Japan is leading with regard to penetration of FTTH and with regard to bandwidths offered. In metropolitan areas bandwidth of 40-100 Mbps is not uncommon (Atkinson, 2006). This is partly due to a number of electrical power companies including K-Opticom (a subsidiary of Kansai Electric Power), which have offered FTTH at prices comparable with those demanded for DSL (Ida, 2006). They are, therefore, in a very good position to take the lead, when cable modem
Adoption of Broadband Services
and DSL technologies become unable to provide VXI¿FLHQWFDSDFLW\WRVDWLVI\FRQVXPHUQHHGV In Japan, 2001 is mentioned as the broadband kick-off year (Murakami, 2003). Before that, Japan was far behind, for instance, the U.S., in terms of broadband penetration. Late in 2000, an 8 Mb ADSL service was introduced by Yahoo BB! at a UDWHRIOHVVWKDQ¼SHUPRQWKDQGWKLVFUHDWHGD boom in the demand for ADSL. In March 2001, 100 0ERSWLFDO¿EHUQHWZRUNDFFHVVZDVLQWURGXFHGDW UDWHVEHORZ¼SHUPRQWKDQG*VHUYLFHVZHUH introduced later the same year. The e-Japan plan was also announced in 2001. The e-Japan strategy was ambitious both with regard to its targets and the budget allocated. The target was to create the most advanced broadband network environment by 2005. In addition to investments in broadband networks, the program also aimed to stimulate demand through facilitation of e-commerce and digitization of the public adminLVWUDWLRQ7KHEXGJHWDOORFDWHGIRUWKH¿UVWSKDVH RIWKHSURJUDPZDVDOPRVWELOOLRQ¼LQDQG PRUHWKDQELOOLRQ¼LQ (KOHUV The remedies used to facilitate development of broadband infrastructures included: • • •
Opening of utility facilities such as poles, ducts, and conduits Facilitation of high-speed Internet access from apartment complexes 2SHQLQJRIXQXVHGRSWLFDO¿EHUVWRSULYDWH use Usual regulatory measures such as unbundling, asymmetric regulation, frequency allocation, and so forth
The ministry has emphasized that the private sector must be the driving force, and that the role of the government should be limited to establish the right environment for the private sector and to play a direct role in nonprivate sectors such as e-government, R&D, and overcoming the digital divide (Ministry of Economy, 2002). In spite of this, the central government actually supports rollout of broadband facilities by offering attractive ¿QDQFLQJVFKHPHVWD[LQFHQWLYHVDQGJXDUDQWHH of liabilities.
In addition to this, local governments engage in cooperative ventures in order to offer broadband to its citizens. One example is in Hyogo, where the Awaji local council has commissioned an ISP provider to install ADSL in the town and provide access to it citizens at a monthly rate of $16.5 (Kelly et al., 2003). Japan seems to have been quite successful with regard to achieving a high penetration of broadband within a short period of time. However, this success still has not yet been followed by similar developments in content production. This lesson is addressed in the continuation of e-Japan (UJapan), which aims to provide ubiquitous access to broadband services, but with more focus on the development of applications than in the previous program. Japan is indeed one of the countries, which has been most successful with regard to broadband provision, but the contribution from the e-Japan program in this respect is far from clear (Murakami, 2003). However, the speed of the growth in penetration compared to other countries could indicate a positive impact.
United States Although the U.S. ranks among the leading nations, it is well behind countries like Korea, Japan, Canada and the Nordic countries with regard to the penetration of broadband services. A comparison of the U.S. position in penetration of broadband with its position as the cradle of the Internet and its total dominance in the development and provision of content indicates that the U.S. has not been very successful in stimulating investments in and demand for broadband networks. The reason seems to be lack of competition, as most areas lack new entrants with aggressive market strategies. In DUHDVZLWK¿HUFHFRPSHWLWLRQEHWZHHQFDEOHDQG telecom companies penetration seem to be higher (Bauer, 2006). The American public sector is considered to have been much less active in their support for stimulating investments and demand than for instance their East Asian counterparts. Although a market based approach is suggested, a wide range of public initiatives promoting penetration
681
Adoption of Broadband Services
of broadband have, however, been taken also in the United States. At the federal level, public broadband policy has in particular focussed on regulation of WKHWHOHFRPPDUNHWV7KH)&&¶VVWUDWHJLFJRDOIRU broadband is “to establish regulatory policies that promote competition, innovation, and investment in broadband services and facilities while monitoring progress toward the deployment of broadband services in the United States and abroad.”3 The government has also in 2003 introduced a “Federal Rural Broadband Access Loan and Loan Guaranty Program,” supporting investments in broadband development in rural and remote areas.4 The government has, furthermore, a “Community Connect Broadband Grant Program,” supporting connections of community oriented institutions like rural schools, libraries, health clinics, and so forth.5 In addition to this, there are many broadband initiatives undertaken at the state or the local government level (Gillet, 2004). These initiatives include measures such as: • • • • • •
Setting up of facilitators (broadband authorities/agencies) Tax credits State-owned (research and educational) networks Demand aggregation programs Funding programs Public/private partnerships
Furthermore, some municipalities are also engaged in broadband development through their ownership of public utility companies investing in broadband. It should be noted that these initiatives are not restricted to the support of a few disadvantaged areas or to disabled citizens. Many of the initiatives are made in order to attract ICT based businesses to a particular state or community. In fact, the states with the highest rates of broadband penetration (New Hampshire, Minnesota, Utah, Wyoming, and Washington) have all implemented some kind of initiatives promoting demand for broadband. These initiatives certainly relate to industrial policy. Both remedies and objectives are similar: Public support is used and the objectives go beyond welfare concerns like those related to
682
universal service obligations. The objective rather seems to be to attract industries—in particular information based business to the region or the state.
CONCLUSION Broadband penetration is growing rapidly in most OECD countries. Broadband is becoming the primary way to access the Internet and is being bundled with offers on broadcasting and voice services. The markets in most low income countries are developing at a slower pace, although it is becoming available in almost every country. The threat for a new digital divide is therefore REYLRXV$GLDOXSFRQQHFWLRQPD\EHVXI¿FLHQW to connect to some basic communication services, but as broadband is becoming more widespread, it ZLOOEHFRPHVWLOOPRUHGLI¿FXOWWRUXQDEXVLQHVV without broadband access. It is therefore important to engage in broadband development also in low income countries. This chapter has, in order to assess the impact of various broadband policies applied in OECD, outlined a number of factors considered to be decisive for penetration of broadband. Within this framework, national policies applied in the EU, South Korea, Japan, and the U.S. are discussed. It is here possible to distinguish between facilitation, regulation, and direct intervention. Facilitation includes improvement of general framework conditions, regulation includes measures stimulating competition, and improving market transparency, while direct intervention includes direct support to infrastructure development and content production. Public policy is an important factor for stimulation of both demand and supply of broadband VHUYLFHV 0RVW FRXQWULHV KDYH GH¿QHG SROLFLHV for promotion of the information society. These include, among others, industrial policy, education, e-government, and telecom regulation. In Europe, the EU has launched the eEurope and the i2010 programs, which are a part of the Lisbon strategy aiming at improving growth and employment in Europe. The objectives of these programs is to
Adoption of Broadband Services
provide modern online public services, dynamic e-business environment, widespread availability of broadband access, and secure information infrastructure and includes actions stimulating network as well as content development (CEC, 2005b). These objectives are also relevant in developing countries, although the technical and economic environment differs in many ways. For instance e-government services may not be accessed from private homes, but from various types of comPXQLW\FHQWHUV7KHSRWHQWLDOEHQH¿WVRIGLVWDQFH learning and telemedicine may be greater, as they can improve accessibility health and educational services. Policy initiatives vary from country to country. Some countries have focused on manufacturing of ICT equipment, while others have put more emphasis on the application of ICT technologies. They also differ in their prioritization and intensity of governmental support. Some countries are SURYLGLQJ¿QDQFLDOVXSSRUWIRUSURMHFWVVWLPXODWLQJ use or production of ICT, while others focus on creation of a competitive environment, for example, through liberalization of the telecom sector, and the remedies therefore ranging from direct subsidies, access (price) regulation, and tax incentives to other less far reaching facilitation measures such as increasing transparency in the market place. While it is—at present—generally accepted that market forces alone are not providing optimal results in the communications sector, too much governmental intervention is also counterproductive. The effectiveness of these policies depends on the local context. Many programs including i2010 and u-Japan stress that it is important that it is the private sector which takes the lead. The reality is, however, that there are several examples of public funding of infrastructure projects providing broadband. Quite a few governments have provided funding for provision of broadband services in certain disadvantaged areas, for instance, in Sweden and &DQDGD RU IRU VSHFL¿F SXUSRVHV HJ UHVHDUFK networks in Canada and in Europe). Support is also provided to the demand side, for instance through tax exemptions for broadband users (as it is done in Denmark) or funding of
broadband access to certain types of organizations. This is done in the U.S., where grants are provided to instance community oriented institutions. The public sector can also stimulate demand by provision of broadband connections to its own institutions such as schools, hospitals, ministries, and so forth. Public funding may be necessary in low-income countries, especially in areas with low population densities. As demonstrated in, for example, Sweden, public funding of infrastructure can easily be combined with a liberal telecom market. Support to content development is most often given through development of content supporting public sector activities. E-government initiatives can be seen as a sort of direct intervention in content production as the public produces its own content, which can stimulate demand for both content and infrastructure. A related kind of support LV¿QDQFLQJRIHGXFDWLRQDOFRQWHQW7KLVLVGRQH in South Korea. In some low income countries, VoIP could be the most important content. Provision of VoIP would JUHDWO\LPSURYH¿QDQFLDOVXVWDLQDELOLW\RIWHOHFHQWers. It has been estimated that rural telecenters in India requires eight times as many customers, if voice is not one of the services provided (Proenza, 2005, p. 2). One particular type of industrial policy is “guideposting,” where a public institution takes the lead in creation of a common vision on future developments. Many national plans for the information society, for example, the e-Japan plan, play this role. In South Korea, the government has issued guidelines for installation of broadband facilities in large apartment complexes and this has contributed to the growth in bandwidths offered. Common visions may include common standards. The national telecom authority or any other governmental body may also play an important role in development of common nonmandated standards. Although, much of the standardization work is done at the international level, there are still important areas, for instance in digital signatures and EDIFACT messages, where development of national standards is necessary.
683
Adoption of Broadband Services
Regulatory bodies may also play an active role in the increase of market transparency, inter alia, by providing information on prices, on availability of products, and on consumer rights. In relation to Broadband access points, NRAs can offer price comparisons, product descriptions, geographical availability and the like. In many cases operators are required by law to provide this information anyway. Better market transparency can be tremendously effective in improving the competitive situation and thereby stimulate investments in broadband facilities. Another form of facilitation measures relate to the settlement of disputes between companies. The competence to make a binding settlement depends on the willingness of the companies to be subject to decisions taken by the authority. This is different from traditional regulation where companies are subject to the decisions regardless of their consent. The public can also facilitate development through upgrade of competences and readiness to take up new technologies, for instance through commission of extensive training programs. Training and education is important both for demand and supply, as both users and producers of broadband VHUYLFHVFDQEHQH¿WIURPWKLV In assessment of policy impact it is necessary to distinguish between factors affecting content and infrastructure, and between factors affecting demand and supply. Some factors cannot be LQÀXHQFHGE\JRYHUQPHQWV PDFURHFRQRPLFHQYLronment, demographics) while others can only be LQÀXHQFHGLQWKHORQJUXQ HJHGXFDWLQJSHRSOH to use broadband services) and some can easily EHLQÀXHQFHG HJVHWWLQJRILQWHUFRQQHFWLRQDQG access regulations by the NRA). Apart from the TXHVWLRQZKLFKIDFWRUVFDQEHLQÀXHQFHGWKHPDLQ LVVXHLVKRZWRSRVLWLYHO\LQÀXHQFHWKRVHIDFWRUV and whether this should be done centrally/locally or market-driven. Income seems to be an important factor in explaining national differences. However, the level of income is not decisive. Germany has a higher per capita income than South Korea, but a substantially lower penetration of broadband. Furthermore it seems that income plays only a minor role in later
684
stages of development. A comparison of incomes among the 20 countries with the highest penetration rates does not show any relationship between income and penetration of broadband. 7KH UROH RI JHRJUDSK\ LV GLI¿FXOW WR DVVHVV Some countries with a low population density such as Canada have a high penetration of broadband services. But often a large share of the population is concentrated in a few high density areas, and broadband might not be available in the remote areas. However, if such countries should remain among the leading countries with regard to broadband penetration, it will be necessary to cover these areas also. Competition seems to be an important parameter. Competition can be at two different levels: between different types of infrastructures and between different operators using the same or the same type of infrastructure. At present, competition between infrastructures takes place mainly between cable modem and DSL services. Here it seems to be important whether the incumbent telecom operator controls the cable infrastructure. The development in both Germany and Denmark indicate that cross ownership of infrastructures has a negative impact on the development. Competition between companies using the same or the same type of infrastructures is more developed in Korea and Canada than in Denmark. However, this does not seem to have had a severe impact on the penetration, but the bandwidth used in Denmark is lower than in the other two countries. In low income countries, where cable and telecom infrastructures are less developed, it may be GLI¿FXOWWRHVWDEOLVKFRPSHWLWLRQEDVHRQXSJUDGH of existing infrastructures. Therefore, wireless service will be even more important. Experiences show that competition is essential in order create a dynamic market. Especially the emergence of new entrants has proved to stimulate both supply and demand. New entrants can be facilitated by making spectrum resources available. Support to community centers may also help competition, if the centers are able to demand broadband connectivity from new entrants. Initially, market and government action lay on the supply side with laying out infrastructure and
Adoption of Broadband Services
believing people would use it. Government intervention with the aim of stimulating the supply of broadband services has played an important role in the success of South Korea. It is, however, clear that the mere availability of broadband access is not enough to stimulate penetration. Many EU countries have achieved a coverage of broadband well above 90% without having achieved more than modest rates of penetration. The public can facilitate development through upgrade of competences and readiness to take up new technologies for instance through commission of extensive training programs. Training and education is important both for demand and supply, as both users and producers of broadband VHUYLFHVFDQEHQH¿WIURPWKLV Technical and economic parameters such as income level play an important role for the development of broadband services. This has been documented through a number of econometric studies. However, successful implementation of broadband also depends on the kind of policy measures to be taken. Most OECD countries have provided active support for stimulating diffusion of broadband and this study indicates that such policies in different countries are important for an explanation of national differences in adoption of broadband.
Global broadband battles. Standford, CA: Stanford University Press. Cava-Ferreruela, I.A.A.-M. (2006). Broadband policy assessment: A cross-national empirical analysis. Telecommunications Policy, 30, 445-463. CEC. (2000, March 23-24). Europe—an information society for all. In Communication on a Commission Initiative for the Special European Council of Lisbon, Commission of the European Communities. CEC. (2002). eEurope 2005: An information society for all. Brussels: CEC. CEC. (2003). Towards an information society for all. Brussels: Directorate-General, CEC. CEC. (2005a). European electronic communications regulations and markets 2005 (11th Report) Annex 2. Brussels, Belgium: CEC. CEC. (2005b). i2010—A European information society for growth and employment. Communication from the commission to the council, the European Parliament, the European economic and social committee and the committee of the UHJLRQV&20 ¿QDO Brussels, Belgium: CEC. CEC. (2006). Bridging the broadband gap, CEC.
REFERENCES Atkinson, R. (2006). U.S. continues to tread water in global broadband adoption (Policy Brief). The Information Technology and Innovation Foundation. Bauer, J.E.A. (2004). Effects of national policy on the diffusion of broadband in OECD countries. In UFL-LBS Workshop, Gainesville, GA. Bauer, J. (2006). Broadband in the United States. In F. Martin (Ed.), Global broadband battles - why the U.S. and Europe lag while Asia leads. Stanford, CA: Stanford University Press. Büllingen, F. (2006). Development of the broadband market in Germany. In M. Fransman (Ed.),
Chung, I. (2006). Broadband, the information society, and national systems: The Korean case. In F. Martin (Ed.), Global broadband battles - why the U.S. and Europe lag while Asia leads. Stanford, CA: Stanford University Press. Crandall, R.W. (2005). Broadband communications. Handbook of telecommunications economics, 2, 156-191. S.K.M.I.V.M. Cave, North-Holland. Daele, P.V., et al. (2005). Broadband in Europe for all: A multidisciplinary approach. Brussels, Belgium: Project Deliverable FP6-IST Program. Distaso, W. (2006). Platform competition and broadband uptake: Theory and emperical evidence for the European union. Information Economics and Policy.
685
Adoption of Broadband Services
DTI. (2002). DTI overseas mission to South Korea. London, England: Brunel University. Ehlers, S. (2003). IT policy in Japan—from hard to soft. ITPC. Falch, M., & Henten, A. (2007). Liberalization of telecommunications and innovation dynamics in Denmark. In P.J.J. Welfens & M. Weske (Eds.), Digital economic dynamics (pp. 91-117). Berlin, Heidelberg: Springer. Falch, M., Saugstrup, D. et al. (2006). How to achieve the goal of broadband for all. In B. Preissl & J. Müller (Eds.), Governance of communication networks (pp. 203-226). Berlin, Germany: Physica-Verlag. Falch, M., & Sigurdsson, H. (2006). Technoeconomic study. Broadband in Europe for all: A multidisciplinary approach. Brussels, Belgium: Project Deliverable FP6-IST Program. Fransman, M. (2006). Global boadband battles - why the U.S. and Europe lag while Asia leads. Stanford, CA: Stanford University Press. Fredebeul-Krein, M. (2006). Developing comeptitive broadband markets: Lessons learned from German and European regulatory approaches on broadband access. In 16th Biennal ITS Conference, Beijing. Gillet, S. (2004). Local government broadband initiatives. Telecommunications Policy, 28(7/8). Henten, A. (1995). European planning for an information society (CTI working paper No. 4). Lyngby, Denmark. Ida, T. (2006). Broadband information society, and the national system in Japan. In F. Martin (Ed.), Global broadband battles - Why the U.S. and Europe lag while Asia leads. Stanford, CA: Stanford University Press. ITU. (2006). World information society report 2006. Geneva, Switzerland: ITU. Kelly, T., Gray, V., & Minges, M. (2003). Broadband Korea: Internet case study. Geneva, Switzerland: ITU.
686
Lee, C. (2004). Competitive advantage of broadband Internet: A comparative study between South Korea and the United States. Telecommunications Policy, 28(9/10), 649-677. /HH+2¶.HHIH50 <XQ. 7KH growth of broadband and electronic commerce in South Korea: Contributing factors. The Information Society, 19, 81-93. Lindmark, S., & Björstedt (2006). The Swedish broadband market. In M. Fransman (Ed.), Global Broadband Battles. Standford, CA: Stanford University Press. Ministry of Economy, T. a. I. (2002). Promotion of the IT revolution and electronic government, commerce and information policy. Tokyo, Japan: Bureau of Ministry of Economy, Trade and Industry. 0LUDOOHV) (I¿FLHQF\RISXEOLFSURPRWLRQV policies in the diffusion of broadband networks. An exploratory analysis. In 34th Research Confernece on Communication, Information, and Internet Policy Arlington, VA (p. 25). Murakami, T. (2003). Establishing the ubiquitous network environment in Japan (NRI Papers no. 66). Tokyo, Japan: Nomura Research Institute. OECD. (2006). OECD broadband statistics to June 2006. Available from OECD Information and Communications Policy. Retrieved July 25, 2007, from http://www.oecd.org/document/9/0,2340,en_ 2649_34223_37529673_1_1_1_1,00.html Prieger, J.E. (2003). The supply side of the equal availability in the broadband Internet access market? Economic Inquiry, 41(2). Proenza, F.J. (2005). The road to broadband development in developing countries is through competition driven by wireless and VoIP. Tomorrow’s network today. St. Vincent (Aosta), Italy. Tadayoni, R., & Sigurdsson, H. (2007). Development of alternative broadband infrastructures - case studies from Denmark. Telematics and Informatics (Forthcoming).
Adoption of Broadband Services
UNCTAD. (2005). Information economy report Geneva: United Nations.
quality of life. The program builds on the same principles as eEurope.
Unit, E.I. (2006). The 2006 e-readiness rankings. A white paper from the economist intelligence unit. London: Economist Intelligence Unit.
IT839 Plan: The IT839 Strategy refers to 20 policies for the development of the Korean IT industry worked out by the Ministry of Information and Communication in 2004. The strategy consists of a series of plans to develop eight telecommunications services, three infrastructure systems, and nine new technologies for growth engines. The plan has been followed by the u-IT839 strategy where the u stands for ubiquity.
Williamson, A. (2006). Comparison of OECD broadband markets: A comparison of cost and performance data for business and residential broadband products in 26 OECD countries. Waitakere City Aotearoa/New Zealand: Wairua Consulting Limited.
KEY TERMS Broadband Policy: Most countries have adopted policy measures within the area of broadband. The most common policy objectives are: development of a broadband infrastructure offering universal access to high bandwidth and development of broadband content service. Lisbon Process: At a summit in Lisbon in March 2000, European leaders set the goal of the EU becoming “the most dynamic and competitive knowledge-based economy in the world” by 2010. The target was a 3% average annual economic growth and the creation of 20 million jobs by 2010. The Lisbon process been followed by numerous initiatives within the ICT area including eEurope and i2010. eEurope: The eEurope program was initiated in 2000 as a part of the Lisbon Process to ensure WKDW WKH (8 ZLOO IXOO\ EHQH¿W IURP RSSRUWXQLties of the information society. The objectives are to make communication services including broadband widely available, to stimulate content development and improve e-literacy, all in a way WKDWVWUHQJWKHQVVRFLDOFRKHVLRQ7KH¿UVWH(XURSH program has been followed by the eEurope 2005 and i2010 programs.
E-Japan and U-Japan: The e-Japan Strategy was launched in January 2001 and intended to PDNH -DSDQ ³WKH ZRUOG¶V PRVW DGYDQFHG ,7 QDtion by2005”. In July 2003 the focus shifted from original infrastructure development to promotion of IT use, and formation of a ubiquitous network society by 2010 as the goal of the infrastructure development for a new IT society allowing connection to networks “anytime, anywhere, by anything and anyone.” Facilitation: The objective of facilitation is to ensure a good environment at the market for broadband services. One example could be establishment of industry for a development of common visions and standards. Regulation: Regulatory measures include obligations on interconnection and local loop unbundling as well as other measures supporting new entrants in their competition with the incumbent operators. Regulation may be an important tool for stimulating investments in communication infrastructures. Direct Intervention: This involves public funding or direct public participation in infrastructure or content development. Direct intervention in infrastructure development is concentrated in rural and less favored regions.
i2010: i2010 is a follow up of the eEurope program. The program objectives are a single European information space, innovation and investment in research, and inclusion, better public services, and
687
Adoption of Broadband Services
ENDNOTES 1 2
3 4
5
688
This study excludes Japan and Korea. http://www.dynamicitkorea.org/koreait_policy/koreait_policy_10910.jsp http://www.fcc.gov/broadband/ http://www.usda.gov/rus/telecom/broadband.htm#info http://www.usda.gov/rus/telecom/initiatives/ index_initiatives.htm#broadband
689
Chapter XLIII
Broadband Diffusion and its Driving Forces Banani Nandi1 AT&T Shannon Laboratories, USA Chandana Chakraborty Montclair State University, USA
ABSTRACT In the light of the emerging consensus on the potential impact of broadband technology on economic growth and development, this chapter analyzes the cross-country differences in growth of broadband technology by examining the key demand and supply factors driving diffusion in the observed countries. In addition, utilizing empirical evidence and country case analyses, the chapter offers tentative policy suggestions for accelerating broadband diffusion under alternative circumstances.
INTRODUCTION Existing literature on endogenous growth theory2 VXJJHVWVWKDWHI¿FLHQWWUDQVIHURINQRZOHGJHDQG information, because of its ability to produce spillover effects, holds the key to high economic growth performance across the world. Extending this conclusion further, modern literature on information and communications technology argues that LQWRGD\¶VLQIRUPDWLRQDJHDFFHVVWRLQIRUPDWLRQ infrastructure services is not only a necessity for socio-economic development but also the absence of such access could become a major constraint for achieving potential development goals.3 In the context of this argument, broadband (BB) technology offers a major step forward towards worldwide access to information and its associated spillover
JURZWKEHQH¿WV4 It is expected that modern BB technology, because of its unique advantage over QDUURZEDQG WHFKQRORJ\ ZLOO SOD\ D VLJQL¿FDQW role in the 21st century information transmission process by becoming the primary vehicle for accessing information on a worldwide scale. Over the past few decades, different components of telecommunications infrastructure have evolved dramatically. Rapid technological progress has occurred in both wired and wireless networks. In wired network, technological development caused the transition from copper and coaxial transmission V\VWHPVWR¿EHURSWLFDQGVLJQDOLQJDQGHQFRGing systems have changed from analog to digital. Further, digital technology has allowed for the introduction of packet transmission technology and facilitated the deployment of IP-based infra-
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Broadband Diffusion and its Driving Forces
structures. Simultaneously, rapid technological progress in the area of radio based telecommunications systems has increased the demand for mobile telecommunications around the world. In recent years, most of the countries with advanced telecommunications systems are paying special attention to building a network that facilitates broadband access to the Internet. The development of high capacity Internet backbone together with WKH :RUOG :LGH :HE KDV VLJQL¿FDQWO\ FKDQJHG the way business, individuals, and government function. A new wave of broadband (high speed) accessibility of the Internet has the potential to revolutionize multimedia data transmission processes by integrating networks of data, voice, and YLGHRLQRQHXQL¿HGFRPPXQLFDWLRQVQHWZRUN7KLV WHFKQRORJ\ZLOOQRWRQO\LQFUHDVHWKHHI¿FLHQF\ of all activities that can be performed using communication, it will also encourage innovation of many new applications and services. With the understanding that deployment of BB network is a key element for increased social and economic development, communication companies and policy makers are debating the best ways to develop and market BB services. In the context of these developments, this chapter focuses on analyzing the current state of BB diffusion rate in various countries across the world while paying close attention to the supply and demand related factors driving such diffusion in the world. In addition, the chapter investigates the possible role of public policy in accelerating BB technology GLIIXVLRQ WKURXJK LWV LQÀXHQFH RQ ERWK GHPDQG and supply factors. The layout of the chapter is as follows. The section that follows introduces the characteristics of broadband access technology and its associated EHQH¿WVLQJHQHUDO&RXQWU\VSHFL¿FGLIIHUHQFHV in performance of broadband access technology are then analyzed while taking account of the underlying structures of demand and supply. Together with outlining a new empirical model, existing approaches to demand and supply estimates of broadband services are summarized. Complementary country case analyses are then presented. Summary and conclusions are provided in the last section.
690
BROADBAND ACCESS AND ITS BENEFITS Broadband technology and its use in offering integrated network services are altogether a new phenomenon in information and communications technology. Much hope is being raised in the literature regarding its potential to effect dramatic changes in information transmission process. The dynamics of its ongoing evolution, however, kept UHVHDUFKHUVIURPSURYLGLQJDGH¿QLWLRQRI%%WHFKnology in precise terms. In operational terms, BB now refers to a set of electronic communications technological solutions. The primary features that distinguish this technology from other competitive transmission technologies include its relatively high bandwidth, always on functionality,5 and capability for high-speed information transmission in both directions—downstream from the Internet to the users and upstream, from users to the Internet. An integrated view of these features, however, does QRWVXJJHVWDVSHFL¿FVSHHGRUDVSHFL¿FVHUYLFH In fact, there is no universally accepted standard regarding the speed of transmission that characterizes “broadband.” Instead, a myriad of speed transmission standards have been proposed in the literature. The U.S. Federal Communications &RPPLVVLRQ )&& LGHQWL¿HVDEDQGZLGWKRI Kbit/s for downstream transmission as the cut off point for BB service and allows for its shift upward over time as technology changes.6 International Telecommunications Union (ITU, 2003a), on the RWKHUKDQGGH¿QHV%%DVDWUDQVPLVVLRQFDSDFLW\ with speed faster than the primary rate of ISDN, indicating that of 1.5 to 2 Mbit/s. Finally, National Research Council (NRC, 2002) proposes keeping its VSHFL¿FDWLRQDOWRJHWKHURSHQDOORZLQJLWWRHYROYH with the technological dynamics of transmission mechanism and changing consumer needs of transmission speed. Empirical observation on BB services around the world indicates that these services can be accessed through a number of different network technologies. These access technologies differ between themselves by their corresponding cost of development and functionality. The commonly known access technology networks of digital
Broadband Diffusion and its Driving Forces
subscriber line (DSL), cable modem, and power communications, refer to evolved stages of existing network infrastructures that require negligible additional sunk cost for implementation. The UHODWLYHO\QHZHUQHWZRUNVRI¿EHUWRWKHSUHPLVHV )773 ¿[HGZLUHOHVVDFFHVV ):$ DQGRWKHU broadband wireless access (BWA) technologies which are part of IMT-2000 or 3G families of QHWZRUNVFDOOIRUUHODWLYHO\VLJQL¿FDQWLQYHVWPHQW before serving as an effective channel for transmission of information. The rapid evolution of BB technology and its widespread appeal as an instrument capable of effecting revolutionary changes in information transmission process has inspired a number of studies in the literature (Gera, Wulong, & Lee, 1998; Loveman, 1994; Nadiri & Nandi, 2001; Roller & Waverman, 1996; Stiroh, 2002; Wildman, 1992) documenting a close link between information and commuication technology (ICT) infrastructure and productivity growth at the industry level as well as at the aggregate economy level. Many of these studies claim that BB network can enhance ICT contribution towards growth by improving the quality of existing ICT infrastructure and affecting an outward shift of the aggregate production function. In addition, these studies persuasively argue that BB capacity can increase ICT contribution by widening the choice of services through applications in voice, audio, video, e-commerce, e-education, e-health, government services, online interaction, and information sharing.7 Yet another group of studies have focused on social welfare gains from BB technology (GreenVWLHQ 6SLOOHU 7KH¿QGLQJVRIWKHVHVWXGies suggest that BB technology produces social surplus from the aggregate economy by allowing SURYLVLRQRIH[LVWLQJVHUYLFHVPRUHHI¿FLHQWO\RQ the one hand and extending the scope for introduction of newer services on the other. Further, they suggest that by improving the speed and quality of information transfer, BB technology facilitates knowledge transfer across the globe and generates spillover growth effects.8 Combined views of all WKHVH¿QGLQJVUHÀHFWV%%WHFKQRORJ\¶VSRWHQWLDO as a catalyst for growth and development in the information age.
'HVSLWHDOOWKHVHSUHGLFWHGEHQH¿WVIURP%% the diffusion rate of BB has increased only modHUDWHO\DURXQGWKHZRUOGDQGYDULHVVLJQL¿FDQWO\ across the countries. While stringent regulatory measures and their impact on investment cost KDYHFRPPRQO\EHHQLGHQWL¿HGDVWKHPDLQIDFtors responsible for such moderate achievement,9 WKH UROH RI FRXQWU\ VSHFL¿F VXSSO\ DQG GHPDQG LQLQÀXHQFLQJ%%GLIIXVLRQUDWHKDVQRW\HWEHHQ fully explored. Beginning with a close look at the regional and country level differences in BB diffusion rate and its associated choice of access platforms, the following section closes this gap by SURYLGLQJDQLQGHSWKDQDO\VLVRIFRXQWU\VSHFL¿F supply and demand of BB services as well as that of their interaction.
BROADBAND DIFFUSION: UNDERLYING DEMAND AND SUPPLY FORCES Summary Statistics Use of Internet services has gained its momentum since the beginning of the 1990s. Based on ,78¶V (2006a) report, the number of Internet users around the world has reached a total of 840 million by the end of 2004, indicating 13.2% of the total population. Increased demand for Internet use has induced innovation and diffusion of BB technology since late 1990s. As reported by ITU (2006a), the total QXPEHURI¿[HGOLQHEURDGEDQGVXEVFULSWLRQVKDG reached 160 million at the end of 2004 accounting for 2.5% of the world population and 38% of world Internet subscribers. This is in comparison WRELOOLRQ¿[HGOLQHVXEVFULEHUVDQGELOOLRQ mobile subscribers (including both second and third generation mobile subscribers) accounting for 19% and 28% of the world population respectively. $WWKHHQGRIWKHWRWDOQXPEHURI¿[HGOLQH broadband subscribers has increased to about 216 million. The distribution of these 216 million BB subscribers by different region around the world is reported in Table 1. It is apparent from Table 1 that more than 38% of the world subscribers have their home in the
691
Broadband Diffusion and its Driving Forces
7DEOH'LVWULEXWLRQRIEURDGEDQGVXEVFULEHUVE\UHJLRQ Region
% of World Subscribers
$VLD3DFL¿F
38.69%
Europe & Central Asia
30.31%
Latin America & Caribbean
3.94%
Middle East & North Africa
0.99%
North America
25.97%
Sub-Saharan Africa
0.10%
TOTAL
100%
Source: ITU World Telecommunication Indicators Database (2006b) )LJXUH%URDGEDQGSHQHWUDWLRQ OHIW %URDGEDQGSHQHWUDWLRQ ULJKW Broadband Penetration, subs per 100 Inhabitants, by Technology, 2000 0.00
2.00
4.00
6.00
8.00
Korea (Rep. of)
8.42 6.67
Hongkong 4.58
Canada 2.51
United States
Netherlands Belgium
1.06
Taiw an, China
1.03
Sw eden
25.24
Korea (Rep. of) Netherlands
25.15
Denmark
24.87 24.76 23.56
Sw itzerland
1.63 1.23
26.54
Iceland
Hongkong
1.89
Denmark
0.00 5.00 10.00 15.00 20.00 25.00 30.00
Liechtenstein
2.38
Austria Singapore
10.00
Broadband Penetration, subs per 100 Inhabitants, by Technology, 2005
Cable DSl Other
0.93
23.13
Finland
22.37
Norw ay
21.46
Canada Sw eden
20.78 20.30 20.21
Macau
0.86
Taiw an, China
Iceland
0.84
Belgium
Japan
0.67
Israel
17.82
DSl
Sw itzerland
0.60
Japan
17.46
Other
Broadband Subs per 100 Inhabitants
19.07
Cable
Broadband Subs per 100 Inhabitants
Source: ITU’s (2006b) World Telecommunications Indicator Database
$VLD 3DFL¿F UHJLRQ :LWK WKHLU VKDUHV RI ZRUOG subscribers at about 30% and 26% respectively, both North America and Europe also play a relatively VLJQL¿FDQW UROH WRZDUGV WKH VSUHDG RI %% VXEscription.10 All other remaining regions, however, contribute only marginally towards subscription of BB technology. Complementary to Table 1, a country-bycountry breakdown of BB diffusion rate and its
692
decomposition by various network platforms for year 2000 and 2005 are presented in Figure 1. 7KHVH WZR ¿JXUHV RIIHU DGGLWLRQDO LQVLJKWV LQWR trends and patterns of diffusion for the most successful (top 15 countries) BB countries around the world. With national diffusion rate measured in terms of number of subscribers per 100 inhabitDQWV$VLD3DFL¿FUHJLRQ¶VOHDGHUVKLSLQWKH\HDU 2000 has mainly been driven by two countries:
Broadband Diffusion and its Driving Forces
7DEOH'LVWULEXWLRQRIZRUOGEURDGEDQGVXEVFULEHUVE\WHFKQRORJ\ Broadband Platform
% of Subscribers
Cable modem
29.5%
DSL
63.9%
Other technology
6.6%
(Satellite, FTTH, Ethernet LANS, etc.)
Source: ITU’s (2006b) World Telecommunication Indicators Database
the Republic of Korea and Hong Kong. The rank order of countries in terms of their BB penetration rate, places these two countries in the top two positions respectively in the referred year. It is, however, interesting to note that though these countries have maintained growth in BB diffusion rate over the period 2000-2005, a few of the Scandinavian countries, especially Iceland, have shown remarkable progress in catching up with their performance in the year 2005.11 Regarding the choice of network technology for delivery of BB services, a majority of the subscribers in the world, as indicated in Table 2, obtained their BB using DSL in 2005. Such prominence of DSL might be indicative of its cost advantage over other competing network services; in addition, this PLJKWSDUWO\EHUHÀHFWLYHRIWKHOLPLWHGFKRLFHXVers face in many countries where cable networks DUHHLWKHUXQDYDLODEOHRULQVXI¿FLHQW The observed difference in diffusion rates and choice of platforms across countries makes it imperative that we take a close look at the country VSHFL¿FGHPDQGDQGVXSSO\UHVSRQVHVKDSLQJWKH BB markets. A number of studies in the literature have directed their focus in identifying success factors FDSDEOHRILQÀXHQFLQJWKHOHYHORI%%SHQHWUDWLRQLQ a country. The International Telecommunications Union (2003a) conducted various studies based on ZRUOGZLGHFRXQWU\OHYHOGDWDDQGLGHQWL¿HGDOLVW of ten success factors for BB uptake: competition, innovation, applications, procompetitive regulation, price, speed, marketing, high ICT usage, urban demographics, and benchmarking. Tajiti and Okazaki (2006) emphasized more on supply
related factors for success with BB diffusion rate and focused mainly on country level availability of existing networks and competition. Directing their attention to demand related factors, Bauer, Kim, and Waldman (2005), in their cross country VWXG\RI%%XSWDNHLGHQWL¿HGSRSXODWLRQGHQVLW\ and preparedness as important determining factors for growth of BB penetration rate in the selected OECD countries. In yet another study, Garcia0XULOORDQG*DEHO LGHQWL¿HGLQFRPHOHYHO SULFHFRPSHWLWLRQDQGSULYDWL]DWLRQDVVLJQL¿FDQW IDFWRUVLQÀXHQFLQJ%%SHQHWUDWLRQ'UDZLQJRQ available country level experience and related statistics, the following two sections critically evaluate WKHUROHRIVXSSO\DQGGHPDQGLQLQÀXHQFLQJ%% technology deployment and adoption and provide a summary account of their macro and micro level implications.
)DFWRUV,QÀXHQFLQJ%URDGEDQG Supply Empirical observations suggest that BB penetration KDVJURZQTXLWHVLJQL¿FDQWO\RYHUWKHODVWGHFDGH Yet it is apparent that while certain economies are advancing at a faster pace, a number of other countries are experiencing serious technological gap with BB penetration rate below even one percent. The reasons are straightforward. High ¿[HG FRVWV RI GHSOR\PHQW RI %% QHWZRUNV DQG the time required in recouping such costs, forces the relatively lower-income countries to shy away from widespread use of BB technology. In addition, investors in these countries face a series of potential risks, for example, demand risk, inter-
693
Broadband Diffusion and its Driving Forces
est risk, regulatory risks, and risk of changes in WHFKQRORJ\ 7KH VLJQL¿FDQFH RI WKHVH ULVNV LQ adding further uncertainty to rates of return on such investment cannot be overemphasized. Recent emphasis on privatization of infrastructure GHYHORSPHQWUHÀHFWVWKHHIIRUWVWRZDUGVWUDQVIHU of these risks to private operators. It is expected that such transfer by inducing the pace of development of new infrastructure and improving that of the existing one would reduce the pressure on government budget. However, to achieve this goal HI¿FLHQWO\UHVSHFWLYHFRXQWU\JRYHUQPHQWVZRXOG need to advocate policy environment conducive for attracting private investment with minimum macroeconomic risk commitment. In countries, or VSHFL¿FDUHDVZKHUHLQVXI¿FLHQWUHWXUQRQLQYHVWment and high risk create hindrances for growth of investment, governments can design appropriate incentives for achieving higher expected returns.12 Country governments can also encourage private investment by introducing risk sharing programs and promoting a healthy regulatory environment. For example, by introducing BB connections in institutions such as schools, hospitals, and convention centers, government can provide a wider customer base for BB service providers and thus can reduce the associated demand risks. Moreover, to attract investment in BB technology, government can provide indirect subsidies to private investors through tax credit and low interest rate loans. In addition to expected return on investment and its potential risks, spatial density of users FDQDOVREHFRQVLGHUHGVLJQL¿FDQWIRULQÀXHQFLQJ supply of BB. Network sharing characteristics of BB technology users predict that high spatial density of users reduces cost of infrastructure per user in urban areas. In addition, concentration of potential users, by causing revenue-cost ratio to be high in urban areas, encourages investments on network deployment in those areas. Utilizing data from OECD countries, Cava-Ferreruela and Alabau-Munoz (2006), in their recent study, found VWDWLVWLFDOO\VLJQL¿FDQWFRUUHODWLRQEHWZHHQOHYHO of urbanization and BB penetration rate. Utilizing theoretical logic and empirical evidence, it may, therefore, be argued that urbanization in general facilitates diffusion of BB technology while
694
keeping investment cost at a relatively low level. Available statistics on population density, level of urbanization, and BB penetration in some of the most successful broadband economies in the world well support this argument. These statistics are presented in Table 3. Clearly, Canada and Iceland, with their low population density and high BB penetration rates, GH¿QH H[FHSWLRQV WR WKH UHVW RI WKH FRXQWULHV included in the panel. The reason behind such DSSDUHQW FRQÀLFW KRZHYHU LV VLPSOH $OWKRXJK both the referred countries have low population density, most of their population is concentrated in the major urban areas. According to ITU (2003a) reports, 78% of the Canadian population lives in cities and towns and more than 50% of them are concentrated in 10 large urban centers. In Iceland, 92 % of the population lives in urban areas with 62 % of them living in the capital City of Reykjavik. Based on these observations, it is quite possible that by targeting deployment of BB technology in major urban areas, public and private funds could EH XWLOL]HG PXFK PRUH HI¿FLHQWO\ WR DWWDLQ KLJK diffusion rate at the national level. Furthermore, market competition plays a sigQL¿FDQWUROHLQLQÀXHQFLQJVXSSO\RI%%VHUYLFHV. The general belief is that government, with proper regulatory measures, can create a competitive environment (Crandall & Sidak, 1995; Gruber & Denni, 2005; Tajiri & Okazaki, 2006) conducive to driving faster deployment of BB technology. Complexity of competition in infant infrastructure makes it harder to identify and establish appropriate policy measures for competition. In fact, many countries are facing the challenge of selecting the best form of competition that can foster advancement of BB technology and satisfy the interests of both investors and consumers. Special characteristics of the telecommunication sector reveal that competition can be introduced primarily through two channels: (1) service-level competition on network facility, or (2) facility based competition. Even though facility based competition seems to be the less controversial one, at least from a regulatory point of view, the policy makers in many countries are encouraging service based competition as a transitional measure for reducing entry costs of
Broadband Diffusion and its Driving Forces
Table 3. Demographics and broadband penetration, 2002 Economy
Broadband penetration (per
Population density (per sq.
Population residing in urban
100 Inhabitants)
km)
areas (%)
Korea (Rep.)
21.29
484
81.9
Hong Kong, China
14.90
6378
100.0
Canada
11.19
3
78.7
Taiwan, China
9.35
624
83.5
Denmark
8.60
125
85.1
Iceland
8.43
3
92.5
Belgium
8.41
338
97.3
Sweden
7.83
20
83.3
Netherlands
7.22
393
89.5
Japan
7.13
338
78.8
United States
6.89
31
77.2
Australia
6.61
97
67.3
Switzerland
6.32
176
67.4
Singapore
5.53
6099
100.0
Malta
4.46
1253
90.9
Germany
3.92
231
87.5
Macao, China
3.87
18391
98.8
St. Kits and Nevis
3.62
180
34.1
Estonia
3.37
30
69.4
Slovenia
2.84
99
49.2
Source: ITU’s (2006b) World Telecommunications Indicators Database
competitors. Many of these countries have mandated certain incumbents to allow competitors use of their networks for provisioning of competitive BB services. The regulatory rule of unbundling local loop of telephone network is an example of such a regulatory measure adopted in many countries; reported experiences suggest that this rule may have fostered competition in BB services that use DSL as its transmission mechanism. The application of such regulatory measures, however, is uneven. “Open access” is the analogous method of making cable networks available to competitive ISPs for offering BB access service. However, only a few countries have mandated cable “open access.”13 The observed rareness of cable openness policy can be explained by the recent trends
in regulatory initiatives. In most of the countries, cable-TV networks have been regulated differently than telecommunication networks and are viewed as facing more competition from satellite and over-the-air video services. As a result, while both cable-TV network and telecommunication network are capable of provisioning BB access services, ZH¿QGDV\PPHWULFUHJXODWRU\VWUXFWXUHEHWZHHQ the two networks which may have constrained BB technology growth in many of these countries (Crandall, Sidak, & Singer, 2002). Recognizing this, ITU, in its recent report (ITU, 2006a), has recommended technologically neutral regulation for BB service provisioning with focus directed towards encouraging construction of and investment on BB access network. It is expected that such
695
Broadband Diffusion and its Driving Forces
ÀH[LELOLW\LQUHJXODWLRQPLJKWHQFRXUDJH%%DFFHVV deployment in rural and underdeveloped areas, and in developing countries, by allowing licensees the choice over a set of network technologies to deliver services. Stated differently, such regulation would allow licensees to choose the delivering network technology that minimizes costs while increasing WKHLU¿QDQFLDOUHWXUQ As we mentioned before, there is no single technology through which BB access service can be delivered. Currently BB access service can be supplied using various technological platforms. Service providers can choose from a variety of platIRUPVLQFOXGLQJ'6/FDEOH79¿EHUWRWKHKRPH (FTTH),14DQG¿[HGZLUHOHVV ): 'XHWRKLJK ¿[HGFRVW%%DFFHVVSURYLGHUV¶FKRLFHRIWHFKQROogy is dependent on the existing network they can use for providing BB service with minimum cost; such choice offers them the scope for extracting maximum value from the legacy network. As noted earlier, this has prompted increased popularity of the DSL platform. DSL protects value of existing copper wired old telephone networks. In contrast, cable providers adopt cable modem service to protect the value of their coaxial cable networks. FTTH network development has been very costly LQWKHSDVW5HFHQWUHGXFWLRQLQWKHFRVWRI¿EHU has made it easier to employ FTTH, but it still remains relatively expensive. However, to protect the value of their existing networks, the telephone and cable companies in most of the countries have fewer incentives to deploy FTTH.15 In addition, the “open access” regulatory policy in many of the observed countries has discouraged investors to invest further on deploying FTTH networks. With regard to wireless BB access service, the principal drivers are the availability, cost of spectrum, and the cost of end-users devices. From the cost point of view, FW access appears to be an economically viable option in the BB market place, especially in rural and remote areas, compared to wire-line and satellite alternatives. However, an important problem faced by BB wireless providers may be the lack of cost effective backhaul infrastructure required for connecting wireless tower to backbone Internet infrastructure (ITU, 2003a). The alternaWLYHVVXFKDVVDWHOOLWHRUSRLQWWRSRLQWZLUHOHVV¿[HG
696
OLQNVDUHVLJQL¿FDQWO\PRUHH[SHQVLYHFRPSDUHGWR the cost of providing a wireless base station for BB access. The ability of BW access to improve over time relies on the amount of spectrum regulators make available for BWA. Thus, the fundamental challenge for spectrum regulators is how to ef¿FLHQWO\DQGHIIHFWLYHO\GLVWULEXWHQHZVSHFWUXP resources. Although regulators can avoid this challenge by granting licensees unlimited technological ÀH[LELOLW\DQGRSHUDWLRQDODXWRQRP\WKH\ZRXOG have to ensure the existence of a proper inter-modal competitive environment for the licensees to make WKHPRVWHI¿FLHQWXVHRIWKHVHVSHFWUXPUHVRXUFHVLQ such cases. The proper choice of regulatory policies on spectrum allocation is still in an experimental stage in many countries. Recently, many countries have explored the scope for using other existing infrastructures capable of providing BB business and residential VHUYLFHV7KLVEHKDYLRUVLPSO\UHÀHFWVHIIRUWVWR reduce immediate costs of BB expansion. In Japan and Iceland, power companies are leveraging their existing infrastructure, which are often made of ¿EHURSWLFFDEOHV7KHH[FHVVFDSDFLWLHVRIWKHVHLQfrastructures are being seen as a potential medium for providing BB services. Following this ideology, India recently is examining the scope for using their widely developed railway infrastructure16 to provide BB Internet access at a cheaper price and a faster speed.
)DFWRUV,QÀXHQFLQJ%URDGEDQG Demand Although availability of BB access service is the primary criterion for BB diffusion, importance RI IDFWRUV VXFK DV XVHUV¶ QHHGV DZDUHQHVV DQG affordability cannot be ignored for their potential impact on demand for BB access. Awareness of users about BB technology and its potential range of services are indeed critical for generating demand. Similar to all new techQRORJ\SURGXFWVDQGVHUYLFHVWKHVSHFL¿FQDWXUH RI %% VHUYLFH PDNHV LW GLI¿FXOW IRU FRQVXPHUV WRDSSUHFLDWHLWVEHQH¿WVDWWKHLQLWLDOVWDJHRILWV development. With increased experience of the EHQH¿WVGHPDQGIRU%%VHUYLFHDFFHVVLVOLNHO\WR
Broadband Diffusion and its Driving Forces
increase gradually. In general, schools and colleges provide an ideal environment for introduction of BB technology and its associated services; in addition WRSURGXFLQJHGXFDWLRQDOEHQH¿WVRILQIRUPDWLRQ access, BB technology access in these institutions promotes awareness of the potential users. Being FRQYLQFHGRIWKHVHGXDOEHQH¿WVPDQ\FRXQWULHV have taken initiatives to introduce BB awareness through its educational programs. The United Kingdom, for example, has funded schools for desired BB connection since 2003. The program is expected to extend coverage for all schools by the end of 2006. Following 1997 FCC directives, E-Rate funds in the U.S. also subsidizes BB access in schools and libraries.15 The Estonian government has also demonstrated support of similar effort. Following a similar policy choice, Singapore has been engaged in raising general awareness of DSSOLFDWLRQVFRQWHQWDQGEHQH¿WVRI%%DFFHVV service and promoting demand through its public service announcement programs (ITU, 2006a). Many other countries are considering bundling of BB with other similar services such as telephone or cable-TV as yet another effective instrument for promoting awareness of BB technology among households. Further, policymakers in almost all countries are exploring the possibility of allowing computer network users to share a BB connection. The expectation is that such sharing, by reducing the effective price per user, will help increase the use of BB access service and, hence, raise its demand. Consequently, awareness of BB technology EHQH¿WVLVDQHVVHQWLDOSUHUHTXLVLWHIRUJHQHUDWLQJ economically viable demand for these services. &RQVXPHUV¶DIIRUGDELOLW\LVHTXDOO\LPSRUWDQW for stimulating demand for BB services. Affordability depends on price of the service and the ability of the consumers to pay the market price. Also, affordability is a major issue for low as well as high-income countries.18 The expectation that affordability and the consequent demand will vary between users has led many policy makers WREHOLHYHWKDWÀH[LEOHSULFLQJVFKHPHVZLOOEHDSpropriate to encourage adoption of BB among large section of the population in both higher and lower income countries. Agreeing with this conclusion, many BB suppliers have introduced tiered pricing
plans that take into account all possible range of connection speeds. Such pricing schemes open up more options to consumers with different affordability and different needs. The expectation, however, is that given more or less similar price of BB access service, demand for BB may differ among countries due to differences in price-income relationships. Summary statistics on price per 100 kbit/s and its share in monthly income for the 20 cheapest broadband countries reported in Table 4 offers evidence in support of such expectation. ITU reported statistics (ITU, 2006b) also show that demand for BB is very much correlated with ICT infrastructure and usage. The higher the Internet usage, the higher is the adoption rate of BB services. South Korea offers the best example in this respect; both Internet usage and BB penetration are high in South Korea (ITU Internet Case Study, 2003b). A similar trend has also been observed in the OECD countries.20 Drawing on such H[SHULHQFHV PDQ\ VFKRODUV KDYH LGHQWL¿HG ,&7 literacy as an important factor driving demand for BB services. In addition, the availability of BB applications and online contents that require high bandwidth have also been considered an imSRUWDQWIDFWRULQÀXHQFLQJGHPDQGIRU%%DFFHVV Consequently, many scholars have argued that government could play an important role in raising BB demand by providing personal computers at a subsidized rate and by supporting development of local content (Tajiri & Okazaki, 2006). In sum, country experiences suggest that government intervention through appropriate policies and incentives has played a key role in creating demand for BB services around the world.
TECHNIQUES OF BB PENETRATION ESTIMATION Existing Models Expecting market tipping of BB technology in the near future, researchers have often used empirical demand and supply functions to estimate BB technology penetration among potential users. Most popular among these are the reduced form
697
Broadband Diffusion and its Driving Forces
Table 4. Broadband price as a percentage of monthly income (20 cheapest countries, 2003) Country
Subscription/month
Price per 100 Kbits/s
Subscription as %
(US$)
(US$)
monthly income
Price per 100 kbit/s as % of monthly income
Japan
$24.19
$0.09
0.87%
< 0.01%
Korea (Rep.)
$49.23
$0.25
5.95%
0.03%
Belgium
$34.41
$1.15
1.78%
0.06%
Hong Kong, China
$38.21
$1.27
1.85%
0.06%
United States
$52.99
$3.53
1.81%
0.12%
Singapore
$33.18
$2.21
1.92%
0.13%
Netherlands
$51.55
$3.36
2.58%
0.17%
Canada
$32.48
$3.25
1.75%
0.17%
Norway
$46.16
$6.56
1.46%
0.21%
Macao, China
$38.34
$2.56
3.20%
0.21%
Germany
$33.93
$4.42
1.80%
0.23%
New Zealand
$40.61
$2.71
3.55%
0.24%
Austria
$45.20
$5.89
2.32%
0.30%
United Kingdom
$32.59
$6.37
1.55%
0.30%
Switzerland
$57.84
$11.30
1.83%
0.36%
Italy
$73.59
$6.13
4.66%
0.39%
Sweden
$44.56
$8.91
2.15%
0.43%
Slovenia
$79.54
$3.88
9.73%
0.48%
France
$51.46
$10.05
2.81%
0.55%
Luxembourg
$91.77
$17.92
2.84%
0.55%
Source: ITU’s (2003a) Internet Report, Birth of Broadband Note: The methodology used by ITU (ITU, 2003a) in estimating the prices displayed in Table 4 is as follows: Broadband prices per month represent general prices of BB in an economy and not the average cost of BB in that country. BB price per 100 kbit/s, reported in column 3, represents the ratio between BB price per month and the down stream speed (maximum advertised download speed) in percentage form. BB price as a percentage of monthly income, shown on column 7, on the other hand, indicates the ratio between BB prices per month and the monthly average of per capita income; the latter variable has been computed from country level data on gross national income and population.19
demand model proposed by Bauer et al. (2005) and the broadband access model proposed by Glenn Gilmour (2002). A summary review of each of these models is outlined below.
Bauer et al. (2005) Model Using cross-country data on BB uptake from 30 2(&'FRXQWULHV%DXHUHWDO LGHQWL¿HGWKH IDFWRUVLQÀXHQFLQJ%%SHQHWUDWLRQDWWKHDJJUHJDWH economy level. Focusing on supply and demand
698
of BB services at the aggregate level, Bauer et al. derive a simple reduced form model for the purpose of empirical estimation. The supply and demand functions used in their model are as follows:
qd = f (price, price of substitute, income, preparedness) qs = g (price, competition, cost - condition)
Broadband Diffusion and its Driving Forces
In equilibrium, the following holds: qd = qs with qd = Local demand for BB qs = Local supply of BB It is important to note that in addition to the price RI%%VHUYLFHVVXSSO\RI%%VHUYLFHVLVGH¿QHG to be a positive function of two other shift factors, namely, intensity of local access competition, and the cost of building a network. While increased intensity of local access is expected to shift the supply curve downward, increased cost of network building is expected to shift the supply curve in an upward direction. Local demand, on the other hand, is taken to be a function of price of BB, the price of substitute technologies, income, and a set of preparedness factors that includes education, familiarity of the service, and attitude towards information technology products and services. The own price movement causes the movement along a demand curve whereas changes in other factors cause a shift in the demand curve. With crossnational data available only at an aggregate level, Bauer et al. (2005) employed the aggregation rule to derive the national level relation as follows:
QD
¦q
d
i
Q
s
¦q
s
i
Qd
Qs
(1)
Application of the linear aggregation rule helps GH¿QH WKH DJJUHJDWH VXSSO\ DQG GHPDQG RI %% services as functions of national averages of price and competitive conditions. Utilizing equilibrium condition (1), the resulting reduced form function of broadband penetration takes the following form: Q = f (price, price of substitutes, income, prepared(2) ness, competition, cost conditions)
$SULRULH[SHFWDWLRQVRQWKHFRHI¿FLHQWVRIWKH model are as follows. The expected sign for the FRHI¿FLHQWVRIH[SODQDWRU\YDULDEOHV³SULFH´DQG ³FRVW´DUHQHJDWLYH H[SHFWHGVLJQVRIFRHI¿FLHQWV for the rest of the explanatory variables including WKRVH LGHQWL¿HG DV SUHSDUHGQHVV YDULDEOHV DUH positive. 7RDFFRXQWIRUFRXQWU\VSHFL¿FGLIIHUHQFHVLQ demand and supply, Bauer et al. (2005) also include the two variables of population density and policy dummy in their model. Population density variable UHÀHFWVWKHFRVWFRQGLWLRQVZKLOHSROLF\GXPP\ variable helps evaluate the impacts of alternative policy regime. The results of the Bauer et al. (2005) model suggest that despite showing an inverse relation with demand for BB services, price of BB services fails to provide any strong impact on BB penetration rate. Further, the results indicate that the variables of income, dial-up price, and competition are staWLVWLFDOO\LQVLJQL¿FDQW7KHFRQFOXVLRQRIWKHPRGHO suggests that the level of BB uptake is explained mainly by the cost conditions of network deployment and the preparedness of a nation to utilize the available technology.
Broadband Access Model (BAM) Deviating from the Bauer et al. (2005) model, Gilmour (2002) analyzes BB take-up rate within Australia by focusing on a variety of demand systems capable of effecting BB penetration within the economy. In particular, the BAM model considers BB Internet access platforms of NB, DSL, and HFC (cable) as explanatory variables for forecasting BB diffusion rate in Australia. Demand for BB penetration or diffusion rate in the BAM model is assumed to be a function of installation price, subscription price, price of substitute products, and expenditure and their respective elasticities. 7KH VSHFL¿F IRUP RI WKH QRQOLQHDU IXQFWLRQ LV given below: FP
ScaleFactor* Pins * Psubs * Pcross * m
699
Broadband Diffusion and its Driving Forces
where FP: Final broadband penetration rate Pins: Installation price with elasticity element D Psubs: Subscriber price with elasticity element E Pcross: Cross price of substitute service with element J m: expenditure with elasticity element G The cross-price effect between narrowband (NB) and broadband (DSL or HFC) in this demand function is assumed to be asymmetric. In other words, the price change effect of BB on NB is expected to be higher compared to that of the effect of NB on BB. This is simply to suggest that people are less likely to migrate back from BB WRQDUURZEDQG%$0¶VGHPDQGPRGHOJHQHUDWHV three logistic diffusion curves for NBI, DSL, and HFC respectively. Different versions of the model are estimated to accommodate market behavior with respect to changing relationships between NB, DSL, and HFC. The diffusion curves of the model are as follows: 1. 2. 3.
Constrained diffusion curve system Unconstrained diffusion curve system Linear approximation of the almost ideal GHPDQGV\VWHP /$,'6 EDVHGRQÀH[LEOH logistic diffusion curve of total penetration of the three Internet service platforms
7KH/$,'6V\VWHPGH¿QHVDQDGGLQJXSUHstriction for the value share (dependent variable) in the equation. The value share indicates the value of either NBI or HFC or DSL as a proportion of total value of household Internet use. Utilizing simulation runs and regression analysis, LA/IDS demand systems are generated for the BAM model. 7KHHVWLPDWHG/$,'6FRHI¿FLHQWVDUHWKHQXVHGDV parameters in the model for forecast purposes. Employing 2000-2002 Australian data on relevant variables, the BAM model shows that BB access platforms are highly income elastic. Own price elasticity estimates of the model suggest that narrow band penetration is relatively more inelastic than broadband penetration.21 Results on cross-price effects, however, remain inconclusive in the model.
700
An Alternative Model and its Microeconomic Implications While appreciating the approaches referred in the above two models, we propose a slightly different model in this section to allow for required government intervention in driving the BB diffusion goal. We consider such formulation as an improvement over the previous models because of its ability to HYDOXDWHFRXQWU\VSHFL¿FSXEOLFSROLF\LPSDFWRQ supply and demand of BB technology adoption. The model assumes two primary players in the process of infrastructure development through private initiatives: (1) network owner company; and UHJXODWRU(DFKKDVLWVRZQVSHFL¿FREMHFWLYH 1HWZRUN RZQHU ¿UP¶V REMHFWLYH LV WR PD[LPL]H SUR¿WV RXW RI LQYHVWPHQW RQ QHWZRUN GHYHORSment through the sales of products and services. 7KHSUR¿WDELOLW\RIWKH¿UPRQWKHRWKHUKDQG LV LQÀXHQFHG E\ WKH IROORZLQJ UHJXODWRU\ environment, (2) network service demand, and (3) technological availability and cost structure. The expectation is that the decision taken by the QHWZRUNRZQHU¿UPJHQHUDWHVQHZSURGXFWVDQG services and network externality, thereby causing increase in social welfare. 7KHUHJXODWRU¶VREMHFWLYHLVWRPD[LPL]HVRFLDO and economic well-being. He can achieve these objectives by creating (1) appropriate incentives for private investment on network development, by (2) controlling ownership structure supporting those incentives, by (3) controlling price of service UHTXLUHGIRUPDLQWDLQLQJHQGXVHUV¶ZHOIDUHDQG UHDVRQDEOHSUR¿WDELOLW\RIQHWZRUNRZQHUVDQGE\ (4) helping stimulate demand through universal service program and other appropriate programs to VKDUHGHPDQGULVNDQGLQFUHDVHEHQH¿WDZDUHQHVV among end-users. $VVXPLQJWKDWD¿UPVHUYHVERWKDVQHWZRUN owner, and network service provider to end-users, and that its objective is to determine optimum level of investment in infrastructure based on the net present value criterion, the demand function IRU%%DFFHVVWHFKQRORJ\FDQEHGH¿QHGLQWKH following form:
Broadband Diffusion and its Driving Forces
f ( I ) g ( X , P, PREP), where f ( I ) ! 0 wg (.) wg (.) 0, !0 and wP wPREP D
where I represents installed level of infrastructure and P represents price of network service. X is a vector of all other explanatory variables that affects demand for BB services. This includes income of user, availability of applications, contents, and so forth. At the regional or country level, it could also include population density as an explanatory variable. Increased density of population in any region is expected to increase the number of potential BB customers in that region. This, by reducing per unit infrastructure cost for provider will then LQFUHDVHWKH¿UP¶VSUR¿WV$VUHÀHFWHGE\WKHGHmand function, BB access demand increases with improvement in infrastructure level and decreases in price. Demand also increases due to improvement in awareness or preparedness factors. )LUP¶VFDVKÀRZRUJURVVSUR¿WS is given by: t
( X , P, I )
P D(.) W ( I t I t 1 ),
where It is the infrastructure level at time t, and W is the unit economic cost of infrastructure installation.22 At the time t0 ¿UPPD[LPL]HVQHWSUHVHQW YDOXHRISUR¿WRUVXPRIGLVFRXQWHGSUR¿WVRYHUWKH time period t = t0 to tf 1HWSUHVHQWYDOXHRISUR¿WLV given by the following equation (see Box 1). 5HJXODWRUV FDQ LQÀXHQFH WKH ¿UP¶V GHFLVLRQ regarding the level of infrastructure installation by GLUHFWO\RULQGLUHFWO\LQÀXHQFLQJW, G, and P and PREP. In addition, regulators can LQÀXHQFHW by regulating ownership structure of infrastructure networks, by providing tax incentives, and so forth. $OVRWKH\FDQLQÀXHQFHWKHDQQXDOGLVFRXQWIDFWRU
G by providing some form of guarantee to share future commercial risks of investing on infrastrucWXUH)XUWKHUWKH\FDQLQÀXHQFHSULFHP in many different ways: by allowing network investors to recover their sunk cost and by allowing them to charge cost-covering service price while keeping control on price through different mechanisms of SULFHVXEVLGL]DWLRQWRSUHVHUYHHQGXVHUV¶LQWHUHVWV DQGUHJXODWRU¶VJRDO*RYHUQPHQWFDQDOVRLQÀXHQFH the PREP variable by proper demand campaign training, promotion and subsidies and help shift the demand curve outwards. 7KRXJKWKHPRGHOGH¿QHVDQLPSURYHPHQWRYHU the previous ones in terms of policy evaluation, the relationship between demand and supply factors is very complex and cannot be easily captured in econometric works. Such estimation also requires much richer data which is very much lacking regarding BB services. Therefore, an increasing number of studies have relied on case study methodologies WRVKHGOLJKWRQWKHIDFWRUVLQÀXHQFLQJGLIIXVLRQ RI %% DQG WKH OLQNDJH EHWZHHQ VSHFL¿F SROLF\ formulation and its impact on the progress of BB penetration. Following these researchers, we analyze the experience of successful BB countries.
BROADBAND DIFFUSION IN SELECTED COUNTRIES: SUCCESS FACTORS AND POLICY SUGGESTIONS In general, any infrastructure development requires DODUJHLQYHVWPHQWIXQGZKLFKSXWVVLJQL¿FDQWSUHVsure on national budgets. For developing countries with limited public funds, the public sector requires VLJQL¿FDQWDPRXQWVRIERUURZLQJWR¿QDQFHVXFK development and thus cause supply inelasticity.
Box 1. tf
PV ( I , X , P, PREP, W )
¦ [ P f ( I ) g ( X , PREP, P ) W ( I t
t
t
t
t
I t 1 )] / (1 )
t t f
t t0
where GUHSUHVHQWV¿UP¶VDQQXDOGLVFRXQWUDWH
701
Broadband Diffusion and its Driving Forces
'ULYHQE\¿VFDOFRQVWUDLQWDQGJURZLQJGLVFRQWHQW with performance of state-provided infrastructure services, more and more governments have turned WRWKHSULYDWHVHFWRUIRU¿QDQFLQJDQGSURYLVLRQLQJ telecommunications services and other infrastructures. Clearly, such investment subjects investors to various risks creating uncertainty on their rates of return. On the other hand, since infrastructure investment often generates substantial amount of network externalities and spillover effects, the social rate of return on such investment always exceeds private rate of return. To optimize social rate of return, government intervention, thus, repeatedly becomes a necessity. Though government does not invest directly, it often shares some of the risks of private investors by providing variRXVJXDUDQWHHV¿QDQFLDOLQFHQWLYHVDQGVXSSRUW structures, and by establishing an effective regulatory environment.23 The case for government intervention in BB technology network, however, cannot be presented following the referred line of reasoning. Since BB is a new technology service, the availability of its infrastructure and the use of such service are not yet at their optimum level for SURGXFLQJVRFLDOEHQH¿W*RYHUQPHQWLQWHUYHQWLRQ in encouraging demand for use of such infrastructure services, thus, needs to precede designing of an optimum BB infrastructure. Notwithstanding the GLI¿FXOW\RIFKRRVLQJWKHULJKWVHWRILQWHUYHQWLRQ policies, examples of government intervention in promoting BB technology growth abound in the literature. Evaluating a variety of government incentive plans used for successful promotion of BB infrastructure, we attempt to provide a tentative policy outline for BB deployment and adoption in the rest of this section. Being aware of the varied extent of regulatory intervention in Eastern and Western countries, we decide to focus mainly on the experiences of South Korea and U.S. with occasional references on those of Iceland and other successful BB economies. According to ITU (2006a), the Republic of Korea is the world leader of BB technology diffusion. Statistical estimates indicate that the country is about 3 years ahead of the global average in terms of converting Internet users to BB.24 A close look at the history of BB technology development in
702
South Korea suggests that the government took many initiatives in encouraging supply and demand of BB services in this country and adopted policies appropriate for fostering growth in BB penetration. South Korea presents a unique case with rapid deployment of BB technology, fast Internet penetration, and large extent of time its people spend online (ITU, 2006a). Rank ordering of countries indicates that with 25.24 subscribers per 100 inhabitants (ITU, 2006b), South Korea is the most wired country in terms of BB penetration. Rapid growth of Internet population and the change in behavior of its population can be attributed to high penetration of BB access at an affordable rate and within a relatively shorter time period (Soung, 2005; Yun, Lee, & Lim, 2002). :LWKUHJDUGWRVSHFL¿FLQLWLDWLYHVWKH.RUHDQ government emphasized deregulation and market privatization in its telecommunications policies. Its telecom sector is one of the most liberalized sectors in Asia offering steady privatization and liberalized market situation. These referred policy initiatives prompted high-speed Internet access infrastructure and facilitated open competition in the high speed Internet market. Intense competition induced a relatively low price of BB access service subsequently causing a rapid increase in demand. In addition to liberal regulatory environment, the Korean Information Infrastructure (KIT) Project was established by South Korea in early 1995 as an avenue for promoting nationwide broadband usage. The ultimate aim of the project was to provide broadband networks to 13.5 million subscribers with the average transmission of 20 Mbit/s by the year 2005. To achieve this, government supported network providers with funding and partially eased their burden of investment in access networks. Greater funding was also reserved for remote areas. In addition, some enterprises were offered a tax exemption for deployment of BB technology infrastructure.25 From the perspective of demand for BB services, the Korean government took important initiatives in promoting growth. To name a few, the government deployed a variety of programs to
Broadband Diffusion and its Driving Forces
boost Internet usage among population, and encouraged IT literacy among the military population, elderly citizens, low income families, and disabled members. Also, the government granted subsidies to private IT/Internet institutions for making BB services available at an affordable price. Moreover, the government introduced the Cyber Building &HUWL¿FDWH 6\VWHP LQ $FNQRZOHGJLQJ WKH preference for high capacity cyber apartments, the JRYHUQPHQWXQGHUWKLVV\VWHPLVVXHGFHUWL¿FDWHV for buildings with high-speed telecommunication FDSDFLW\ 7KHVH FHUWL¿FDWHV LQGXFHG EXLOGHUV WR enhance BB access platform of newly developed apartments (Yun et al, 2002). These new buildings, in turn, led to increased density of urban populaWLRQ7RJHWKHU.RUHD¶VKRXVLQJSDWWHUQDQGLQGXFHG XUEDQL]DWLRQFRQWULEXWHGVLJQL¿FDQWO\WRZDUGVLWV rapid deployment of BB technology. Broadband penetration is also very high among the Icelanders. With 26.54 subscribers per 100 LQKDELWDQWV,FHODQGLQWKH\HDUUDQNHG¿UVW in terms of its BB penetration rate. However, if we analyze the experience of Iceland UHJDUGLQJ%%GHYHORSPHQWZH¿QGDYHU\GLIIHUHQWH[SHULHQFH7KHSULPDU\UHDVRQIRU,FHODQG¶V success rests on its high urbanization that reduces cost of deployment of BB per inhabitant. In Iceland, 92% of the population lives in urban areas with 62% in the capital Reykjavik. Iceland also has been very innovative in terms of its use of power line capacity for providing BB services quickly ZLWKRXWDQ\VLJQL¿FDQWFDSLWDOH[SHQVH$OVRWKH Iceland government has provided active support for rapid BB diffusion in the entire country. The universal service concept is indeed an effective way to accelerate demand and increase penetration rate. Even though no country has adopted BB service as part of a universal service program, countries like Australia and Iceland have made data transmission part of universal service an obligation. As of February 2003, 98% of Iceland households had access to ISDN with minimum speed of 128 Kbit/s. Recently a study group has proposed that BB should be available to all households in Iceland at the same price and has advised the government to establish a fund to subsidize the difference in cost of 2 Mbit/s BB connections between rural and
urban areas. Iceland's ministry is considering the proposal for possible implementation. In contrast to the experiences of South Korean and Iceland, government support in advancing BB technology has been weaker in the United States. In addition, the government emphasized the need for creating competitive environment in the BB market. Following the 1996 Telecom Act, U.S. regulators introduced rules for unbundling local loops. Based on these rules, incumbents were mandated to share their local loops with competitors and were permitted to compete for provisioning of long distance services. Unbundling regulation was further extended for provisioning of BB services. U.S. regulators also controlled the price of the unbundled element. While many economists argued in favor of such price control, the regulations may have resulted in negative incentives for BB investment in the U.S. The pricing of unbundled BB elements is, of course, key. In this respect, the FCC required the price to be based on total element long-run incremental cost (TELRIC), a concept that has raised serious debate among economists. While one group claims that TELRIC based price discourages BB deployment by not allowing incumbents for sunk cost and risk and uncertainty adjustment,26 the other group argues in favor of such price by suggesting that it enhances infrastructure investment by fully compensating incumbents for their supply of interconnection and unbundled network elements.27 In addition to unbundling of local loops and control of price, the U.S. has used asymmetric regulation of the provisioning of BB Internet access service until early 2003.28 Although highspeed Internet service through cable and DSL are similar services, providers of these services followed completely different regulations. While cable network owners were not required to share their cable network, incumbent local carriers who offered DSL service encountered price regulation as well as the obligation to share their BB network with competitors at a wholesale or unbundled price. The social cost of these asymmetric regulations has by now become familiar in the literature (Crandall et al., 2002). Consequently, many countries are
703
Broadband Diffusion and its Driving Forces
advocating technology neutral regulatory environment. Echoing this popular regulatory choice, ITU study (ITU, 2003a) has also recommended that BB regulation should be technologically neutral and that licensing and authorization policies should ZHOOUHÀHFWWKHQHXWUDOLW\RIWKHUHJXODWLRQ ,QÀXHQFHGE\,78UHFRPPHQGDWLRQDQGFRPparative assessment of BB penetration at home,29 FCC in the U.S. has recently allowed for some ÀH[LELOLW\LQWKHLUUHJXODWRU\UXOHV )&& ,Q particular, the commission has allowed loosening of the unbundled loops for the entrants by eliminating the requirement for the incumbent to share the network at a relatively low price. Following )&&UHFODVVL¿FDWLRQVFKHPHEURDGEDQGVHUYLFHV SURYLGHGRYHUERWK¿[HGORFDOORRSDQGFDEOH79 networks are now included under the category of “information services.” In addition, these services DUHQRORQJHUUHTXLUHGWRIXO¿OOWKHDFFHVVUHJXODtion. Similar revisions have also been made for the telecom companies. These companies will no longer be required to offer competitors access to new BB investment. This has caused increased incentives to invest in new infrastructure such as ¿EHUWRWKHKRPH )77+ LQWKH86 In sum, broadband is increasingly being seen as a vehicle for social and economic development in the information age. More and more economies are focusing on implementing BB access for the social DQGHFRQRPLFEHQH¿WVRIWKHLUQDWLRQV(YDOXDWLRQ of country experiences suggests that the right interaction between public and private sector can bring about success in BB adoption rate by stimulating underlying demand and supply factors. Regulatory controls may only be required when market fails to generate healthy competition. As regards to BB infrastructure development, ITU recommendation calls for technologically neutral regulation. However, granted that market take-up requires a critical mass of users, promotion of demand should receive adequate attention in government policy choices. A balance between supply and demand is the key for successful BB deployment and adoption, and national governments can, indeed, play DVLJQL¿FDQWUROHLQWKLVUHVSHFW
704
CONCLUSION 6WDUWLQJ RXW ZLWK D EULHI RXWOLQH RI EHQH¿WV RI broadband technology access in general and its heterogeneous usage across different countries, this chapter explains the observed variation in BB techQRORJ\SHUIRUPDQFHE\DQDO\]LQJFRXQWU\VSHFL¿F differences in the structure of demand and supply underlying broadband technology services. $QDO\VLVLQGLFDWHVWKDW¿[HGFRVWVRIGHSOR\ment, time to recoup deployment costs, and associated risks hinder provision of broadband services in many countries. In addition, lack of incentives for private investment in broadband deployment keeps the growth rate of broadband access supply from achieving its desired goals in these countries. Further, the disparity in provision rate within and across countries widens with differences in population density among regions and among countries. Accordingly, the analysis suggests that country VSHFL¿F JRYHUQPHQWV FDQ HQKDQFH SURYLVLRQ RI broadband access technology by inducing private investment through indirect subsidies including tax credits and low interest loans. The analysis also suggests that designing an appropriate regulatory structure is equally important for fostering market competition. With respect to demand, the analysis draws atWHQWLRQWRFRQVXPHUV¶DIIRUGDELOLW\DQGDZDUHQHVV as factors explaining country level differences in realized demand for broadband access technology. Flexible price mechanism that allows for a range of different demand intensities is seen to hold promise for moderating demand gap between high and low income areas and countries. Review of a number of empirical models on demand and supply of broadband access suggests that both cost conditions for deployment and awareness to new technology services are statistically sigQL¿FDQWIRUHVWLPDWLQJEURDGEDQGDFFHVVGHPDQG Further, the estimates of the models suggest that cost differences between regions and countries provide opportunities for substitution between narrowband and broadband access technology. Alternative model formulation indicates that reguODWRUVFDQLQÀXHQFHGHPDQGDQGVXSSO\UHVSRQVHV by designing appropriate policy frameworks that
Broadband Diffusion and its Driving Forces
target a combination of controlled price of access IRUFRQVXPHUVDQGUHDVRQDEOHSUR¿WDELOLW\IRU¿UPV making provisions of such access services. The above conclusions are supported by complementary case analyses. Carefully designed government support programs seem to have played DVLJQL¿FDQWUROHLQGULYLQJSHQHWUDWLRQUDWHIRUZDUG in South Korea and Iceland. In contrast, lack of supportive regulatory framework and government initiatives may have constrained broadband growth diffusion in the U.S. Clearly, experience of all these countries demonstrate that bringing the proper balance between supply and demand is the key for successful deployment and adoption of broadband technology. Country governments can SOD\DVLJQL¿FDQWUROHLQHIIHFWLQJWKLV LQWHQGHG balance. To conclude, the discussion on broadband diffusion and effective policy choices included in this chapter has a broader implication regarding the current state of technological applications and market conditions in the world. Increasingly, broadband access platforms are moving away from wire-line to wireless systems. The choice of the latter platform is gaining popularity in the developing countries and in remote areas. Also, convergence of computer, communications, and broadcasting technology is causing a rapid change in the characteristics of communication and information technology in general. Consequently, successful adoption and use of communications and information technology in any country will depend much on readiness of the country governPHQWVWRGHVLJQÀH[LEOHSROLF\FKRLFHVFDSDEOHRI accommodating continuous changes in technology DQGXVHUV¶SUHIHUHQFHV
ACKNOWLEDGMENT We express our sincere thanks to Richard N. Clarke and Sam H. Parker of AT&T, and two other anonymous referees for their helpful suggestions and comments. We also thank Aparajita Nandi of Queens College, City University of New York, for her editorial comments.
REFERENCES Alleman, J., & Noam, E. (Eds.). (1999). The new investment theory of real options and its implications for telecommunications economics. Kluwer Academic Publishers. Barro, R.J. (1990). Government spending in a simple model of endogenous growth. Journal of Political Economy, 98(5), 103-125. Bauer, J.M., Gia, P., Kim, J.H., Muth, T.A., & Wildman, S.S. (2002). %URDGEDQGEHQH¿WVDQGSROLF\ challenges (research report). Michigan: Quello Center of Michigan State University. Bauer, J.M., Kim, J.H., & Wildman, S.S. (2005). Effects of national policy on the diffusion of broadband in OECD countries (Unpublished Paper). Michigan: Department of Telecommunications Information Studies and Media, Michigan State University. Cava-Ferreruela, I., & Alabau-Munoz, A. (2006). Broadband policy assessment: A cross-national empirical analysis. Telecommunications Policy, 30(8-9), 445-463. Crandall, R.W., & Alleman, J.H. (Eds.). (2002). Broadband: Should we regulate high speed Internet access? Washington, D.C.: AEI-Brookings Joint Center for Regulatory Studies. Crandall, R.W., & Jackson, C.L. (2001, July). The $500 billion opportunity: The political economic EHQH¿WV RI ZLGHVSUHDG GLIIXVLRQ RI EURDGEDQG Internet access. Criterion Economics. Crandall, R.W., & Sidak, G.J. (1995, July). Competition and regulatory policies from interactive broadband networks. Southern California Law Review, 68(5). Crandall, R.W., Sidak, J.G., & Singer, H.J. (2002). The empirical case against asymmetric regulation of broadband Internet access. Berkeley Technology Law Journal, 17(1). 953-987. David, P.A. (2001, June). Knowledge, capabilities and human capital formation in economic growth (New Zealand Treasury Working Paper, 01/13).
705
Broadband Diffusion and its Driving Forces
Dudley, L. (1999). Communication and economic growth. European Economic Review, 43, 595619.
late high speed Internet access? (pp. 197-222). Washington D.C.: AEI-Brookings Joint Center for Regulatory Studies.
Federal Communication Commission (FCC). (2003). FCC Press Release. Retrieved July 26, 2007, from http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-231344A1.doc
International Telecommunications Union (ITU). (2003a). Internet report: Birth of broadband. Geneva, Switzerland: ITU.
Firth, L., & Mellor, D. (2005, March-April). BroadEDQG %HQH¿WVDQGSUREOHPVTelecommunications Policy, 29(2-3), 223-236. Gera, S., Wulong, S., & Lee, F.C. (1998). Information technology and productivity growth: An empirical analysis for Canada and the United States (Microeconomic Analysis Working Paper, No 20). Canada. Gilmour, G. (2002), Broadband demand modeling (Report). Australia: Department of Communications, Information Technology and the Arts of the Government. Gracia-Murillo, & Gabel, D. (2003, September). International broadband deployment: The impact unbundling. Paper presented at the 31st Research Conference on Communications, Information and Internet Policy, Arlington, VA. Greenstien, S.M., & Spiller, P.T. (1996). Estimating the welfare effects of digital infrastructure (NBER Working Paper, No. 5770). Cambridge, MA: National Bureau of Economic Research, Inc. Gruber, H., & Denni, M. (2005). The diffusion of broadband telecommunications: The role of competition (Unpublished paper). University of Rome, IL. Retrieved July 26, 2007, from http://ssm. com/absrtact-829504 Hausman, J. (1999). The effect of sunk costs in telecommunications regulation. In J. Alleman & E. Noam (Eds.), The new investment theory of real options and its implications for telecommunications economics. Boston, MA: Kluwer Academic Publishers. Hazlett, T.W. (2002). Regulation and vertical integration in broadband access supply. In Crandall and Jackson (Eds.), Broadband: Should we regu-
706
International Telecommunications Union (ITU). (2003b, March). Broadband Korea: Internet Case Study. Retrieved July 26, 2007, from http://www. itu.int/ITU-D/ict/korea/material/CS_KOR.pdf International Telecommunications Union (ITU). (2006a). Trends in telecommunications reform 2006: Regulation in the broadband world. Geneva, Switzerland: ITU. International Telecommunications Union (ITU). (2006b). World telecommunications indicator database. Retrieved July 26, 2007, from http://www. itu.int Loveman, G. (1994), An assessment of the productivity impact of information technologies. In J.A. Thomas & S. Michel (Eds.), Information technology and corporation of the 1990s: Research studies (pp. 84-110). Oxford, England: Oxford University Press. Lucas, R.E. (1998). On mechanics of economic development. Journal of Monetary Economics, 22, 3-42. Mankiew, N.G., Romer, D., & Weil, D. (1992). A contribution to the empirics of economic growth. Quarterly Journal of Economics. Mayo, J.W., Baumol, W.J., Willig, R.D., Hubbard R.G., Lehr, W.H., & Clarke, R.N. et al. (2003). Pricing based on economic cost - the role and mechanics of TELRIC (A collection of Essays published in FCC website, Document_ID=6515382451). Retrieved July 26, 2007, from http://gullfoss2.fcc.gov/prod/ ecfs/retrieve.cgi?native_or_pdf=pdf&id_document=6515382451 Mokyr, J. (2000). Knowledge, technology and economic growth during the industrial revolution. In B. Van Ark & G. Kuper (Eds.), Technology and productivity growth (pp. 253-292). The Hague, Netherlands: Kluwer Academic Publisher.
Broadband Diffusion and its Driving Forces
1DGLUL0, 1DQGL% %HQH¿WVRIFRPmunications infrastructure capital in U.S. economy. Economics of Innovation and New Technology, 10(2-3), 89-107.
Wales, C., Sacks, G., & Firth, L. (January, 2003). Killer applications vs. killer attributes in broadband demand. In PTC Conference, Hawaii (pp. 18-23).
National Research Council. (2002). Broadband: Bringing home the bits (Report of the Committee of Broadband Last Mile Technology, Computer Science and Telecommunications Board), 62-81. Washington, D.C.: National Academy Press.
Wildman, S. (1992). Information technology, productivity and trade implications (Working Paper No 521). New York: Columbia Institute for Tele-Information.
Precursor Group. (2001, February). How broadband deployment skews economic/business growth. Rappaport, P., Kridel, D., Taylor, L., Alleman, J.H., & Duffy-Demo, K. (2003). Residential demand for access to the Internet. In G. Madden (Ed.), The handbook of telecommunications, volume II (pp. 55-72). Cheltenham, U.K.: Edward Elgar. Roller, L.H., & Waverman, L. (1996). The impact of telecommunications infrastructure and economic development. In P. Howitt (Ed.), The implication of knowledge based growth of micro-economic policies. Calgary, Canada: University of Calgary Press. Romer, P.M (1989). Capital accumulation in the theory of long-run growth. In R.J. Barro (Ed.), Modern business cycle theory. Cambridge, MA: Harvard University Press. Romer, P.M. (1990). Endogenous technological change. Journal of political economy, 98(5), 71102. Sein, M.K., & Harindranath, G. (2004). Toward understanding the role of ICT in national development. The Information Society, 20(1), 15-24. Soung, M. (2005). The emerging business models LQEURDGEDQGEDVHG¿[HGPRELOHFRQYHUJHQFHHUD KT case. Paper presented at ITS 2005 Conference, Berlin, Germany. Stiroh, K.J. (2002). Are ICT spillovers driving the new economy? Review of Income and Wealth, I(48), 33-57. Tajiri, N., & Okazaki, T. (2006). Policies for broadband diffusion in developing countries (Working Paper). Tokyo, Japan: Waseda University, GITS.
Yun, K., Lee, H., & Lim, S.-H. (2002). The growth of broadband Internet connections in South Korea: Contributing factors (Working Paper). 6WDQIRUG8QLYHUVLW\ $VLD3DFL¿F5HVHDUFK&HQter. Retrieved July 26, 2007, from http://APARC. stanford.edu ;DYLHU 3 6KRXOG EURDGEDQG EH SDUW RI universal service obligations? ,QIR(1), 8-25.
KEY TERMS Information and Communications Technology (ICT): Technology of information retrieval, conversion, storage, protection, processing, and transmission. Broadband (BB) Technology: Refers to WUDQVPLVVLRQFDSDFLW\ZLWKVXI¿FLHQWEDQGZLGWKWR permit combined provisioning of voice, data, and YLGHR %URDGEDQG GRHV QRW UHIHU WR DQ\ VSHFL¿F lower limit of speed of transmission or bit-rate as it is evolving with technological dynamics and consumer demand. Broadband (BB) Access Networks: Broadband access network, often shortened to “broadband internet”, is a high data transmission rate Internet connection. DSL and cable modem, which are popular consumer broadband access technologies, are typically capable of transmitting faster than a dial-up modem. Broadband (BB) Penetration: A measure for extent of access to broadband communications within the population of a particular location where location could refer to a region, a country, or the entire world. Normally, it is computed as
707
Broadband Diffusion and its Driving Forces
a ratio between number of subscribers and total population for a particular location. Often, this ratio is multiplied by 100 to represent broadband penetration per 100 inhabitants. Broadband Diffusion: Refers to the spread of deployment and adoption of broadband technology.
3
Broadband Deployment: Refers to development of broadband networks or infrastructure through which broadband services can be delivered. Broadband Demand: Consumer willingness to use and pay for broadband services. Unbundling of Local Loop Regulation: Unbundling local loop is a regulatory process of requiring incumbent operators to provide access to some or all of the disaggregated elements and/or IXQFWLRQDOLW\RIODVWPLOHV WHOHSKRQHH[FKDQJH¶V FHQWUDORI¿FHWRFXVWRPHU¶VSUHPLVH RIWKHLUOHJDF\ telephone networks to their competitors so as to enable the latter to deliver services to their own customers.
4
Technology Neutral Regulation: Refers to D VSHFL¿F UHJXODWRU\ SURFHVV XQGHU ZKLFK UXOHV and regulations prevent service providers from preferring one type of technology over another in provisioning of their services.
5
ENDNOTES
6
1
7
2
708
The views expressed in this chapter (without any reference) are solely those of the authors. As documented in the literature on new growth theory (Barro, 1990; David, 2001; Lucas, 1998; Mankiew, Romer, & Weil, 1992; Romer, 1989, 1990), human capital and accumulated stock of knowledge accelerate economic growth by inducing technical change in the growth process. Access to knowledge, however, is dependent on availability of information and communications infrastructure systems. While a well-devel-
oped information infrastructure can serve as a catalyst to the growth process by facilitating rapid diffusion of knowledge, inadequacy of such an infrastructure can constrain growth by limiting access to sources of technical change. Dudley (1999) provides evidence on contribution of communications technology towards the economic growth of the western world in the past two centuries. Mokyr (2000) provides VLPLODUHYLGHQFHRQWKHVLJQL¿FDQFHRIFRPmunications technology and its potential for diffusing knowledge across the world. Sein and Harindranath (2004) argue that ICT serves as a powerful agent for economic and social development by improving productivity and quality in a number of production sectors. Granted the existing controversies on meaVXUHPHQWRIEHQH¿WV )LUWK 0HOORU WKHSRWHQWLDOVRFLDODQGHFRQRPLFEHQH¿WVRI BB technology at individual, organizational, and national level have been explored in Wales, Sacks, and Firth (2003), Precursor *URXS DQG;DYLHU With traditional “dial-up” Internet access, users plug a computer into their phone line YLDPRGHPDQGGLDOVRPHVSHFL¿FWHOHSKRQH number to access Internet connection. With its “always on” characteristic, broadband internet connection on the other hand, eliminates the need for dial up service altogether. See http://www.fcc.gov/cgb/broadband.html for further details. Even though policy makers of different countries are becoming aware of possible channels IRUWUDQVPLWWLQJ%%QHWZRUNEHQH¿WVYHU\ few attempts have yet been made to quantify VXFK SRWHQWLDO EHQH¿WV 2I WKRVH IHZ ZKR sought to measure potential gains from BB enabled activities, the study done by Crandall and Jackson (2001) with experimental data on U.S. economy deserves special attention. 7KHLUUHVXOWVLQGLFDWHWKDWEHQH¿WVIURP%% could range from $272 billion to $520 billion.
Broadband Diffusion and its Driving Forces
8
9
10
11
12
13
14 15
16
17
18
The availability of BB infrastructure in the host country also attracts FDI and encourages job migration from foreign companies to host nations (out sourcing). See Bauer, Gai, Kim, Muth, and Wildman (2002) and Crandall and Alleman (2002). The difference in subscription share noted between Asia and the other two regions of North America and Europe may be accounted for by the difference in their population. It is important to note that even though Iceland and S. Korea, in Figure 1(right), score at the top for BB penetration rate in 2005, in absolute terms, U.S. has the highest number of BB subscribers in the world. Such hindrances are generally observed in the rural or remote areas where the lack of demand for BB services often reduce the exSHFWHGSUR¿WPDUJLQVRQGHSOR\PHQWUHODWHG investment. Recent review of technology regulation reveals that Canada is one of the very few major economies with a cable openness policy; the results of this policy, however, have been limited (http://www.crtc.gc.ca/eng/publications/ reports/PolicyMonitoring/2006/tmr2006. pdf). FTTH is equivalent to FTTP. This, however, is no longer the case in the U.S. FCC (2003), as per its press release (February 20, 2003), now provides substantial XQEXQGOLQJ UHOLHI IRU ORRSV XWLOL]LQJ ¿EHU facilities. In a plan proposed in 2000, Indian Railroad Internet project aimed to make use of 65,000 kilometers of unused cable infrastructure DOUHDG\LQSODFH7KHVHVLJQL¿FDQWFDEOHVUXQ along the train tracks and have large amount of spare capacity. The referred plan proposed to use these cables for transmitting Internet WUDI¿F WR RXWVLGH DUHDV ZKLOH DYRLGLQJ WKH cost of laying a new cable network. See http://www.universalservice.org/sl/about/ overview-program for more information. Ability to pay depends on both price and monthly per capita income of users.
19
20
21
22
23
24
25
26
27
28
29
For further information regarding methodology of estimating monthly price per 100 Kbits/s, see ITU (2003). Cava-Ferreruela and Alabau-Munoz (2006) provide statistical evidence in this respect. High own price elasticity of U.S. BB services has also been reported in Rappaport, Kridel, Taylor, Alleman and Duffy-Demo (2003) and Crandall et al. (2002). )LUP¶V FDVK ÀRZ GHSHQGV RQ GHPDQG IRU network services, prices of services, and features of available new services. Each of these explanatory factors, however, encounter uncertainty due to changes in technology and HQGXVHUV¶SUHIHUHQFH Japan offers support for private broadband investors by designing incentives under three GLIIHUHQW FDWHJRULHV ¿QDQFLQJ V\VWHP WD[ EHQH¿WLQFHQWLYHVDQGJXDUDQWHHVRIOLDELOLties (ITU, 2006a). It is interesting to note that with government initiatives, Iceland, in the recent years, has PDGHVLJQL¿FDQWSURJUHVVLQLQFUHDVLQJ%% penetration rate among large section of its population (ITU, 2006a). As reported by ITU (2006a), similar to Korea, the Japanese government also offered supSRUWWRQHWZRUNSURYLGHUVE\¿QDQFLQJ%% deployment through no/low interest loan, tax EHQH¿WVDQGJXDUDQW\RIOLDELOLWLHV Hausman (1999) provides evidence on linkage between price regulation and reduced incentive for BB technology deployment in the U.S. Hazlett (2002) supports this conclusion by reporting that mandatory unbundling discourages investment in risky activities. See chapters by Economides, Baumol, Clarke and Pelcovits in Alleman and Noam (1999). See also Mayo et al. (2003). )&&¶V:LUH/LQH%URDGEDQGRUGHULQ lifted all asymmetric regulations on DSL and conforming its regulation to that of cable. &RPSDUDWLYH HYDOXDWLRQ UHÀHFWV WKDW %% penetration in Canada is about 60% higher than that in the U.S. Since both the countries are similar in terms of their demography and population density, the difference in the ob-
709
Broadband Diffusion and its Driving Forces
served BB penetration rate can be attributed to the difference in their regulatory requirements. Canada has a less rigorous unbundling requirement for local telephone companies than U.S. and virtually no mandated network sharing for competitive BB supplies.
710
711
Chapter XLIV
Diffusion of Broadband Access in Latin America Arturo Robles Rovalo Universidad Nacional Autónoma de México, Mexico, and Universidad Politécnica de Madrid, Spain Claudio Feijóo González Institute for Prospective Technological Studies, JRC, EC, and Universidad Politécnica de Madrid, Spain José Luis Gómez-Barroso Universidad Nacional de Educación a Distancia, Spain
ABSTRACT The “geographic” digital divide is obvious when comparing more developed countries to the rest. Its ¿UVWDQGPRVWREYLRXVVLJQLVWKHGLIIHUHQFHLQWKHGLIIXVLRQRIEURDGEDQGDFFHVVHV+RZHYHULWLVFOHDU that there are also lines of separation in smaller geographic ranges: between countries of a same area, LQVLGHHDFKFRXQWU\DQGVRPHWLPHVLQHDFKVSHFL¿FUHJLRQ7KLVFKDSWHUVKRZVWKLVVLWXDWLRQE\VWXG\LQJ the broadband access diffusion in Latin America on a three level basis (regional, national, and local). At the national level, a few explanatory variables of the different situations presented by the countries chosen for the study are researched. Additionally, a description of the environment (market and public action) where this diffusion is occurring is also included.
INTRODUCTION The generalized use of the services and applications provided on broadband infrastructures is considered a critical element for the full incorporation of individuals and societies into the information society. Setting aside the social or cultural implications this carries,1 it is broadly accepted that these infrastructures will be (or are already) a key element for the growth of productivity and
competitiveness in any economy. The statistics regarding the progress in the number of broadband connections is starting to be used, in this fashion, in indicators of current and future economic vitality of countries. Availability and interest for its adoption are, as in the case of any other good or service, the two stages that are necessary (and consecutive) for the diffusion of broadband networks to progress.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Diffusion of Broadband Access in Latin America
$FFHVVLVWKH¿UVWDQGsine qua non condition. ,QYHVWPHQWLQEURDGEDQGZKLFKUHTXLUHVDVLJQL¿cant improvement of the existing infrastructures or even a new network deployment, will mainly come from the private sector. However, many zones do not provide the necessary market conditions to awaken the interest of private initiative. 7KHVSHFL¿FVLWXDWLRQYDULHVZLWKWKHRURJUDSKLFDO and sociodemographic conditions, and, basically, with the level of economic development. In the most developed countries, the access problem is restricted to rural or isolated zones. In many other countries, the picture is a negative of the above: the “islands” are not the areas without connection but the connected ones instead. 2QFHWKH¿UVWUHTXLUHPHQWLVPHWWKHVHFRQGRI the factors appears: adoption. This factor depends in turn on a broad set of variables: prominently, the affordability of the service, but also the adaptation of the services and applications to the interests and needs of the users in addition to a broad range of cultural, social, and educational circumstances. The result obtained from adding up these elements does not mitigate the disparities already present in access. On the contrary, it accentuates them. From the market perspective, the issues of access and adoption are inextricably interwoven: adoption is impossible without access, but access LV HFRQRPLFDOO\ GLI¿FXOW WR SURYLGH ZLWKRXW WKH prospect of rapid and widespread adoption (HolOL¿HOG 'RQQHUPH\HU 7KXVWKHGLJLWDOGLYLGHKDVD¿UVWDQGREYLRXV sign: the geographical one. The various deployment and adoption rhythms of broadband threaten to increase the distance separating developed and developing countries. However, on the other hand, inside countries, separating lines are also being traced, at regional or local scales. This chapter intends to move forward in the knowledge of these divisions in Latin America, both those separating the region from other geographical areas, and those inside the region itself. In order to meet this objective, we will start by studying, with a double perspective, what the environment where the development of broadband occurs is like. On the one hand, it is necessary to generally be aware of what has been the evolution
712
of the telecommunication services sector and what is the situation of the markets. On the other hand, there is an intense public activity that intends to complete the private activity and the knowledge of which is essential. Subsequently, the article will focus on the situation of the penetration of broadband in Latin America2 at three different levels: continental, naWLRQDODQGORFDO$WWKH¿UVWOHYHOZHZLOOFRPSDUH the present Latin American situation (and its recent evolution) to that of other regions. For the second level of study, we have chosen six countries from the main economic blocks: Argentina and Brazil, Mercosur3 member countries, Chile and Mexico, APEC4 economies, and Peru and Colombia, members of the Andean Community of Nations.5 Despite the fact that the nature of the article is basically descriptive, this section provides some investigation on what some of the causes that explain the different levels of penetration are. Last, the local detailed study must obviously restrict itself to a VSHFL¿FFDVHVWXG\ 0H[LFRKDVEHHQFKRVHQ The conclusions are structured as an examination of the future and, based on the described situation, as orientations towards what the best policies to face this future should be.
BACKGROUND To proclaim the wonders that will be brought on by the knowledge society, and subsequently label the promotion of its development as vital for both regions and individuals, has become a cliché in the political rhetoric. It would be interesting to know whether behind every statement there exists DFOHDUDZDUHQHVV RUHYHQEHWWHUDUHÀHFWLRQ RQ what model should be adopted and what are the determining factors allowing the progress to occur. It is a fact that there are more variables involved, undoubtedly, than those stated most of the times. But what is really unquestionable is that one of those factors, probably the most important one since it is a prior condition, is the availability of an infrastructure allowing the access to the information that names the new society. In the midterm, these infrastructures will surely tend to form the basic
Diffusion of Broadband Access in Latin America
mesh that will support a great deal of activities of the knowledge-based economy, and thus, the generalized availability of broadband connectivity represents a qualitative leap in the road towards the information society.6 Thus, telecommunications networks have become something of a fetish of economic growth, a “catalyst” of economic success (Falch & Henten, 2000). The fact, as such, is not new: telecommunication service access universalisation is a concern WKDW VXUYLYHV VLQFH WKH ¿UVW QHWZRUNV VWDUWHG WR be deployed. However, the generalized aspiration referred, now, to the advanced infrastructures, takes on a new dimension. The guarantee of indiscriminate access to broadband networks seems to become the guarantee that the opportunity for not missing the train to the future will reach everyone. The argument is not less true despite it being trite. However, it does sometimes require a more rigorous treatment. Progressing in the diffusion of broadband networks raises a series of challenges and opportunities for both the private and public sectors. On the one hand, companies, mainly responsible for the development of the broadband market, must adapt their traditional infrastructures and business models to this new paradigm, while being aware of the major investments this requires. The SXEOLFVHFWRU¶VLQLWLDOWDVNLVWRHVWDEOLVKIDYRUDEOH conditions for the market to effectively develop adequately. What can be expected from private activity? Market limitations for deploying truly universal broadband networks are present even in the most economically developed countries, where the universalization of telephone networks is a fact and the processes of deregulation and introduction of competition in the telecommunications sector are more or less mature. This is stated in the works of Gómez-Barroso and Pérez-Martínez (2007), Grubesic (2004), and Strover (2003). As expected, these problems increase in less rich countries. Most countries in the “second global pack” (where Latin America can be generally included) have faced the process of liberalizing their telecommunications markets, which has usually contributed to a dynamic development of the
industry. The number of lines has grown substantially during the past few years. In particular, the number of wireless connections has been subject to impressive growth rates. Nevertheless, even in the most successful cases, the task of achieving an authentically universal network just for voice communications is still far from being achieved. The delay in the deployment of broadband is much more obvious. This is not at all surprising, indeed. The market DQVZHUVWREHQH¿WV7 And in areas with low population densities, even with high levels of service penetration, the minimum threshold of demand covering the costs of developing the access infrastructures may not be reached. When the result of this equation is negative, public intervention becomes necessary,8 if the intention is for networks to reach those places where the offer will probably not allow for them, or where they will arrive with an unacceptable delay. This intervention is generally a part of the programs for the development of the information society designed in many countries (which are actually convinced of the importance of its progress or that simply follow what seems to be a “universal WUHQG´ 7KHVSHFL¿FPHDVXUHVXVXDOO\ DQGVKRXOG address both the offer and the demand side, attempting to generate a virtuous circle of growth for both. Indeed, the deployment of the network must be directly (with public investments) or indirectly favored, promoting the demand so as to achieve WKDW VRPH QRQSUR¿WDEOH UHJLRQV PDQDJH WR JHW past the business threshold required by operators to invest and provide service. The promotion of the adoption is necessary from this perspective but, even with universal networks, would also continue to be necessary for “convincing” those WKDWKDYLQJDFFHVVGRQRW¿QGDQ\DGYDQWDJHVLQ connecting. Obviously, corrective policies must be based on the correct assessment of the situation to be corrected and the adaptation to that reality (GómezBarroso & Feijóo, 2006). This is not always so. And the results, as a consequence, are not always as positive. Their assessment is not easy due to the lack of adequate statistics.
713
Diffusion of Broadband Access in Latin America
With this framework as a reference, the evolution and situation of the diffusion of broadband infrastructures in Latin America is studied below, starting with the market and public environment where this evolution is occurring.
EVOLUTION OF THE TELECOMMUNICATIONS SECTOR IN LATIN AMERICA The current situation in the telecommunications services markets is a generalized process of gradual opening-up of a sector that has been traditionally reserved to a public monopoly. Latin American countries, some since the 1980s but mainly during the 1990s, also face this stake. As in any other country requiring the development of their infrastructures, the reforms have tried to achieve a balance between the introduction of an actual and sustainable competition and the progress towards the objectives of universal service. Each country in the region has developed its own mechanisms and created sundry incentives. Although Latin America is occasionally considered a homogeneous region, the truth is that the strategies, and thus the results, have been quite disparate. The varied programs designed for expanding the network, although not identical, coincide in their basic pillars: sector liberalization and reform, privatization of the monopoly, introduction of FRPSHWLWLRQ LQERWK¿[HGDQGPRELOHWHOHSKRQ\ and several contributions to the universal service by the operators. Generally, the liberalization-privatizationregulation triangle has contributed in the short term to an important advance in network penetration, especially immediately after the privatization of the previously monopolistic operator. The competition/ growth pairing has not achieved the fair penetration of telephone coverage. A liberalization of the market that has been too deep, when said market was still in a stage of basic telephone penetration, could have undermined the effectiveness of the universal service programs. Other factors can be DGGHGWRWKLVRQHVXFKDVWKHTXHVWIRUSUR¿WDELOLW\ in the short term by the operators in a complicated
714
¿QDQFLDOHQYLURQPHQW ZLWKWKHVXEVHTXHQWHIIHFW of investment reductions) and also, occasionally, the uncertainty regarding the regulation that should cover the new services (even new technologies) that are being commercially developed. In the transformation of the sector, the best results have been achieved by those countries, such as Brazil, that established methods with a certain GHJUHHRIÀH[LELOLW\DQGEDVHGRQWKHSUHVHQFHRI LQFHQWLYHVWKDWLVWKDWFUHDWHGVSHFL¿FJRDOVDQG deadlines which, in case of being met compreKHQVLYHO\DOORZHGWKHRSHQLQJXSRUPRGL¿FDWLRQ of the legislation in order to favour the awarded companies (Robles Rovalo, Gómez Barroso, & Feijóo González, 2006). Obviously, this is not the best scenario for the development of broadband. As we will see below, “traditional” infrastructures (and particularly, the telephone network) are, at least today, waiting for the commercial boom of other alternative technologies, the means through which the offer of a broadband access reaches the users. Thus, DQLQVXI¿FLHQWGHSOR\PHQWRIWKHEDVLFQHWZRUNV initially burdens the possibilities of developing advanced networks.
PUBLIC PROMOTION OF BROADBAND IN LATIN AMERICA Actions in the Regional Scope As occurs frequently in other regions, broadband promotion in Latin America is generally clasVL¿HGLQWKHEURDGHUSURJUDPVWDUJHWHGWRZDUGV developing the information society. There is no single regional program or common policy for its development although there are declarations of intent in the supranational framework. The Florianopolis Statement9 of 2000 already set forth the intention of the signing governments to fully “integrate” into the information society and admitted that the preparation of national broadband policies was necessary (since “leaving the evolution” of access and services of the information society exclusively in the hands of the market could entail “risks” related basically to the
Diffusion of Broadband Access in Latin America
growth of the digital divide). This statement was followed, in 2001, by the preparation of the Connectivity Agenda for the Americas10 and, later on, in 2003, coinciding with the regional preparatory conference for the World Summit on the Information Society, the Bavaria Statement11 was signed. This statement agreed to “promote and strengthen” proactive national strategies, admitting the need for the governmental authorities, the private sector, DQGWKHFLYLOVRFLHW\WRSDUWLFLSDWHLQWKHGH¿QLWLRQ of said strategies. At a subregional level, the Information Society in the Andean Community program (ASETA & 81(6&2 DOVR HVWDEOLVKHV VSHFL¿F SULQciples, objectives, and projects for the construction and development of the information society in the region.
National Policies for Broadband Development To go into some level of detail, it seems necessary WRUHIHUWRVSHFL¿FSURJUDPV:LWKWKLVSXUSRVH the same six countries that will be subject later on to the characterized study on the degree of broadband diffusion have been chosen. Namely, these are Argentina, Brazil, Chile, Colombia, Mexico, and Peru. These six countries (as almost any other one of the region, with no exception) have designed national strategies for the progress of the information society. In three of them (Chile, Colombia, DQG0H[LFR WKLVVWUDWHJ\LVDSDUWRIWKHFRXQWU\¶V comprehensive development plans. Public responsibility is always admitted, in some cases even explicitly.12 Many of the actions set forth in these plans affect, directly or indirectly, the spreading of broadband. As regards promoting the offer, we can underline the following measures: •
Plans for Internet Universalization, which HVSHFLDOO\LQWKHLU¿UVWVWDJHVDUHEDVLFDOO\ focused on providing access (Chile and ArJHQWLQD KDYH VSHFL¿F SURJUDPV IRU DFFHVV provision since 1998)
•
•
Investment in access networks for advanced services: Argentina (PSI, National Program for the Information Society), Colombia (Connectivity Agenda), Mexico (Connectivity section in the e-Mexico Program) and Peru (Huascarán Project) contribute resources for direct investment in broadband networks for community accesses Connection of educational networks (Colombia: Compartel Program; Mexico: Pl@ zas Comunitarias Program), as well as for the Public Healthcare sector
Demand promotion measures appeared a little later in the region (Brazil released a “Green Paper” in 1999; Colombia, Mexico, and Peru joined the trend during the following years) although they have been progressively and gradually consolidated. Generally, governments have focused their efforts on the services which are mostly used by the overall population: •
•
The development of e-Government has been based on three basic aspects: { Electronic administration has experienced a major boost in the region and has achieved, with the exceptions of 3HUX DQG &RORPELD TXLWH VLJQL¿FDQW breakthroughs13 in services such as tax payment, social security, and public tenders { Likewise, access to governmental electronic information has been helped E\ D VSHFL¿F OHJLVODWLRQ $UJHQWLQHDQ Public Information Access Decree of 2003; General Archive Act of Colombia, 2000; Openness and Access to Public Information Act of Mexico, 2002) { Last, some intergovernmental network projects connect the different administrative organizations (for example, the Chilean INTERLEGIS network) E-market development has been boosted E\SURPRWLQJVSHFL¿FOHJLVODWLRQSURYLGLQJ security (5460/01 Act in Brazil; 19223 Act in Chile; Legislative Decree in Peru on Computer offences) and validity (Digital Signature
715
Diffusion of Broadband Access in Latin America
•
•
Act in Argentina, 2001; Digital Signatures DQG&HUWL¿FDWHVLQ3HUX WRHOHFWURQLF transactions Training of human resources and educational portals have been promoted (Argentina launched Educ.ar and Mexico did the same with the e-education portal of e-Mexico) Business usage of the advanced telecommuniFDWLRQVHUYLFHVKDVEHHQSURPRWHGZLWK¿VFDO incentives (Mexico, Peru, and Chile apply a zero tariff for importing new technologies) and direct subsidies (Brazil and Argentina have reduced rates for the connection of VSHFL¿FXVHUV
In between the promotion of the offer and the demand, the “public access points” or telecenters are one of the most usual tools (ECLAC, 2005) used in the new universalization programs. Indeed, their installation intends to meet several objectives simultaneously. In cases where there is no other broadband alternative in the locality, their construction can be included in the section dedicated to network extension. They also promote
the digital literacy of marginal groups and stimulate the usage of advanced services, thus boosting future demand. The development of telecenters has been so important that some studies (IIRSA, 2003; ITU, 2003a) state that community accesses, be they private (Internet café) or public (telecenters), were at the time the main point of access to advanced services in Latin America.
BROADBAND DIFFUSION IN LATIN AMERICA Broadband in Latin America: Macro-Regional Level According to the world telecommunication/ICT development report (ITU, 2006), in Latin America, Internet penetration (10.5%) is below the world average (13.2%) and very far from regions such as Europe (31.1%), the United States and Canada (62.7%). Figure 1 shows that most of these users connect through the conventional narrowband
Figure 1. Broadband penetration by region. Prepared by the authors based on data adapted from the ,78:RUOG7HOHFRPPXQLFDWLRQ,QGLFDWRUV'DWDEDVH
716
Diffusion of Broadband Access in Latin America
telephone line. However, the growth of broadband accesses in Latin America has been almost exponential in recent years: a 63.61% compound annual growth rate (CAGR) for the 2001-2004 period.14 Despite this fact, penetration (0.76% during 2004) was very low as compared to the world average (3.93%). At the time Figure 1 was prepared, Latin America was above Africa (0.02%), but quite far from regions such as Europe (5.17%), the United States, and Canada (12.4%). Even when compared to regions with a similar degree of development, be they Asia (1.59%) or Oceania (3.63%), Latin America is left behind in terms of penetration. As regards the world total, Latin America held only 2.8% of the broadband connections of the planet (Figure 2). Almost globally, broadband offer is extraordinarily dependent on DSL technology and, to a smaller extent, on cable networks. The global veracity of this statement can be backed by the data of the DSL Forum15 estimating that by the end of 2004 there were 98 million DSL connections, 51 million cable connections, and 9.3 million using other technologies.160RUHVSHFL¿FDOO\LQWKH OECD countries, in December 2004, 60.08% of the broadband subscribers had chosen DSL technology, 33.48% cable-modem, and the remaining 6.44% other technologies. In fact, should we exclude the
United States, the ADSL percentage would represent 69% of the total (OECD, 2005). Knowing the broadband access deployment situation in Latin American countries faces the problem of the lack of data. Those that the national authorities in charge of regulating the telecommunications sector provide regarding broadband DUHQRWFODVVL¿HGE\WHFKQRORJ\RULQVRPHFDVHV simply do not exist. The available data prove the authority of ADSL (according to the DSL Forum, by the end of 2004, DSL connections represented 83% of all the broadband accesses in the region). Wherever cable networks were not widespread, ADSL was the clear leader; thus, in December 2004, Mexico had 545,000 ADSL subscribers, 228,000 cable subscribers, and 43,000 using other technologies. However, in other countries, cable is still in the lead. In Colombia, in June 2004, there were 66,881 cable subscribers as opposed to 18,403 ADSL subscribers. Yet, the cases where cable prevails could simply answer to the delay in the launch of the commercial ADSL offer. Chile, which on the other hand is the leading country in the region as regards broadband connections per number of households and also the one that provides the most comprehensive evolution data, is an example of this statement. As can
Figure 2. Global distribution of broadband subscribers. Prepared by the authors based on data adapted from ITU World Telecommunication/ICT Development Report (2006)
717
Diffusion of Broadband Access in Latin America
Figure 3. Broadband access evolution in Chile (December 2001 – December 2004). Prepared by the authors based on data from SUBTEL
Figure 4. Broadband evolution in selected Latin American countries. Prepared by the authors based on GDWDIURPWKH,787HOHFRPPXQLFDWLRQ,QGLFDWRUV'DWDEDVH ,78 2(&'%URDGEDQGVWDWLVWLFV (OECD, 2006), INDEC, ANATEL, SUBTEL, CRT, COFETEL, and OSIPTEL Broadband Access Evolution Latin America (01-06)
Broadband access / 1000 inhab
50 45 40 35 30 25 20 15 10 5 0 Argentina
Brazil
Chile
Colombia
Mexico
Peru
2000 2001 2002 2003 2004 2005 2006 p
be seen in Figure 3, although cable was the most popular technology at the beginning of the decade, in late 2003 ADSL took the lead. In fact, while cable PDLQWDLQVDOLQHDUJURZWKWUHQG$'6/¶VJURZWKLV exponential. “Other technologies” are losing their VKDUHVLQFHWKHVSHFL¿FDFFHVVIRUORFDOQHWZRUNV and leased lines are being replaced by any one of the two majority technologies.
718
Until 2004, there were still countries in the subcontinent where there was no commercial broadband offer for homes17 or microenterprises (ITU, 2006).
Diffusion of Broadband Access in Latin America
Broadband in Latin America: National Scope The evolution of broadband has not been uniform in the different countries of the region. The analysis of the broadband access penetration levels carried out in the countries chosen for this study (Argentina, Brazil, Chile, Colombia, Mexico, and Peru) shows that until 2000, the penetration of broadband was quite modest and was basically limited to the connections hired by companies and public organizations (Figure 4). It is in 2001 when the commercial high speed accesses for individuals and small and medium size enterprises starts to have some degree of success. Subsequently, and also probably thanks to the public initiatives covered above,18 the density of broadband connections in the region has grown substantially. Chile is the country with the best broadband density rates (46.76 accesses/1000 inhabitants) and has become a reference model for some of their neighbors. Argentina (28.78) and Brazil (26.64) have followed similar paths reaching
LQWHUPHGLDWH¿JXUHV0H[LFRWKDQNVWRWKHQRWDEOH advances achieved in recent years (growth rates of 67.7% and 69.6% in 2005 and 2006, respectively), has joined this group (27.51). In Peru, broadband is experiencing a strong and continuous growth (75.12% CARG in the 2001-2006 period) but their penetration levels are still quite low (14.55). Last, &RORPELD LVVWLOOLQWKH¿UVWVWDJHVRIGHYHOopment (Table 1). :KDW YDULDEOHV LQÀXHQFHG WKH GLIIHUHQFHV LQ these results? Particularly noticeable is the fact that the two FRXQWULHVWKDWKROGWKH¿UVWSRVLWLRQVRIWKHFODVVL¿FDWLRQ&KLOHDQG$UJHQWLQDZHUHWKH¿UVWWR establish a national strategy for the development of the information society, in 1998. The next country to follow this path was Brazil, in 1999, also coinciding with their third position until 2005 in broadband penetration.19 This could indicate that the countries with the greatest growth are those where the public and private sectors have interacted during a longer period.
Table 1. Broadband penetration (broadband access/100 inhab.) in selected Latin American countries. Source: Prepared by the authors based on data from the ITU Telecommunication Indicators Database ,78 2(&'%URDGEDQGVWDWLVWLFV 2(&' ,1'(&$1$7(/68%7(/&57&2)(7(/ and OSIPTEL COUNTRY
2000
2001
2002
2003
2004
2005
2006 p
Argentina
0.111
0.421
0.410
0.532
1.116
2.043
2.827
Brazil
0.025
0.186
0.376
0.690
1.108
2.101
2.563
Chile
0.052
0.415
1.097
2.135
3.108
4.018
4.776
Colombia
0.018
0.032
0.074
0.104
0.210
0.708
0.733
Mexico
0.014
0.100
0.300
0.400
0.988
1.657
2.756
Peru
0.008
0.050
0.151
0.351
0.838
1.310
1.455
p=preliminary data
719
Diffusion of Broadband Access in Latin America
)LJXUH%URDGEDQGDFFHVVDQGWHOHSKRQH ¿[HGSOXVPRELOH DFFHVVJURZWKHYROXWLRQ Prepared by the authors based on data from the ITU Telecommunication Indicators Database (ITU, 2(&'%URDGEDQGVWDWLVWLFV 2(&' ,1'(&$1$7(/68%7(/&57&2)(7(/DQG OSIPTEL Broadband Access vs Telephone (fixed and cellular) Access 2000- 2005
Broadband access per 1000 inhabitants
45 40 35 30 25 20 y = 0.3758x - 8.9867 2 R = 0.6727
15 10 5 0 -5
0
10
20
30
40
50
60
70
80
90
100
-10 fixed and cellular access lines per 100 inhabitants Argentina
Brazil
Chile
Colombia
Mexico
Peru
p= preliminary data
The extension of telephony, that is, mobile DQG¿[HGDFFHVVHVFRPELQHG20 is, naturally, one of the factors that present a positive correlation with the broadband penetration rates (Correlation &RHI¿FLHQW 5 FRQVLGHULQJ WKH period).21 7KH FRXQWULHV ZLWK JUHDWHU ¿[HG DQG mobile telephone line density, Chile (89.83%) and Argentina (80.07%), are also those that lead WKHEURDGEDQGFODVVL¿FDWLRQ )LJXUH $QGWKH density of telephone lines is related, in turn, to initial economic parameters and the success of the liberalization-privatization-regulation triangle as well as the establishment of a sustainable competition model in the mid to long term. Not surprisingly, the economy per capita of each state and the affordability of the broadband service are also factors that affect noticeably the extension of broadband. Figure 6 shows the close relationship between the gross national income (GNI) per capita and broadband penetration (corUHODWLRQFRHI¿FLHQW5 2QWKHRWKHUKDQG
720
the comparison between the price of broadband and the monthly income per capita (monthly price per 512kbs/power purchase parity (PPP) per month) shows a negative relationship (correlation FRHI¿FLHQW5 )LJXUH ,QRWKHUZRUGV the lower penetration levels are present in countries such as Peru (1.45%) and Colombia (0.73%), where the price of broadband represents a greater participation in the monthly income rate (40.6% and 36.5%, respectively). Last, the analysis shows that the development of the economy does not affect directly the evolution of broadband in the countries studied. There are few indications of agreement between the development of the gross domestic product and the number of broadband accesses per inhabitant (correlation FRHI¿FLHQW5 IRUWKHSHULRG On the other hand, medium-size economies such as Chile (GDP = 115 thousand million USD, 2005) and Argentina (GDP = 183 thousand million USD, 2005) show better broadband levels than major
Diffusion of Broadband Access in Latin America
Figure 6. GNI per capita and broadband penetration (access per 100 inhabitants) in selected Latin $PHULFDQ&RXQWULHV3UHSDUHGE\WKHDXWKRUVEDVHGRQGDWDIURPWKHDSSURSULDWHQDWLRQDOUHJXODtory bodies (INDEC Argentina; ANATEL, Brazil; SUPTEL Ecuador; SUBTEL Chile; CRT, Colombia; COFETEL, Mexico; OSIPTEL, Peru; INDOTEL, Dominican Republic and CONATEL, Venezuela) Broadband access penetration vs GNI per capita 6.0% Broadband access per 100 inhab
y = 6E-06x - 0.0051 2 R = 0.5584
5.0%
Chile
4.0% Argentina
3.0%
Mexico
Brazil
2.0% Venezuela
Peru
1.0% Colombia Dominican Rep
Ecuador
0.0% 0
1000
2000
3000
4000
5000
6000
7000
8000
GNI per capita (US dollars) 2005
Figure 7. Broadband penetration vs. monthly broadband price as a percentage of the GNI in selected /DWLQ$PHULFDQ&RXQWULHV3UHSDUHGE\WKHDXWKRUVEDVHGRQGDWDIURPWKHQDWLRQDOUHJXODWRU\ bodies and main service providers Broadband access vs ADSL price/GNI per capita
Broadband access per 100 inhab
6.0% 5.0%
Chile
4.0% Mexico
3.0%
y = -0.077x + 0.035 2 R = 0.5946
Argentina Brazil
2.0% Venezuela
Peru
1.0%
Dominican Republic
Colombia Ecuador
0.0% 0%
10%
20%
30%
40%
50%
60%
-1.0% % Monthly ADSL Price (512kbps) / GNI per month 2005
721
Diffusion of Broadband Access in Latin America
economies such as Brazil (GDP =794 thousand million USD) and Mexico (GDP = 768 thousand million USD). Peru (GDP = 78 thousand million USD, 2005), the smallest economy, shows twice as many accesses per inhabitant when compared to Colombia (GDP = 122 thousand million USD, 2005). This analysis matches the global trend (World Bank, 2006) that shows that for similar levels of development, smaller economies tend to be more network-ready than bigger ones.22
Broadband in Latin America: Intranational Scope Going down a step further in the analysis, it also seems predictable beforehand that the national ¿JXUHV GLVJXLVH GLVSDUDWH VLWXDWLRQV LQVLGH HDFK country. In fact, there is proof that the investment and deployment of infrastructures by private operators is concentrating around the most densely populated areas (Hilbert & Katz, 2003). Capital
and major cities are still favored since they often have more than one option for accessing advanced services, while most of the rural areas continue to be in a “digital isolation” situation, with no type of access whatsoever (OECD, 2004). Even in the most advanced markets (Chile, Mexico, and Brazil), the offer is limited to major cities. Colombia concentrates 87% of its broadband accesses in the cities of Bogota and Medellin. The problem increases when analyzing in detail the conditions (geographic, economic, and cultural) SUHVHQWLQWKHUHJLRQWKDWLQÀXHQFHWKHH[SDQVLRQ of infrastructures (Garcia-Murillo, 2003; IDRC, 2005) and it is apparent that, as opposed to what occurred in developed countries,23 the market will not provide this access at least in the short or medium term. Mexico can represent an adequate case study for analyzing the internal disparities since their diversity of economic, geographic, and demographic characteristics covers almost all the possible situa-
Table 2. Zones types characteristics. Source: Prepared by the authors based on data from the XII General Population and Housing Census 2000 and the Municipal SIMBAD Database System of INEGI (Mexico) ZONE
Population
Household
Zone
Population
Region
Population
density (hab/
density (HH/
size
included per
FODVVL¿FDWLRQ
included per
km2)
km2)
(km2)
zone %
High dense urban
A
12,000
2000-3000
175
26.4
Dense urban
B
6,000
1300-1600
175
21.0
Urban
C
1,000
150-300
125
11.7
Suburban
D
250
50-60
125
1.9
Suburban- ex rural
E
125
25-40
80
13.7
Rural
F
30
6-8
25
8.2
Isolated rural
G
12
3-4
25
17.2
722
region
Urban
59.1
Suburban
15.6
Rural
25.4
Diffusion of Broadband Access in Latin America
WLRQVRIWKHVXEFRQWLQHQWDQGUHÀHFWVTXLWHIDLWKIXOO\ the sociodemographic reality of the region. Table 2 shows a characterization of the different regional types that can be found in Mexico. 7KLV FODVVL¿FDWLRQ KDV EHHQ PDGH DFFRUGLQJ WR the sociodemographic characteristics and leads to GH¿QLQJVHYHQGLIIHUHQWDUHDV $* Using this division, Table 3 shows the infraVWUXFWXUHGHSOR\PHQWDQGVHUYLFHDFFHVV¿JXUHVIRU each type of zone. Towns with population under 15,000 inhabitants have no residential broadband service coverage.24 That is, the homes of all the towns considered as suburban or rural, where 40% of the overall population lives, are not covered by a commercial broadband network. Isolation makes the situation worse: in rural communities near urban or suburban centers (located at less than 2.5 Km. from a community of at least 15,000 inhabitants) the penetration of telephone lines in homes (8.45%) is over twice that of the rural zones that do not meet this requirement (“isolated,” 3.32 %).
CONCLUSIONS: FUTURE TRENDS AND GUIDELINES If from the description presented in the previous section, the “static” situation resulting from the
analysis of the latest data available is taken into consideration, the results are not very optimistic. Indeed, the availability of broadband in Latin America still shows very low levels and is gathered DURXQGWKHPRVWSUR¿WDEOHDUHDVOHDYLQJWKHYDVW majority of the population out. This new broadband divide threatens to separate the region even more from the developed areas. This general scenario, however, offers numerous nuances when examined in further detail. The extension of broadband is quite different from one country to another. The number of accesses is basically connected to the national income per capita levels, the legacy basic infrastructures, and the affordability of the service. The situation is different when observing the data from a time-based perspective. In the dynamic analysis, the broadband accesses in Latin American are in a phase of accelerated growth. What is more important, it seems that this growth will continue for the next few years. What can be expected in the mid-term? It is obvious that the progress of broadband will always be stronger in countries with higher income rates (GNI per capita) or in those that have high mobile DQG¿[HGWHOHSKRQ\SHQHWUDWLRQUDWHV7KHDIIRUGability problem can also be fought by showing the SRSXODWLRQWKHEHQH¿WVRIDQDGHTXDWHDFFHVVWR
Table 3. Broadband availability in homes (Mexico). Source: Prepared by the authors based on data from the national enquiry of household ICT use and availability, 2004 (INEGI, 2004) Nat. Zone G
% of national population living in the area
% households with telephone service
% household with telephone & Internet % households with telephone, Internet & broadband
% household without broadband
Zone F
Zone E
Zone D
Zone C
Zone B
Zone A
Avrg
17.2
8.2
13.7
1.9
11.7
21
26
-
3.42
8.45
30.5
36.6
40.4
78.5
89.2
46.4
0.64
1.58
5.72
6.87
7.58
14.7
16.7
8.70
0.00
0.00
0.00
0.00
2.23
4.33
4.92
2.22
100
100
100
100
97.7
95.6
95.1
97.7
723
Diffusion of Broadband Access in Latin America
the Information Society. Obviously, the price that the users are willing to pay for the service depends RQWKHXVHIXOQHVVWKH\¿QGLQLW,QWKLVVHQVHLW is important to underline that, although the price of mobile telephony in Latin America is higher as compared to that of the developed regions and can represent 18% of the monthly income (IDRC, 2005), mobile density is achieving acceptable growth levels. What will be the extension of this progress? The economic limitations present in the region seriously condition the geographic extension of the networks. Under the current perspective, once WKH SUR¿WDEOH DUHDV WKRVH ZKHUH WKH SRSXODWLRQ ¿QGLQJ WKH VHUYLFH DIIRUGDEOH OLYH DUH FRYHUHG the growth would probably stop. One factor can change this forecast: the commercial development of new wireless technologies. Wireless networks reduce considerably the deployment costs and will stir the interest of private operators for reaching new regions. Despite the above, it is quite uncertain whether the service will reach depressed and isolated areas unless there is some sort of public intervention. The poverty suffered in most of the rural areas leads to supposing that the predictable price reduction generated by the lower costs of installing wireless technologies and the increase of the competition ZLOO QRW VXFFHHG LQ VLJQL¿FDQWO\ LPSURYLQJ WKH situation. In any case, it seems obvious that the PRVW HI¿FLHQW ZD\ RI SURYLGLQJ DFFHVV LQ DUHDV lacking coverage is through wireless networks capable of providing a broadband that is adaptable to the progressive increase of users. Therefore, it is necessary to continue studying the feasibility of using these technologies, an effort which must be backed by the governments who should provide WKH WRROV UHTXLUHG WR ¿QG WKH RSWLPXP VROXWLRQ (spectrum, funds for research, regional block strategies, and agreements leading to a good use of scale economies). The role of governments must however go be\RQGLIWKHJRDOLVWR¿QGDQDQVZHUWRWKHGLI¿FXOW equation of promoting broadband development to the levels of more developed regions. Public intervention seems essential in Latin America: on one hand, by extending the networks (be it by laying
724
out the network, taking on the role of access provider, or facilitating the economic resources for a SULYDWHRSHUDWRUWRSURYLGHDFFHVVLQQRQSUR¿WDEOH zones) and on the other hand, by insisting on digital literacy programs and promoting adoption. In any case, not to be forgotten is the fact that access to broadband networks is one of the necessary elements for the development of the information society, but not the only one, particularly in places where economic (extreme poverty), cultural (low levels of education and skills), and geographic (dispersion and isolation) limitations obstruct even more the incorporation to the information society. Therefore, network extension must be merged with a joint development strategy. A program of investments in ICT, neglecting other critical developmental priorities would probably achieve no result at all. Such a strategy should move in parallel with the actions implemented as regards the great deal of additional causes that hamper economic development: improving other basic infrastructures, opening up markets, pursuing an effective legal and regulatory system, and providing education for all. To sum up, it is necessary to stress that broadband development is part of a more ambitious plan for the progress of the information society, and the ODWWHULQWXUQPXVW¿QGLWVSODFHLQWKHJHQHUDOORQJ term strategic planning of the country.
ACKNOWLEDGMENT The authors would like to thank the support of the &iWHGUDGHO&ROHJLR2¿FLDOGH,QJHQLHURVGH7HOHcomunicación (COIT, Spain), and of the Consejo Nacional de Ciencia y Tecnología (CONACYT, Mexico) to one of the authors to embrace his PhD studies.
REFERENCES ASETA & UNESCO. (2003). Information society in the Andean community report. Asociación de Empresas de Telecomunicación de la Comunidad Andina. Retrieved July 27, 2007, from http://infolac.
Diffusion of Broadband Access in Latin America
ucol.mx/observatorio/wsis/reunion/05_InformeASETA.pdf Bauer, J.M., Gai, P., & Kim, J. (2002). Broadband: %HQH¿WVDQGSROLF\FKDOOHQJHV. East Lansing, MI: The James H. and Mary B. Quello Center for Telecommunication Management and Law Michigan State University. Cava-Ferreruela, I., & Alabau-Muñoz, A. (2006). Broadband policy assessment: A cross-national empirical analysis. Telecommunications Policy, 30(8-9), 445-463. CRT. (2005). 3URPRFLyQ\PDVL¿FDFLyQGHODEDQGD ancha en Colombia. (Working Paper, Vol. 2). Bogotá, Colombia: Ministerio de Comunicaciones. ECLAC. (2005). Benchmarking the plan of action of the world summit on the information society (WSIS) in Latin America and the Caribbean (Working Paper, Version 3.0). Santiago, Chile: Economic Commission for Latin America and the Caribbean. Falch, M., & Henten, A. (2000). Digital Denmark: From information society to network society. Telecommunication Policy, 24, 377-394. )LUWK/ 0HOORU' %URDGEDQG %HQH¿WV and problems. Regional development and business prospects for ICT and broadband networks. Telecommunications Policy, 29(2-3), 223-236. García-Murillo, M. (2003). Patchwork adoption of ICTs in Latin America. Electronic Journal on Information Systems in Developing Countries, (1),1-9. Gómez-Barroso, J.L., & Feijóo, C. (2006). Public policies against the digital divide: A necessary adaptation to different degrees of development. International Journal of Internet and Enterprise Management, 4(3), 257-268. Gómez-Barroso, J.L., & Pérez-Martínez, J. (2005). Public intervention in the access to advanced telecommunication services: Assessing its theoretical economic basis. Government Information Quarterly, 22(3), 489-504.
Gómez-Barroso, J.L., & Pérez-Martínez, J. (2007). ADSL deployment in the community of Madrid: Investigating the geographical factors of the digital divide. Telematics and Informatics, 24(2), 101-114. Grubesic, T.H. (2004). The geodemographic correlates of broadband access and availability in the United States. Telematics and Informatics, 21(4), 335-358. Hilbert, M., & Katz, J. (2003). Building an information society: A LatinAmerican and Caribbean perspective. Santiago, Chile: Economic Commission for Latin America and the Caribbean (ECLAC). +ROOL¿HOG &$ 'RQQHUPH\HU -) &UHDWLQJGHPDQG ,QÀXHQFLQJLQIRUPDWLRQWHFKQROogy diffusion in rural communities. Government Information Quarterly, 20(2), 135-150. IDRC. (2005). Digital poverty: Latin American and Caribbean Perspectives. Ottawa, Canada: International Development Research Center. IIRSA. (2003). Tecnologías de información y comunicación al servicio de la competitividad y la integración sudamericana: Plan de Acción (Working Paper, Vol. 1). Integración de la Infraestructura Regional en América del Sur – Banco Interamericano de Desarrollo Iniciative. INEGI. (2004). Disponibilidad y uso de tecnologías de información en los hogares de México 2001, 2002 y 2004. Instituto Nacional de Estadística, Geografía e Informática. Estadísticas de Ciencia y Tecnología. Retrieved July 27, 2007, from http://www.canieti. RUJDVVHWV¿OHVGH0D\R'& 3%ADa%20Mundial%20de%20Internet.pdf ITU. (2003a). World telecommunication development report: Access indicators for the information society. Geneva, Switzerland: International Telecommunication Union. ITU. (2003b). Internet report 2003: Birth of broadband. Geneva, Switzerland: International Telecommunication Union. ITU. (2005). ITU world telecommunication indicators database. Geneva, Switzerland: International Telecommunication Union.
725
Diffusion of Broadband Access in Latin America
ITU. (2006). World telecommunication/ICT development report 2006: Measuring ICT for social and economic development. Geneva, Switzerland: International Telecommunication Union. Lanvin, B. (2006). Information, knowledge and competitiveness: Lessons from a worldwide assessment of e-strategies & knowledge strategies. 3DSHUSUHVHQWHGDWWKH,QWHUQDWLRQDO)RUXPRQ,&7¶V Strategies and Investment, Marrakech, Morocco. The World Bank. OECD. (2004). The development of broadband access in rural and remote areas. Committee for Information, Computer and Communications Policy. Working Party on Telecommunication and Information Services Policy: Organization for Economic Co-operation and Development. OECD. (2005). OECD broadband statistics, December 2004. Organization for Economic Co-operation and Development. Retrieved July 27, 2007, from http://www.oecd.org/document/ 60/0,2340,en_2649_37409_2496764_1_1_1_ 37409,00.html OECD. (2006). OECD broadband statistics, December 2005. Telecommunications and Internet Policy. Directorate for Science, Technology and Industry. Retrieved July 27, 2007, from http://www. oecd.org/document/60/0,2340,en_2649_34225_ 2496764_1_1_1_1,00.html Robles Rovalo, A., Gómez Barroso, J.L., Feijóo González, C., & Rojo, D. (2005). Convergencia ¿MRPyYLO3ULPHUDVHYLGHQFLDVHQ/DWLQRDPpULFD AHCIET Journal, 24(101),40-48. Robles Rovalo, A., Gómez Barroso, J.L., & Feijóo González, C. (2006). Service universalization in Latin America: Network evolution and strategies. In B. Preissl & J. Müller (Eds.), Governance of communication networks: Connecting societies and markets with IT (pp. 25-39). Heidelberg: Physica-Verlag, Springer. Strover, S. (2003). The prospects for broadband deployment in rural America. Government Information Quarterly, 20(2), 95-106.
726
World Bank. (2006). Global trends and policies. Information and communications for development. Washington, D.C.: The World Bank.
KEY TERMS Broadband: Any access (last-mile connection) providing transmission speeds above those offered by a conventional line. Generally, in Latin America and consequently, in this chapter, transmission capabilities exceeding 128 kbps are considered. Diffusion: Process by which a service or technology is propagated or adopted over time along the units (individuals) of a system (region or population). Latin America: Geographic and cultural region comprised of Spanish-speaking and Portuguesespeaking countries in North, Central, and South America. Information Society: Means of development or economic and social stage based on the capability of obtaining, sharing, and transmitting information by making the most of the opportunities provided by new technologies. Developing Country or Region: Low and middle-income area where most people have a lower standard of living with access to fewer goods and services than most people in high-income countries.
ENDNOTES 1
2
The broadband generalization not only brings RQEHQH¿WV,WFDQFDXVHDQHJDWLYHLPSDFW as well; “problems” as stated by Firth and Mellor (2005), in different economic, social, and labor-related aspects. The concept of broadband is dynamic. For the ITU (2003b), in its “current condition,” broadband must be considered as any Internet FRQQHFWLRQSURYLGLQJVSHHGV³VLJQL¿FDQWO\´ faster than those achieved with conventional
Diffusion of Broadband Access in Latin America
3
4
5
6
7
WHOHSKRQ\WHFKQRORJLHV,QOLQHZLWKWKLVGH¿nition, in Latin America, broadband is considered any access allowing transmission speeds exceeding those offered by a conventional line (CRT, 2005) and, in some cases, as with the Mexican Regulatory Agency (COFETEL), it LVGH¿QHGDVSURYLGLQJ³WUDQVPLVVLRQFDSDbilities exceeding 128 kbps.” We will use this EURDGEDQGVHUYLFHGH¿QLWLRQIRUWKLVFKDSWHU However, it is convenient to underline that the commercial offers are quickly evolving towards quite faster speeds. The Southern Common Market (MERCOSUR) is the commercial block integrating several South American economies: Argentina, Brazil, Paraguay, Uruguay, and Venezuela (incoming member). Refer to http://www. mercosur.int/msweb/ 7KH$VLD3DFL¿F(FRQRPLF&RRSHUDWLRQRU APEC, is the premier forum formed by 21 economies for facilitating economic growth, cooperation, trade, and investment in the $VLD3DFL¿FUHJLRQ0H[LFRLVDOVRDPHPEHU of the North American Free Trade Agreement (NAFTA), the trade block integrated by Canada, the United States, and Mexico, the three countries that integrate North America. Refer to http://www.apec.org/content/apec/ about_apec.html and http://www.nafta-secalena.org/ The Andean Community of Nations (CAN, Comunidad Andina de Naciones) is the cooperation bloc comprising the Latin American countries of Bolivia, Colombia, Ecuador, Peru, and Venezuela (retiring member). Refer to http://www.comunidadandina.org/endex. htm If access to the information and communications technologies (ICT) is the prior and essential requirement for achieving an inclusive information society, broadband access (hereinafter simply referred to as broadband) is “the base,” as per the expression used by Bauer, Gai, and Kim (2002) or the “infrastructure of” (paraphrasing the ITU, 2003b), of the knowledge economy. Cava-Ferreruela and Alabau-Muñoz (2006) provide a cross-national empirical analysis
8
9
10
11
12
13
14
15
16
WKDWH[SORUHVWKHIDFWRUVLQÀXHQFLQJEURDGband supply, demand, and adoption. Public action must be based on one or several of the causes that justify state intervention on the activity of the sector. Gómez-Barroso and Pérez-Martínez (2005) analyze the presence of these “justifying” motives (“market failures”) in telecommunications in general and in new infrastructures and services in particular, and conclude that most of the arguments could justify public intervention. Statement of the Governments gathered in Florianopolis (Brazil) during June 2000. Ministerial Meeting on Information and Communication Technologies. See http:// www.eclac.cl/publicaciones/secretariaejecuWLYDOFOÀRULDQRSROLVKWP The Connectivity Agenda for the Americas is the result of the Americas Summit held in Quebec. See http://www.summit-americas. org/Documents%20for%20Quebec%20City %20Summit/Quebec/connecting-Span.htm Statement of the preparatory Regional Ministerial Conference before the World Summit on the Information Society held in Bavaro, Dominican Republic. See http://www.eclac. FOSUHQVDQRWLFLDVQRWLFLDV %DYDUR¿nalesp.pdf The Mexican (e-Mexico National System) and Chilean (Digital Action Group) programs, literally state that the government provides a “catalytic role” in the promotion and adoption of digital technologies. Chile (22), Mexico (31), Brazil (33), and Argentina (34) are among the 35 best positions RIWKH8QLWHG1DWLRQVFODVVL¿FDWLRQRQHJRYernment (2005). On the other hand, Mexico is seventh in the so-called e-participation section. Comparable to the accelerated increase of mobile telephony subscribers. Refer to Robles Rovalo, Gómez Barroso, & Feijóo González (2005). DSL Forum: http://www.dslforum.org/ dslnews/pr.shtml 6SHFL¿F WHFKQRORJLHV QRQZLUHOHVV IRU WKH construction of local area networks and
727
Diffusion of Broadband Access in Latin America
17
18
728
particularly addressed to business customers (“leased” lines), are included here, as well as other more expensive ones, such as GHSOR\LQJRSWLF¿EUHWRWKHVXEVFULEHU¶VGRRU (with some importance in countries such as Japan or Korea). This means that wireless solutions, which are particularly adequate IRUSODFHVZLWKLQVXI¿FLHQWLQIUDVWUXFWXUHV were hardly even developed at the time. Currently, Honduras still has no offer targeted at households. Recently, the remaining markets have started offering service: In Bolivia (COMTECO), broadband service was not available until 2005; in Guatemala the latest data available show that there are only satellite providers (Java Networks) and some WHVWVZLWK:L0$;
19
20
21
22
23
24
This relationship is not direct in the remaining cases; Peru has been the last country of the six considered in launching their national strategy, in 2003; however, it broadly exceeds the rates of Colombia who implemented their program three years earlier. Mexico has a National Plan since 2001. In the latest ITU Basic Indicators/ICT statistics Database 2006 report, telephone accesses DUHGH¿QHGDVWKHJURXSRIPRELOHDQG¿[HG subscribers. It is necessary to underline that the correlation is greater should only the mobile telephony rates be considered where the value R=0.87 is used. In this section, the World Bank (Lanvin, 2006) stresses the cases of Chile, Thailand, and Estonia due to the fast progress of their network-readiness levels. In this sense, one of the main conclusions of the OECD (2004) is that “Competition in the provision of broadband access is emerging in rural areas, and governments should take this into account before embarking on programs to subsidize infrastructure.” Satellite connections providing national coverage are not included in this section.
729
Chapter XLV
Diffusion Forecasting and Price Evolution of Broadband Telecommunication Services in Europe Dimitris Varoutas University of Athens, Greece Christos Michalakelis University of Athens, Greece Alexander Vavoulas University of Athens, Greece Konstantina Deligiorgi University of Athens, Greece
ABSTRACT This chapter is concerned with the methodologies for the study of the diffusion patterns and demand estimation, as well the pricing schemas for broadband telecommunication services in Europe. Along with the introduction of diffusion models and price indexes which can represent broadband convergence and diversity, a description of the theoretical models and methodologies are given and application of these models in European telecommunication market is performed. Evidence from Europe outlines telecom market behavior and contributes to better understanding of broadband diffusion worldwide. To this direction, a price index is constructed regarding the ADSL technology.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Diffusion and Price of Broadband in Europe
INTRODUCTION The worldwide expansion of the Internet and other network-based services forced the participant agents of the area of telecommunications to reconsider their strategic plans, in order to be able to meet the continuously increasing demand for required bandwidth as well as the continuously increasing numbers of users. On the other hand, as network infrastructures are being constructed to support the increasing demand, many multimedia applications will be UHDOL]HG 1RZDGD\V FXVWRPHUV¶ GHVLUHV FRQVLVW of high-speed Internet access, as much as interactive, bandwidth—consuming multimedia applications. Following these considerations, the area of telecommunications merits a continuous improvement and development towards the direction of the quantity and quality of the offered services. Contemporary technology allows extended network capabilities and the development of new products, which in turn increase the quality of services offered to customers. However, convergence of telecommunication services often disorientates customers and regulators; the former concerning their potential selections among the offered products and the latter regarding market monitoring and regulation. The price at which a service or DSURGXFWLVRIIHUHGLVDVXEVWDQWLDOO\LQÀXHQWLDO factor, as it is related to its future diffusion among the potential adopters. In addition, from an economics point of view, telecommunication services obviously reveal the characteristics which strongly relate them with the network externalities effects, the phenomenon that a good becomes more valuable to each user the more other consumers use the same or a compatible product. Summarizing the above considerations, regarding demand evolution and pricing shapes of SURGXFWVDQGWKHUHODWHGLQÀXHQWLDOSDUDPHWHUVWKLV chapter is devoted to the provision of answers/outlines concerning the following questions:
•
•
The rest of the chapter is organized as follows: a short overview of diffusion theory and diffusion models followed by a study of the evaluation of diffusion models over selected cases. The factors affecting broadband access development in whole Europe are then presented. An introduction of telecommunication services and ADSL high-speed Internet connections is given in the next section and then theoretical background to econometric PHWKRGVDQGHPSLULFDOPRGHOVDUHJLYHQ7KH¿QDO section proceeds with a concluding discussion and suggestions for future development in the telecommunications area.
BACKGROUND Diffusion Theory and Diffusion Models As it is evident that innovations in the area of telecommunications and high technology, in genHUDODUHIDFLQJDVLJQL¿FDQWDQGTXLWHSURPLVLQJ expansion, corresponding research activities focus on the study of the diffusion process dynamics. Their main targets aim towards the direction of developing and applying methods to provide comprehensive analyses concerning the demand for telecommunications products, services, and technology (Fildes & Kumar, 2002). The main areas of research interest include: •
•
730
How can we model the diffusion of broadband networks and services?
How can we model the impact of countries/ regions with higher broadband penetration to those with lower penetration levels? How can estimate prices for broadband services and products that enter the market IRUWKH¿UVWWLPHRUKDYHEHHQPRGL¿HGFRQverged? &DQZHGHWHUPLQHDXQL¿HGSULFHLQGH[IRU WKHVHSURGXFWVLQDVSHFL¿FSHULRGDQGZKDW do these prices tend to become over time?
•
The study and development of demand models The development of mathematical methods for WKHHVWLPDWLRQRIWKHPRGHOV¶SDUDPHWHUV
Diffusion and Price of Broadband in Europe
•
The evaluation of the results related to the PRGHOV¶SHUIRUPDQFHRYHUFDVHVRILQWHUHVW
The above cases are not exhaustive, as in a more detailed analysis, the study of the diffusion of multigenerational products (Rai, 1999), the LQÀXHQFHRQGHPDQGRIFURVVQDWLRQDOGLIIXVLRQ (Kumar & Trichy, 2002), as well as the development of methods for estimating diffusion without or with limited prior data should be included (Venkatesan & Kumar, 2002). The latter cases constitute quite common characteristics in the area of telecommunication innovations and in the wider area of high technology products. Diffusion theory is a methodological approach used for estimating the adoption of technological innovations or other products or services. Diffusion models are, consequently, mathematical functions of time which are used to estimate the parameters RIWKHGLIIXVLRQSURFHVVRIDSURGXFW¶VOLIHF\FOHDW an aggregate level, without taking in consideration WKHXQGHUO\LQJVSHFL¿FSDUDPHWHUVWKDWGULYHWKH process (Bass, 1969). They are constructed such that they have a main target of capturing the general WUHQGVRIWKHPDUNHWUHDFWLRQVWRDQLQQRYDWLRQ¶V introduction. These models are of major importance LQGHWHUPLQLQJWKHSURGXFW¶VH[SHFWHGOLIHF\FOH and the associated parameters such as maximum penetration. The cumulative diffusion shapes of innovations are often described by sigmoid patterns, the so called S-shaped growth patterns, which originate from a number of early buyers (innovators). These initial adopters of the services and the size of the “critical mass” needed are of VLJQL¿FDQWLPSRUWDQFHIRUWKHUHVWRIWKHGLIIXVLRQ process, especially for the saturation level and the VDWXUDWLRQ WLPH ,QQRYDWRUV¶ GHFLVLRQV WR DGRSW the service are independent from the decisions of the rest of the social system population. Apart from the innovators, there is another category of adopters, the imitators, who proceed to the adopWLRQRIWKHVHUYLFHLQÀXHQFHGE\WKHLQWHUDFWLRQ with innovators (word-of-mouth) and by external LQÀXHQFHVXFKDVPDVVPHGLDFRPPXQLFDWLRQDQG other communication channels. Finally, the market reaches maturity, when the maximum number of adopters among the considered population is met (market saturation).
Based on the concepts of diffusion theory presented above, the present section attempts to provide an insight concerning estimation and forecasting of broadband technology diffusion underlying mechanics. This purpose is the main reason for employing a number of existing diffusion models, the results of which can be considered DVDQLQWXLWLYHFRQ¿GHQFHLQWHUYDOGHVFULELQJWKH H[SHFWHGYDOXHVRIWKHVSHFL¿FPDUNHW¶VXOWLPDWH potential. These aspects constitute the main functional utility of the present work.
Diffusion Models The most widely used representatives of the aggregate models developed for diffusion estimation are the Bass model (Bass, 1969), the Fisher-Pry model (Fisher & Pry, 1971), logistic family models (Bewley & Fiebig, 1988), and the Gompertz model (Rai, 1999). Logistic models and variations of the Gompertz model provide “S-shaped” curves which are used in common in forecasting diffusion of products or services. S-shaped patterns derive from the following differential equation:
dY(t) = dt
(1)
In (1), Y(t) represents total penetration at time W6WKHVDWXUDWLRQOHYHORIWKHVSHFL¿FWHFKQRORJ\ DQGįLVDFRQVWDQWRISURSRUWLRQDOLW\WKHVRFDOOHG FRHI¿FLHQW RI GLIIXVLRQ 3HQHWUDWLRQ LV GH¿QHG as the proportion of the population that uses the product or service being examined. In that sense, the diffusion rate of a product is proportional to the already recorded penetration as well as to the UHPDLQLQJSRWHQWLDORIWKHPDUNHW¶VXVHUV As observed in (1), diffusion speed, as a function of time, is proportional to a product of two factors: the population that has already adopted the service, denoted by Y(t), as well as to the remaining market potential represented by S-Y(t). It is easily derived, given t>0, that as the time variable increases Y(t) also increases towards S, and this can be interpreted as an increase of the “pressure” of the adopters over the nonadopters.
731
Diffusion and Price of Broadband in Europe
Based on the above assumption, a number of diffusion models have been developed for diffusion estimation and forecasting purposes. The most widely used ones include the Bass model (Bass, 1969), the Gompertz model, and models of the logistic family (Bewley & Fiebig, 1988). The linear logistic model, a representative diffusion model which belongs to the logistic family models, was used for evaluating the cases presented in this work. The linear logistic model is also known as FisherPry model (Fisher & Pry, 1971). It is described by the following equation:
Y(t) =
S 1+e-a-bt
(2)
Cross-National Diffusion In many cases, the same product is introduced either simultaneously, or after a time lag, into a number of different markets (Kumar, Ganesh, & Echambadi, 1998). This is quite frequent especially in the markets of high technology products because they usually target international markets. In such cases, it is quite important to study the diffusion proceGXUHLQZHOOGH¿QHGJURXSVRIPDUNHWVDFFRUGLQJ to their characteristics, and the interaction among WKHP7KHVHPDUNHWJURXSVFDQEHGH¿QHGHLWKHUDV a group of neighboring countries, as a number of areas within the boundaries of the same country, or any other kind of geographical segmentation in general, according to various geographical or social characteristics imposed (Gatignon, Eliashberg, & Robertson, 1989; Michalakelis, Dede, Varoutas, & Sphicopoulos, 2005; Takada & Jain, 1991). The effects of simultaneous introduction are UHODWHGWRWKHLQÀXHQWLDOEHKDYLRUEHWZHHQXVHUV of the corresponding markets as a result of people interaction (Bass, 1969). This fact is usually not taken into account when estimating the diffusion process of the product and the penetration among studied population. Thus, the effect of market LQWHUDFWLRQ DQG WKH FRQVHTXHQW FRLQÀXHQFH LQ the diffusion rates is overlooked, although it can be able enough to modify the initially estimated diffusion process.
732
EVALUATION OF DIFFUSION ESTIMATION AND FORECASTING In this section, evaluation results of the diffusion theory introduced in the preceding sections are presented, with respect to the ADSL technology. The datasets used for this evaluation describe penetration of ADSL connectivity across the European area. The corresponding data sources were (XURVWDW¶V DQG 2(&'¶V :HE VLWHV DV LQGLFDWHG while presenting the datasets. The cases considered are the average values of penetration in the European Union (15 countries) and in the OECD countries. They were selected in order to depict the worldwide dynamics in ADSL adoption. In DGGLWLRQWZRVSHFL¿F(XURSHDQFRXQWULHVDUHDOVR included, Portugal and the United Kingdom. This particular choice was made in order to examine the diffusion process between two diverging countries; one considered developing (Portugal) and one considered developed (UK), together with their relative position as compared to EU and OECD mean values of ADSL technology penetration. The historical data are semiannual values, as indicated by the values next to the year (e.g., 2002m10 refers to October of 2002) and they refer to penetration of ADSL technology for residential access only. It is obvious that the same evaluation can be performed over data regarding business connections. In the presented diagrams the horizontal axis (Years) depicts time, whereas the vertical (S) refers to the penetration value, over the referenced population. At this point, it should be noted that the term “population” has a meaning that varies, depending on each examined case, which must be FODUL¿HGHDFKWLPH7KXVLQWKLVSDUWLFXODUFDVH the term population refers to individuals rather than households. A quite accurate approximation for performing a transformation between households and people, as this derives from demographic data (source: Eurostat), is the ratio 1:3. In this way, the results can be easily converted to depict penetration among households. The estimation and forecasting evaluation process was performed by using the linear logistic model (Fisher & Pry, 1971) while the estimation of the parameters was performed by employing
Diffusion and Price of Broadband in Europe
nonlinear regression (NLS). Moreover, the accuracy of the results provided and the consequent effectiveness of the model was based on appropriate statistical measures, such as the mean square error (MSE), the mean absolute error (MAE) and the mean absolute percentage error (MAPE). Another SUDFWLFHDGRSWHGIRUGHWHUPLQLQJWKHPRGHO¶VGDWD ¿WWLQJDELOLW\ZDVWRKROGEDFNVRPH\HDUVRIGDWD DQGXVHWKHPLQRUGHUWRGHWHUPLQHWKHPRGHO¶V forecasting power, based on the earlier portion of
WKHGDWDVHWWKHWUDLQLQJGDWD0RUHVSHFL¿FDOO\ four simulations have been performed by holding back one, two, three, and four points of data, respectively. The evaluation results are presented in the following tables (Table 1 through Table 4) and LQ WKH FRUUHVSRQGLQJ ¿JXUHV )LJXUH WKURXJK Figure 4). Table 5 summarizes the saturation levels, as these were estimated by the model, together with
7DEOH$FWXDOHVWLPDWHGDQGIRUHFDVWHGYDOXHVRI$'6/SHQHWUDWLRQRYHU(8 &RXQWULHV 1 Year
EU 15 – Actual
EU 15 - Est.
2002m07
2,3
2,4
2002m10
2,7
2,9
2003m01
3,4
3,5
2003m07
4,5
4,2
2003m10
5,1
5,0
2004m01
6,0
6,0
2004m07
7,6
7,2
2004m10
8,4
8,6
2005m01
9,9
10,2
2005m07
12,0
12,0
2005m12
14,2
14,1
2006m01
16,5
2006m10
19,1
2006m12
22,0
2007m04
25,2
)LJXUH$'6/GLIIXVLRQ (XURSHDQ8QLRQ FRXQWULHV ADSL Diffusion - EU15
30,0
25,0
15,0
10,0
5,0
0,0
20 02 m 07 20 02 m 10 20 03 m 01 20 03 m 07 20 03 m 10 20 04 m 01 20 04 m 07 20 04 m 10 20 05 m 01 20 05 m 07 20 05 m 12 20 06 m 01 20 06 m 10 20 06 m 12 20 07 m 04
S
20,0
Years EU 15 - Actual
EU 15 - Est.
733
Diffusion and Price of Broadband in Europe
Table 2. Actual, estimated, and forecasted values of ADSL penetration for OECD countries2 Year
OECD - Actual
OECD - Est.
2001
2,9
2,9
2002
4,9
4,8
2003
7,3
7,3
2004
10,2
10,2
2005
13,6
13,6
2006
17,2
2007
20,8
2008
24,4
2009
27,9
Figure 2. ADSL diffusion: OECD countries ADSL Diffusion - OECD Countries
30,0
25,0
S
20,0
15,0
10,0
5,0
0,0 2001
2002
2003
2004
2005
2006
2007
2008
2009
Years OECD - Actual
the expected year of reaching them, in each case. In addition, Figure 5 provides a comparison view among the evaluated cases. As it can be observed, the diffusion process in the European area is moving towards a rapid growth, and it is expected to have covered a percentage of approximately 76%
734
OECD - Est.
until year 2012. A somewhat slower but also steady growth is the process of ADSL technology adoption among the OECD countries. The mean European Union diffusion is expected to reach a higher level of saturation than the OECD mean value earlier in time than the latter.
Diffusion and Price of Broadband in Europe
Table 3. Actual, estimated, and forecasted values of ADSL penetration in Portugal3 Year
Portugal - Actual
Portugal - Est.
2002m07
1,5
1,6
2002m10
1,8
2,0
2003m01
2,5
2,6
2003m07
3,6
3,2
2003m10
4,1
4,0
2004m01
4,8
5,0
2004m07
6,4
6,0
2004m10
7,2
7,3
2005m01
8,2
8,6
2005m07
10,1
10,0
2005m12
11,5
11,4
2006m01
12,9
2006m10
14,2
2006m12
15,5
2007m04
16,6
Figure 3. ADSL diffusion: Portugal ADSL Diffusion - Portugal
18,0 16,0 14,0
10,0 8,0 6,0 4,0 2,0
20 07 m 04
20 06 m 12
20 06 m 10
20 06 m 01
20 05 m 12
20 05 m 07
20 05 m 01
20 04 m 10
20 04 m 07
20 04 m 01
20 03 m 10
20 03 m 07
20 03 m 01
20 02 m 10
0,0 20 02 m 07
S
12,0
Years Portugal - Actual
Portugal - Est.
735
Diffusion and Price of Broadband in Europe
Table 4. Actual, estimated, and forecasted values of ADSL penetration in UK4 Year
UK - Actual
UK - Est.
2002m07
1,6
1,7
2002m10
2,1
2,1
2003m01
2,6
2,7
2003m07
3,7
3,5
2003m10
4,4
4,4
2004m01
5,3
5,6
2004m07
7,4
7,0
2004m10
8,8
8,7
2005m01
10,3
10,8
2005m07
13,5
13,2
2005m12
15,9
15,9
2006m01
19,0
2006m10
22,3
2006m12
25,8
2007m04
29,3
Figure 4. ADSL diffusion: UK ADSL Diffusion - UK 35,0 30,0 25,0
S
20,0 15,0 10,0 5,0
20 02 m 07 20 02 m1 0 20 03 m 01 20 03 m 07 20 03 m 10 20 04 m0 1 20 04 m 07 20 04 m 10 20 05 m0 1 20 05 m 07 20 05 m 12 20 06 m 01 20 06 m1 0 20 06 m 12 20 07 m 04
0,0
Years UK - Actual
UK - Est.
7DEOH6DWXUDWLRQOHYHOVRI$'6/SHQHWUDWLRQDQGH[SHFWHGWLPHWRVDWXUDWLRQ
736
EU
OECD
Portugal
UK
Saturation Level
76,02
35,46
22,01
54,61
Year
2012
2014
2019
2018
Diffusion and Price of Broadband in Europe
)LJXUH$'6/GLIIXVLRQ &RPSDULVRQDPRQJ(82(&'DQGVHOHFWHG(XURSHDQFRXQWULHV ADSL diffusion - comparison 35,0 30,0 25,0
S
20,0 15,0 10,0 5,0
20 07 m 04
20 06 m 12
20 06 m 10
20 06 m 01
20 05 m 12
20 05 m 07
20 05 m 01
20 04 m 10
20 04 m 07
20 04 m 01
20 03 m 10
20 03 m 07
20 03 m 01
20 02 m 10
20 02 m 07
0,0
Years EU-15
OECD
Regarding the evaluation results concerning the two participating countries, Portugal and UK, although diffusion was at the same levels at the initial stages, the results indicate that in UK the diffusion rate is expected to be higher and be met earlier. Consequently, the saturation level is expected to reach a quite higher level in UK than in Portugal. This is strongly related to the rate that high technology products penetrate in each country, together with other reasons such as demographic, cultural, and economic ones. By inspecting Figure 5, the dynamics of ADSL technology penetration can be derived. More VSHFL¿FDOO\LWFDQEHREVHUYHGWKDWLQWKH2(&' countries ADSL has met a higher level of penetration than in EU, from the year 2002 to the year 2006. However, the diffusion rate was higher in EU (the slope of the corresponding curve), which resulted in EU mean penetration becoming higher than the corresponding one of the OECD countries. According to this estimation, a higher saturation level is expected to be recorded in EU than in OECD, almost two times higher. In addition, in the United Kingdom penetration was steadily above the European mean value and after the year 2005 revealed a substantial diver-
Portugal
UK
sion above it. Exactly the opposite is the case of Portugal, which was below the European mean diffusion during all the time of study. After the year 2005 diverged negatively from the European mean, it retains a steadily incremental rate. Concluding, the performed evaluation provided an estimation of the so-far recorded diffusion of the ADSL technology among the studied markets and a forecasting for the expected future process. These results, especially for the part related to forecasting, were derived based on the current dynamics of the diffusion process. This means that they can be considered more as a worst case scenario of the expectations for ADSL adoption than as absolute values of penetration. To be PRUH VSHFL¿F LI JRYHUQPHQWDO DQG RWKHU SXEOLF or private initiatives are supported towards this direction, the diffusion of the ADSL technology should be expected to grow faster and saturate at a much higher level. In any case, the recorded penetration among the referenced market must be fed into the diffusion model and the estimation VKRXOGEHUHFDOFXODWHGLQRUGHUWRSURYLGHUH¿QHG forecasting estimations. As mentioned above, the drivers affecting diffusion are related to a number RIGHPRJUDSKLFVRFLDODQG¿QDQFLDODVSHFWVSURE-
737
Diffusion and Price of Broadband in Europe
ably one the most important of them being pricing policies. This is the main subject of consideration of the following section.
GDP Per Capita and Investments in Whole Europe A review in the European status shows a variation in telecom sector investments and Eastern Europe countries are trying to balance Western Europe countries. Overall, the level of existing investment in Eastern European countries is quite less than that of Western countries. However, the rate at which Eastern countries are investing in telecommunications remains amazing. The strongest per capita subscriber growth comes from Denmark, Austria, Norway, the Netherlands, Finland, Luxembourg, Sweden, and the United Kingdom. In each of these countries more than six subscribers per 100 inhabitants were added, during the past year. Generally, investments expected to be made during each year, or a period of time, depend on the required equipments and network infrastructure. However, a crucial role to these investments is the fact that the equipment prices are decreasing over time due to the mass production (Broadwan, 2WKHUVLJQL¿FDQWSDUDPHWHUVLQWKHLQYHVWment plans construction are the expected costs that
FACTORS AFFECTING BROADBAND ACCESS DEVELOPMENT IN THE WHOLE OF EUROPE 0DQ\ IDFWRUV LQÀXHQFH WKH GLIIXVLRQ SURFHVV LQ the telecommunication sector and, consequently, broadband access development in all of Europe. The amount of investments in telecommunications networks is one of the most important factors. In addition, liberalization and privatization in Europe is of great interest while the increased number of operators and carriers has a great impact to the whole market. Furthermore, the globalization of the economy in Europe provokes the necessity of a regulatory framework by the European CommisVLRQ)LQDOO\WKHVXEVFULEHUV¶SUR¿OHLVH[DPLQHG while the governmental policies and guidelines for the telecom market are also of considerable importance.
Figure 6. Broadband penetration and GDP per capita Broadband penetration 30
GDP per capita 70.000 Broadband penetration (subscribers per 100 inhabitants, June 2006) GDP per capita (USD PPP, 2004)
25
60.000 50.000
20
40.000 15 30.000 10
20.000
pu bl ic Gr ee
ce
y
nd Re ak ov
ar
la Po Sl
ng
lan d
Hu
ub lic ep
hR ec Cz
I re
l ga rt u Po
n
ly I ta
ai
y an
Sp
ce rm Ge
st r
rg
an Fr
Au
m ou mb
xe
om Be lg i u
Lu
d Un
i te
738
ng d
Ki
Fi
er l a
it z Sw
er
nm th
De
Ne
Source : OECD
ia
0 nd n la nd No rw ay Sw ed en
0 ar k
10.000
la n ds I ce la n d
5
Diffusion and Price of Broadband in Europe
will probably appear at the studied period, such as maintenance and repair costs of the existing network infrastructure, licenses costs (software and hardware), marketing costs for promoting a SURGXFWDQGFXVWRPHUV¶LQVWDOODWLRQ0DQ\LQYHVWments plans have been proposed in order to estimate WKHVHSDUDPHWHUV+RZHYHUWKHPRVWGLI¿FXOWSDUWWR estimate are the running costs, since new services appear and in replacement of others. In addition, it has been noticed that higher-income households are using Internet more often, for a longer period of time (Dickenson & Ellison, 1999) and for a variety of services (e.g., goods purchase, banking services, services under creditcard, video, audio, etc.) while people who live in urban areas are more likely to use Internet more often than people who live in suburban and rural areas. Thus, there is a higher investment risk in the latter areas with a long period of pay back, which in turn leads in a growing inequality of the Internet access capability among different geographical regions (Monath, Elnegaard, Cadro, Katsianis, & Varoutas, 2003).
From a Monopoly Situation to the Liberalization Telecommunications operated under governmental provision and monopolistic status quo for a long period of time. The main reason, besides the governmental strategic perception, was military security. However, in the last two decades of the 20th century, governments reconsidered this option which resulted in the start of liberalization in telecommunications. In Table 6, the year of liberalization for several countries across Europe is provided. Legislation alone does not seem to be enough for the liberalization process to begin. The change from a central planned economy into a free market LVGLI¿FXOWDQGFXVWRPHUVKDYHWROHDUQUHDFWDQG interact in order to take advantage. When liberalization started in the telecommunications market, D¿HUFHFRPSHWLWLRQHPHUJHGDPRQJFRPSDQLHVDV they tried to attract as many customers as possible by offering packages in quite competitive prices. As a consequence, tariffs started to decrease and be-
ing more accessible to potential customers. Within West Europe we can distinguish a difference in behavior between the North and the South. In some European countries, telecommunication sector followed a great development, but as it is believed they were “forced” by European Community to change to liberalization and privatization facts in order to have investments in their countries and become more competitive. Sweden can be considered as a characteristic example, the leader of all European countries, where no monopolies exist. However, privatization raised other problems such as staff reduction in telecom companies, which is DQDFFXVWRPHGZD\WRLQFUHDVHSUR¿W2QWKHRWKHU side, the growing competition led to advantages for consumers who enjoyed lower tariffs. Countries of Eastern Europe have to overcome major economic problems which are preventing them to adopt European legislation more quickly.
Rapid Technological Change The telecom sector is changing rapidly in the last decades, due not only to the fast technological improvement in software and hardware but also in network infrastructure. The development of cooperation between telecommunication industries across Europe was of substantial importance for information and knowledge sharing. As a result, YDULRXV FKDQJHV LQ FRXQWULHV¶ DQG FRPSDQLHV¶ SUR¿OHWRRNSODFH&RPSDQLHVQHHGHGUHVRXUFHV and this translated into people, engineers, and investments. Another reason for this international cooperation derived from the mass production of several, continuously smaller, components. These parts are constructed in countries which have cheap “labor hands” and lower operational costs. In Eastern Europe the quality of networks is generally low, with digitalization in low levels in Ukraine, Moldova, Belarus, and Bulgaria (Broadwan, 2004). Telecom operators are expected to provide broadband communication services in all geoJUDSKLFDODUHDVGXHWRVXEVFULEHUV¶SUHIHUHQFHVDQG the growing political interest (e.g., e-Europe 2005 guideline), necessitating the technological upgrading as inevitable. As a result, network operators
739
Diffusion and Price of Broadband in Europe
Table 6. Year of liberalization for countries of Europe6 Country
Liberalization
United Kingdom
1991
Sweden
1996
Nertherlands
1997
Italy
1997
Belgium
1998
Austria
1998
France
1998
Germany
1998
Spain
1998
Luxenbourg
2000
Ireland
2000
Portugal
2000
Greece
2001
Czech Republic
2001
Estonia
2001
Poland
2001
Slovenia
2001
Hungary
2002
Latvia
2003
F.Y.R.O.M. (Former Yugoslavian Republic of
2003
Macedonia) Bulgaria
2006
Lithuania
2006
Romania
2006
are enforced to increase the available bandwidth in order to give more services to customers. This is achieved by replacing parts of existing networks QHWZRUN ZLWK ¿EHU LQIUDVWUXFWXUH HVSHFLDOO\ LQ DUHDV ZLWK VSHFL¿F FKDUDFWHULVWLFV VXFK DV KLJK density of population—potential subscribers. In addition, they have to identify these characteristics and make corresponding plans in order to ensure a VHFXUHSUR¿WDIWHUXSJUDGLQJWKHLUQHWZRUN0RUH VSHFL¿FDOO\WKH\KDYHWRH[DPLQHWKHH[LVWHQFHRI these characteristics in urban, suburban, and rural areas. These include the necessary cable length for the expected connections, the number of subscribers per square km, the number of buildings in each VSHFL¿FDUHDDVZHOODVWKHQXPEHURISRWHQWLDO subscribers per building (Monath et al., 2003). As DFRQVHTXHQFHDGHWDLOHGVWXG\DQGLGHQWL¿FDWLRQ of the necessary network infrastructure in order WRVDWLVI\,QWHUQHWVHUYLFHVDQGWUDI¿FFKDUDFWHUistics is extracted. It is worth mentioning that in North Europe, the differences between rural and urban areas concerning the use of Internet are quite small, while the differences in south Europe are observable (Kalhagen & Olsen, 2002). Rural areas of Northern Europe have low population density, which leads to very high costs of network infrastructures. On the contrary, in South Europe the situation is quite different. In rural areas, with high population density, which is an advantage for the network infrastructure and equipment cost, the demand for broadband services is low as a result of social and economic differences between them and the urban areas. Generally, as observed in Figure 7, Iceland, Sweden, and Norway (Northern Europe) met a high level of broadband penetration, even if in these areas population density is low, where in countries like Greece, Poland, and Germany (Central and Southern Europe) where population density is higher, the corresponding broadband penetration is low.
Operators, Carriers, and Competition Network operators have the responsibility to install, manage, and operate telecommunication transmission network so as to offer public tele-
740
Diffusion and Price of Broadband in Europe
Figure 7. Broadband penetration and population densities7 Broadband penetration 30 25
Population density Broadband penetration (subscribers per 100 inhabitants, June 2006) Population density (inhab/km2, 2004)
20
450 400 350 300 250
15
200
10
150 100
5
50 0
Den
m Net ark herl and I ce s S w la n itz e d r la n F in d la n d Nor way Uni Swede ted n Kin gdo Be m lg Lux i u m emb ourg Aus tria Fran c Ger e man y Sp ai n Italy Por Cze tugal ch R epu blic Irel an Hun d gary P Slov olan ak R d epu blic Gre ece
0
Source : OECD
phone or network services. They are distinguished in local operators, which offer services to users ZKROLYHLQVSHFL¿FDUHDVDQGQDWLRQDORSHUDWRUV which offer services no matter where the user lives. They were “forced” by the EU Commission to provide a minimum set of telecommunication VHUYLFHVDFFRUGLQJWRVSHFL¿FWHFKQLFDOVWDQGDUGV This means that EU Commission applies similar conditions in similar circumstances to companies which provide similar services by the meaning of the same quality. Different number of operators that have license or authorization to offer network services with different entrance times in the market can be found across member states. Thus, operators are developing new tariff policies with different customers paying different prices for the same service, depending on the time they signed a contract. In Eastern Europe, only a few operators are PDNLQJPXFKSUR¿WEHFDXVHHYHQLIWKHUHLVDKLJK Internet penetration rate; the usage of Internet is far away from Southern European countries (Gunther, 2001). Low Internet penetration constitutes a brake
on their market. In addition, according to Table 7, it is obvious that in EU candidate countries which are Bulgaria, Croatia, F.Y.R.O.M., Romania, Turkey, and the rest of European countries there is a VLJQL¿FDQWXVDJHJURZWKDFURVV\HDUVZKHUHDVWKH Internet penetration compared to the population is quite less yet. In addition to the information provided in Table 7, carriers are promoted in all of Europe. Carriers are companies that buy lines in wholesale and are independent of technical changes as they can act quicker. They buy big capacities of lines at lower prices and they rent them. This will lead to a continuing reduction of prices since customers can choose among several carriers. In order to ensure high quality and reliability of Internet services, telecom operators offer a number of choices according to the customer demands and paying abilities. As a result, there is D¿HUFHFRPSHWLWLRQDPRQJWKHPWRDWWUDFWPRUH subscribers (Goff, 2002). Some of them offer contracts with free of charge extra time to an Internet connection, the duration of which depends on the
741
Diffusion and Price of Broadband in Europe
Table 7. Internet usage in Europe8 Europe European Union EU candidate countries Rest of Europe
Population (2006
% population
Internet users,
Penetration (%
% Usage of
Use growth
estimation)
of world
Latest Data
population)
world
(2000 – 2006)
462.371.237
7.1 %
239.881.917
51,9 %
22,1 %
157,5 %
110.206.019
1.7 %
24.983.771
22,7 %
2,3 %
622,1 %
234.711.764
3.6 %
43.847.215
18,7 %
4,0 %
417,5 %
offered bandwidth. There is also a great difference between residential and business packages offered. Currently, operators offer ADSL packages with varying data rates. There are low prices for Internet connections that are designed to attract residential customers, while the most expensive ones, providing higher speed connections, are designed for business customers. On the other hand, some operators offer free unlimited access to Internet, provided that telephone companies agree to pay these providers (Haan, 2001). It is REYLRXV WKDW DV WKH FRPSHWLWLRQ DPRQJ ¿UPV LV growing, customers will enjoy lower prices with better services.
Delivery and Repair Times of Telecommunication Services by Companies The time period from the date when the user PDNHVDUHTXHVWIRUDVSHFL¿FW\SHRIDWHOHFRPPXQLFDWLRQVHUYLFHXQWLOWKHUHTXHVWLVIXO¿OOHGLV called delivery time. The determination for each FRXQWU\LVGLI¿FXOWPDLQO\GXHWRWKHODUJHQXPber of operators. However, it seems that serious delays still remain in some countries, even if there are improvements. This is a major problem for the telecommunication sector and especially the broadband expansion, since the rapid accommodation is a key component for the ongoing diffusion of broadband services. On the other hand, the intermediate period of time from when a failure message has been given to the responsible unit within the organization, ZKLFK SURYLGHG WKH VSHFL¿F W\SH RI WHOHFRPmunication service, until its connection has been
742
UHHVWDEOLVKHGDQGQRWL¿HGEDFNLQRSHUDWLRQWRWKH user is called repair time. There are remarkable differences across member states at repair times too. This points out an important impact on the market again because the quality of service and the response time for repairing it, if any damage happens, shows a continuing competition among FRPSDQLHVVRWKDWFXVWRPHUVDUHVDWLV¿HGDQGDV a result, prices decline.
6XEVFULEHUV¶3UR¿OH It is evident that people across Europe are accessing Internet at a growing rate. Northern European countries such as Denmark, the Netherlands, Iceland, Switzerland, and Finland continued their advance with high broadband penetration rates (Table 7). ,QWHUQHWSHQHWUDWLRQLPSRVHVVLJQL¿FDQWGLYHUVLRQV DFURVV (XURSHDQ FRXQWULHV EHLQJ LQÀXHQFHG E\ several characteristics such as the age, the educational level, the level of the income, and the type of the family (Dickenson & Ellison, 1999). It is well NQRZQWKDWGLIIHUHQWQHHGVDUHVDWLV¿HGLQGLIIHUHQW ages. For example, young people are interested in Internet for playing games, listening to the radio, or enjoying educational services. On the contrary, an adult has other interests like e-mail accessing, health and tourism information, electronic banking, and purchase of goods. Moreover, the educational level of the subscriber is an important factor, since higher-educated people are more familiar and interested in new technologies (Lu, 2003) and are more willing to use Internet for their needs. In addition, there is a strong relationship between income and Internet use (Madden & Coble-Neal, 2005). Households with a higher income have a
Diffusion and Price of Broadband in Europe
7DEOH%URDGEDQGVXEVFULEHUVSHULQKDELWDQWV9 2001
2002
2003
2004
2005
Austria
3,6
5,6
7,6
10,1
14,1
Belgium
4,4
8,7
11,7
15,5
18,3
Czech Republic
0,1
0,2
0,5
2,5
6,4
Denmark
4,4
8,2
13,0
19,0
25,0
Finland
1,3
5,5
9,5
14,9
22,5
France
1,0
2,8
5,9
10,5
15,2
Germany
2,3
4,1
5,6
8,4
13,0
Greece
0
0
0,1
0,4
1,4
Hungary
0,3
0,6
2,0
3,6
6,3
Iceland
3,7
8,4
14,3
18,2
26,7
Ireland
0
0,3
0,8
3,3
6,7
Italy
0,7
1,7
4,1
8,1
11,9
Luxembourg
0,3
1,5
3,5
9,8
14,9
Netherlands
3,8
7,0
11,8
19,0
25,3
Norway
1,9
4,2
8,0
14,8
21,9
Poland
0,1
0,3
0,8
2,1
2,4
Portugal
1,0
2,5
4,8
8,2
11,5
Slovak Republic
0
0
0,3
1,0
2,5
Spain
1,2
3,0
5,4
8,1
11,7
Sweden
5,4
8,1
10,7
14,5
20,3
Switzerland
2,0
5,6
10,1
17,5
23,1
United Kingdom
0,6
2,3
5,4
10,5
15,9
stronger effect on technology adoption, as they are expected to have already used some kind of such technology in the past, obtaining the necessary familiarity. It is obvious that in developing countries possible customers have higher income, which means greater payment ability, and they can afford to pay higher prices for technology equipment and services. A last characteristic of WKHVXEVFULEHUV¶SUR¿OHLVWKHW\SHRIIDPLO\7KH demand in Internet usage seems to be higher in families with children, instead of families with no children, or one person households. This is because all members of a family are accessing Internet for different reasons and preferences. Northern European countries such as Denmark, the Netherlands, Iceland, Switzerland, and Finland have continued their advance with high broadband penetration rates (Table 8).
Globalization of the Economy In the decade of acceleration, globalization brings chained reactions to legislation, economies, and social movements, which are formed at an international level. As the competition is growing, company merging becomes the solution for the dominant market share acquisition. This led to a continuing reduction of service prices and there is a trend in all companies to charge the same prices for the same services. The global economic slowdown has created a negative balance for the telecom companies and the telecom market is quite IUDJLOH$VDUHVXOWDXQL¿HGUHJXODWRU\IUDPHZRUN for telecom companies is derived by the European Commission in order to protect consumers and to promote a healthy competition among them.
743
Diffusion and Price of Broadband in Europe
Governments’ Policy and Instructions Governments are trying to promote broadband communications since the information being traded through Internet includes health, tourism, entertainment, and commerce-oriented subjects (Kalhagen & Olsen, 2002). In general, the promotion of broadband services has a positive effect to economic activity and speeds up the economic development (Kim & Galliers, 2004) while such policies will improve the quality of life and modernize the less developed areas by offering these services. Consequently, as the governments recognizes the need for special attention, treatment, and confrontation in rural areas, they proceed to adopt active measures by creating network infrastructures, which leads to high speed Internet access and so all citizens can participate in an information society.
TELECOMMUNICATION SERVICES: ADSL HIGH-SPEED INTERNET CONNECTIONS %URDGEDQGVHUYLFHVDQGDSSOLFDWLRQVDUHFODVVL¿HG according to the offered data rate. The domination of ADSL technology for broadband access across Europe during the last years demonstrated the high-speed Internet access and IP-telephone as the most common broadband services. However,
Figure 8. Typical ADSL service basket
744
a wide range of new services is now present in the mass market with an increased penetration of new WHFKQRORJLHV HJ )77[ $'6/ :L0$; (Ecosys, 2004). Table 9 summarizes typical broadband services that are expected to dominate in the near future, together with the corresponding demand in bandwidth. In order to specify the basic basket of broadband services an extended survey of providers across European countries took place, focused on the services offered, the pricing policy as well as the development in broadband market. As a result, the typical ADSL service basket was determined as a combination of main and additional services according to Figure 8. Main services are distinguished into “horizontal” and “vertical” according to the QXPEHURI¿[HGYDULDEOHVDPRQJVXSSRUWHGGDWD rate (DR), maximum consumed data volume (V) and maximum allowed minutes on line (T). Additional services include a number of e-mail DGGUHVVHV:HEVSDFHIRU:HEKRVWLQJDQGRU¿OH storage, and optional free local and nationwide phone calls and static IP addresses. The choice of the appropriate combination for each operator LVGHSHQGLQJRQWKHVSHFL¿FEXVLQHVVSODQDVZHOO as the techno-economic model parameters and assumptions. The evolution of broadband technology offers new and challenging options. The EU CommisVLRQ¶V³%URDGEDQGIRUDOO´SROLF\LVH[SHFWHGWR grow the interest for broadband in the next years
Diffusion and Price of Broadband in Europe
and to enforce the infrastructure competition among providers. As a result of this competitive environment, the provision of enhanced broadband services with reduced tariffs is expected to LQFUHDVH VLJQL¿FDQWO\ WKH QXPEHU RI EURDGEDQG subscribers.
THEORETICAL MODEL Econometric Methods Econometric methods have been used for price index calculation for a long period of time: cars (Griliches, 1961), refrigerators (Triplett & Mc Donald, 1977), and computers (Cole, 1986) are some examples. Furthermore, indices about information technology can be found in Cartwright and Smith (1988) and Moreau (1991). In addition, statisticians use econometric methods in the U.S., but the root of the hedonic approach, which is a part of economic research, goes back to Waugh (1928), Court (1939), and Stone (1954,1956). There are two types of econometric methods, hedonic methods and matched model methods, each of which have both advantages and disadvantages. One choice is to apply the “hedonic methods,” such as two-period method, single-period method, twoperiod method with an indicator for new models or single-regression method. Such indices are commonly used for products which undergo rapid technological changes.
Hedonic methods refer to regression models in which product prices are related to product characteristics and the observed price of a product (service) is considered to be a function of its characteristics. Generally, hedonic methods are based on the idea that a service (product) is a bundle of characteristics and that consumers just buy bundles of product characteristics instead of the product itself. These methods can be used to construct a quality-adjusted price index of a service. Berndt (1991) and Triplett (2000) described an overview on hedonic price equations. Rosen (1974) states that from a large amount of product varieties, a consumer chooses ZLWKRXWLQÀXHQFLQJSULFHV7KHUHIRUHFRQVXPHUV PD[LPL]HXWLOLW\DQGSURGXFHUVPD[LPL]HSUR¿WV In hedonic studies, it is possible to adjust the price of a service for its quality not quantity. All of them DUHEDVHGRQVRPHHVWLPDWHGFRHI¿FLHQWVWKDWDUH LQÀLFWHGRQWKHFKDUDFWHULVWLFVRIWKHSURGXFWVLQ both periods: m and m+1. Someone can estimate WKHFRHI¿FLHQWVIRUHYHU\\HDUVHSDUDWHO\RUFDQ have observations of two or all years together and HVWLPDWHDFRPPRQVHWRIFRHI¿FLHQWV7KHDGYDQtage of this method is that calculations are easy and fast. Indeed, hedonic methods are very fast to apply but the disadvantage is that index price can change even if no new products are existed or all prices remain the same. Another choice is to apply a matched model method such as chained Laspeyres (LCPI, LPI) or chained Paashe (PCPI, PPI) or chained Fisher or chained Tornqvist or chained geometric-mean
Table 9. Required bandwidth for typical broadband services Service
Bandwidth (Mbps)
+LJKGH¿QLWLRQ79 +'79
16 – 20
Telemedicine
6
Video on demand (VoD)
6 – 18
Internet access
1,5
Video conferencing
1,5
Telecommuting
3
Multiple digital TV
6 – 24
745
Diffusion and Price of Broadband in Europe
(Okamoto & Sato, 2001). A classic LPI cannot deal with such complexity due to rapid technological changes or the introduction of new products (services). With LPI, an index shows how much the product would cost in period m+1 relatively to what it cost in period m. Other price indices function in the same way with slight differences. The hedonic price indices are commonly used as approximations to the true cost-of-living indices (COLI), which indicate how much money a consumer would need in period m+1 relatively to the amount of money he needed in period m so as to keep the same level of utility in period t as in period t0 (Jonker, 2001). The solution is to GHWHUPLQHFRQVXPHU¶VSUR¿OHVRDVWRUHDFWWRD varied and fast-changing supply of products. But KRZFDQWKLVSUR¿OHEHGHWHUPLQHGZKHQHYHU\RQH has different needs and requirements? No matter ZKDW SUR¿OH LV GHFLGHG LW ZLOO EH D K\SRWKHVLV DQGDQDVVXPSWLRQWKDWZLOOUHVSRQGDWDVSHFL¿F model. In addition to the above, someone can see WKDWFRQVXPHU¶VGHVLUHLVQRWVWDEOHDQGWKLVLVQRW unreasonable because there is a great offer as the “goods” of technology become more and more attractive. However, according to this approach the price index is constructed only using the prices of products, which are available in two adjacent periods.
Matched Model Method (Laspeyres Method) According to Laspeyres, in order to create a price index at a time, someone observes the number of units sold in a period m (for example a month) and the average unit price in the period m and m+1. These data are used in the following formula:
Hedonic Method The term “hedonic methods” refers to a “hedonic function” f(X), which is used in economic measurement, Pi = f(Xi)
(4)
where Pi is the price of a variety (or a model), i of a product, and Xi is a vector of characteristics associated with the variety. The hedonic function is then used, for different characteristics among varieties of the product, in calculating the price index. As soon as it is determined which characteristics have to be considered, then the equations (13) and (14) are estimated for N telecommunication products in period m and m+1: ln(pim) = b0+ b1 X1i + b2 X2i + uim , i = 1,…,N (5) ln(pim+1) = b0+ b1 X1i + b2 X2i + b3+ uim+1
(6)
where bi DUH VRPH FRHI¿FLHQWV WKDW KDYH WR EH estimated.
True and Candidate Model by Using the Hedonic Function
n
¦p
ferred to a previous period among units sold in the same period. The problem with such a price index is that, as it is known, a basket of products does not remain the same over time. Furthermore, it is noticeable that some products disappear from the market (especially in the telecom market) and some others DUH PRGL¿HG VR VRPHRQH KDV WR LQWURGXFH QHZ products in order to keep the basket indicative of FXVWRPHUV¶SUHIHUHQFHV,IWKHTXDOLW\FKDQJHVDUH ignored, the resulting price index will be biased.
q
im 1 im
I m 1/ m
i 1 n
(3)
¦ pim qim i 1
Price indices are measured, as it is mentioned above, by the matched model method of Laspeyres with chaining average unit prices, which are re-
746
In this approach, there is a set of consumers who have preferences over some characteristics of a service. The construction of a price index is complicated by product-pricing limits such as different FKDUJHV IRU YDULRXV FKDUDFWHULVWLFV VR WKH ¿UVW WKLQJVRPHRQHKDVWRGRLVWRGH¿QHDEDVNHWRI services (products). The model which describes the
Diffusion and Price of Broadband in Europe
attributes (characteristics) of a product and their prices is given by the function:
So, the suitable model from a variety of candidate models via the AICc by using ( fˆ , bˆ, ˆ 2 ).
Pi = f(Xb)+u
Assumptions for AICc Criterion
(7)
where X =(x1,x2,……xn) is an nup matrix of random regressors values, xi and b are pu1 vectors and f(Xb) is an nu1 vector with i-th component f(xiǯE), (i=1,2,3,….n), u for given X=x is distributed as 1 ı2 I nun DQG ı LV DQ XQNQRZQ VFDODU )XUthermore, we assume that f is an unknown function, which means that its shape is unknown and b estimators have a unit norm. In a similar way to the true model, we construct the candidate models E\WKHGH¿QLWLRQRIFKDUDFWHULVWLFVWKDW³SOD\´D VLJQL¿FDQW UROH WR WKH FRQVWUXFWLRQ RI WKH SULFH index. In order to estimate the distance between the true and candidate models, is necessary the single-index model to be described.
Single-index models can be estimated by using iterative or direct methods. Whatever method will be followed, the result will be the same. In the iterative method nonparametric regression is applied in order to be calculated the mean regressor b. By using the iterative method the computation LV GLI¿FXOW EHFDXVH LW LV UHTXLUHG DQ HVWLPDWH RI nonparametric mean regression at each data point in order to compute a representative function f. In contrast, by using the direct method the computation is easier because the b relative weights are estimated by sliced inverse regression (SIR) (Li, 1991) and it is not necessary the estimation of function f. Once the relative weights bi are estimated, an index z XbˆSIR is constructed and the nonlinear link price fˆ ( z ) is estimated by applying local polynomial regression (LPR) (Simonoff, 1996). ˆ ˆ are estimated by iterative or As soon as fandb direct method we compute:
{8 fˆ ( Xbˆ)}c {Y fˆ ( Xbˆ)} n
2 d(( fˆ , bˆ, ˆ ) = E0{-2logf(Y)}
(9)
where f(Y) shows the possibility for the candidate model and E0 shows expectation under the true model. So, AICc
log ˆ 2
1 tr ( Hˆ p Hˆ np Hˆ p Hˆ np ) / n 1 {tr ( Hˆ Hˆ Hˆ Hˆ ) 2} / n p
np
p
np
(10) where Hˆ p Vˆ (Vˆ cVˆ ) Vˆ c, Vˆ is obtained by replacˆ ˆ , and ing b* and f .inV with estimators bandf ˆ ˆ H np isH np evaluated at Xb Xb . %HFDXVH RI WKH GLI¿FXOW FRPSXWDWLRQ VLQFH there is an unknown function, without having great inclination, the following assumptions facilitate the computation of the AICc value: 1
Single-Index Model
ˆ2
,QRUGHUWR¿QGRXWZKLFKLVWKHEHVWPRGHOIURPD variety of candidate models that describe a product with a set of characteristics, the following equations are used:
(8)
Assumption 1: f ( Xb*) # H npY Assumption 2: Similarly, E0 { f ( Xb*)} # f 0 ( Xb*) Assumption 3: In addition to the above, fˆ ( Xbˆ) f ( Xb*) # V (bˆ b*) # H p {Y fˆ ( Xb*)} where H p V (V cV ) 1V c,V wf ( Xb) / wb |b b* f .( Xb*) X A problem which is posed is the selection of the best model, so in order to estimate prices, a sliced inverse regression (SIR) is performed, without knowing the shape of the function (Li, 1991). Then a local polynomial regression (LPR) is applied and a possible shape of the hedonic function is extracted. Among several candidate hedonic models the best one is derived by applying the Akaike information criterion (AIC) (Naik & Tsai, 2001).
747
Diffusion and Price of Broadband in Europe
bSIR
EMPIRICAL EXAMPLE Models with Nonlinear Functions /HW¶VKDYHDQH[DPSOHLQZKLFKWKHFKDUDFWHULVWLFV of the product are determined, so as to show the quality of this product. As soon as the characteristics of a product are determined, by applying the hedonic method, someone estimates its price. Otherwise the price of a product can be estimated “manually” by comparing the new product with the most similar old one. The main assumption is that ADSL connections have four main physical characteristics: supported DR (up and down), maximum V, and maximum T. These four characteristics are widely used from telecom operators for valuating and selling leased lines across Europe. Regarding the ADSL connections data, which are collected from year 2003 to 2005 for both residential and business connections, it is clear that there is a large variation of values in the variables and even more the price level is also varying from country to country, for the same service considered. First of all, the hedonic model is estimated, described by (7). The data are sorted by Pi and then are divided into three slices. Without specifying the unknown link function we derive:
(0.581899, 0.78326, 0.00011, 0.21886)
The above results imply that the price is strongly related only with the downlink DR (Triplett, 2004). As it can be seen after inspecting the diagram (Figure 12), the described relationship can be easily expressed as a linear model. By having four characteristics, we take under consideration 24-1 nested candidate models. For each of the nested models SIR estimates are obtained (Figure 9) and then byapplying the LPR (Figure 10) the link function f k (k 1, 2,3,..15). Figures 9 and 10 also show that some individual WDULIIVDUHH[LVWLQJZKLFKGHFOLQHVLJQL¿FDQWO\IURP the main cluster. In order to examine the relationship between the price of an ADSL connection and their main characteristics, such as the supported data rate and the maximum consumed data, several candidate link functions are applied. Across the candidate models in several shapes of link function, the one which gives the smallest AICc (0,2125746 vs. 7,846617) value from (10), is described by: ln(P) = b0 +b1•DR(down)+b2 •DR(up)+b3 •(V)+b4 •(T)
(11) Even link functions such as hyperbolic sine or hyperbolic cosine give almost the same results
Figure 9. Plot of Pi against SIR directions 20000
0 0 -20000
-40000
-60000
-80000
-100000
748
20
40
60
80
100
120
140
160
180
200
Diffusion and Price of Broadband in Europe
Figure 10. Local polynomial regression with kernel smoothing 20000
0
-20000
C2
-40000
-60000
-80000
-100000
1
2
3
4
5
C3
Figure 11. Residuals 100 80
Residuals of linear model on price
60
Residuals of linear model on ln-ned price
40 20 0 -20
1
2
3
4
5
-40 -60
with the nonlinear functions. On the contrary, the AIC value of the linear model, after considering the natural logarithm of the dependant variable, the price, turned out to be the smallest calculated. This gave an advantage to the linear model for selecting it to construct the price index. Working in the logarithmic scale using a linear PRGHOUHVXOWVVKRZVDEHWWHU¿WWKDQDOOWKHRWKHU models because the residuals from the log-linear
model are all around zero (Figure 11) (Triplett, 2004) and have less standard error (0,5845 vs.26,58).
Constructing the Price Index Finally, when there are N telecommunication products in period m and m+1, the proposed hedonic
749
Diffusion and Price of Broadband in Europe
price index can be calculated by the following equation:
I m 1/ m
f m 1 (bi * X i ) f m (bi * X i )
(12)
Using data such as those presented in Table 10 for the case of Europe, it can be observed in Figure 12 that it is not easy to compare prices for different data such as data rate, consumed volume, allowed time on line, and so forth, but there are similarities and patterns that must be evaluated. It is obvious that the more a consumer demands for a product with upper services, the more the prices of this product are increased, even if there is a remarkable trend to be kept prices in low levels. Because of WKLVFRQVXPHUV¶³EHKDYLRU´WKHUHDUHQRLPSOLFLW prices for all characteristics. 7KLVEHKDYLRU¿WVWRWKHKHGRQLFDSSURDFKDQG it can be observed by calculating the hedonic price index from (12). In Figure 13, this index and its evolution are presented for the case of European countries broadband connections market, even if WKHUHLVQ¶WKLVWRULFDOGDWD
ASPECTS AND CONCLUSIONS As observed in the preceding sections, diffusion methodology was quite capable of describing the diffusion process of ADSL technology over the studied cases. According to the results, there is obviously an increasing rate in the adoption of ADSL services, as was probably expected. As indicated, the used actual data refer to the penetra-
tion of ADSL technology over 100 inhabitants that access this kind of broadband networks. If these numbers were expressed in terms of households, considering an average number of three people in a household, it is concluded that the penetration has already reached a substantially high level and it is expected to grow even more. In the context of the newcomer countries in the European Union, there is an increased interest in considering the application of such methodologies in order to estimate the expected process of adoption of broadband technologies. In this case, the application of the corresponding cross-national methodology is also expected to provide interesting and useful results, to be used as indicative drivers to the performance of the pertinent technoeconomic DQDO\VLV0RUHVSHFL¿FDOO\LQWKHFRQWH[WRI$'6/ networks there is always the case of the need of a precise estimation of the network that should be developed in order to support the increasing future demand of the offered services. This example reveals that the accuracy of the forecasting results is an extremely critical factor, mainly because of the strong relationship with future investments that will be required. An even more detailed methodology, which is one of the major objectives in the corresponding area of research, should include the economical and social indicators in the areas of interest. Although the methodology is quite robust, demand can be expected to rise even more, depending on the levels of some important economical and social indicators in the markets of reference. The most representative of these indicators include the GDP per capita, the industrialization level of the country, the unemploy-
Table 10. Data for ADSL connections Data rate(Down)
Data Rate(Up)
Volume
(Kbps)
(Kbps)
(Mbps)
59,95
10000
512
15.360
14,95
256
64
307,2
Subscription Price
750
21,95
1536
256
1.536
29,95
800
256
-
74,95
12000
1024
8.192
Diffusion and Price of Broadband in Europe
Figure 12. Evolution data rate (down) in Europe 4500 4000 3500 3000 Down
2500
Up 2000
Subscription prices Linear (Down)
1500 1000 500 0 2003
2004
2005
Years
Figure 13. Hedonic price index evolution for the case of ADSL market in Europe 1,2 1
y = 0,5339x + 0,0221 R2 = 1
indexes
0,8 0,6 0,4 0,2 0 2004/2003
2005/2004 years
PHQWOHYHOWKHSRWHQWLDOXVHU¶VSHUFHLYHGULVNRI FKRLFHWKHWLPHRIWKHSURGXFW¶VLQWURGXFWLRQWKH education level of population, and the competition OHYHORIWHOHFRPPXQLFDWLRQV¶PDUNHW7KHVHIDFWRUVDPRQJRWKHUVDUHPDMRUGULYHUVIRUGH¿QLQJ the demand function of the product. In addition, a study of factors that affect the broadband expansion and development, the domiQDQWFKDUDFWHULVWLFVWKDWLQÀXHQFHWKHWDULIISROLF\ RI$'6/VHUYLFHVDQG¿QDOO\WKHFRQVWUXFWLRQRI price index for ADSL connections is presented. It has been shown that the general trend is the decrease
of subscription prices, as expected. It is evidenced that customers are taking advantage from operaWRUV¶FRPSHWLWLRQ7KLVLVH[WUDFWHGIURPWKHIDFW that they are interested in the offered downstream data rates, since these parameters designate the quality of the subscribed services. In addition, this analysis is to foresee the trend of prices of telecommunication services (products) and especially ADSL connections over time across all countries of Europe, applying a hedonic method IRU VRPH GH¿QHG FKDUDFWHULVWLFV 7KLV PHWKRG works better when there is a variety of important
751
Diffusion and Price of Broadband in Europe
characteristics but less satisfactorily when these change rapidly over time. ADSL connections have LPSRUWDQWDQGVSHFL¿FFKDUDFWHULVWLFVLQGHHGDQG their prices vary slowly over time. The results give a view of telecommunication prices over time and VKRZKRZWKHSULFHVZLOOÀXFWXDWHWKHQH[W\HDU The application of these econometric methods, IROORZLQJWKHGH¿QLWLRQRISURGXFWV¶FKDUDFWHULVWLFV provides a reliable and accurate method able to produce an exact estimate of prices both for new products and over next years. The validity of the model and the appropriate selection of the functional form that has been chosen to relate price to characteristics can be validated over the next years and more observations. However, as the factors affecting broadband tariffs can be numerous, a more detailed study that takes into account additional parameters is needed.
REFERENCES Baltas, G., & Freeman, J. (2001). Hedonic price methods and the structure of high-technology industrial markets, an empirical analysis. Industrial Marketing Management, 30, 599-607. Bass, F.M. (1969). A new product growth model for consumer durables. Management Science, (5), 215-227. Berndt, E.R. (1991). The practice of econometrics: Classic and contemporary. Cambridge, MA: Addison-Wesley Publishing. %HZOH\5 )LHELJ'* $ÀH[LEOHORJLVWLF growth-model with applications in telecommunications. International Journal of Forecasting, 4(2), 177-192. Bonnetain, P. (2003). A hedonic price model for islands. -RXUQDORI8UEDQ(FRQ 368-377. Montagne, R. (2005, April 21). Broadband access roadmap based on market assessment and technical-economic analysis. EU FP6 IST project BROADWAN. Deliverable D15.
752
Budde, P. (2004). Eastern European infrastructure, ¿[HGYRLFHDQGGDWDPDUNHWUHSRUWCommunication Pty Ltd. Buiter, (2004). To purgatory and beyond. In Conference on challenges for central banks in an enlarged EMU. Vienna, Austria. Cartwright, D.W., & Smith, S.D. (1988, November). 'HÀDWRUV IRU SXUFKDVHV RI FRPSXWHUV LQ *13 Revised and extended estimates 1983-88. Survey of Current Business. Central and Eastern European Countries Synthesis of Master Reports. (2000). Regulatory Developments. European Survey of Information Society Projects and Actions (ESIS). Available from: http:// www.eu-esis.org/esis2reg/synthCEEC7.htm Cole, R. (1986). Quality-adjusted price indexes for computer processors and selected peripheral equipment. Survey of Current Business. Bureau of Economic Analysis, U.S. Department of Commerce. Communications Committee of European Commission. (2001a). Final 2001 report on performance in the supply of leased lines pursuant to Directive 92/44/EC (Working Document). Communication from the Commission to the Council, the European Parliament. (2001b). The economic and social committee and the committee of the regions. In Seventh Report on the Implementation of the Telecommunications Regulatory Package. Commission of the European Communities. Communication from the Commission to the Council, the European Parliament. (2001c). Status RIOLFHQVLQJDQGIHHVIRU¿[HGQHWZRUNVDQGVHUvices. In Seventh Report on the Implementation of the Telecommunications Regulatory Package. Commission of the European Communities. Communication from the Commission to the Council, the European Parliament (2002). The economic and social committee and the committee of the regions. In Eighth Report on the Implementation of the Telecommunications Regulatory Package. Commission of the European Communities.
Diffusion and Price of Broadband in Europe
Court, A.T. (1939). Hedonic price indexes with automotive examples. In The dynamics of automobile demand (pp. 99-117). New York: The General Motors Corporation. Dewenter, R., Haucap, J., Luther, R., & Rotzel, P. (2004). Hedonic prices in the German market for mobile phones (Discussion Paper No.29). Hamburg, Germany: University of the Federal Armed Forces. Dickenson, P., & Ellison, J. (2001). Plugging in: The increase of household Internet use continuous into 1999. Connectedness Series, Statistics Canada. Diewert, (2003). Services and the new economy: Data needs and challenges. Vancouver, Canada: University of British Columbia. Ecosys, (2004). Overview of demand forecasts IRUWKH¿[HGDQGPRELOHQHWZRUNVDQGVHUYLFHVLQ Europe. Deliverable 2. Fildes, R., & Kumar, V. (2002). Telecommunications demand forecasting - a review. International Journal of Forecasting, 18(4), 489-522. Fisher, J.C., & Pry, R.H. (1971). Simple substitution model of technological change. Technological Forecasting and Social Change, 3(1), 75-88. Fixler, D., Greenless, J., & Lane, W. (2001). Telecommunications indexes in the U.S. Consumer price index. In Sixth Meeting of the International Working Group on Price Indices. Canberrra, Australia.
Gunther, J. (2001). Regulations of telecommunications in Europe. Krems, Austria: Danubue University. Haan, M. (2001). The economics of free Internet access. Journal of Institutional and Theoretical Economics, (3), 359-379. Harald, E. (2005). Do hedonic price indexes change KLVWRU\"7KHFDVHRIHOHFWUL¿FDWLRQ Working paper VHULHVLQHFRQRPLFVDQG¿QDQFH6WRFNKROP School of Economics. Harhoff, D., & Moch, D. (1997). Price indexes for PC database software and the value of code compatibility. Research Policy, 26, 509-520. Jonker, N. (2001). Constructing quality adjusted price indexes: A comparison of hedonic and discrete choice methods. WO Research Memoranda (discontinued) 673. The Netherlands: Netherlands Central Bank, Research Department. Kalhagen, K.O., & Olsen, B.T. (2002). Provision of broadband services in non-competitive areas in Western Europe countries. In Proc. ISSLS 2002, Seoul, Korea. Kam, Y.U. (2001). An elementary price index for Internet service providers in Canada: A hedonic study. In Sixth Meeting of International Working Group on Price Indices. Canberrra, Australia. Kim, C., & Galliers, R.D. (2004). Toward a diffusion model for Internet systems. Internet Research Journal, 14(2), 155-166.
Gatignon, H., Eliashberg, J., & Robertson, T.S. (1989). Modeling multinational diffusion patWHUQV²DQHI¿FLHQWPHWKRGRORJ\Marketing Science, 8(3), 231-247.
Kumar, V., Ganesh, J., & Echambadi, R. (1998). Cross-national diffusion research: What do we know and how certain are we? Journal of Product ,QQRYDWLRQ0DQDJHPHQW(3), 255-268.
Goff, D.H. (2002). An assessment of the broadband media strategies of Western European telecoms. In Proceedings of the Fifth World Media Economics Conference. Turku, Finland.
Kumar, V., & Trichy, K. (2002). Research note multinational diffusion models: An alternative framework. Marketing Science, 21(3), 330.
Griliches, Z. (1961). Hedonic price indexes for automobiles: An econometric analysis of quality change. In The price of the federal government (General Series No. 73) (pp. 137-196). New York: Columbia University Press for NBER.
Lee, H.S., Park, K., & Kim, S.Y. (2003). Estimation of information value on the Internet: Application of hedonic price model. Electronic Commerce Research and Applications, 2, 73-80.
753
Diffusion and Price of Broadband in Europe
Li, K.C. (1991). Sliced inverse regression for dimension reduction (with discussion). J. Am. Statist. Assoc., 86, 316-342.
Rai, L.P. (1999). Appropriate models for technology substitution. -RXUQDORI6FLHQWL¿F ,QGXVWULDO 5HVHDUFK(1), 14-18.
Lu, J. et al. (2003). Technology acceptance for wireless Internet. Internet Research Journal, 13(3), 206-222.
Rosen, S. (1974). Hedonic prices and implicit markets: Product differentiation in pure competition. Journal of Political Economy, 92, 34-55.
Madden, G., & Coble-Neal, G. (2005). Australian residential telecommunications consumption and substitution patterns. Review of Industrial Organization, 26, 325-347.
Silver, M. (2000). Hedonic regressions: An application to VCRs using scanner data. Omega, 28, 399-408.
Michalakelis, C., Dede, G., Varoutas, D., & Sphicopoulos, T. (2005). Impact of cross-national diffusion process in telecommunications demand forecasting. In 3URFHHGLQJVRI1$(&Garda, Italy. Monath, T., Elnegaard, N.K., Cadro, P.H., Katsianis, D., & Varoutas, D. (2003). Economics of ¿[HGEURDGEDQGDFFHVVQHWZRUNVWUDWHJLHVIEEE Commun. Mag., 41, 132-139. Montella, M., Mostacci, F., & Zanolini, G. (2001). Consumer price indexes for telecommunication services in Italy: Work in progress. In Sixth Meeting of the International Working Group on Price Indices. Canberrra, Australia. Moreau, A. (1991). A price index for microcomputers in France. Document de Travail de la Direction des Statistiques Economiques, 9109. Naik, P.A., & Tsai, C.L. (2001). Single-index model. Biometrica Trust, 88(3), 821-832. Nerlove, M. (1995). Hedonic price functions and the measurement of preferences: The case of Swedish wine consumers. Eur. Econ. Rev., 39, 1697-1716. Nicholson, J.L. (1967). The measurement of quality changes. The Economic Journal, 77, 512-530. Okamoto, M., & Sato, T. (2001). Comparison of hedonic method and matched models method using scanner data: The case of PCs, TVs, and digital cameras. In Sixth Meeting of International Working Group on Price Indices. Canberra, Australia.
754
Simonoff, J.S. (1996). Smoothing methods in statistic. New York: Springer. Simonoff, J.S., & Tsai, C.L. (1999). Semiparametric and additive model selection using an improved Akaike information criterion. J. Comp. Graph. Statist. 8, 22-40. Stiroh (2001). The economic impact of information technology. Federal Reserve Bank of New York, Academic Press. Stone, R. 5RZH'$&RUOHWW:-+XUVW¿HOG5 & Potter, M. (1954). The measurement of consumer behavior and expenditure in the United Kingdom, 1920-1938. The American Economic Review, 44(3), 426-433. Stone, R. (1956). Quantity and price indexes in national accounts. Organization for European Economic Cooperation. Paris: OEEC. Takada, H., & Jain, D. (1991). Cross-national analysis of diffusion of consumer durable goods LQ3DFL¿F5LPFRXQWULHVJournal of Marketing, (2), 48-54. Taneda, M.A., Takada, J., & Araki, K. (2001). The problem of the fading model selection. IEICE Transactions on Communications, E84b(3), 660-666. Triplett, J.E. (2004). Handbook on hedonic indexes and quality adjustments in price indexes: Special application to information technology products. OECD Science, Technology and Industry Working Papers 2004/9. OECD Directorate for Science, Technology, and Industry.
Diffusion and Price of Broadband in Europe
Triplett, J.E. (2000). Draft copy handbook on quality adjustment of price indexes for information and communication technology products. Paris: OECD.
ENDNOTES 1
2
Triplett, J.E., & McDonald, R.J. (1977). Assessing the quality error in output measures: The case of refrigerators. Review of Income and Wealth, 137-176. Venkatesan, R., & Kumar, V. (2002). A genetic algorithms approach to growth face forecasting of wireless subscribers. International Journal of Forecasting, 18, 625-646.
3 4 5 6
7 8 9
Waugh, F.V 4XDOLW\ IDFWRUV LQÀXHQFLQJ vegetable prices. Journal of Farm Economics, 10(2), 185-196.
Source for actual data: Eurostat (available online at: http://ec.europa.eu/eurostat/). Source for actual data: OECD (available online at: http://www.oecd.org/). Source for actual data: Eurostat Source for actual data: Eurostat. Source: OECD Source: Yankee Group Europe (1997); Regulatory Developments (2000). Source: OECD Source : Internet World Stats Source: OECD broadband statistics to June 2006.
755
756
Chapter XLVI
Explaining Patterns of Broadband Development in OECD Countries Inmaculada Cava Ferreruela Regional Government of Valencia, Spain
ABSTRACT The aim of this chapter is to provide some insights about the explaining patterns of broadband deployment and adoption. This problem is addressed by examining them in the light of the results of an exhaustive cross-national empirical analysis that uses a comprehensive panel data set from the 30 OECD countries with more than 40 features. The results suggest that technological competition and the low cost of deploying infrastructures on one side, and the predisposition to use new technologies as well as some social indicators on the other, appear to be the key drivers for broadband deployment and adoption, respectively.
INTRODUCTION In recent years, it has generally been accepted that the development of broadband as a means to promote new interactive and advanced applications is supposed to be the basis of what is called the knowledge-based economies and societies. NevHUWKHOHVVUHFHQWVWXGLHVFRQ¿UPWKDWEURDGEDQG SHQHWUDWLRQUDWHVDUHRQO\VLJQL¿FDQWLQGHYHORSHG countries and, what is more, that broadband adoption lags behind in rural and remote areas worldwide, even in developed countries (OECD, 2003c). As a consequence, the economic and social gap EHWZHHQWKRVHZKRDUHDEOHWRHQMR\WKHEHQH¿WV of these new technologies and those who do not is widening throughout the world.
This fact has led most governments to consider broadband availability and affordability as an objective of social cohesion, and most of them released action plans in the 1990s. The analysis of these policies shows that there is a wide range of alternatives, mainly due to the novelty of policies of this type and heterogeneous types of factors that should be considered (economic, social, and technical factors) (Cava & Alabau, 2003; Cava & Alabau, 2006; Falch, 2004). This situation leads WRGLI¿FXOWFKRLFHVIRUSROLF\PDNHUVWKDWXVXDOO\ lack practical experience on this issue. In this context, the aim of this chapter is to provide some useful insights about the main constraints and motivators for broadband development and thus help to identify which policies could be
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Patterns of Broadband Development in OECD Countries
most effective to promote broadband deployment and adoption. To this end, this chapter starts from the historical experience and uses a set of cross-national empirical data from the 30 Organization for Economic Cooperation and Development (OECD) countries from the years 2000 to 2002. The empirical study extends the work done in Cava and Alabau (2006). In particular, it follows a methodological approach to alleviate some of the weaknesses found, and introduces the use a nonlinear regression technique, multivariate adaptive regression splines (MARS). This technique allows accurate nonlinear and robust estimates and fully interpretable results. This chapter presents the main components of this research study, which are structured as follows. In the second section, existing literature is reviewed and some weaknesses of previous works DUH LGHQWL¿HG 7KHQ WKH WKLUG VHFWLRQ GHVFULEHV the characteristics of the panel data and, in the forth section, the methodology followed and the multivariate adaptive regression splines regresVLRQWHFKQLTXH 0$56 DUHSUHVHQWHG7KH¿IWK section examines the results obtained from the empirical analysis. This section is divided into two parts which are the analysis of supply-side factors in broadband deployment and the analysis of the MRLQWLQÀXHQFHRIVXSSO\DQGGHPDQGVLGHIDFWRUV in broadband adoption. Finally, in sixth section are the results discussion, along with some suggestions for future research.
LITERATURE REVIEW A growing number of studies about the development of broadband networks and services have become available in latest years. Two main research approaches can be distinguished, the case study method and the empirical analysis. The case study PHWKRGH[SORUHVWKHIDFWRUVWKDWFRXOGKDYHLQÀXenced broadband development from a qualitative point of view. Some research studies analyze only one success case such as South Korea to explore what are the key explaining factors (Lau, Kim, & Atkin, 2005). Other studies however, use this research approach to obtain some conclusions af-
ter the comparison of different cases. This is the methodology followed in extensive cross-national analysis such as those presented in OECD (2001) and BDRC Ltd. (2001), where the situation in GLIIHUHQWFRXQWULHVLVFRPSDUHGWR¿QGFRPPRQ patters that could explain broadband development. This is also the research method in Lee and ChanOlmsted (2004a) to identify the key explaining IDFWRUV WKDW KDYH FRQWULEXWHG WR 6RXWK .RUHD¶V competitive broadband environment, compared to the United States. 5HJDUGLQJ WKH LQÀXHQFH RI SROLWLFDO FKRLFHV some studies deal with regulatory and political implications on broadband. In a recent work (Frieden, 2005) Canada, Japan, Korea, and the United States best practices are examined with an eye toward determining the optimal mix of legislative, regulatory, and investment initiatives. In Lee and Chan-Olmsted (2004b), the impact of different policies to encourage broadband awareness, availability, and adoption in rural Scotland DUHDQDO\]HGWKURXJKLQÀXHQFHGLDJUDPV The studies that follow the empirical approach broach the broadband problem by using empirical data, modelling, and statistical techniques. Some of them use econometric analysis to estimate broadband and Internet demand considering that broadband infrastructure is already available (Rappoport, Kridel, Taylor, & Alleman, 2003). Other studies such as Gabel and Kwan (2000) and Grubesic and Murray (2002) examine the role of both supply and demand factors using a statistically valid sample of customer locations. Other studies explore the geographical differences in broadband provision and adoption considering the effects of socioeconomic and demographic differences, both at regional level (Grubesic, 2004), or at cross-national level (Bauer, Kim, & Wildman, 2003; Garcia-Murillo & Gabel, 2003). The political and regulatory implications are only indirectly analyzed in empirical studies by including the effects of the public actions as variables in the models. Thus in Gillet, Lehr, and Osorio (2004), a taxonomy of local government initiatives in the United States is offered, and some preliminary empirical results for a sample of communities with municipal electric utilities are
757
Patterns of Broadband Development in OECD Countries
provided. In Distaso, Lupi, and Maneti (2006), an econometric model is used to explore the effects of inter and intraplatform competition on the diffusion of broadband, and test the model using data from 14 European countries. The chapter follows an empirical approach to identify the key explaining patters of broadband deployment and adoption, and continues the work done in cross-national empirical studies. Nevertheless, some methodological innovations have been introduced to alleviate some of the weaknesses found in previous works. First, in some works it KDVEHHQREVHUYHGVRPHPRGHOVWKDWFOHDUO\RYHU¿W to the training data (and thus they cannot generalize or extrapolate to unseen data). This problem can be explained as broadband modeling based on cross-national data is an ill-posed problem where the rate of available sample is very low compared to the number of potentially useful variables. We have alleviated this situation by using 3 years of observation interpreted as independent observations. 6HFRQGWKHVHOHFWLRQRISRVVLEOHLQÀXHQFLQJ factors is generally based on previous and usuDOO\QRWDOZD\VZHOOMXVWL¿HGDVVXPSWLRQVZKLFK could lead to discarding some potentially relevant variables for the problem. However, in this study, rather than relying on subjective criteria, an objective feature selection process based on statistical data analysis is used to select the features to be included in the model. This, in turn, also improve the accuracy results as the model is developed using statistically relevant information only. Finally, in most of the available papers, the assumption of linear relationships between the independent and dependent variables is made. This, however, constitutes a too simplistic way of looking at the problem as it is clear (and widely accepted) that the underlying system is too complex to be linear, and thus it can not be accurately modeled using linear methods. To eliminate this drawback, a nonlinear regression technique, the so-called multivariate adaptive regression splines (MARS), is used to identify the key factors inÀXHQFLQJEURDGEDQGGHYHORSPHQW7KLVPHWKRG in addition, provide high performance and good interpretation capabilities.
758
DATA This study uses a comprehensive panel data set from the 30 OEDC countries with more than 40 features. Although most data refers to telecommunications, data for social, economic, and demographic indicators were also included. The database from this empirical study was built by gathering information mainly from databases of various international organizations. The period of observations for this study is from 2000 to 2002 (ITU, 2003a; OECD, 2002b; OECD, 2003b; World Bank, 2003). These three years of observation were pooled and interpreted as independent observations for each variable. As data from 30 OECD countries were used, this method allowed the authors to have a maximum number of 90 observations for most variables instead of 30 (for the 30 countries in one year) and WKXVUHGXFHWKHSUREOHPRIRYHU¿WWLQJLQPXOWLSOH regression models. Moreover, some variables were introduced with a one-year lag (these variables are PDUNHGZLWKWKHSUH¿[/LQWKHLUQDPHV LQRUGHU WRWDNHLQWRDFFRXQWWKHGHOD\HGLQÀXHQFH Apart from the consolidation of data, other data processing tasks were carried out in order to obtain a clean and useful database. The main tasks were the recovery of lost data by means of a linear estimation and the calculation of the value of newly composed variables, which are relevant to the analysis. Although each variable represents a VSHFL¿FFKDUDFWHULVWLFRIWKHFRXQWU\VRPHFDWHJRULHVRIYDULDEOHVZHUHGH¿QHGWRUHSUHVHQWDPRUH JHQHUDOIHDWXUH7KXVQRWRQO\FDQWKHLQÀXHQFHRI RQHYDULDEOHEHLGHQWL¿HGEXWDOVRWKHHIIHFWRIWKH HQWLUHFDWHJRU\)LUVWWKHYDULDEOHVFDQEHFODVVL¿HG as supply-side and demand-side related features. Supply-side factors characterize the availability and development of telecommunication services in a country. This category includes the following indicators by groups: (Group 1) infrastructure availability, (Group 2) infrastructure investment, (Group 3) market competition, and (Group 4) prices. Demand-side factors characterize the inclination of the society to adopt telecommunication services, including broadband. This category includes the following indicators by groups: (Group 5) tele-
Patterns of Broadband Development in OECD Countries
communication services penetration, (Group 6) Internet indicators, (Group 7) economic indicators, (Group 8) demographic indicators, (Group 9) education indicators, and (Group 10) social indicators. Regarding the variables representing broadband, WKHGH¿QLWLRQRIWKH2UJDQL]DWLRQIRU(FRQRPLF Cooperation and Development in (OECD, 2001) has been adopted, and consequently, broadband implies transmission speeds equal to or greater than 256 kbit/s for downstream connections, and equal to or greater than 64 kbit/s for upstream connections.1 In terms of infrastructure, broadbandLVGH¿QHG as the last mile connection to the end-user that is able to provide a broadband service. Finally, while there are different types of broadband technology, this chapter considers only the most popular ones, namely DSL and cable modem. The variables that represent broadband supply are those included in Group 2 which are: DSLCOV, which is the percentage of DSL enabled local loops, and CTVHPAS, which is the percentage of homes passed by a cable TV network. Although CTVHPAS does not represent the availability of broadband cable infrastructure, it indicates the availability of the network infrastructure that can be upgraded to provide broadband services (HFC networks). Finally, the variables that represent broadband penetration, (Group 11) are: DSLSB100, which is the number of DSL subscribers per 100 inhabitants, and CABSB100, which is the number of cable modem subscribers per 100 inhabitants. Only DSL and cable modem technologies were considered, as these technologies are the leading choices that are currently available to most residential users in OECD countries (OECD, 2001).
consider as many potentially valuable features as possible, without any a priori assumption on their relevance. Second, an objective data analysis is intended to yield the most relevant subset of features LQÀXHQFLQJEURDGEDQGGHYHORSPHQW7KLVH[WHQVLYH approach is based on the results of previous studies that point out that many diverse features could LQÀXHQFHEURDGEDQGDGRSWLRQ %DXHUHWDO /L ;X In this context, it is worth noting that such an approach requires an extensive data collection but also a set of fully interpretable techniques for data analysis and modeling. Actually this research study has considered a general framework for statistical data analysis and linear modeling (Hair, Anderson, Tatham, & Black, 1998). The aim of this framework is to alleviate some of the problems found in previous broadband empirical analyzes. The learning scheme is organized in three stages: statistical analysis, correlation analysis, and multivariate adaptive regression splines.
Statistical Analysis of the Data This stage is devoted to gaining knowledge about the characteristics of the data to be used in later VWDJHVRIWKHDQDO\VLV7KLVWDVNFRQVLVWVRI¿UVW getting basic statistics of the data, taking into account all observations and then dividing the total number of cases into groups according to the value of a factor variable. The objectives of this analysis are the following: to remove abnormal samples that could disturb a linear model and to identify UHODWLRQVKLSVEHWZHHQYDULDEOHVWREHFRQ¿UPHG in further steps.
Correlation Analysis METHODOLOGY FOR THE EMPIRICAL ANALYSIS This research study follows a systematic approach for data analysis and modeling aimed at obtaining the main constraints and motivators for broadband penetration. First, the entire range of possible LQÀXHQFLQJ IHDWXUHV ZDV LQLWLDOO\ WDNHQ LQWR DFcount. The underlying idea on this approach is to
This stage consists of calculating the Pearson biYDULDWHFRUUHODWLRQFRHI¿FLHQWEHWZHHQDVHOHFWHG set of factors and the variable under study. The objectives of this analysis are the following: to EXLOGDQLQLWLDOUDQNLQJRIIDFWRUVWKDWLQÀXHQFH WKHGHSHQGDQWYDULDEOHWRLGHQWLI\FURVVLQÀXHQFHV between independent features and, to help the feature selection process for multivariate adaptive regression splines analysis.
759
Patterns of Broadband Development in OECD Countries
As the correlation analysis only takes into account bivariate relationships between variables, the conclusions obtained need to be completed using another technique to determine the relative importance of each independent variable within a group. Multivariate adaptive regression splines analysis provides a way to resolve this problem.
Formulation The structural model built by MARS considers a set of expansions in piecewise linear basis functions (BFs) of the form (x - t)+ and (t - x)+ , where the subscript “+” means “positive part,” and they DUHGH¿QHGDVIROORZV
Multivariate Adaptive Regression Splines (MARS) Analysis
ha ( x t ) ( x t )
This stage consists of modeling the deployment and adoption of broadband, with the multivariate adaptive regression splines regression technique (MARS). MARS is an adaptive procedure for function approximation and regression and has some very attractive and potentially useful properties for model development and explanation (Friedman, 1991). For instance, MARS is well suited for high-dimensional problems (those with a large number of variables, features, or attributes), and also when dealing with ill-posed problems (the rate of available sample is very low compared to the number of potentially useful variables), such as the one treated here. For the nonexpert reader, we can say that MARS can be viewed as a generalization of the familiar stepwise linear regression, and constitutes one of the common techniques for UHJUHVVLRQ LQ WKH ¿HOG RI PDFKLQH OHDUQLQJ DQG data mining. The MARS analysis is the last step to identifying WKH NH\ IDFWRUV LQÀXHQFLQJ EURDGEDQG GHYHORSment, as this regression technique provides with a ranking of variable importance taking into account nonlinear relationships. However this criteria was complemented with the previous knowledge of the problem, so new models were obtained by reducLQJWKHQXPEHURIYDULDEOHV7KH¿QDOREMHFWLYHRI WKLVIHDWXUHSURFHVVZDVWR¿QGXVHIXODQGHDVLO\ interpretable models. In this section, we review the formulation of this technique and we analyze how this technique deals with outliers, missing values, and the important issues of regularization FRPSOH[LW\FRQWURO DQGRYHU¿WWLQJ
and
x!t t ® ¯ 0 otherwise (1) x t x ® ¯ 0 otherwise (2)
hb ( x t ) (t x)
An alternative way to represent these basis functions are ha(x,t) = max(x - t,0) and hb(x,t) = max(t - x,0), respectively. In Figure 1, we depicted DQLOOXVWUDWLYHH[DPSOHRIWKHFRXSOHG RUUHÀHFWHG functions (x - 0.5)+ and (0.5 - x)+ . Each function is piecewise linear, with a knot at the value t. The idea underlying MARS is to conform a combinaWLRQRIUHÀHFWHGSDLUVIRUHDFKLQSXWxj with knots at each observed value xij of that input. Therefore, the collection of basis functions is given by:
C {( x j t ) (t x j ) }j 1 2… p and t {x1 j x2 j … xNj } (3) where N indicates the number of knots to split the space of each variable xj , and p represents the number of variables. It is worth noting here that (1) there are 2Np basis functions, and (2) in spite the fact that each basis function depends only a single variable xj , it is considered a function of the whole input space Rp. The MARS model is very similar to a forward stepwise linear regression but, instead of using the original inputs, we are allowed to use functions from the set C and their products. Thus the model has the form: M
yˆ
f (X )
o ¦
m 1
760
h (X )
m m
(4)
Patterns of Broadband Development in OECD Countries
Figure 1. Example of basis functions used by MARS
0.5
Basis functions
0.4 (tx) 0.3
(x t) +
+
0.2 0.1 0 0.1
0
0.2
0.4
0.6
0.8
1
x
where each hm(X) is a function in C, or a product of two or more such functions. As a result, the model performs non-linear (piecewise linear) regression. Note that a simple multiple regression analysis constitutes a particular case in (7) when hm(X) = X. Given a choice for the hm WKHFRHI¿FLHQWVEm are estimated by minimizing the mean square error (MSE), that is, by standard linear regression:
MSE
1 n ( yi yˆ i ) 2 ¦ ni1
where yi and yˆ i are the dependent variable and its prediction, respectively. However, the important point here is the construction of the functions hm (•) . Basically, adjustment of MARS weights (Em) and thresholds ( tij ) model is carried out in a two-step procedure. In the ¿UVWVWHSDPRGHOLVgrown by adding basis functions until an overly large model is found. In the second step, basis functions are pruned in order to obtain an optimal balance of bias and variance. Note that the model can potentially trace any arbitrary shape for the dependent function, as the user can force the use of a great number of knots. This, however, FRXOGOHDGWRRYHU¿WWLQJDQGWKXVWKHFRQWURORI
MARS capacity constitutes an important issue in GHYHORSLQJWKHPRGHO7KHSUREOHPRIRYHU¿WWLQJ is controlled in the aforementioned second step, that is, the backward deletion procedure, which consists on removing the term with lower impact on the quality of the model.
Characteristics of MARS In this section, we emphasize the good characteristics of MARS for regression by analyzing in detail some important aspects of model development. In particular, we analyze the validation procedure followed in this work, the procedure to control de capacity of the model, how to deal with missing values of a given variable(s), and how to measure the relevance of the variables from a built model. Free Parameters ,Q RUGHU WR EXLOG D ¿QDO 0$56 PRGHO VHYHUDO free parameters have to be adjusted. First, the maximum number of basis functions (MNBF) to be explored. Second, the intensity of the search for knot location (I); as I increases, more possible knots are evaluated. Finally, the possibility to introduce a maximum number of interactions among
761
Patterns of Broadband Development in OECD Countries
EDVLVIXQFWLRQV 0$;,17(5 ,QDOORXWWHVWVZH ¿[HG01%) WRFRQWUROWKHFRPSOH[LW\RIWKH model, I = 2 in order to avoid computational cost SUREOHPV DQG 0$;,17(5 DV VXJJHVWHG elsewhere (Friedman, 1991; Salford Systems; Steinberg, Colla, & Martin, 1999). Evaluating Model’s Performance An important issue in regression problems is DERXWKRZWRPHDVXUHPRGHO¶VDFFXUDF\1RWHWKDW MNBF models are explored and only one has to be selected. In this chapter, we used the generalized cross validation (GCV) measurement, which ZHLJKWVWKHDFFXUDF\ 06( RUJRRGQHVVRI¿W R2) criteria with a factor depending on model complexity. This is measured through the number of effective parameters, which accounts both for the number of terms in the models and the number of parameters used in selecting the optimal positions of the knots. More details can be obtained from Friedman (1991). In this work, we additionally used the V-fold cross validation method to select the best free parameters through a more reliable measurement. This method consists on randomly splitting the available data in V folds, train with V - 1 folds and validate the results with the remaining fold. The procedure is repeated V times, thus generating V measurements of the least obtained error and accuracy (in our case, the GCV-R2). Then, in all developed models in this chapter, the average GCV-R2 is given. This way, results are more robust WRRYHU¿WWLQJWRWKHLQVDPSOHGDWD Dealing with Missing Values MARS can deal with missing values in the features, something that it is not readily available for stepwise linear regression. In such a case, MARS generates a missing variable X _ mis for a variable X, and uses the variable X only when interacted with the X _ mis=0 basis function. As a result, MARS effectively substitutes 0 for all missing values. Notationally, we will indicate in the (interacting) basis functions both possibilities: data present (“>.”) or data absent (“=.”).
762
Evaluating Variable Relevance An additional good characteristic of the MARS model is that the basis functions are easily interpreWDEOHDVWKH¿QDOPRGHOFDQEHYLHZHGDVDOLQHDU regression of the transformed variables through “max” functions. Therefore, the associated weight LVWKH¿QDOUHOHYDQFHIRUDJLYHQYDULDEOHRQO\ZKHQ it becomes activated, while the knot value indicates the threshold for which the variable becomes to be relevant in the model. This will allow us to extract useful conclusions from the built models, but also will allow us to generate rankings of variables through sensitivity (or saliency) analysis. In this context, the least important variable is the one with the smallest impact on the model quality (GCVR2); similarly, the most important variable is the RQHWKDWZKHQRPLWWHGGHJUDGHVWKHPRGHO¿WWKH most. The model variables are ranked from most to least important and displayed in the variable importance table. Data Analysis Limitations It is worth pointing out that, despite good experimental results, this analysis poses some problems concerning the adopted empirical methodology and assumptions. The main assumption is that the same process links the dependent with the independent variables for all the countries studied. Although the 30 countries included in the study belong to the OECD, different patterns are present and, thus, this assumption could be misleading. For instance, European countries and South Korea cannot be modeled under the same pattern. This problem could be reduced by including more observations from the same countries (time series modeling), or by clustering similar countries and applying dedicated models to each cluster. These approaches, though very promising, could not be followed because of the reduced number of available data. Another limitation of this approach is the use of data at a national level. The lack of available data at a regional or local scale limited the ability to develop more accurate models for these administrative levels. This is an important problem that has also been pointed out by other authors in the literature (Garcia-Murillo & Gabel, 2003).
Patterns of Broadband Development in OECD Countries
RESULTS In this section, the results obtained from the empirical analyzes aimed at identifying the key facWRUVLQÀXHQFLQJEURDGEDQGVXSSO\DQGEURDGEDQG DGRSWLRQDUHSUHVHQWHG2QO\WKHWDEOHVDQG¿JXUHV WKDWUHÀHFWWKHPRVWSHUWLQHQW¿QGLQJVKDYHEHHQ included in this section.2 MARS parameters have EHHQ¿[HGWRWKHIROORZLQJYDOXHV PD[LPXPLQteractions allowed as 1, maximum BFs as 30, speed factor as 2, and 10 V-fold cross-validation.
Analysis of Broadband Deployment In ordeUWRLGHQWLI\WKHPDLQIDFWRUVWKDWLQÀXHQFH broadband deployment, the effects of the following variables were studied: the supply-side variables (Groups 1-3), the economic level indicators (Group 7), and the population distribution indicators (Group 8), in DSL coverage (DSLCOV). The results of the empirical analysis suggest WKDW WKH PRVW LQÀXHQWLDO IDFWRUV IRU EURDGEDQG infrastructure supply are as follows: (a) the economic level of the country (Group 7); (b) the level of competition in the telecommunications market (Group 3); (c) the demographic distribution of the population (Group 8); and (d) the availability and investment in telecommunications infrastructure *URXS 7KHLQÀXHQFHRIWKHVHJURXSVRIIDFWRUV are the following.
Economic Level Indicators (Group 7) The results of the variable importance analysis of MARS show that DSL coverage in a country is mainly explained by its economic level, which is represented by the gross national income per capita variable (GNICAP). This variable is the PRVW LPSRUWDQW RQH IRU WKH EHVW ¿WWHG PRGHOV (Models 1-4) and also for any model that includes this variable. The statistical analysis of the data shows that the number of access lines that can be enabled to provide DSL technology (DSLCOV) is, on average, more than three times greater for countries with higher income levels than for countries with ORZHURQHV1HYHUWKHOHVVWKHFRUUHODWLRQFRHI¿FLHQW
GRHVQRWFRQ¿UPFOHDUO\WKLVSRVLWLYHUHODWLRQVKLS (0.555) which could indicate a non linear relationship between these two variables. 7KLVLVFRQ¿UPHGE\WKHDQDO\VLVRIWKHUHJUHVVLRQVORSHVIRU*1,&$3DVWKHLQÀXHQFHRQ'6/ coverage is not positive for all the variable range. $OOWKHPRGHOVFRLQFLGHLQWKDWWKLVLQÀXHQFHLV strongly positive for the region, that is from 5,250 USD and the knot around 14,300 USD which is approximately the GNICAP of Spain (2001, 2002) and New Zealand (2002) with relatively low economic level but high DSL coverage (more than 7KHQWKHLQÀXHQFHLVVOLJKWO\QHJDWLYHIRU higher values of GNICAP. These results indicate that an increase of the JURVVQDWLRQDOLQFRPHSURGXFHVDVLJQL¿FDQWLQcrease of DSL coverage for countries with income below 14,230 current USD. However, for countries with national income above this value, the results indicate that the relationship between GNICAP and DSL coverage is not so clear. Actually there are countries with more than 30,000 USD of national income per capita as Norway (2000) or Luxemburg (2001), with relatively low DSL coverage level (20% and 65% respectively), whereas others with national income around 20,000 USD, have almost complete DSL coverage such as Belgium (95%, 2002).
Level of Competition in the Telecommunications Market (Group 3) The empirical analysis suggests that the level of competition in the telecommunications market represented by Group 3 of variables is the second most important driver for broadband deployment. Nevertheless, it also reveals that the competition between different technologies (technological competition) has a greater impact on broadband availability than the competition between operators of the same technology (market competition). ,QWKHHPSLULFDODQDO\VLVWKHLQÀXHQFHRIWHFKnological competition is represented by: DSLVSCAB, which is a dummy variable that indicates the existence of competition between DSL and cable modem,3 the coverage of cable networks CTVHPAS, and the number of operators of other broadband technologies (OPUMTS, OPWLL,
763
Patterns of Broadband Development in OECD Countries
23),;36723&(/02% 2QWKHRWKHUKDQGWKH LQÀXHQFHRIPDUNHWFRPSHWLWLRQIRU'6/WHFKQRORJ\LVUHSUHVHQWHGE\WKHQHZHQWUDQW¶VVKDUHRI PSTN access lines (LINNEWEN) and by the numEHURI¿[HGVZLWFKHGWHOHFRPPXQLFDWLRQQHWZRUN 3671 RSHUDWRUV 23),;367 2WKHULQGLFDWRUV such as the number of operators offering broadband services over DSL and cable modem technologies, would have been useful for the analysis, but they were not available for all OECD countries. Technological and Market Competition Both the statistical and the correlation analysis RI WKH GDWD FRQ¿UPHG WKH VLJQL¿FDQW DQG SRVLWLYH LQÀXHQFH RI WHFKQRORJLFDO FRPSHWLWLRQ LQ the availability of DSL access networks. DSL coverage is dramatically higher for countries with competing cable infrastructures (an average of 10 times greater), and the existence of DSL and cable competition is the variable with the highest corUHODWLRQFRHI¿FLHQWZLWK'6/FRYHUDJH On the other hand, according to the variable importance analysis of MARS, the availability of competing infrastructures, either cable infrastructures (represented by DSLVSCAB) or PSTN access lines (represented by LINNEWEN), seems to have some relevance to explain DSL coverage. If these
two types of competition between infrastructures are compared (Model 2), the technological competition with cable networks appears as the most important factor after the gross national income per capita, whereas the market share in the local loop is in fourth position, after the percentage of urban population. The analysis also shows that the mere existence of market competition between ¿[HG3671RSHUDWRUV 23),;367 DVZHOODVWKH number of operators of other broadband technoloJLHVGRQRWVLJQL¿FDQWO\LQÀXHQFHWKHOHYHORI'6/ coverage (Models 1-4). The analysis of the regression slopes reveals WKDWWKHLQÀXHQFHRIWKHH[LVWHQFHRIWHFKQRORJLcal competition is always positive, whereas the LQÀXHQFHRI3671DFFHVVOLQHVFRPSHWLWLRQLVQRW positive for all the variable range. According to 0RGHOWKLVLQÀXHQFHLVRQO\VWURQJO\SRVLWLYH IRUWKH¿UVWDQGVHFRQGUHJLRQIRUPDUNHWVKDUH bellow 1.5%. This is the case of Korea where the new entrants share of access lines increased form 0.3% in 1999 to 1.5% in 2001, and the DSL coverage was relatively high (70%). )RUKLJKHUYDOXHVRIPDUNHWVKDUHWKHLQÀXHQFH of this variable is strongly negative for the third region limited by the knot value of 7.5%. This change of tendency is due to some cases such as the U.S. (2000) or the Netherlands (2000) with more
Table 1. Model 2, predicted variable: DSL coverage (DSLCOV) Variable
Importance
GNICAP
638.984
DSLVSCAB
519.491
DSLVSCAB_mis
482.029
73.013 ||||||||||||||||||||||||||||||||||||||||
URBPOP
349.498
37.260 ||||||||||||||||||||
URBPOP_mis
349.498
37.260 ||||||||||||||||||||
LINNEWEN
341.860
34.075 ||||||||||||||||||
LINNEWEN_mis
341.860
34.075 ||||||||||||||||||
POPDENS
324.321
25.285 |||||||||||||
Naive Adjusted: 0.742 GCV R-square: 0.717
764
Cost of Omission
100.000 ||||||||||||||||||||||||||||||||||||||||||||||||||||||| 80.283 ||||||||||||||||||||||||||||||||||||||||||||
Patterns of Broadband Development in OECD Countries
Figure 2. MARS regression slopes (a) for gross national income per capita (GNICAP) (DSL coverage, Model 1-3); (b) for number of ISDN channels per 100 inhabitants (ISDNCH100) (DSL coverage, Model 3); (c) for existence of competition between DSL and cable technologies (DSLVSCAB) (DSL coverage, Model 2); (d) for existence of competition between PSTN access lines (LINNEWEN) (DSL coverage, Model 2) Curve 1: Pure Ordinal
Curve 3: Pure Ordinal
60
25
50
20
40
15 30
10 20
5
10
0
0 0
10000
20000
30000
40000
0
50000
10
20
50
Curve 6: BF25--Categorical-Ordinal Interaction
LINNEWEN_mis, DSLVSCAB_mis
30
50
40
(b)
(a) Curve 3: BF5--Categorical-Ordinal DSLVSCAB_mis
30
ISDNCH100
GNICAP
40
20 30 20
10 10
0 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
DSLVSCAB
(c)
than 7% of market share but low DSL coverage, 36% and 40% respectively. Nevertheless, the cable network coverage is very high for these countries, with 94% and 97% of homes passed by a cable TV network respectively. Actually the success in the development of PSTN facilities-based competition in countries such as United States and the United Kingdom (19.8% and 10% of new entrants market share in UK and U.S. respectively, in 2001), both early starters in introducing telecommunications competition, can be partly explained by the development of telephony services provided by cable networks operators. This reason could also explain the increase of new entrants share of access lines
0
2
4
6
8
10
12 14 16 18 20 22 24
LINNEWEN
(d)
in other countries such as Spain (2.3% in 2000 to over 5.2% in 2001) (OECD, 2003c). )LQDOO\WKHLQÀXHQFHRIFRPSHWLWLRQLQWKHORFDO loop infrastructures is slightly positive for values higher than 7.5%, mainly due to some cases such as Denmark (2001) with 12% of market share and 90% of DSL coverage. Nevertheless, in Model 1 a QHJDWLYHLQÀXHQFHFDQEHREVHUYHGIRUWKLVUHJLRQ due to cases such as United Kingdom (2002) or Hungary (2001) with 19.8% and 21% of market share in the local loop, and low level of DSL coverage, 66% and 20% respectively. These results can be interpreted to mean that the mere existence of market competition does
765
Patterns of Broadband Development in OECD Countries
QRWKDYHDVLJQL¿FDQWLQÀXHQFHRQEURDGEDQGGHvelopment, either positive or negative. Moreover, only the existence of competing infrastructures is a key factor to explain broadband coverage. This LQÀXHQFH LV FOHDUO\ SRVLWLYH IRU WKH FRPSHWLWLRQ between infrastructures of different broadband technologies, in concrete DSL and cable. However WKHLQÀXHQFHRI3671IDFLOLWLHVEDVHGFRPSHWLWLRQ seems to be only strongly positive for countries that DUHVWDUWLQJWRLQWURGXFHFRPSHWLWLRQDQGWKHLQÀXence is even negative when there is a high market share of new entrants in the local loop.4 Local Loop Unbundling, Line Sharing Strategies to open the local loop are viewed as being fundamental in promoting a fast roll-out of broadband services (ITU, 2003b). Overall, a greater number of higher-income economies have embraced local loop unbundling (LLU) policies than have developing economies. Nevertheless, the effectiveness of LLU is under discussion and, to date, the major criticism of unbundling or line sharing is that such policies could discourage investment in new infrastructure. In order to shed light over this issue, the effects of local loop unbundling have been included in the empirical analysis by the following indicators: the existence of the regulation of LLU (ULLREG), and the percentage of unbundled local loops (ULLLIN). The statistical analysis shows that DSL coverage is higher for countries where LLU is mandatory5 (1.8 times greater on average) and where there is a higher percentage of unbundled local loops (2.8 times greater). It has also been observed that a high number of unbundled local loops is always associated to high levels of DSL coverage; whereas low number of unbundled local loops, this does not imply necessarily low coverage. Despite this apparent relationship, the correlaWLRQDQDO\VLVVKRZVWKDWWKHFRUUHODWLRQFRHI¿FLHQW LVSRVLWLYHEXWQRWVLJQL¿FDQWDQGDFFRUGLQJWRWKH MARS variable importance analysis, the implementation of LLU has no relevance to predict DSL coverage (Model 1-4). These results thus suggest that there is not a clear relationship between the
766
implementation of LLU and the deployment of broadband infrastructures, either positive or negative.
Demographic Cistribution of the Population (Group 8) The empirical analysis suggests that the demographic distribution is among the most important factors for broadband infrastructure availability. This issue is represented by the variables in Group 8, which include the household density (HOUSDENS), the population density (POPDENS), and the percentage of urban population (URBPOP). The statistical analysis shows that DSL coverage is, on average, 3.3 times greater for countries with higher levels of urbanization. The correlaWLRQDQDO\VLVFRQ¿UPVWKDWWKHUHLVDSRVLWLYHDQG VLJQL¿FDQWUHODWLRQVKLSEHWZHHQWKHSHUFHQWDJHRI urban population (URBPOP) and household density (HOUSEDENS) with DSL coverage. According to MARS models, an increase of population density implies an increase of DSL coverage for all the range of values of this variable (Models 1, 2). This is different for the percentage of urban population as this variable only has a SRVLWLYHLQÀXHQFHRQ'6/LQIUDVWUXFWXUHGHSOR\ment for values under 67% for cases (Models 1, 3, 4). This result suggests that the percentage of urban population is only a key factor to explain the deployment of broadband infrastructures, when this concentration is bellow this threshold value, or equivalently, when the percentage or rural population is high (i.e., Germany, Slovak Republic, Ireland, Finland, Greece, etc.). 7KH LQÀXHQFH RI GHPRJUDSKLF GLVWULEXWLRQ indicators can be explained by the direct relationship between the cost of broadband infrastructure deployment and the spatial density of users. Thus, in urban areas with high household and population density, the operators can take the maximum EHQH¿W IRU WKH LQIUDVWUXFWXUH GHSOR\PHQW FRVW as the number of possible customers covered is also high. Nevertheless, in rural or remote areas, which are characterized by low household or low SRSXODWLRQGHQVLW\WKHFRVWEHQH¿WUDWLRLVZRUVH than in urban areas for two reasons. First the cost
Patterns of Broadband Development in OECD Countries
Figure 3. MARS regression slopes (a) for population density (POPDENS) (DSL coverage, Model 1, 2); (b) for urban population (URBPOP) (DSL coverage, Model 1, 3, 4) Curve 2: Pure Ordinal
Curve 5: BF10--Categorical-Ordinal DSLVSCAB_mis, URBPOP_mis
25
60 20
50 40
15
30 10 20 5
10
0 0
100
200
300
400
500
POPDENS
(a)
of deploying broadband infrastructures is higher for areas of this type (long distances to cover with ¿EHUFDEOHRUUDGLROLQNVDQGQHZHTXLSPHQWLQWKH ORFDORI¿FH $QGVHFRQGWKHH[SHFWHGEHQH¿WVLQ rural areas are lower because there is usually less demand for broadband services. Thus, generally speaking, it can be assessed as an initial hypothesis that low levels of urbanization and low household and population density are constraints for broadband deployment.
The Availability and Investment in Telecommunications Infrastructure (Groups 1, 2) In the case of DSL technology, another constraint for infrastructure deployment has also been idenWL¿HGE\WKHHPSLULFDODQDO\VLV7KLVLVWKHOHYHO RIPRGHUQL]DWLRQRIWKH¿[HGWHOHSKRQHQHWZRUN represented by the number of access lines connected to digital exchanges, DIGACLIN, and the number of ISDN channels per inhabitant, ISDNCH100 (Group 1). At the end of 2001, the process of digitalizing WKH¿[HGWHOHSKRQHQHWZRUNZDVQHDULQJFRPSOHtion across the OECD area, with an average of 97%. Nevertheless, there are still some countries where the proportion of digital access lines is a little under 90%. These countries are Hungary, Korea, the Slovak Republic, Spain, and Turkey (OECD, 2003c).
0 40
50
60
70
80
90
100
URBPOP
(b)
According to the statistical analysis, DSL coverage is lower for countries whose PSTN networks have a lower digitalization level (2.3 times greater on average). Nevertheless, there are outliers such as Spain whose rate of digitalization is relatively low while its DSL coverage is relatively high (87% and 81% in 2001, respectively). In any case, the countries that lead the way for DSL coverage (90% or more) are fully digital, like Belgium and Denmark. The correlation analysis also shows a VLJQL¿FDQWSRVLWLYHFRHI¿FLHQWEHWZHHQGLJLWDOL]Dtion and DSL coverage. The analysis of the regression slopes of MARS models shows that an increase in the number of ISDN channels per 100 inhabitant, corresponds to an increase of DSL coverage for countries with less than 11.8 ISDN channels per100 inhabitants (Model 3). This result suggests that the deployment of DSL infrastructures is positively related to the modernization of the PSTN networks, especially for countries with low network digitalization level. In summary, the empirical analysis of broadband supply suggests that broadband coverage is mainly explained by the economic level of the country, although the relationship between these two variables is not always positive. Nevertheless, it is also clear that there are other key factors that FRXOG VLJQL¿FDQWO\ LQÀXHQFH EURDGEDQG GHSOR\ment. The existence of strong competition between
767
Patterns of Broadband Development in OECD Countries
telecommunications infrastructures, especially between broadband technologies seems to be the main driver for broadband deployment. However, low urbanization levels and consequently the high cost of network deployment per user appear to be the main constraints for building networks of this type. The empirical analysis also allows inferring DSRVLWLYHLQÀXHQFHRIWKHOHYHORIPRGHUQL]DWLRQ of the PSTN network. Finally, the presence of unbundled access to local loop seems to have not a clear relationship with the deployment of broadband infrastructures, either positive or negative.
Analysis of Broadband Adoption It has generally been accepted that both the supplyside and demand-side factors have a determining LQÀXHQFHRQEURDGEDQGSHQHWUDWLRQDQGWKHUHIRUH both should be taken into account to stimulate broadband adoption. Nevertheless, some have argued that government policies should be focused mainly on demand stimulation measures (i.e., UK). Others consider that more direct involvement by governments in broadband infrastructure availability and affordability would be more effective (i.e., Sweden and Japan). To shed light on this issue, the empirical analysis was carried out to study WKHMRLQWLQÀXHQFHRIVXSSO\DQGGHPDQGIDFWRUV on broadband adoption, both for DSL and cable modem technologies (DSLSB100 and CABSB100 variables). 5HJDUGLQJ WKH IDFWRUV WKDW FRXOG LQÀXHQFH broadband demand, evidence from many countries indicates that sociodemographic factors such as income, education level, gender, or age are major determinants (Savage, 2001). To explore this DVVXPSWLRQ DQ DQDO\VLV RI WKH LQÀXHQFH RI WKH group of demand-side variables (Groups 5-10) in broadband penetration (DSLSB100 and CABSB100 variables) was carried out (Models 1, 3). However, supply-side features such as the availability of EURDGEDQG LQIUDVWUXFWXUHV REYLRXVO\ LQÀXHQFH broadband adoption as well, so two supply-side variables were also included in the models which are DSL and cable coverage (LDSLCOV and LCTVHPAS variables, Group 1) (Models 2, 4). Finally, the DSL penetration has been included as a vari-
768
able in cable modem penetration and vice versa, LQRUGHUWRFDSWXUHWKHLQÀXHQFHRIWKHDGRSWLRQ of competing technologies. The results of the empirical analysis suggest WKDW WKH PRVW LQÀXHQWLDO IDFWRUV IRU EURDGEDQG adoption are: (a) the economic level of the country (Group 7); (b) the level of infrastructures availability (Group 1); (c) social indicators (Group 10); (d) the usage and adoption of Internet (Group 6); and (e) the education level (Group 9). Besides this, the adoption rates for both DSL and cable technologies are strongly and positively related (Group 11). Actually, the MARS models to predict DSL or FDEOHDGRSWLRQDUHSRRUO\¿WWHGLIDGRSWLRQUDWHRI the competing technology is not included in as a SUHGLFWRU7KHLQÀXHQFHRIWKHVHJURXSVRIIDFWRUV are the following.
Economic Level Indicators (Group 7) As for the broadband deployment, the results of the variable importance analysis of MARS show that broadband adoption are mainly explained by the economic level or the country, which is represented by the gross national income per capita variable (GNICAP). 7KLVUHVXOWLVKRZHYHUQRWFRQ¿UPHGE\WKH correlation analysis as none of the features considered as economic indicators in this study apparHQWO\KDYHDVLJQL¿FDQWFRUUHODWLRQZLWKEURDGEDQG penetration. Nonetheless, a more detailed statistical study revealed that the presence of outliers such as Korea in the upper part of broadband penetration (more than percentile 75) and a low income level (GNI less than percentile 25) could explain the ORZFRUUHODWLRQFRHI¿FLHQW7KLVSRLQWZDVWHVWHG E\FDOFXODWLQJWKHFRUUHODWLRQFRHI¿FLHQWEHWZHHQ national income and broadband penetration without taking into account the outliers. The results con¿UPHGWKDWLQFRPHOHYHOKDVDSRVLWLYHLQÀXHQFH on broadband penetration. The analysis of the regression slopes for *1,&$3DOVRUHYHDOVWKHLQÀXHQFHRIWKHFDVHRI .RUHD)RUWKHEHVW¿WWHGPRGHOVWKHLQÀXHQFHRI this variable is strongly positive before the knot of 9,010 USD, which is approximately the gross national income per capita of Korea, followed by
Patterns of Broadband Development in OECD Countries
Table 2. Model 2, predicted variable: DSL subscribers (DSLSB100) considering DSL coverage Variable
Cost of
Importance
Omission
GNICAP
0.753
LDSLCOV_mis
0.491
100.000 ||||||||||||||||||||||||||||||||||||||||||||||||||||||| 77.431 ||||||||||||||||||||||||||||||||||||||||||
EMPWOMR
0.366
63.950 |||||||||||||||||||||||||||||||||||
CABSBS10
0.331
59.688 ||||||||||||||||||||||||||||||||
LDSLCOV
0.228
44.637 ||||||||||||||||||||||||
PC100
0.160
30.926 |||||||||||||||||
PUNDSEC
0.148
27.802 |||||||||||||||
SITES
0.109
13.327 |||||||
Naive Adjusted: 0.971 GCV R-square: 0.979
a strongly negative slope until the knot of 11,430 USD. Finally, the last region presents a decreasing slope (Model 2). As for the models for DSL coverage, these results indicate that an increase of the gross national income produces an increase of DSL adoption for countries with low income level, whereas this relationship is not so clear of even negative for countries with higher incomes.
Infrastructures Availability (Group 1) The results of the variable importance analysis show that the availability of broadband infrastructures, represented by the percentage of DSL enabled local loops (LDSLCOV) and the percentage of homes passed by a cable TV network (LCTVHPAS) (both in the previous year), is a key factor to explain broadband penetration. 7KHDQDO\VLVDOVRLQGLFDWHVWKDWWKHLQÀXHQFH of infrastructure availability is positive both for DSL and cable modem penetration. Whereas for '6/WHFKQRORJ\WKHLQÀXHQFHLVSRVLWLYHIRUDOO the values range, it is only positive for values higher than 75% of homes passed in the case of cable technology. These results suggest that the availability of infrastructures is still a key factor to consider in understanding broadband adoption in OECD countries.
Social Indicators (Group 10) The analysis indicates that some social indicators, especially the employment rate of women (EMP:205 FRXOGKDYHVRPHLQÀXHQFHWRH[SODLQWKH tendency of the society to adopt broadband, both for DSL and cable modem technologies. This result could be interpreted to mean that open-minded societies, possibly characterized by a high level of women incorporated into the labor market, could have a tendency to adopt new technologies such as broadband.
Internet Usage (Group 6) The results of the empirical analysis show that the number of PCs per 100 inhabitants (PC100) and the number of local Web sites (SITES) are WKH YDULDEOHV RI WKLV JURXS WKDW PRVW LQÀXHQFH broadband demand for both DSL and cable modem technologies. The analysis of the regression slopes RI0$56UHYHDOVWKDWWKHLQÀXHQFHRIWKHQXPEHU of local Web sites is positive for all the range of values of this variable, and for values higher than 9.5 PCs per 100 inhabitants (Model 1). 7KHLQÀXHQFHRIWKHQXPEHURI3&VLVFRQ¿UPHG by the MARS variable importance analysis only for DSL penetration, whereas the number of local Web sites appears as a relevant predictor both for DSL
769
Patterns of Broadband Development in OECD Countries
Figure 4. MARS regression slopes (a) for lagged DSL coverage (LDSLCOV) (DSL subscribers, Model 2); (b) for lagged cable TV coverage (LCTVHPAS) (Cable modem subscribers, Model 4) Curve 6: BF14--Categorical-Ordinal Interaction
LDSLCOV_mis
Curve 4: Pure Ordinal 1.6
2.5
1.4 1.2
2.0
1.0 1.5
0.8 0.6
1.0
0.4 0.5 0.2 -10
0
10
20
30
40
50
60
70
80
90 100
LDSLCOV
(a)
DQGFDEOHPRGHP7KLVFRQFOXVLRQLVFRQ¿UPHG by the statistical analysis as DSL penetration is almost four times greater in countries where there are more local Web sites and this number increases WR¿YHIRUFDEOHPRGHPSHQHWUDWLRQ1HYHUWKHOHVV WKH LQÀXHQFH RI WKLV YDULDEOH LV QRW FRQVLVWHQWO\ FRQ¿UPHGE\WKHFRUUHODWLRQDQDO\VLV These results suggest that broadband demand is higher when people have some kind of predisposition to use Internet and new technologies as well as the skills to use them, especially for DSL technologies. It also reveals that the local content availability represented by the number of local content, could be a driver for broadband adoption.
Education (Group 9) ,Q WKH HPSLULFDO DQDO\VLV WKH LQÀXHQFH RI WKLV indicator is represented by PUNDSEC, that is the percentage of population that has not attained secondary level of education, PUPSEC, that is the percentage of population that has attained upper secondary and postsecondary level of education, and PTER, that is the percentage of population with a tertiary education level (PTER), all of them for people aged between 25 and 64. The statistical analysis reveals that the education level has a positive relationship on broadband
770
-10
0
10
20
30
40
50
60
70
80
90
100
LCTVHPAS
(b)
penetration. For countries with higher education levels, DSL and cable modem penetration is 5.6 DQGWLPHVJUHDWHUUHVSHFWLYHO\7KLVLQÀXHQFH is only weakly demonstrated by the correlation analysis but appears clearly demonstrated in the variable importance analysis of MARS especially for DSL penetration. The analysis of the regression VORSHVFRQ¿UPVWKHSRVLWLYHLQÀXHQFHRIWKHSHUcentage of population with education level higher than secondary (PTER and PUPSEC, Model 1), DQG WKH QHJDWLYH LQÀXHQFH RI WKH SHUFHQWDJH RI population with lower education level (PUNDSEC, Model 2). Nonetheless, some of the regression VORSHVDUHQRWORJLFDO VXFKDVWKHSRVLWLYHLQÀXence of PUNDSEC in DSL penetration in Model 1), which indicates that a further analysis should be done over this issue. In summary, the empirical analysis of broadband adoption suggests that it is mainly explained by the economic level of the country, although there is not a direct relationship between these two values. It also reveals that the availability of broadband infrastructures is still a key factor to explain the penetration of these technologies. Besides this, this results suggest that broadband demand is higher when local content is available and the people have some kind of predisposition to use new technologies such as the PC. Finally, the
Patterns of Broadband Development in OECD Countries
)LJXUH0$56UHJUHVVLRQVORSHV D IRUQXPEHURI3&VSHULQKDELWDQWV 3& '6/ subscribers, Model 1, 2); (b) for number of local Web sites (SITES) (DSL subscribers, Model 2); (c) for percentage of population with a tertiary education level (PTER) (DSL subscribers, Model 1); (d) for percentage of population that has not attained secondary level of education (PUNDSEC) (DSL subscribers, Model 2) Curve 5: Pure Ordinal
Curve 9: BF29--Categorical-Ordinal LDSLCOV_mis
3.0
0.8
0.6
2.0
0.4 1.0 0.2
0.0
0.0 0
10
20
30
40
50
60
0
70
10
20
30
40
50
PC100
SITES
(a)
(b)
Curve 2: Pure Ordinal
60
70
80
90
Curve 5: Pure Ordinal 1.4
50
1.2
40
1.0 30
0.8
20
0.6 0.4
10
0.2
0 0
10
20
30
40
50
0.0 10
20
30
40
50
60
PTER
PUNDSEC
(c)
(d)
analysis also shows that other social features that could indicate open-mindedness of high educaWLRQOHYHOKDYHDSRVLWLYHLQÀXHQFHRQEURDGEDQG demand.
DISCUSSION AND CONCLUSION The results of the empirical study show what could EHWKHPDLQIDFWRUVWKDWLQÀXHQFHERWKEURDGEDQG supply and demand in OECD countries. Regarding broadband supply, the existence of competition between telecommunication infra-
70
80
90
structures, especially between broadband technologies, seems to be the main driver for broadband deployment. The empirical analysis also allows XVWRGHGXFHDSRVLWLYHLQÀXHQFHRIWKHOHYHORI digitalization of the PSTN network on DSL coverage. On the other hand, low urbanization levels, and consequently the high cost of network deployment per user, appear to be the main constraint for broadband provision. Finally, the presence of unbundled access to local loop seems to have not a clear relationship with the deployment of broadband infrastructures, either positive or negative. Regarding broadband adoption, the analysis suggests that demand is higher when local Web
771
Patterns of Broadband Development in OECD Countries
content is available and when the use of new technologies, such as the PC, is generalized among the population. Also, social features that could indicate open-mindedness and high education level, seem WRKDYHDSRVLWLYHLQÀXHQFHRQEURDGEDQGGHPDQG The analysis also reveals a positive relationship between broadband supply and demand, as the availability of broadband infrastructures is still a key factor to explain the penetration in many countries. Once the empirical study has provided which are the main constraints and motivators for broadband deployment and adoption, these results can be used to shed light on the effectiveness of the different broadband policy alternatives, which is WKH¿QDOREMHFWLYHRIWKLVUHVHDUFK$FFRUGLQJWR Cava and Alabau (2006), these alternatives can be URXJKO\FODVVL¿HGDV D VRIWLQWHUYHQWLRQVWUDWHgies, (b) medium-intervention strategies, and (c) hard-intervention strategies. Soft-intervention strategies are characterized by low government involvement in broadband infrastructure deployment as they rely on market forces to ensure broadband supply. Some of the most representative examples of this orientation are the strategies of the United Kingdom, Switzerland, Denmark, and New Zealand. Soft-intervention strategies are strongly recommended by OECD (OECD, 2002a) and organizations such as the European Competitive Telecommunications Association (ECTA, 2004). The main arguments for this approach are that the development of broadband infrastructure is still at an early stage and an excessive government involvement may distort competition and affect future market development. On the contrary, hard-intervention strategies are characterized by very proactive government involvement in broadband infrastructure deployment. This is the case of countries such as the Republic of Korea, Norway, and Singapore, where there is a strong tradition of state intervention in the investments of the private sector. As an intermediate option, medium-intervention strategies are characterized by a relatively proactive government involvement in broadband infrastructure deployment to complement market functioning in geographic areas where there is a
772
lack of broadband infrastructures. The rationale for these strategies is that, under market conditions, there are certain areas that are scarcely populated such as rural or remote areas where broadband services might not be available. This type of public actions is very common in the European Union framework as member states have LQWURGXFHG ¿QDQFLDO VWUDWHJLHV XVLQJ VWUXFWXUDO funds to facilitate deployment in remote and rural areas in countries such as Italy, France, and Spain (European Commission, 2004). There are also similar strategies in the U.S. such as the Rural Utilities Service Telecommunications Program to subsidize the development of plain service telephone networks (PSTN) in rural areas that are prepared for broadband.6 According to the results of the empirical analysis, the most effective policy for accelerating broadband deployment would be the promotion of technological competition (i.e., between DSL and cable networks), which is in the foundations of the soft and medium-intervention strategies. On the other hand, since it has also been proven that one of the main constraints for broadband provision is the high cost of network deployment per user in rural or scarcely populated areas, it is reasonable to conclude that medium-intervention strategies that provide public funding for infrastructure supply in these areas, could be more suitable for a balanced broadband coverage in the whole territory. It is worth noting that, in the light of the empirical analysis, other competition policies such as the promotion of unbundled access to local loops seems to have no impact on broadband provision, either negative or positive. Nevertheless, it can be argued that this policy could encourage competition DQGWKXVKDYHRWKHUEHQH¿FLDOHIIHFWVVXFKDVWKH reduction of access prices and the improvement of ¿QDOVHUYLFHTXDOLW\ Regarding the policies to promote demand, the analysis suggests that soft-intervention strategies aimed WRLQFUHDVHSHRSOH¶VXVHRIQHZWHFKQRORgies and Internet could be very effective. To this purpose, public actions to raise the awareness of EURDGEDQGEHQH¿WVDQGWRLPSURYHVNLOOVIRUXVLQJ PCs or other user equipment could be useful to promote demand. However, the strong relationship
Patterns of Broadband Development in OECD Countries
between broadband supply and demand also allows the conclusion that medium-intervention policies aimed to guarantee the availability and quality of broadband networks could be useful to promote broadband demand. In conclusion, it can be stated that the crossnational empirical analysis can provide some useful information to design effective broadband policies that support the drivers and attenuate the constraints to broadband development. However, it should be pointed out that the inherent limitations of the empirical method do not allow us to infer strong political recommendations. In order to overcome these limitations, some improvements can be put in practice to tackle the problem of broadband develRSPHQWPRGHOLQJLQWKHIXWXUH ¿UVWWRLQFRUSRUDWH more data for the modeling process, both from a temporal perspective (time-series data) and from a spatial perspective (data at a regional or local scale), and second, to capture the effects of other W\SHVRIVRFLDOIHDWXUHVVXFKDVWKHLQÀXHQFHRI the public policies to promote broadband. To this purpose, the coding of government initiatives and their inclusion as independent variables should be an effort made in future work.
REFERENCES Bauer, J.M., Kim, J.H., & Wildman, S.S. (2003). Broadband uptake in the OECD countries: Policy lessons and unexplained patterns. Paper presented at the 14th European Regional Conference of the International Telecommunication Society, Helsinki, Finland, August 23-24, 2003. BDRC Ltd. (2001). The development of broadband access platforms in Europe. Technologies, services, markets. Brussels: European Commission. Retrieved July 29, 2007, from http://europa. eu.int/eeurope Cava, I., & Alabau, A. (2003). Comparative analysis of broadband policy. A methodological approach. Paper presented at the 14th European Regional Conference of the International Telecommunication Society, Helsinki, Finland, August 23-24, 2003.
Cava, I., & Alabau, A. (2006). Broadband policy assessment: A cross-national empirical analysis. Telecommunications Policy, 30, 445-463. Distaso, W., Lupi, P., & Maneti, F.M. (2006). Platform competition and broadband uptake: Theory and empirical evidence from the European Union. Information Economics and Policy, 18, 87-106. ECTA. (2004). ECTA News. Retrieved July 29, 2007, from http://www.ectaportal.com European Commission. (2004). Connecting Europe at high speed: National broadband strategies. Retrieved July 29, 2007, from http://europa. eu.int/information_society/eeurope/ Falch, M. (2004). How to achieve the goal of broadband for all. Paper presented at the ITS Conference, Berlin, Germany, September 14-17, 2004. Frieden, R. (2005). Lessons from broadband development in Canada, Japan, Korea, and the United States. Telecommunications Policy, 29, 595-613. Friedman, J.H. (1991). Multivariate adaptive regression splines (with discussion). Annals of Statistics, 19(1), 82-91. Gabel, D., & Kwan, F. (2000). Accessibility of Broadband Telecommunication Services by Various Segments of the American Population. Paper Prepared for the Telecommunications Policy Research Conference, Queens College, CUNY. Retrieved July 29, 2007, from http://itc.mit.edu/ itel/docs/MISC/gabel_kwan_tprc.pdf Garcia-Murillo, M., & Gabel, D. (2003). International broadband deployment: The impact of unbundling. Paper presented at the 31st Telecommunications Policy Research Conference, Arlington, Virginia. Gillet, S.E., Lehr, W.H., & Osorio, C. (2004). Local government broadband initiatives. Telecommunications Policy, 28, 537-558. Grubesic, T.H. (2004). The geodemographic correlates of broadband access and availability in the United States. Telecommunications Policy, 21, 335-358.
773
Patterns of Broadband Development in OECD Countries
Grubesic, T.H., & Murray, A.T. (2002). Constructing the digital divide: Spatial disparities in broadband access. Papers in Regional Science, 81(2), 197-221.
OECD. (2002b). Broadband access for business. Paris: Head of Publications Service, OECD. Retrieved July 29, 2007, from http://http://www. oecd.org/
Hair, J.F., Anderson, R.E., Tatham, R.L., & Black, W.C. (1998). Multivariate data analysis. Upper Saddle River, NJ: Prentice-Hall International, Inc.
OECD. (2003a). OECD telecommunication database 2003 on CD-ROM.
ITU. (2003a). World telecommunication indicators database on CD-ROM (7th ed.). Geneva: International Telecommunication Union. Retrieved July 29, 2007, from http://www.itu.int/ITU-D/ict/publications/
OECD. (2003b). Society at a glance: OECD social indicators 2002 Edition. OECD. OECD. (2003c). OECD communications outlook. Paris: Head of Publications Service, OECD. Retrieved July 29, 2007, from http://www.oecd.org/
ITU. (2003b). ITU Internet reports 2003: Birth of broadband. Geneva: International Telecommunication Union. Retrieved July 29, 2007, from http://www.itu.int/broadband
Rappoport, P.N., Kridel, D.J., Taylor, L.D., & Alleman, J. (2003). Residential demand for access to the Internet. In G. Madden (Ed.), The international handbook of telecommunications economics, Vol. II (pp. 55-72). Cheltelham, UK: G. Madden.
Lau, T.Y., Kim, S.W., & Atkin, D. (2005). An exDPLQDWLRQRIIDFWRUVFRQWULEXWLQJWR6RXWK.RUHD¶V global leadership in broadband adoption. Telecommunications Policy, 22, 349-359.
Salford Systems. MARS v1.0, © copyright. All rights reserved worldwide. Retrieved July 29, 2007, from http://www.salford-systems.com
Lee, C., & Chan-Olmsted, S.M. (2004a). Competitive advantage of broadband Internet: A comparative study between South Korea and the United States. Telecommunications Policy, 28, 649-677.
Savage, J. (2001). International public programs to provide broadband access to the Internet. Benchmarking analysis. Canada National Broadband Task Force. Retrieved July 29, 2007, from http:// www.broadband.gc.ca
Lee, C., & Chan-Olmsted, S.M. (2004b). Competitive advantage of broadband Internet: A comparative study between South Korea and the United States. Telecommunications Policy, 28, 649–677.
Steinberg, D., Colla, P.L., & Martin, K. (1999). MARS user guide. San Diego, CA: Salford Systems.
/L: ;X/& 7KHSROLWLFDOHFRQRP\RI privatization and competition: Cross-country evidence from the telecommunications sector. Journal of Comparative Economics, 30, 439-462.
World Bank. (2003). World development indicators 2003 on CD-ROM. Washington, DC: World Bank Group.
OECD. (2001). The development of broadband access in OECD countries. Paris: Head of Publications Service, OECD. Retrieved July 29, 2007, from http://http://www.oecd.org/pdf/M00020000/ M00020255.pdf
KEY TERMS
OECD. (2002a). Broadband infrastructure deployment: The role of government assistance. Paris: Head of Publications Service, OECD. Retrieved July 29, 2007, from http://www.oecd.org/sti/working-papers
774
Multivariate Adaptive Regression Splines Regression (MARS): An adaptive procedure for function approximation and regression. The structural model built by MARS considers a set of expansions in piecewise linear basis functions (BFs) of the form (x - t)+ and (t - x)+ , where the subscript “+” means “positive part” whose corresponding weighs are adaptively tuned.
Patterns of Broadband Development in OECD Countries
Multivariate Linear Regression: A regression method that assumes that the relation of the response to the explanatory variables is a linear function. The basic assumptions of multivariate regression are (1) multivariate normality of the residuals, (2) homogenous variances of residuals conditional on predictors, (3) common covariance structure across observations, and (4) independent observations. Digital Subscriber Line (DSL): A family of technologies that provide digital data transmission over the wires of a local telephone network. DSL takes advantage of the unused bandwidth of the local loop (frequencies above 3.4KHz) by creating 4312.5 Hz wide channels starting between 10 and 100 for digital data transportation. Asymmetric digital subscriber line (ADSL) is the most chosen broadband option in the world, holding more than 60% of the broadband market. The access device is a DSL modem. Hybrid Fiber Coaxial (HFC): A type of QHWZRUN EDVHG RQ ERWK RSWLFDO ¿EHU DORQJ ZLWK coaxial cable to provide interactive broadband VHUYLFHV 7KH ¿EHU RSWLF QHWZRUN H[WHQGV IURP WKH FDEOH RSHUDWRUV FHQWUDO RI¿FH WR WKH DUHD¶V KHDGHQGRXWWRDQHLJKERXUKRRG¶VKXEVLWHDQG ¿QDOO\WRD¿EHURSWLFQRGHZKLFKVHUYHVIURPWR KRPHV7KH¿EHURSWLFQRGHFRQYHUWVRSWLFDO signals to electronic and vice versa, and from this node to the user premises, the network is based on coaxial cables that connect 100 to 2000 homes in a WUHHDQGEUDQFKFRQ¿JXUDWLRQ%\XVLQJIUHTXHQF\ division multiplexing, an HFC network may carry a variety of services, including noninteractive (TV services) and interactive services such as broadband access to Internet. The access device is a cable modem. Local Loop Unbundling (LLU): A regulatory requirement that enables a competing telecommunications operator to use the twisted-pair telephone FRQQHFWLRQVIURPWKHWHOHSKRQHH[FKDQJH¶VFHQWUDO RI¿FHWRWKHFXVWRPHUSUHPLVHV7KLVORFDOORRSLV owned by the incumbent local exchange carrier. Broadband Digital Divide: The gap between those with effective access to broadband services
and those without. The term refers not only to the unavailability of physical access to broadband QHWZRUNVEXWDOVRWRWKHGLI¿FXOWLHVLQXVLQJWKHVH technologies because of lack of skills and literacy. Given that broadband is seen as the key driver for future social and economic development, the digital divide widens the gap in economic divisions around the world. Digitalization of Telephone Network: The replacement of analogue exchanges by digital exchanges. It allows the extension of ISDN coverage and improves the reliability and quality of the network.
ENDNOTES 1
2
3
4
5
6
This is because this is the most basic speed being offered by telecommunications providers in OECD countries and allows the comparison and contrast of the initial speeds available for broadband access. Nevertheless, ISDN technologies are considered not to constitute broadband. The material related to this empirical research is available by request. There are very few OECD countries that have two or more competing providers using different infrastructures on a widespread basis for broadband access (OECD, 2002b). This result coincides with (Distaso et al., 2006). $WWKHHQGRIRQO\¿YHFRXQWULHVKDG not mandated LLU in one form or another, Mexico, New Zealand, the Slovak Republik, Switzerland and Turkey (OECD, 2001). Overall, a greater number of higher-income economies have embraced LLU policies than have developing economies. Switzerland and New Zeland are two exceptions as LLU is not mandatory and DSL coverage is very high. This condition mainly consist of limiting the maximum local loop length to allow broadband technologies as ADSL or even VDSL.
775
776
Chapter XLVII
ICT Statistics for Broadband Promoting Regulatory Policy 'LDQD.RUVDNDLWơ Communications Regulatory Authority, Lithuania Tomas Lamanauskas Communications Regulatory Authority, Lithuania
ABSTRACT This chapter introduces the statistical analysis of a number of information and communication technologies (ICT) market indicators as a means to develop sound regulatory policies aiming to promote broadband take up. The chapter provides analysis of the concept of broadband, statistical analysis of ICT indicators, time series and cross-country series against broadband penetration, determination of FRQWHUPLQRXVIDFWRUVZLWKVWDWLVWLFDOO\VLJQL¿FDQWLQÀXHQFHRYHUEURDGEDQGSHQHWUDWLRQLQ(XURSHDQ VWDWHVDQGGHULYDWLRQRIFDOOVIRUUHJXODWRU\SROLF\SDUWLFXODUDFWLRQVRXWRIVWDWLVWLFDOO\VLJQL¿FDQW relations. The overall aim of the chapter is to offer a way how ICT statistical data could be employed to suggest preconditions for possible practical solutions in the domain of broadband promotion and in this way to bring the rhetoric of statistics down to an operational level.
INTRODUCTION The 21st century economy, or information economy, is based on information society on one hand and inIRUPDWLRQLQIUDVWUXFWXUHRQWKHRWKHUKDQG 7RIÀHU citing Rytel, 2001). As Professor W. Melody argues, this economy is founded on information/communication networks and dependent on effective reforms in the telecom sector (Melody, 2003). With this in mind, we can state that broadband functions as a conductor or link delivering information and communication technologies (ICT) gains to every individual.
Much attention is delivered to the issue of statistics on the development of the ICT sector and adoption of ICTs in the highest political forums—the World Summit on the Information Society could be just one out of many examples. However, statistical data in the ICT domain is quite often used for comparison purposes solely, which does not allow employing its potential to the full extent. We would say that the true value of statistical data is delivered when statistics, properly applied, suggest preconditions for possible solutions. Therefore keeping this principle in mind, delivering due regard to importance of broadband
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
ICT Statistics for Broadband Promoting Regulatory Policy
development for the information economy, and understanding the importance of regulatory intervention on the progress of the aforementioned processes, we will employ statistical data on ICT development and adoption to deduce eligible and possible regulatory policy guidelines fostering broadband development. The overall aim of the research is to propose a way that ICT statistical data could be employed to suggest preconditions for possible practical solutions in the domain of broadband promoting regulatory policy, in this way bringing the rhetoric of statistics down to operational level. The chapter LVRUJDQL]HGDVIROORZV ¿UVWDQDO\VLVRIEURDGEDQG concept is performed; second, statistical analysis of ,&7LQGLFDWRUV¶WLPHVHULHVDQGFURVVFRXQWU\VHULHV is executed; third, conterminous factors having VWDWLVWLFDOO\VLJQL¿FDQWLQÀXHQFHRYHUEURDGEDQG penetration in European states are determined; fourth, calls for regulatory policy actions out of WKLV LQÀXHQFH DUH LGHQWL¿HG ¿QDOO\ JXLGHOLQHV for broadband promoting regulatory policy are provided. Our research will be focused mainly on the EU states. This geographical and political area was chosen with regard to several arguments: ¿UVWWKHUHLVDFRPPRQSROLWLFDODQGUHJXODWRU\ framework in this area; second, development of the common market and observed processes of evolving pan-European ICT networks assimilate business modes or supply side of the sector; third, the demand side of the sector or usage patterns of consumers in the area are or are about to be harmonized as well. All this constitutes that the choice of EU area underlies avoidance of possible VLJQL¿FDQW GLVFUHSDQFLHV GXH WR VSHFL¿FLWLHV RI national ICT policies and practices (this would be incurred if states from very different regions were chosen). We used the following methods: literature review, analysis and synthesis of statistical data, graphical, logical, and empirical analysis.
BACKGROUND The broadband adoption and regulatory policy promoting the broadband take up is studied in a
number of researches. The role of the government in the development of alternative broadband networks is studied in the work by Gillett, Lehr, and Osario (2004), who provides a taxonomy distinguishing four categories of government action based on the QDWXUHRIWKHJRYHUQPHQW¶VUROH JRYHUQPHQWV DV EURDGEDQG XVHU LQÀXHQFLQJ WKH GHPDQG RI EURDGEDQG JRYHUQPHQWDVUXOHPDNHULQÀXencing the regulatory framework that affects the commercial deployment of the infrastructure and VHUYLFHV JRYHUQPHQWDV¿QDQFLHULQÀXHQFLQJ ¿QDQFLDOLQFHQWLYHVWRLQYHVWDQGGHSOR\QHWZRUNV and acquire services, and (4) government as infraVWUXFWXUHGHYHORSHULQÀXHQFLQJWKHVXSSO\VLGHRI EURDGEDQGPDUNHWSODFH%DVHGRQWKH¿QGLQJVRI the research on numerous U.S. cases, the authors conclude that public authorities play an important role for the wider adoption of broadband and that involvement in the development of broadband infrastructure is continuing to grow (Gillett et al., 2004). Chlamtac et al. discusses possible models as to how the public sector can be involved in broadband SURMHFWV UDLVLQJDZDUHQHVVRIWKHEHQH¿WVRI broadband, (2) carrying out tasks related to assessment of potential demand, (3) usage of public funds effectively in broadband projects, and (4) determination of level of involvement and modeling broadband deployment (Chlamtac et al., 2005, as cited in Tadayoni & Sigurdsson, 2006). Quite similarly, roles of councils/local governments in promoting broadband are distinguished by van Winden and Woets (2004), who investigate actions of local governments in different cities and GLVFXVVWKHSDUWLFXODUEHQH¿WVDQGULVNVRISXEOLF sector involvement in broadband infrastructure deployment. Tadayoni and Sigurdsson (2006) further applied and elaborated the taxonomy of Gillett et al. (2004) in their investigation of the technological, economic, and political/regulatory drivers and barriers of alternative broadband infrastructures, including the role of government in fostering their H[LVWHQFH)ROORZLQJWKH¿QGLQJVRISUHVHQWHGFDVH studies, the authors argue that lower market entry barriers and public sector involvement are the main two factors acting as drivers for alternative broad-
777
ICT Statistics for Broadband Promoting Regulatory Policy
EDQGLQIUDVWUXFWXUHV¶GHSOR\PHQW)XUWKHUPRUHLW is noted that alternative broadband operators tend to employ more open, operator-independent business models in opposition to traditional broadband market players. A study conducted by Choudrie and Dwivedi (2005) on methods used to examine technology adoption issues both in the context of organizations and households that are employed in the papers published in peer reviewed journals between 1985 and 2003 proposes that survey and case study methods are most frequently used in the context in organizations; authors have as well found that in the case of technology adoption by households, a much greater variety of research methods employed is observed, including surveys, multimethods, ethnographic studies, time use diaries, and second data analysis. The authors conclude that the choice of a particular method to be employed is relevant to the extent to which a researcher is a part of the context under study. We admit that most analyses on government policies in the domain of broadband are based on case studies. However, government institutions make and argument their choices with statistics, while deciding upon policies. Therefore, the chapter suggests the statistical approach to formulation of broadband promoting regulatory policy.
ANALYSIS OF THE CONCEPT OF BROADBAND &ODUL¿FDWLRQRIWKHFRQFHSWRIEURDGEDQGLVLPSRUtant for academic exercise purposes as well as for numerous groups of stakeholders—for consumers, service providers, application and content developers, policy makers and regulators, and public interest groups. As Flournoy (2004) points out, broadband is a technological term having to do with the quantity and quality of information to be exchanged over a distance using channels of telecommunications. But the term includes much more—broadband is a business and social goal to be achieved, a way of expressing corporate and public aspirations for a more perfect society EDVHGRQSHRSOH¶VDELOLW\WRJHWPRUHRIZKDWWKH\
778
want, and faster. Here we dedicate our efforts to GH¿QLQJWKHEURDGEDQGQRWLRQ The European Commission distinguishes the following main features of common understanding of the term “broadband:” (1) it is Internet connection, (2) it is “always on,” (3) it provides a speed that LVVLJQL¿FDQWO\IDVWHUWKDQGLDOXSDQG LWVXSports delivery of innovative content, applications and services (Connecting Europe, 2004). Prima IDFLHWKH¿UVWIHDWXUHFRXOGVHHPGRXEWIXODVPDQ\ technological applications are not limited to one network, but this feature cannot be ignored—the current interest in broadband is largely due to the Internet, permitting familiar services to be delivered in unfamiliar ways; moreover, the Internet until now has been generally delivering traditional and new services at lower quality with less reliability than conventional networks but broadband access promises to change all that (Broadband and the Internet, 2001). The second feature, “access 24 hours a day,” is highlighted by many sources in XQLVRQIRUH[DPSOHWKH1DWLRQDO2I¿FHIRUWKH Information Economy (Australia) (Inquiry, 2004), Kim, Bauer, and Wildman (2003), Telstra (Inquiry, 2004), (Always on, 2002), and others. 0DQ\WHFKQRORJLFDOGH¿QLWLRQVOLQNWKHFRQFHSW of broadband to the transmission capacity, probably because broadband was originally an engineering term referring to an amount of information that could be carried between a sender and a receiver through a communications channel, with the implication that a broadband network can carry a lot more information than the traditional methods of accessing the Internet, typically referred to as “narrowband” or “dial up,” using a telephone line and modem (Inquiry, 2004). The Electronic Industries Alliance notes that the concept of broadband captures the notions of greatly increased transmission speed along with transmission capacity and it also enables interactive services (Broadband, 2004). However, there is little consensus as to what “a lot” of information means or which of the evolving technologies are attributed to broadband ones. The European Commission (European Electronic Communications, 2004), for comparison purposes, considers broadband as “capacity equal to, or higher than, 144 kbps;” the Federal Communica-
ICT Statistics for Broadband Promoting Regulatory Policy
tions Commission (FCC) and Canadian Broadband Task Force consider broadband to deliver data rates above 200 kbps (FCC, 2004; Kim at al., 2003); in Australia, a rate of 200 kbps has been accepted (Household broadband, 2001; Inquiry, 2004); OECD goes for as much as 256 kbps downstream capacity (Connecting Europe, 2004). Another approach to determining data rate speed is related to services provided: certain services can dictate the data transfer rate required for broadband. For example, a data transfer rate of 1.544 Mb/s is suitable for full motion videoconferencing, whereas for business applications 384 kb/s is the typical data rate; companies offering residential high speed Internet access over phone or cable modem typically advertise a data transfer rate of 1 Mb/s and up for downstream data and 120 kb/s and up for upstream data (Understanding broadband, 2003). Such a variety of possible approaches does not allow IRUPXODWLQJDVLQJOHGH¿QLWLRQ.LPHWDODUJXH that broadband refers to a general-purpose comPXQLFDWLRQVWHFKQRORJ\UDWKHUWKDQRQHVSHFL¿F technological solution; therefore, common features of broadband additionally include the reliance on some form of packet switching and superiority to traditional networks in terms of latency and jitter (Kim et al.; Best practice options, 2004). Another approach to the broadband notion suggests that broadband is the ability of a single access line or wireless or satellite link, connected
to telecommunications network, to provide support for a fast, always-on access to digital content, applications, and a range of services (Best practice options, 2004), some or all of which can occur simultaneously (Inquiry, 2004). Flournoy (2004) argues that broadband is a two-way communication FRQFHUQHGZLWK³ODVWPLOH´DQG³¿UVWOLQH´DVZHOO In Canada, broadband refers to two-way access to a variety of services via a high-speed connection to the public data network and/or Internet (Understanding broadband, 2003). Systematic analysis of the approaches to broadband shows that broadband is analyzed as an access rather than a transmission technology. There are other aspects important for full understanding of the broadband concept, such as addressability, different applications and content, sophisticated network design/architecture. The main characteristics to be included into the notion of broadband are shown in Figure 1. Following the discussion on broadband concept above, we conclude that broadband is to be viewed as a local access link performance, which does not OLPLWXVHU¶VFDSDELOLW\WRUXQWRGD\¶VDSSOLFDWLRQV delivered over public data networks which encourages the development of new applications. :KLOH¿QLVKLQJWKLVGLVFXVVLRQRQEURDGEDQG notion, it is important to mention technologies for broadband provision. Fundamentally, there are two groups of technologies—wired and wireless
Figure 1. Layout of broadband technological aspects
Access to public data networks (the Internet)
Network design & architecture
Access applications & content
Addressability
Always-on
Bi-directional communication
Latency & Jitter
Speed
BROADBAND
Composed by the authors
779
ICT Statistics for Broadband Promoting Regulatory Policy
technologies. Wired technologies include xDSL, FDEOH PRGHP ¿EHU RSWLFV )77[ DQG SRZHU line communications (PLC) (Point, Skouby, & Van Daele, 2004). Wireless technologies propose greater variety, including the following groups: wireless local area network, metropolitan area network, wide area network, cellular world wide area network technologies, satellite, and high altitude platform stations (Flournoy, 2004; Point et al., 2004).
STATISTICAL ANALYSIS OF ICT INDICATORS General Remarks on Broadband Development In 2003, there were 98.8 million broadband subscribers in the world (Informatization, 2004). In July, 2006, the worldwide number passed 247.1 million broadband lines (World Broadband Statistics, 2006), and 400 million subscribers are forecasted by 2009 (IMS Research, as cited in Telecomredux, 2005; World Broadband Statistics, 2005). As statistical evidences from different sources show, xDSL is the leading platform of choice for broadband, accounting for about 67% of world subscribers; Cable modem follows with 23%, and all other technologies gain as little as 10% of world subscribers; North America is the only region where broadband via cable modem VLJQL¿FDQWO\RYHUFRPHV['6/ :RUOG%URDGEDQG Statistics, 2006). In structural view, xDSL is H[SHFWHG WR JDLQ HYHQ PRUH VLJQL¿FDQW VKDUH LQ future (IMS Research, as cited in Telecomredux, 2005; Burke, 2004). However, in this context it has to be admitted that dependency on xDSL is not a SDQDFHDIRUFRXQWU\¶VEURDGEDQGPDUNHW²['6/ is not available at every single geographical point of a territory (rural, remote, low density areas, etc.) and is mostly provided by the incumbent entities. 7KLV ZD\ WKH JURZLQJ OLRQ¶V VKDUH DFTXLUHG E\ xDSL is to be considered reservedly. In the world, countries have different broadband penetration ranging from 30% (Iceland) to 0% (developing countries) (World Broadband Statistics,
780
2006). 7KHVSHHGRIVXEVFULEHUOLQHV¶JURZWKLQ countries varies much as well. A comparison of “top ten countries” by number of lines (as of June 30, 2006) and by percentage of growth of number of lines (second quarter, 2006 and 2005) indicates, that the biggest markets are not necessarily the fastest growing ones—fast growth is more related to low penetration rate. Five European countries fall in the list of top 10 having the biggest number of broadband lines in the world (Germany, UK, Spain, France, Italy), and four European countries fall in the list of top 10 best performing in broadEDQGOLQHV¶JURZWK LQWKH¿UVWDQGVHFRQGTXDUWHU of 2006) (Germany, UK, Spain, France) (World Broadband Statistics, 2006). Lithuanian growth LQOLQHV¶QXPEHUZDVWKH¿IWKIDVWHVWLQWKHZRUOG LQ¿UVWTXDUWHU :RUOG%URDGEDQG6WDWLVWLFV 2005), and together with Slovenia and Estonia is valuated as driving broadband growth in the whole Eastern Europe region in the second quarter, 2006 (World Broadband Statistics, 2006). The broadband sector is one of the fastest growing sectors in EU since its emergence, and the yearly growth of broadband lines counted up to 41.75% (July 2005 – July 2006) (Broadband access, 2006). Although in relative terms this JURZWKLVORZHUWKDQLQWKHSUHYLRXVPRQWKV¶ period (61.86%), in absolute numbers it is higher than the growth recorded between July 2004 and July 2005 (Broadband access, 2006). Penetration of broadband varies much in EU member states (MS)—the ratio of penetration between highest-penetrated Denmark (29.42%) and lowest-penetrated Greece (2.68%) is 11 times. There might be a number of reasons causing such a difference and Figure 2 deals with determination RISRVVLEOHIDFWRUVLQÀXHQFLQJEURDGEDQGSHQHWUDtion and development.
Determination of Statistically 6LJQL¿FDQW,QÀXHQFHVRQ%URDGEDQG Penetration in European States :H KDYH XVHG D FRUUHODWLRQ FRHI¿FLHQW DV D WHVW LQVWUXPHQW LQ RUGHU WR GHWHUPLQH IDFWRUV LQÀXencing broadband penetration. We have checked broadband penetration against the indicators used
ICT Statistics for Broadband Promoting Regulatory Policy
)LJXUH*URZWKLQEURDGEDQG¿[HGDFFHVVOLQHVLQWKH(XURSHDQ8QLRQ 80000 435% 70000
400% 350%
361%
60000 278% 50000
300% 250% 200%
208%
40000 133% 30000
82%
150% 100%
36%
20000
500% 450%
10000
50% 0%
0 JAN 2003
JUL 2003
JAN 2004
JUL 2004
JAN 2005
JUL 2005
JAN 2006
JUL 2006
Total BB fixed access lines % variation ince January 2003
Source: Broadband access, 2006.
Figure 3. European Union broadband penetration rate as of July 1, 2006 35 30 25 20 15 10
Denmark
Netherlands
Finland
Sweden
Belgium
UK
France
Luxemburg
Estonia
Austria
Spain
Italy
Portugal
Malta
Slovenia
Czech Rep
Lithuania
Ireland
Hungary
Latvia
Cyprus
Slovakia
Poland
Greece
0
Germany
5
Source: Broadband access, 2006
by international organizations to evaluate ICT development with the presumption that similar IDFWRUVPLJKWKDYHLQÀXHQFHIRUEURDGEDQGSHQHWUDtion as for the other communication services. We have limited our research to 25 EU member states, taking all the states as a single region (EU25). We have also split the 25 states into two subregions in order to analyze whether there are differences in statistical relations. The EU15 subregion refers to the “old” member states (EU before enlargement of May 1, 2004) and the EU10 subregion refers to
the “new” member states (those entered the EU on May 1, 2004). Our analysis was horizontal as well as vertical (i.e., we have made spread in countries and dynamics in time), and serves as an evaluation of liaisons between broadband penetration and different ICT development parameters for the whole region. We take note that the statistical data used comes for the period since 2001 December 31 to 2006 July 1, 2006, while broadband penetration statistics was used as of July 1, 2006, July 1, 2005, and July 1, 2004.
781
ICT Statistics for Broadband Promoting Regulatory Policy
Figure 4. Assessment of relation of categories of indicators to broadband penetration by average rankLQJRIFRUUHODWLRQFRHI¿FLHQWV 5
EU10, rank
EU15, rank
EU25, rank
4 3 2 1
EQP 04 EQP 05 EQP 06 INFR 04 INFR 05 INFR 06 MRKT 04 MRKT 05 MRKT 06 HUM 04 HUM 05 HUM 06 ENV 04 ENV 05 ENV 06 EQP 04 EQP 05 EQP 06 INFR 04 INFR 05 INFR 06 MRKT 04 MRKT 05 MRKT 06 HUM 04 HUM 05 HUM 06 ENV 04 ENV 05 ENV 06 EQP 04 EQP 05 EQP 06 INFR 04 INFR 05 INFR 06 MRKT 04 MRKT 05 MRKT 06 HUM 04 HUM 05 HUM 06 ENV 04 ENV 05 ENV 06
0
Calculated and composed by the authors
We have performed a desk research of statistics on ICT development and logical analysis while choosing the indicators of ICT development to be used in our research. Subsequently, 77 ICT LQGLFDWRUV ZHUH JURXSHG LQWR ¿YH FDWHJRULHV L equipment indicators showing the penetration of different end user devices; (ii) infrastructure indicators, showing the legacy infrastructure in an economy; (iii) market indicators, showing the SRWHQWLDORIWKHQDWLRQ¶VHFRQRP\DQG,&7PDUket; (iv) human capabilities indicators, showing the general and e-literacy of population; (v) and environment indicators, showing the context of broadband development in a particular economy. A full list of the grouped ICT indicators used for our analysis is provided in Appendix 1. We have taken three directions to determine FRUUHODWLRQ FRHI¿FLHQWVRI(8UHJLRQLQRUGHUWR GHWHUPLQHWKHPDLQLQÀXHQFLQJIDFWRUVDQGFRHI¿FLHQWVRI(8VXEUHJLRQDQG(8VXEUHJLRQLQ order to compare and determine whether there are GLIIHUHQFHVRILQÀXHQFHEHWZHHQVRFDOOHG³ROG´ and “new” member states (detailed data on correlation is provided in Appendix 1). Every indicator in every category was ranked from 5 to 1, subject to how high the correlation of broadband to a particular indicator was: correla-
782
WLRQFRHI¿FLHQWDERYHGHOLYHUHGUDQNWRWKH LQGLFDWRU XQGHU TXHVWLRQ FRUUHODWLRQ FRHI¿FLHQW between 0.65 and 0.85 delivered rank 4 to the inGLFDWRU FRUUHODWLRQFRHI¿FLHQWEHWZHHQDQG 0.65 delivered rank 3 to the indicator; correlation FRHI¿FLHQWEHWZHHQDQGGHOLYHUHGUDQN WRWKHLQGLFDWRU DQGFRUUHODWLRQFRHI¿FLHQWEHORZ 0.2 delivered rank 1 to the indicator. This ranking was performed in accordance to Kardelis (2002) V\VWHPRIFRUUHODWLRQVLJQL¿FDQFHQDPHO\³YHU\ KLJK´ FRUUHODWLRQ VLJQL¿FDQFH LV LQ WKH FDVH LI FRUUHODWLRQFRHI¿FLHQWLVDERYH³+LJK´FRUUHODWLRQ VLJQL¿FDQFH PHDQV WKDW WKH FRUUHODWLRQ FRHI¿FLHQW LV EHWZHHQ DQG ³0LGGOH´ FRUUHODWLRQ VLJQL¿FDQFH PHDQV WKDW WKH FRUUHODWLRQFRHI¿FLHQWLVEHWZHHQDQG³/RZ´ FRUUHODWLRQVLJQL¿FDQFHPHDQVWKDWWKHFRUUHODWLRQ FRHI¿FLHQWLVEHWZHHQDQG³9HU\ORZ´ FRUUHODWLRQVLJQL¿FDQFHPHDQVWKDWWKHFRUUHODWLRQ FRHI¿FLHQWLVEHORZ7KHDVVHVVPHQWRIDFDWHJRU\¶VWRWDOUDQNZDVPDGHDVDVLPSOHDYHUDJH mean of ranks of indicators within the category under question. 7KH LQWHUSUHWDWLRQ RI WKH ¿QGLQJV RI WKLV UHsearch is as follows. In EU25 member states, the PRVWVLJQL¿FDQWFRUUHODWLRQRIEURDGEDQGSHQHWUDtion is with equipment category of indicators. PC
ICT Statistics for Broadband Promoting Regulatory Policy
penetration in households and PC equipped employees in all enterprises are the most correlated to broadband penetration; however, PC-equipped employees in different types of enterprises correlate to broadband penetration almost the same OHYHO 1RWHZRUWK\ WKH EURDGEDQG SHQHWUDWLRQ¶V correlation of PC penetration shows growing tendency over the years of investigation. It means WKDW SHRSOHV¶ VWUDLQ WR XVH D SDUWLFXODU VHUYLFHV depends on technical means at their disposal, and QDPHO\SHQHWUDWLRQRIEURDGEDQGLVLQÀXHQFHGE\ penetration of end user means to get the service. 7KHVHFRQGPRVWVLJQL¿FDQWFRUUHODWLRQRIEURDGband penetration is with the human capabilities category of indicators; the strongest relationship is determined between broadband and attainment of people to formal education, that is, 25-64 year olds participating in education and training, 25-64 year olds with at least upper secondary education, and 25-64 year olds with tertiary attainment and individuals using Internet for formalized education activities (school, university, etc.). Again, relationship between broadband and human capabilities indicators gets stronger over the years. It means that broadband penetration is highly dependent RQHGXFDWHGVRFLHW\ZKLFKH[SHULHQFHVEHQH¿WV IURPXVLQJWKHVHUYLFH7KHWKLUGPRVWVLJQL¿FDQW correlation of broadband penetration is with the environmental category of indicators—number of Internet hosts (means a variety of content available to users), research and development expenditure (means searching for innovative technologies and applications), ICT expenditure (on IT hardware, equipment, software, services), which means acquiring means to use more sophisticated services, e-government supply and e-government demand as well as e-commerce supply and demand. Summarizing, favorable environment to provide and to receive electronic communications services, in particular environment-facilitating demand for broadband as well as user understanding of technological achievements, is the group of factors sigQL¿FDQWO\LQÀXHQFLQJWKHEURDGEDQGSHQHWUDWLRQ Among the infrastructure category of indicators, WKHPRVWVLJQL¿FDQWDUHLQWHUQDWLRQDOEDQGZLGWK per inhabitant (high relation) and international bandwidth per economy (medium relation). These
mean potential and actual technical quality of broadband services in terms of speed; however both are interdependent with broadband penetration. Indicator of investments into infrastructure shows that amount of investments per subscriber matters, but amount of investment per revenue does not. Noteworthy, technological variety of broadband does not have a strong relation to broadband penetration in EU25 region (according xDSL as % of all subscriptions), but we think this is due to VLJQL¿FDQWGLIIHUHQFHVEHWZHHQ³QHZ´DQG³ROG´ European countries with regard to broadband development stage. It seems that the importance of technological variety increases with the growth of broadband penetration. Therefore, greater technological variety from the beginning of broadband development gives more impetus to the growth of penetration in later stages of the development. This means that from a technological point of view, penetration of broadband is related to the quality of individual service to subscriber. In the set of market indicators, the strongest relation to revenues LQWHOHFRPPDUNHWDQGHFRQRP\¶V*'3 LQGLFDWRUV of wealth), prices for wholesale leased lines of high capacity (necessary to service providers without their own networks) are of medium importance in EU25. Summarizing, the positive general trends of economy development and the market developPHQWDUHQHFHVVDU\EXWQRWVXI¿FLHQWIDFWRUVIRU broadband development. In EU15 member states, equipment and human capacity categories are the ones to be mostly emSKDVL]HG,QWKH¿UVWVHWSHUFHQWDJHRIKRXVHKROGV with home PCs has the strongest relation to broadband penetration. These categories are followed by environment set of indicators and infrastructure set of indicators, but market set of indicators does not VHHPWREHVLJQL¿FDQWO\LQÀXHQWLDORYHUEURDGEDQG penetration. In EU10 member states, indicators in environmental set has the highest rank. Among the indicators, e-government and e-commerce indicators show the highest correlation, followed by R&D expenditure and ICT expenditure indicators. Infrastructure and market indicators are of similar VLJQL¿FDQFH,QWKHVHFDWHJRULHV,QWHUQHWSHQHWUDtion, total teledensity, international bandwidth
783
ICT Statistics for Broadband Promoting Regulatory Policy
(infrastructure) and telecom revenue (market) could EHPHQWLRQHG,QWKHKXPDQFDSDFLWLHVLQGLFDWRUV¶ group, the highest relationship is observed to individuals taken computer courses, and people involved in tertiary education. Noteworthy, huPDQFDSDFLWLHV¶LQGLFDWRUVVKRZWKHGLPLQLVKLQJ relationship over time to broadband penetration in EU10. In the Equipment category, which shows the weakest relationship to broadband penetration in EU10, the noticeable relationship is observed to PC equipped inhabitants, but not to PC equipped employees.
IDENTIFICATION OF BASIS FOR REGULATORY POLICY ACTIONS 7KH¿QGLQJVRIUHVHDUFKVKRZWKDWWKHUHLVYDULRXV intensity of interrelations between different factors against broadband penetration. The very concept of WKHFRUUHODWLRQFRHI¿FLHQWSURYLGHVWKDWWKHKLJKHU WKHFRHI¿FLHQWLVWKHVWURQJHULQÀXHQFHRIDIDFWRU on the observed phenomenon might be expected, EXWH[LVWHQFHRILQÀXHQFHper se is determined by logic—that is by deciding, if a particular factor can KDYHLQÀXHQFHRQWKHREVHUYHGSKHQRPHQDRUWKH correlation is not more than coincidence. So we KDYHFKHFNHGHYHU\IDFWRUSUHVXPDEO\LQÀXHQFLQJ the broadband, against criteria of logic and r2. As regards logic, we have checked whether the particuODUIDFWRUFDQLQÀXHQFHRUZKHWKHULWLVLQÀXHQFHG by the broadband penetration. As regards the r2 criteria, we have selected the factors, presumably causing 25% or more of the broadband penetration LH FRUUHODWLRQ FRHI¿FLHQW LV DQG KLJKHU )LQDOO\ZHJRWIDFWRUVLQÀXHQFLQJEURDGEDQG and 7 more factors with bidirectional relations to the broadband. Based on these factors we have derived several guidelines for regulatory policy encouraging the broadband development. 7DNLQJWKHLGHQWL¿HGIDFWRUVKDYLQJXQLGLUHFWLRQDODQGELGLUHFWLRQDOLQÀXHQFHRYHUEURDGEDQG penetration, the guidelines for regulatory policy promoting broadband development can be grouped into four broad domains: regulatory policy action lines at the society level, regulatory policy action lines at the household level, regulatory policy ac-
784
tion lines at the market level (wide approach), and regulatory policy action lines at the market level (narrow approach). Regulatory policy action lines at the society level include: (i) to implement the computerization of schools and other establishments of education and training; (ii) to promote the life-long learning and education style of personal development; (iii) with these and presumably other action lines, to create the social-cultural environment, favorable to the growth of broadband demand. Early and overall computerization of education establishments, provision of relevant software and training ZLOOVHUYHLQGLUHFWLRQV ¿UVWWKHFDSDELOLWLHVRI people to use PC and e-environment will be formed; second, habits of people to use e-environment for different purposes will be molded, and people will be ready for more sophisticated applications; third, e-environment will be a natural part of everyday life. The life-long learning system will encourage people to use different sources of information, including the inexhaustible sources on the global network; usage of network on narrowband access would turn to usage of network on broadband access due to natural demand to get better services with time. All this will inspire a networked, information society, where broadband access is a must. Regulatory policy action lines at the household level include: (i) to provide motivation for people to acquire broadband related technical means; (ii) to encourage people to connect to the Internet using different motifs; (iii) to aggregate purchasing of broadband by the state. Encouraging people to acquire the technical means to get broadband services directly to the home would lower the initial barriers to become a broadband subscriber. For example, the Lithuanian initiative to allow deduction of income tax, if PC, software and / or Internet access is DFTXLUHGLVDJRRGH[DPSOHRISURYLGLQJ¿QDQFLDO incentives important for broadband penetration. Noteworthy, not every place or region of a state can be connected to network with broadband access under pure business incentives; here aggregated purchases by state for its establishments in the region might serve as a factor lowering the initial price of connection for private entities (when the main installment is purchased by the state for state
ICT Statistics for Broadband Promoting Regulatory Policy
Table 1. The most important factors and suggested regulatory policy action line
The most important factors interrelated with broadband penetration
Y/B1 logic
r2
Proposed regulatory action line
analysis
PC, per 100 inhabitants
Y
56.3%
Motivating people to acquire broadband related
PC, per 100 households
Y
68.4%
technical means
PC equipped employees, small enterprises
Y
49.6%
PC equipped employees, medium enterprises
Y
59.3%
Motivating business to acquire broadband
PC equipped employees, large enterprises
Y
57.9%
related technical means
PC equipped employees, all enterprises
Y
60.8%
International bandwidth, per inhabitant
B
60.1%
Internet penetration, per 100 inhabitants
B
48.9%
Internet penetration, per 100 households
B
74.5%
B
77.8%
B
27.1%
16-74 year olds individuals regularly using the Internet, once a week, within the last 3 month Internet penetration, small enterprises Fixed penetration, per 100 inhabitants
Y
26.0%
Telco revenues, USD per subscriber
B
32.2%
Telecom investment, USD per inhabitant
B
25.7%
GDP, $ per capita
Y
44.2%
25-64 year olds participating in education and training, %
Y
51.2%
25-64 year olds with at least upper secondary education, %
Y
52.6%
25-64 year olds, tertiary attainment, %
Y
46.6%
Y
35.2%
Individuals using Internet for formalized education activities (school, university, etc.), % Enterprises using e-learning applications for training and
Encouraging people to connect to Internet; Favorable general environment to broadband market development and demand growth Encouraging business to connect to Internet Promoting infrastructure competition; Promoting services competition Promoting competition Promoting competition and the legacy Infrastructure provision 1RQHVSHFL¿F *HQHUDOPDFURHFRQRPLFSROLF\
Promoting the life-long learning and education; Computerization of schools and other places of education and training
Y
28.8%
Internet hosts (total)
Y
38.6%
Internet hosts (per 10.000 inhabitants)
Y
61.0%
e-government usage (demand side)
Y
61.2%
e-government availability (supply side)
Y
30.9%
Y
61.9%
Y
60.5%
education of employees, small and medium enterprises, %
Promoting competition
Promoting content development Promoting e-government usage and egovernment supply (content); Provision of e-
ICT expenditure, IT hardware, equipment, software, services, % of GDP R&D expenditure, as % of GDP
government services Support programs dedicated to development of new technologies, applications, safety; Promoting content development
<±WKHIDFWRULQÀXHQFHVWKHEURDGEDQGSHQHWUDWLRQ %±WKHIDFWRUDQGWKHEURDGEDQGSHQHWUDWLRQKDYHWKH ELGLUHFWLRQDOLQÀXHQFH
1
785
ICT Statistics for Broadband Promoting Regulatory Policy
Table 1. (cont.) Enterprises received orders on-line over the last calendar year, % of small enterprises Enterprises received orders on-line over the last calendar year, % of medium enterprises Enterprises received orders on-line over the last calendar year, % of large enterprises Enterprises purchased on-line over the last calendar year, % of large enterprises Individuals ordered goods or services over the Internet, for private use, in the last 3 months
Y
44.8%
Y
44.9%
Y
33.5%
Y
26.7%
Y
56.3%
Favorable general environment to e-commerce market development and e-commerce demand growth
Composed by the authors
establishment (hospital, library, etc.), acceding private users would get better prices). Regulatory policy action lines at the market level (wide approach) include: (i) to promote content development; (ii) to ensure full range of e-governPHQWVHUYLFHV¶VXSSO\ LLL WRLQLWLDWHDQGVXSSRUW programs dedicated to develop new technologies, applications, safety. In order to use broadband applications, rich content is a must—including but not limited to e-government, e-health, e-libraries, electronic commerce, interactive entertainment, and so forth. National content is the most important factor here. Research and development (R&D) is an inherent part of modern applications; so if a state wants to be a developer and exporter of services, it must turn to invention processes. The security and trust in usage of e-environment is a prerequisite to broadband penetration, so we see the role of regulatory policy in promoting the esecurity culture. Regulatory policy action lines at the market level (narrow approach) include: (i) to promote competition in the market; (ii) to promote the infrastructure competition, (iii) to promote legacy infrastructure provision, (iv) with these action lines and others, to create the political-regulatory environment, favorable to the growth of broadband market and to development and investments into new forms of rich content. Competition is the choice bringing to users better choices in terms of better value-for-money and ICT market is not
786
an exclusion. Here, competition means two asVHUWLRQV ¿UVWFRPSHWLWLRQLQWKHPDUNHWDPRQJ service providers, second, infrastructure-based competition. Usage of legacy infrastructure is not of less importance—traditionally telecoms own the well developed areas of infrastructure, the provision of which will attract new comers and the market development. However, opening of access to infrastructure should not undermine the importance of ensuring infrastructure-based competition in the long run. All this requires a predictable, transparent, and nondiscriminatory regulatory environment and the developed system of decision-making, which is oriented to long run promotion of investment into infrastructure and content development.
CONCLUSION: GUIDELINES FOR REGULATORY POLICY The importance of the broadband phenomena cannot be overestimated and it stands high on the agenda of the most important international forums as well as national policy agendas. In contrast to academic research practice (mostly employing case study analyses for proposing roles of government to promote broadband), the public policy actors develop and reason their policies relying on broadband statistics, which is used mainly for comparison and illustrative purposes. In this chapter, we
ICT Statistics for Broadband Promoting Regulatory Policy
attempt to bring down the rhetoric of statistics to the operational level and propose a way in which ICT statistics can be employed to derive possible suggestions for practical solutions in the domain of broadband promoting regulatory policy. The analysis of different approaches to broadband notion has led us conclude that broadband is to be viewed as a local access link performance, ZKLFKGRHVQRWOLPLWXVHUV¶FDSDELOLW\WRUXQWRGD\¶V applications delivered over public data networks and encourages the development of new ones. From a structural point of view, xDSL broadband is prevailing worldwide and is expected to gain an even bigger market share than it has now. However, the growing share of xDSL connections has twofold consequences—while increasing the number of connected users, due to its technological features xDSL might contribute to the enlargement of digital divide—and therefore has to be regarded deliberately. Statistical analysis has shown that penetration of broadband in the enlarged EU25 is mostly related to the favorable environment with regard WRHQGXVHUV¶HTXLSPHQWDQGKXPDQFDSDELOLWLHV In the “old” member states, broadband correlation with equipment settings and human capacity are RIVLPLODUVLJQL¿FDQFHDVLQWKHZKROHHQODUJHG EU, while favorable environment has the highest correlation to broadband in the “new” member states. However, we think that “new” member VWDWHVPLJKWH[SHFWWKHVDPHIDFWRUVWRLQÀXHQFH positively their broadband in longer run as in “old” member states due to time lag and broadband penHWUDWLRQ,QWKLVFRQWH[WEDVHGRQRXU¿QGLQJVZH argue that special attention shall be given to create a favorable environment to provide broadband services and create demand, including demand for appropriate devices; however, solely technological solutions and commercial supply must be supported E\SHRSOH¶VDELOLWLHVDQGFOHDUPDWHULDOLQWHUHVWLQ use broadband access. 2QWKHEDVLVRI¿QGLQJVRIVWDWLVWLFDODQDO\VLV of ICT indicators, we propose some general guidelines for regulatory actions to foster broadband: (i) improvement of political-social environment, including investment into research and science, (ii) raising of human capabilities, including full
and early computerization of schools and lifelong learning encouragement, (iii) promotion of demand for broadband through rich content and affordability of relevant devices, and (iv) building of competition-stimulating regulatory environment, with special emphasis on infrastructure based competition.
REFERENCES Access indicators for the information society. (2003). World Telecommunication Development Report. Geneva: International Telecommunications Union. The always-on pervasive Internet: Why broadband means more than bits (n.d.). Research paper. San Francisco, CA: Harbor Research, Inc. Best practice options for improving and extending access to electronic communications in Lithuania. (2004). Rotterdam: World Bank, ECORYS, TNO. Broadband access in the EU: situation at 1 July 2006. (2006). Communications Committee Working Document (Unpublished report). Brussels. Broadband and the Internet (n.d.). Retrieved July 29, 2007, from http://itu.int/ osg/spu/ni/broadband/ background.html Burke, D. (2004, December). Broadband in Latin America. Paper presented at DSL Forum, Orlando, Florida. Retrieved July 29, 2007, from http://ww.dslforum.ord/PressRoom/orlando_summit_burke.pdf Choudrie, J., & Dwivedi, Y.K. (2005). Investigating the research approaches for examining technology adoption issues. Journal of Research Practice, 1(1), Article D1. Retrieved July 29, 2007, from http://jrp. icaap.org/index.php/jrp/article/view/4/7 Connecting Europe at high speed: National broadband strategies. (n.d.). Communication from the Commission to the Council, the European Parliament, the European Economic and Social Committee and the Committee of Regions). Retrieved
787
ICT Statistics for Broadband Promoting Regulatory Policy
July 29, 2007, from http://ec.europa.eu/information_society/eeurope/2005/doc/all_about/broadband/com_broadband_en.doc Electronic Industries Alliance. (2004). Broadband. Retrieved July 29, 2007, from http://www.eia. org/new_policy/ broadband.phtml European electronic communications regulation and markets 2003. (n.d.). Ninth Report. Retrieved July 29, 2007, from http://ec.europa. eu/information_society/policy/ecomm/implementation_enforcement/annualreports/9threport/index_en.htm European electronic communications regulation and markets 2004.(n.d.). Tenth Report. Retrieved July 29, 2007, from http://ec.europa. eu/information_society/policy/ecomm/implementation_enforcement/annualreports/10threport/index_en.htm European electronic communications regulation and markets 2005. (n.d.). Eleventh Report. Retrieved July 29, 2007, from http://ec.europa. eu/information_society/policy/ecomm/implementation_enforcement/annualreports/11threport/index_en.htm Eurostat database, 2004-2006. (n.d.). Retrieved July 29, 2007, from http://epp.eurostat.cec.eu.int/ Federal communications commission releases data on high-speed Internet access services (n.d.). Federal Communications Commission News. Retrieved July 29, 2007, from http://www.fcc. gov/Bure aus/Common_Carrier/Reports/FCCState_Link/IAD/hspd 1204.pdf Flournoy, D.M. (2004). The broadband millenium: Communication technologies and markets. Chicago, IL: International Engineering Consortium. Gillett, S.E., Lehr, W.H., & Osario, C. (2004). Local government broadband initiatives. Telecommunications Policy, 28(7/8), 537-558. Global broadband booms to 150mn users: New IMS data says VDSL will be stronger later (n.d.). Telecomredux news release. Retrieved July 29, 2007, from http://www.telecomredux.com/index2. php?opti on=com_content&do_pdf=1&id=788
788
Household broadband access in Australia (n.d.). Research note. Canberra, Australia: Information and Research Services. Retrieved July 29, 2007, from from http://www.aph.gov.au/library/pubs/ rn/2001-02/02rn34.pdf Informatization: Broadband IT Korea. (2004). Seoul: National Computerization Agency. Inquiry into competition in broadband services. (2004). Report. Canberra, Australia: The Senate, Environment, Communications, Information Technology and the Arts References Committee. ITU statistical database. (n.d.). Retrieved July 29, 2007, from http://www.itu.int/ITU-D/ict/statistics/index.html Kardelis, K. (2002). &0RNVOLQLǐ W\ULPǐ PHWodologija ir metodai. Kaunas: Judex. Kim, J.H., Bauer, J.M., & Wildman, S.S. (2003, Setember). Broadband uptake in OECD countries: Policy lessons from comparative statistical analysis. Paper prepared for presentation at the 31st Research Conference on Communication, Information and Internet Policy, Arlington, Virginia. Retrieved July 29, 2007, from http://tprc.org/papers/2003/203/Kim-Bauer-Wildman.pdf Melody, W.H. (2003, July). Broadband development in Ireland. Paper presented to Joint Oireachtas Committee on Communications, Marine and Natural Resources, Dublin, Ireland. OECD fact book 2006 - economic, environmental and social statistics. (n.d.). Retrieved July 29, 2007, from http://oberon.sourceoecd.org/vl=7329999/ cl=11/nw=1/rpsv/factbook/data/02-01-01-t02.xls OECD statistics database for information and communications technology. (n.d.). Retrieved July 29, 2007, from http://www.oecd.org/topicstatsportal/ 0,2647,en_2825_495656_1_1_1_1_1,00.html Point, J.C., Skouby, K.E., Van Daele, P. (Eds.). (2004). First combined report on the multitechnological and multidisciplinary analysis of the broadband for all concept (Report). Brussels, Belgium: BREAD.
ICT Statistics for Broadband Promoting Regulatory Policy
Rytel, T. (2001). ,QIRUPDFLQơVYLVXRPHQơVSOơWUD Japonijoje: Y.Masudos vizija. Vilnius: Mokslo aidai. Tadayoni, R., & Sigurdsson, H.M. (in press). Development of alternative broadband infrastructures—case studies from Denmark. Telematics and Informatics. Retrieved July 29, 2007, from http://www.viskan.net/publications/061031_TogI_ draft_Halldor.pdf Understanding broadband wireless access. (n.d.). Report. Ottawa, Canada: Industry Canada Broadband Technical Resource Team. United Nations statistical database. (n.d.). Retrieved July 29, 2007, from http://www.un.org/popin/functional/statistics.html van Winden, W., & Woets, P. (2004). Urban broadband internet policies in Europe: A critical review. Urban Studies, 41(10), 2043-2059. :RUOGEURDGEDQGVWDWLVWLFV 4&RPPHQWV PHWKRGRORJ\DQG¿JXUHV (n.d.). Report. London: Point Topic Ltd. World broadband statistics: Q2 2006. Comments, PHWKRGRORJ\DQG¿JXUHV (n.d.). Report. London: Point Topic Ltd. World telecommunication development report: Access indicators for the information society. (n.d.). Geneva, Switzerland: International Telecommunication Union.
Yearbook of Statistics, Telecommunication Services, Chronological Time Series 1994-2003. (n.d.). Geneva, Switzerland: International Telecommunication Union.
KEY TERMS Broadband: A local access link performance ZKLFK GRHV QRW OLPLW D XVHU¶V FDSDELOLW\ WR UXQ WRGD\¶V DSSOLFDWLRQV GHOLYHUHG RYHU SXEOLF GDWD networks and which encourages the development of new applications. Regulatory Policy: Any government intervention in the market, particularly to limit or control behavior of market actors. Correlation: The simultaneous change in value of two numerically valued random variables. &RUUHODWLRQ&RHI¿FLHQW &RHI¿FLHQWLQGLFDWLQJ the strength and direction of a linear relationship between two random variables. xDSL: Family of technologies that provide highspeed digital data transmission over the PSTN. Cable Modem: Devices allowing high-speed access to the Internet via a cable television network. Penetration: The degree to which a population adopts a particular service (device), usually estimated as number of services (devices) per 100 inhabitants or per 100 households.
789
ICT Statistics for Broadband Promoting Regulatory Policy
APPENDIX &RUUHODWLRQFRHI¿FLHQWVRIEURDGEDQGSHQHWUDWLRQLQDQGDJDLQVWGLIIHUHQWVWDWLVWLFDO LQGLFDWRUVLQ(8(8DQG(8UHJLRQV &RUUHODWLRQFRHI¿FLHQWRIEURDGEDQGSHQHWUDWLRQDJDLQVW«
EU25
EU15
EU10
-0,033
-0,500 -0,585
T ECH NOLOGY/ EQU I PM EN T GROU P I N DICATOR S 2003 … TV equipped households, %
… PC equipped households, %
… PC equipped inhabitants, %
… PC equipped employees in Small enterprises, %
… PC equipped employees in Medium enterprises, %
… PC equipped employees in Large enterprises, %
… PC equipped employees in all enterprises, %
0,115
2002
-0,024
-0,115
2001
-0,006
-0,006
n.a.
2005
0,827
0,841
0,207
2004
0,762
0,770
0,183
2003
0,846
0,846
n.a.
2005
0,750
0,681
0,440
2004
0,768
0,680
0,640
2003
0,687
0,505
0,795
2005
0,704
0,696
0,383
2004
0,690
0,679
0,291
2003
0,460
0,460
n.a.
2005
0,770
0,746
0,306
2004
0,754
0,752
-0,65
2003
0,643
0,643
n.a.
2005
0,761
0,730
0,373 0,119
2004
0,754
0,705
2003
0,573
0,573
n.a.
2005
0,780
0,737
0,343
2004
0,774
0,776
0,008
2003
0,703
0,703
n.a.
0,405
I N F R AST RUCTU R E GROU P I N DICATOR S
… Internet penetration, % of inhabitants
… Internet access having households, %
… Internet access having small enterprises, %
… Internet access having medium enterprises, %
… Broadband access having households, %
790
2005
0,699
0,765
2004
0,703
0,810
0,718
2003
0,714
0,705
0,705
2005
0,863
0,835
0,504
2004
0,693
0,726
0,206
2003
0,785
0,785
n.a.
2005
0,521
0,430
0,315
2004
0,648
0,438
0,686
2003
0,377
0,377
n.a.
2005
0,155
0,142
-0,154
2004
0,413
0,214
0,190
2003
0,283
0,283
n.a.
2005
0,952
0,972
0,737
2004
0,925
0,935
0,790
2003
0,977
0,977
n.a.
ICT Statistics for Broadband Promoting Regulatory Policy
(cont.)
… Broadband access having small and medium enterprises, %
… Broadband access having large enterprises, %
… Total telephone penetration, % of inhabitants
… Fixed telephony penetration, % of inhabitants
… Mobile telephony penetration, % of inhabitants
… Mobile telephony penetration growth, % points
… Cable TV penetration, % of households
… Satellite TV penetration, % of households
2005
0,597
0,643
0,571
2004
0,516
0,459
0,541
2003
0,680
0,680
n.a.
2005
0,597
0,643
0,571
2004
0,516
0,459
0,541
2003
0,680
0,680
n.a.
2005
0,487
0,061
0,612
2004
0,532
0,089
0,613
2003
0,466
0,019
0,297
2005
0,510
0,140
0,384
2004
0,565
0,267
0,418
2003
0,508
0,207
0,239
2005
0,173
-0,058
0,340
2004
0,360
0,082
0,412
2003
0,371
-0,066
0,287
2004-2005
-0,361
-0,349
-0,100
2003-2004
-0,159
0,141
-0,075
2002-2003
-0,262
0,106
-0,179 0,293
2003
0,371
0,478
2002
0,430
0,493
0,357
2001
0,361
0,441
0,064
2003
0,236
-0,218
0,891
2002
-0,012
-0,099
-0,144
2001
-0,080
-0,003
-0,094
2003
0,134
-0,081
-0,516
2002
0,130
-0,105
-0,523
2004
0,442
0,273
-0,445
… Terrestrial TV penetration, % of households
… International bandwidth, total, Mbps
… International bandwidth, bps per inhabitant
… xDSL as % of all broadband subscriptions
2003
0,432
0,240
-0,302
2002
0,479
0,155
-0,766
2004
0,775
0,744
0,588
2003
0,771
0,738
0,193
2002
0,826
0,783
-0,285
2006
-0,070
-0,494
-0,379
2005
-0,118
-0,511
-0,504
2004
-0,290
-0,508
-0,350
2003
0,228
-0,028
-0,411
2002
0,201
-0,144
-0,472
2003
0,507
0,364
0,198
2002
0,610
0,378
0,181
2003
0,469
0,461
0,022
… Telecom investment, total, mln. USD
… Telecom investment, USD per inhabitant
… Telecom investment, USD per subscriber 2002
0,426
0,114
-0,015
2003
-0,025
0,285
-0,225
2002
-0,163
0,088
0,129
… Telecom investment, as % of revenue
791
ICT Statistics for Broadband Promoting Regulatory Policy
(cont.) M A R K ET/ ECONOM Y DEV ELOPM ENT GROU P I N DICATOR S
… GDP per capita, USD
… GDP per capita, USD, PPP
2005
0,665
0,433
2004
0,591
0,279
0,269 0,210
2003
0,600
0,245
0,068
2005
0,520
0,183
0,277 0,355
2004
0,511
0,142
2003
0,429
-0,009
0,159
2003
0,212
-0,080
-0,470 -0,489
… Telecom revenue, total, USD 2002
0,196
-0,075
2003
0,403
-0,306
0,123
2002
0,365
-0,071
-0,287
2003
0,586
0,188
0,463
… Telecom revenue, USD per employee
… Telecom revenue, USD per subscriber 2002
0,582
0,325
-0,180
2003
-0,233
-0,306
0,552
2002
-0,285
-0,198
0,115
2006
-0,263
-0,230
-0,398 -0,552
… Telecom revenue, as % of GDP
… Broadband market share by the incumbent, %
«1XPEHURI¿[HGOLQHRSHUDWRUVRIIHULQJVHUYLFHV
… Number of mobile network operators offering services, #
«0RQWKO\WRWDOFRVWSHUIXOOXQEXQGOHGORFDOORRS¼
«0RQWKO\WRWDOFRVWSHUVKDUHGDFFHVV¼
«
«
… Costs of Internet connection, $ per 20 hours
2005
-0,373
-0,410
2004
-0,163
-0,154
-0,213
2005
0,277
0,301
-0,377
2004
0,388
0,379
-0,351
2003
0,358
0,189
-0,256
2005
0,443
0,511
0,083
2004
0,287
0,272
0,010
2003
0,330
0,158
0,146
2005
-0,218
0,062
-0,346
2004
-0,215
-0,147
-0,189
2003
-0,143
-0,143
n.a.
2005
-0,391
-0,148
-0,385
2004
-0,376
-0,221
0,080
2003
-0,211
-0,211
n.a.
2005
-0,211
0,231
-0,172
2004
-0,184
0,474
-0,572
2003
-0,304
0,169
-0,580
2005
-0,095
-0,324
0,797
2004
-0,356
-0,465
-0,526
2003
-0,399
-0,582
-0,363
2003
-0,193
-0,385
-0,178
H U M A N C A PA B I L I T I E S G R O U P I N D I C A T O R S
… 25-64 year olds participating in education and training, %
792
2005
0,722
0,703
0,289
2004
0,695
0,717
0,201
2003
0,657
0,631
0,221
ICT Statistics for Broadband Promoting Regulatory Policy
(cont.)
… 25-64 year olds with at least upper secondary education, men, %
… 25-64 year olds with at least upper secondary education, women, %
… 25-64 year olds, tertiary attainment, %
… Individuals used Internet, in the last 3 months, for formalized education activities (school, university, etc.), %
… Individuals used Internet in the last 3 months for other educational courses UHODWHGVSHFL¿FDOO\WRHPSOR\PHQWRSSRUWXQLWLHV
… Individuals used Internet in the last 3 months for post educational courses, %
… Individuals who have taken a computer course in the last 12 months, %
… Individuals who have never taken a computer course, %
… Enterprises using e-learning applications for training and education of employees, small enterprises, %
… Enterprises using e-learning applications for training and education of employees, medium enterprises, %
… Enterprises using e-learning applications for training and education of employees, large enterprises, %
2005
0,763
0,749
0,294
2004
0,738
0,755
0,276
2003
0,689
0,667
0,210
2005
0,688
0,668
0,273
2004
0,659
0,685
0,145
2003
0,626
0,598
0,222
2003
0,683
0,554
0,110
2002
0,691
0,531
0,475
2001
0,662
0,480
0,946
2005
0,593
0,721
0,195
2004
0,419
0,487
0,634
2003
0,199
0,199
n.a.
2005
0,206
0,170
0,117 -0,149
2004
0,163
0,006
2003
0,068
-0,068
n.a.
2005
0,214
0,556
0,150
2004
0,263
0,004
0,458
2003
0,117
0,117
n.a.
2005
0,273
0,517
0,636
2004
0,687
0,714
0,455
2003
0,663
0,663
n.a.
2005
-0,487
-0,425
-0,248
2004
0,390
0,484
-0,485
2003
0,694
0,694
n.a.
2005
-0,511
-0,446
-0,128
2004
-0,536
-0,438
-0,151
2003
-0,374
-0,374
n.a.
2005
-0,563
-0,539
-0,251
2004
-0,594
-0,521
-0,322
2003
-0,429
-0,429
n.a.
2005
-0,454
-0,464
0,041
2004
-0,523
-0,450
-0,347
2003
-0,456
-0,456
n.a.
EN V I RON M ENT GROU P I N DICATOR S
… R&D expenditure, total, EUR
… R&D expenditure as % of GD
… Telecom equipment exports, USD
2004
0,304
0,071
-0,362
2003
0,288
0,029
-0,392
2002
0,244
-0,214
-0,453
2004
0,778
0,684
0,442
2003
0,758
0,682
0,156
2002
0,711
0,516
0,099
2003
0,383
0,296
0,504
2002
0,379
0,189
-0,014
2001
0,262
0,046
0,472
793
ICT Statistics for Broadband Promoting Regulatory Policy
(cont.)
… Telecom equipment imports, USD
… ICT expenditure, telecom hardware, equipment, software, services, % of GDP
… ICT expenditure, IT hardware, equipment, software, services, % of GDP
…Internet hosts, total
…Internet hosts, per 10000 inhabitants
… 16-74 aged individuals regularly using the Internet, once a week, within the last 3 month
… e-government availability (supply side)
… e-government usage (demand side)
… e-commerce as % of total turnover by enterprises
… Enterprises received orders on-line over the last calendar year, % of small enterprises
… Enterprises received orders on-line over the last calendar year, % of medium enterprises
… Enterprises received orders on-line over the last calendar year, % of large enterprises
… Enterprises purchased on-line over the last calendar year, % of small enterprises
… Enterprises purchased on-line over the last calendar year, % of medium enterprises
… Enterprises purchased on-line over the last calendar year, % of large enterprises
794
2003
0,278
0,104
0,375
2002
0,349
0,063
-0,463
2001
0,338
0,080
-0,528
2005
-0,378
-0,124
0,205
2004
-0,374
-0,125
0,213
2003
-0,344
-0,139
0,370
2005
0,787
0,818
0,362
2004
0,795
0,813
0,307
2003
0,752
0,708
0,094
2005
0,621
0,510
-0,333
2004
0,611
0,495
-0,352
2003
0,515
0,356
-0,425
2005
0,781
0,764
0,490
2004
0,762
0,744
0,427
2003
0,713
0,672
0,598
2005
0,882
0,907
0,640
2004
0,872
0,897
0,519
2003
0,871
0,871
n.a.
2005
0,556
0,343
0,898
2004
0,586
0,222
0,855
2003
0,269
0,269
n.a.
2005
0,782
0,785
0,428
2004
0,809
0,829
0,024
2003
0,855
0,855
n.a.
2005
0,049
-0,215
0,139
2004
0,067
-0,131
0,134
2003
-0,284
-0,284
n.a.
2005
0,669
0,575
0,532
2004
0,710
0,595
0,478
2003
0,370
0,370
n.a.
2005
0,670
0,541
0,466
2004
0,710
0,567
0,501
2003
0,337
0,337
n.a.
2005
0,579
0,506
0,515
2004
0,712
0,582
0,415
2003
0,500
0,500
n.a.
2005
0,448
0,224
0,504 0,830
2004
0,412
0,163
2003
0,366
0,366
n.a.
2005
0,428
0,195
0,449
2004
0,421
0,130
0,822
2003
0,223
0,223
n.a.
2005
0,517
0,215
0,532
2004
0,494
0,194
0,871
2003
0,257
0,257
n.a.
ICT Statistics for Broadband Promoting Regulatory Policy
(cont.) … Individuals ordered goods or services over the Internet, for private use, in the last 3 months
Perceived barriers to e-commerce, “No need”
Perceived barriers to e-commerce, “Too expensive”
Perceived barriers to e-commerce, “Security concerns”
Perceived barriers to e-commerce, “Privacy concerns”
2005
0,750
0,655
0,008
2004
0,759
0,666
-0,010
2003
0,617
0,617
n.a.
2005
-0,072
0,138
0,844
2004
0,083
0,289
0,547
2003
0,302
0,302
n.a.
2005
-0,152
-0,404
0,886
2004
-0,370
-0,148
-0,512
2003
-0,158
-0,158
n.a.
2005
0,315
0,350
0,242 -0,021
2004
0,172
0,238
2003
0,344
0,344
n.a.
2005
0,183
0,189
0,172
2004
-0,142
-0,108
0,030
2003
-0,114
-0,114
n.a.
Calculated and composed by the authors
795
796
Chapter XLVIII
Impact of Broadband VoIP on Telecoms: A Cross-Country Analysis Bardo Fraunholz Deakin University, Australia Chandana Unnithan Deakin University, and La Trobe University, Australia
ABSTRACT VoIP is a technology that has received much attention over the past few years. Speculations are rampant that it will be “the technology” for telecommunications of the future, as broadband gains mass market penetration in every nation. It holds the promise of ubiquity and eliminates the need for a separate infrastructure for telecommunications. In this chapter, we have undertaken a cross country analysis of two economies, Germany and India, at varied levels of broadband voice over internet protocol (VoIP) diffusion, to examine the future potential of this technology in the respective nations and their telecommunications industries. Our brief analysis revealed some valuable insights regarding the impact of VoIP in both economies which may prove to be useful for other economies and telecommunication industries.
Introduction Voice over internet protocol (VoIP) is touted to be a technology that has the potential to change the telecommunications industry across the globe. The basic infrastructure for this technology is the diffusion of broadband into the mass market. In developed nations with high broadband diffusion, VoIP holds the promise of ubiquitous communication, eliminating the need to maintain a separate telecommunications infrastructure. Lately, some reports also seem to suggest that
developing economies such as Nigeria are trying to “leap frog” into new technologies, eliminating the need for building a new infrastructure for telecommunications (Olarunda & Olarunda, 2006). Speculations continue to emerge that VoIP will be an inexpensive and pervasive technology that will eventually eliminate the need for a separate telecommunications infrastructure (Kouroupas, 2006). However, research on the impact of VoIP in developed and developing economies is still emerging and fragmented. This chapter proposes
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Impact of Broadband VoIP on Telecoms
to address this gap in research by taking a closer look at two economies: Germany and India. Germany is a developed nation with a strong telecommunication infrastructure. The telecommunications market is almost completely deregulated (some exceptions such as the recent VDSL legislation still exist) and competitive. Consumer demand and uptake of broadband Internet services is high within the nation with OECD (2005) statistics indicating 10.7 million DSL subscribers as of December 2005, which translates into almost one-third of all households. The rising demand for cheaper communications and the ubiquity offered by the VoIP solution has driven its growth to mature stages of acceptance. Competition continues to rise among the existing 80 plus providers for broadband VoIP services (GerVoip, 2006). Research also indicates that the overall European telecommunications market is more vulnerable to VoIP providers because of per minute tariffs (Richardson, 2005). Conversely, the developing nation of India has had a disparate telecommunication infrastructure. Although consumer demand for Internet services is rising, broadband uptake is low and fragmented within the nation. While Search Agency (2006) reported that in March 2005, India had a total of 815 thousand broadband households; TRAI (2006) reported a total of 7.5 million subscribers for broadband in the country, as of November 2005. The telecommunications, although privatized, are still regulated strictly via the Telecom Regulatory Authority of India (TRAI, 2006). Indirectly, the market with varied service providers is still dominated by the older monopoly telecom companies which hold the high market share. Within the nation, the rural areas which house a high volume of population often do not have a telecommunications infrastructure that lends itself to per household access of broadband. Urban areas are densely populated by young households that cannot afford to own a PC or broadband connection per household. The initiatives of the government to connect the whole nation have resulted in communication via public telephone booths and lately cybercafés across the nation. The lack of infrastructure in rural areas is currently be-
ing addressed via availability of low cost mobile communications. In this climate, in an attempt to address the disparities with the promise of the ubiquitous communications that VoIP offers, the JRYHUQPHQWRI,QGLDLVPRYLQJWRZDUGVDXQL¿HG licensing for the use of VoIP (Hindustan Times, 2006). As a result, there are approximately 70 VoIP providers in India who are competing in various regions, but are still regulated by TRAI which restricts the usage of VoIP for domestic calls. 7KLVFKDSWHU¶VPLVVLRQLVWRSUHVHQWDSUHOLPLnary analysis on the possible impact of VoIP in the two countries. The chapter is structured as follows. We begin the next section by presenting an overview of VoIP, placing it into the context of both the nations concerned. Subsequently, a section on broadband is presented as a facilitator for VoIP. Drawing from these sections, a brief discussion follows which analyzes the impact of VoIP on the telecommunications sector in both nations. Finally, an outlook for both economies is presented, covering the future of VoIP and its impact on telecommunications and on the industry. We expect that this chapter will provide an opportunity for both economies to learn from each other, while appreciDWLQJWKHHFRQRP\VSHFL¿FIDFWRUV&RQYHUVHO\ZH also expect the chapter to be of relevance to other developed/developing economies, policy makers, forums, telecom industries, vendors of equipment, and academics as it reveals unique and common factors that are applicable globally.
VOIP: A SYNOPSIS VoIP is an IP based telephony term for a set of facilities that are used to manage the delivery of voice over the Internet (Wallingford, 2005). It involves sending voice based information in digitized format in discrete packets rather than using the traditional circuit committed protocols as in the case of public switched telephone networks (PSTN). As Ledford (2006, p. 4) explains, “in simple terms, it is a way to have a telephone conversation using the Internet, rather than a traditional telephone line.” For a basic telephone call, the phone is connected to a jack in the wall which connects to the
797
Impact of Broadband VoIP on Telecoms
public switched telephone network (PSTN). When the phone is picked up to dial a particular number, the call is routed across a line that goes across the national or international infrastructure. According to Ledford (2006), The difference of VoIP is the way that it facilitates the call. Instead of being connected a phone jack in the wall, the telephone is connected (possibly using an adapter) to an Internet connection. When a call is then made it is transmitted via the Internet. (p. 4) 6SHFL¿FDOO\ OHW XV WDNH WKH FDVH RI PDNLQJ a call using a VoIP service with a conventional telephone. The telephone is connected to an analogue adapter, which in turn is connected to a router that provides the Internet connection. The analogue adapter converts the voice into data, divides it into small packets which are sent over the Internet to the destination of the call, regardless whether the person at the destination has traditional telephony, known as POTS. The packets sent are encoded with information indicating how they should be arranged and decoded, on reaching their destination. Subsequently, they are decoded at the receiver end of the call and reassembled so that the receiver gets to hear it as if it was a traditional phone call (Ledford, 2006, pp. 4-5). There are two types of protocols that affect VoIP, that is, signaling and device controlled. As the name suggests, signaling protocols help set up communications between two end points and gateways. Conversely, device controlled protocols control the functions of the device. H.323 is the oldest of signaling protocols, recom mended by the Inter national Telecommunications Union. It was designed initially to provide teleconferencing with voice, video, and data capacities on packet switching networks (VoIP Foro, 2006). However, its current version governs point-to-point communications such as telephone calls and multipoint communications such as video or Web conferences. The protocol ensures interoperability between vendors by governing the endpoints of communications—terminals, gateways, gatekeepers, and multipoint control
798
units (Ledford, 2006). ,WKDVEHHQGH¿QHGVRWKDW WKHPDQXIDFWXUHUVFDQDGGWKHLURZQVSHFL¿FDWLRQV WR WKH SURWRFRO DQG GH¿QH RWKHU VWUXFWXUHV DQG standards which in turn let the devices acquire new features or capacities (VoIP Foro, 2006a). However, the most popular protocol is the sessions initiation protocol (SIP) developed by the IETF MMUSIC Working Group. It is the proposed standard for initiating, modifying, and terminating an interactive user session that involves multimedia elements such as video, voice, instant messaging, online games, and virtual reality. The latest publication of this protocol is known as the RFC 3261, which was released in 2002. The main objective of SIP is the communication between multimedia devices (VoIP Foro, 2006b). It helps FUHDWHDFRQQHFWLRQE\HVWDEOLVKLQJDFRPPRQ¿HOG on which two separate connections can meet, even though one end may not have VoIP. This is achieved by using user agents or end systems along the network that receive requests and return responses for the user, and the agent which processes the information and sends it out across the Internet (Ledford, 2006). SIP is an application-layer control protocol, a signaling protocol for Internet telephony. It can establish sessions for features such as audio/ videoconferencing, interactive gaming, and call forwarding to be deployed over IP networks. This in turn enables service providers to integrate basic IP telephony services with Web, e-mail, and chat services. This is because it is based on request and answer messages and reuses many concepts of previous standards like HTTP and SMTP (VoIP Foro, 2006b). SIP is the most popular protocol with references often implying that SIP and VoIP are the same (AT&T, 2006b). Gartner expects SIP to be broadly adopted throughout 2007 for establishing voice, video, and Web conferencing sessions with instant messaging networks being forced into native adoption of SIP and its multimodality derivatives (Cain & Munch, 2005). However, the authors also comment that there is much work to be done on the development of SIP, interoperability, and construction of various SIP extensions. Nevertheless, it offers the promise
Impact of Broadband VoIP on Telecoms
of supporting a wide range of services beyond basic telephony, including instant message (IM), presence management, and voice-enabled Webbased e-commerce (AT&T, 2006b). Some potential advantages (AT&T, 2006b) are illustrated with the following examples: •
•
Customer attempting to contact a representative through a Web portal simultaneously initiates an e-mail, an IM, and a VoIP call to the appropriate contacts to maximize responsiveness Instant messaging session between two parties can become a session among four before transforming into an ad hoc audio and video conference to resolve an issue when audio DORQHLVQRWVXI¿FLHQW $QHQGXVHU¿QDOL]LQJDGRFXPHQWDVSDUWRI DSURFHVVGH¿QHGZLWKLQDZRUNÀRZPDQDJHment platform pulls the original author into a document-sharing session (with voice) where subtle changes are made and the document is updated in the process
Some device control protocols include the media gateway controller (MEGACO) also known as H.248, and the skinny client control protocol (SCCP), which was developed by Selsius Corporation. Currently, no single protocol directs all the VoIP activity resulting in a lack of common standards. H.323 and SIP are the two broadest reaching protocols (Ledford, 2006); however, Cisco uses SCCP for communication with its call manager. 7KHUHDUHEURDGO\WZRFODVVL¿FDWLRQVRI9R,3 7KH¿UVWPHWKRGLVWRIDFLOLWDWHLWYLDWKHFRPSXWHU (computer to computer), for example, through instant messaging software such as Yahoo, Skype, or Google Talk which are proprietary protocols. Alternatively, it is possible to use a softphone client on a PC which facilitates regular calls using the SIP standard (Ledford, 2006). VoIP is the family of technologies that allow IP networks to be used for voice applications, such as telephony, voice instant messaging, and teleconferencing and it includes specialized voice applications such as Skype (Wallingford, 2005).
The proprietary VoIP technology organization “Skype” is a company that was founded in 2003 by Niklas Zennström and Janus Friis, who had SUHYLRXVO\SLRQHHUHG33¿OHVKDULQJWHFKQRORJLHV (Skype, 2006a). They applied the technology to create a P2P communication or instant messaging tool, which would be easy and fun to use. It set an example of a proprietary VoIP service that crosses all geographic boundaries, enabling unlimited voice and video communication for no cost between users. However, being a closed protocol, it cannot be used by other providers. Nevertheless, it has become almost an alternative protocol to SIP, through its popularization via Skype-to-Skype PC based user base (Skype, 2006a) and the Skype out telephony service. Skype has given a new dimension to global communications via its innovative VoIP services. As a VoIP service, it is available in 27 languages and has approximately 100 million users worldwide. The company headquartered in Luxembourg has RI¿FHVLQ(XURSH$VLDDQGWKH86$DQGZDV acquired by eBay in October 2005. With these developments, Skype was able to provide its users extensive voice and video calling features, some of which are unavailable on traditional telephones, as ZHOODVLQVWDQWPHVVDJLQJDQG¿OHVHQGLQJIHDWXUHV The other features are SkypeOut (call from Skype to traditional landlines or mobiles), SkypeIn (a phone number your friends can call, you answer in Skype), and Skype voicemail (takes your calls ZKHQ\RXDUHEXV\RURIÀLQH ²ZKLFKDUHQRWIUHH but are often still cheaper than regular telecommunications (Skype, 2006a). Manufacturers of VoIP phones have introduced Skype-enabled phones, without the need for a computer (Skype, 2006b). 7KHUHVHDUFK¿UP(YDOXHVHUYHUHSRUWVWKDW6N\SH LVVHWWREHFRPHDPDMRUWKUHDWWR(XURSH¶VWHOHcommunications as more people use the Internet to make phone calls that are free between Skype users (Richardson, 2005). The main requirement or pivot on which VoIP revolves is the access point, which is the Internet, through which it can be enabled. Before we explore the infrastructure for VoIP in detail, it is necessary to understand the varied possibilities of access.
799
Impact of Broadband VoIP on Telecoms
7KH ¿UVW LV WKH SURSULHWDU\ 9R,3 VHUYLFHV DV illustrated in the previous paragraphs, such as via instant messaging services, that is, Skype, Yahoo, or Googletalk. These are based on computer to computer based Internet connections. The second type of facilitation is via replacing the existing telephony systems with a VoIP system. Effectively, this type of VoIP partially replaces WKH3276EXWVWLOOSOXJVLQWRLW6SHFL¿FDOO\DV illustrated in the earlier section, a conventional telephone is connected to an analogue adapter, which in turn is connected to a router that provides the Internet connection. This option has been taken up by many businesses across the globe, mainly due to the ability to cut costs. The third possibility is via 3G mobile networks, which enable Internet access. This technology is increasingly becoming popular, especially where mobile phones have almost replaced traditional ¿[HG OLQH WHOHSKRQ\ V\VWHPV 6SHFL¿FDOO\ LQ many developing economies, mobile telephony KDVDOPRVWUHSODFHG¿[HGOLQHWHOHSKRQ\DVWKHUHLW eliminates the need of building expensive, logistically complicated infrastructure. As against the 2G/GPRS bases mobile telephony, the 3G technologies are expected to provide cheaper alternative for Internet access. Another possibility of access is via using phone LQVWUXPHQWVHQDEOHGE\:L0$;RUVLPLODUWHFKnologies. If a telephone instrument is equipped with this technology, using perhaps the SIP standard will be capable of enabling VoIP.
VoIP Risks The very nature of the Internet makes VoIP vulQHUDEOHWRVHFXULW\ULVNV$7 7 D FODVVL¿HV the threats into three main areas: denial of service (DoS), fraud and abuse, andGDWDFRQ¿GHQWLDOLW\ and privacy. Denial of service (DoS) attacks on the Internet occur due to software vulnerabilities or implementation problems. It is an attempt to tie up network resources or interfere with service protocols so that the level of service is denied to subscribers. Although they are aimed primarily at networks, they can also target customer equip-
800
ment. DoS attacks can occur when a network or GHYLFHLVRYHUORDGHGZLWKPHDQLQJOHVVWUDI¿FRU VHQWDVSHFL¿FFRPPDQGWKDWZLOOGLVDEOHLWUHQdering the network unavailable (Maslog-Levis, 2004). An example is sending repeated hang-up FRPPDQG WR HDFK KDQGVHW ZKLFK LV GLI¿FXOW WR GHWHFWRUSUHYHQW6SHFL¿FDOO\DVYRLFHLVVKDULQJ a network with the data, it becomes susceptible to this type of attack that has been used against data networks for years. VoIP is susceptible to viruses and therefore requires an adequate management framework and even virus protection on handsets (AT&T, 2006a). Fraud is also known as theft of service where users are able to access more services and resources than their entitlement. It can cover a wide spectrum from complete theft, where a service is used without authorization by non subscribers, to partial theft, where a subscriber is using more services than paid for. Abuse in general is any behavior against the acceptable use policy for the service. Protection mechanisms include physical security WROLPLWGLUHFWDFFHVVDQGVHFXUHFRQ¿JXUDWLRQVWR limit remote access. It is recommended that logs be monitored to frequently detect unauthorized access. Further, for service providers, new frauddetection algorithms need to be developed (AT&T, 2006a). Data privacy issues are related to protecting the rights of VoIP user data (AT&T, 2006a). Typical attacks made against privacy include compromizing network elements or databases that contain subscriber data. For example, private data in the signaling protocol may include phone numbers called by the subscriber. Unprotected voice and signaling transported over public or shared IP QHWZRUNV PD\ EH VXVFHSWLEOH WR FRQ¿GHQWLDOLW\ (eavesdropping) attacks. An example is the risk of KDYLQJDXVHU¶VSKRQHEDQNLQJGHWDLOVLQFOXGLQJ account number or pin numbers, intercepted across the network because of non encryption. Lemon (2006) pointed out that banks who are switching their phone systems to VoIP are making themselves vulnerable to phishing attacks for which there are currently no effective detection or prevention tools. The attacks allow hackers to steal personal data, including credit card numbers and
Impact of Broadband VoIP on Telecoms
bank account information. In certain scenarios, when a customer phones a bank, the service line has already been taken over by hackers. Intrusion detection systems are still not capable of detecting these attacks although there are tools such as SIPhallis which allow security managers to manage 6,39R,3SDFNHWVRQWKHLUQHWZRUNV6SHFL¿FDOO\ the tool provides an interface to create and send VoIP packets and allows monitoring of the packets (Lemon, 2006). A KPMG report (Maslog-Levis, 2004) had also questioned the integrity of VoIP as the VoIP packets, similar to data packets, are vulnerable to loss as the voice and data are sharing the same medium. However, over the last few years, it has been found that this loss rarely occurs. As the VoIP technology continues to gain momentum, there is much accent on the need for adequate infrastructure security, encrypted transmission, maintaining end point security, and monitoring for new threats as they arise.
VoIP Limitations vs. Advantages In the mid 1990s, when the concept of VoIP emerged, only businesses had tested it as a potential service that offered cost savings (Wallingford, 2005). However, the poor quality of audio, delayed, and unsynchronized calls led to much frustration. For example, dial up Internet services being slow often resulted in proportions of conversation being dropped or the call being prematurely terminated (Ledford, 2006). It was at the beginning of 2004, with the growing diffusion of broadband technologies that businesses had begun re-testing its usage. While businesses with a broadband infrastructure found the ubiquitous nature of the technology offered cost savings by eliminating the need for another communications infrastructure, consumers began using it for saving money for long distance calls (Barkins, 2004). With features such as Voice mail, caller ID, and three-way calling included in the basic price of software or subscription, it was especially appealing to the customers with broadband connections. Forrester Research surveyed 714 businesses in North America and Europe regarding their
approach to VoIP (Herrell, 2006). Among them, 45% had deployed the technology and 30% were HYDOXDWLQJRUSLORWLQJLW7KHVLJQL¿FDQWIDFWRUVIRU consideration were reliability in North America and costs in Europe. Forrester research also pointed out that European enterprises are more likely to deploy 9R,3ZLWKLQWKHQH[W\HDUVUHÀHFWLQJWKHKLJKHU acceptance of using external service providers (Herrell, 2006). The primary advantage of VoIP seems to be the reduction of communication costs as it enables conversations with multiple people, anytime, irrespective of locations with added features such as Web cameras for visual communication. The combination of data communication with voice allows easy exchange of messages, videos, and documents along with conversations (Intertangent, 2006). During November and December 2005, Continuity Central (2006) conducted a survey into the business implications of VoIP implementation. Among the businesses surveyed, 24% of responGHQWV VDZ 9R,3 RIIHUHG WKH EHQH¿W RI UHGXFHG risks while 41% saw an increase in the risks due to its implementation. Table 1 lists the highlights from this survey which captures the sentiment of businesses regarding VoIP. Some businesses that are moving towards VoIP tend to retain at least a stripped down version of their traditional telecommunications infrastructure. These are often businesses which are still vary of new technologies and, therefore, tend to maintain a backup infrastructure. This has an impact on the cost-benefit equation. 6SHFL¿FDOO\KDYLQJWRPDLQWDLQRQHLQIUDVWUXFWXUH which facilitates Internet and telecommunications and lowers costs is the main advantage of VoIP deployment by businesses. However, considering the vulnerabilities and risks, some businesses continue at least skeletal telecommunications alternatives or a separate wireless communication network in place, which in fact adds to at least the short term costs of a business. Nevertheless, as the VoIP technologies are becoming stabilized, most businesses are willing to make a complete switch from conventional telephony to VoIP. Frost and 6XOOLYDQ SUHGLFWVLJQL¿FDQWJURZWKLQWKH global conferencing and collaboration market, as
801
Impact of Broadband VoIP on Telecoms
Table 1. Business sentiments on VoIP (adapted from Continuity Central, 2006) ³9R,3SUHVHQWVERWKDPLWLJDWLRQDQGDULVNLQEXVLQHVVFRQWLQXLW\WHUPV$VLJQL¿FDQWFRPSRQHQWRIRXUEXVLQHVVFRQWLQXLW\ strategy is to enable as many people as possible to work anywhere. Whenever an incident occurs most of our people switch WRKRPHZRUNLQJWKHUHVWWRDQDOWHUQDWHRI¿FH9R,3PHDQVWKHLUSKRQHDQGQXPEHUWUDYHOZLWKWKHP7KHULVNLVWKDWLIWKH incident also led to a major network outage, our phones are massively impacted. This is mitigated in two ways, one obviously is to build resilience into our internal network. In addition, all key personnel have a mobile phone and our contact lists include VoIP, mobile, and home contact details.” ³0RYLQJWR9R,3LQFUHDVHVRXUGHSHQGHQF\RQWKH,7QHWZRUNZKLFKPD\PDNHGLVDVWHUUHFRYHU\PRUHGLI¿FXOWIRUH[DPSOH following a major virus attack.” “There will be more outages in a VoIP system than in the old voice network.” “VoIP brings IT risks to the telecommunications arena. Consider a denial of service attack. If this happens for data only; bad news. Consider for data and voice; very, very bad news.” “We have a single call center with a switch that can not be colocated. VoIP will allow us to deploy staff to multiple locations if the primary site goes down.” “I think the key dependency is whether or not there are backup facilities for telephony. Without backup, risks are increased. With backup, the risks are mitigated.” ³,QP\RSLQLRQ9R,3LVDJRRGZD\WRUHGXFHFRVWDQGDWWKHVDPHWLPHPDNHEXVLQHVVSURFHVVHVPRUHHI¿FLHQW7KHULVNDWOHDVW LQ*HUPDQ\ZLOOVWD\DWWKHVDPHRUULVHVOLJKWO\ GHSHQGHQFHRQRQO\RQHWHFKQRORJ\ &RQFXUUHQWO\DVLPSOL¿FDWLRQRIWKH infrastructure also offers a possibility to implement effective defenses.”
companies migrate from standalone systems to integrated collaboration solutions. Conference calling is a feature that is most used in telecommunications and an example where VoIP can deliver unique business value (AT&T, 2006b). Being an HVVHQWLDOWRROIRUWRGD\¶VKLJKO\GLVWULEXWHGPRELOH workforce, the popularity of the teleconferencing feature has made it one of the biggest line items in a typical business budget. VoIP has also had its many advantages for consumers (PUC, 2006). From its initial stages it has been experimented with as a cheaper alternative for long distance calls (Wallingford, 2005). In Australia for example, using a VoIP service, a consumer can make untimed calls to all capital cities at local call charges (MyNetFone, 2006). The expense of traditional interstate calls is eliminated. In many economies, broadband Internet is in its mature stages of diffusion. The ubiquity offered by VoIP services eliminates the need to have another ¿[HGOLQHWHOHSKRQHVXEVFULSWLRQ6SHFL¿FDOO\LI a consumer has an already established broadband Internet connection, it is cheaper to subscribe to
802
a VoIP service. This is due to the fact that VoIP SURYLGHUVGRQRWQHHGWREXLOGRUOHDVH¿EHURSWLF networks. Further, VoIP is a largely unregulated “information service” rather than a phone service which means that the providers do not charge acFHVVFKDUJHVDQGIHHVWKDWJHWVDGGHGRQWR¿[HG line phone bills, making VoIP even cheaper by comparison. Some providers allow checking of voice messages from a computer and easy management/sharing of messages similar to e-mail. Many providers allow a range of features that takes advantage of the important difference between Internet calling and traditional calling services (PUC, 2006). One of the other key advantages of VoIP is locaWLRQLQGHSHQGHQFH6SHFL¿FDOO\LIDFXVWRPHUKDV a VoIP subscription, it can be used from anywhere in the world, via a broadband Internet connection. It is location independent in that the user can make or receive calls from a local phone number. The ubiquity of the service, that is, having a phone number for life often gets overlooked. Conversely, when we consider the location independence in
Impact of Broadband VoIP on Telecoms
terms of emergency services, to a small degree, it is a disadvantage. Specifically, with the conventional PSTN services, when an emergency call is made, it is automatically routed to the closest location from where the phone call is made. VoIP services being location independent do not have this provision. The current regulatory framework in most economies requires that VoIP service provision is given to individuals or businesses having a local address, so that emergency calls are traceable (Marsen, 2006). However, if a person has a VoIP number that is registered in Germany and if the same person is now in India, using the VoIP number that is portable and dials an emergency number, the call is routed to the closest emergency services to home in Germany. As a result, the emergency call becomes redundant. However, this is considered a small issue which is expected to get resolved as the technology matures. The growing popularity of broadband VoIP services has drawn some regulatory responses around the world, resulting in two approaches: light touch regulation and technology neutral. 7KH ¿UVW DSSURDFK LV PDQGDWHG LQ FRXQWULHVUHgions such as the European Union, Hong Kong, Malaysia, Singapore, and the United States aimed at promoting the new service as well as emergency
service provision. The second approach is adopted in countries such as Canada where VoIP is regarded as another voice service that should be subject to similar regulations as PSTN voice services. In yet other countries such as Korea and Japan, a mixed approach has been resulting in light regulation of VoIP services (V&D, 2006). Some broad limitations of VoIP as compared to traditional telephone services are summarized in Table 2. There is currently only limited standardization of the telephone numbers issued by VoIP service providers (Taylor & Coombs, 2006). In particular, this means that number portability does not exist in the same way as for PSTN numbers and, therefore, the users who wish to change service providers will have a new telephone number allocated by their new service provider, rather than simply being able to transfer their existing number. Yet another disadvantage is that there is no universal standard for VoIP services as yet and all the existing SURWRFROV DUH VWLOO QRW IXOO\ UH¿QHG :LWK DOO LWV limitations, risks and constraints VoIP seems to be gaining popularity. The necessary infrastructure for adoption of VoIP is the diffusion of broadband into the PDVV PDUNHW :LWK WKLV SHUVSHFWLYH ZH EULHÀ\ summarize the current and future possibilities of
Table 2. Broad limitations of VoIP (adapted from PUC, 2006) If the broadband Internet connection goes down the VoIP phone may not work. Some providers offer a service that will forward incoming calls to an alternate number (such as a cell phone) in the event the VoIP connection fails. /DQGOLQHWHOHSKRQHVDUHSRZHUHGE\WKHORFDOSKRQHFRPSDQ\¶VQHWZRUNVRWKH\NHHSZRUNLQJZKHQSRZHULVORVW6LQFHPRVW9R,3 services are tied in with computers, they shut down when the power goes out. A backup power supply can keep the computer and the phone working, but if the broadband service is provided by a cable company, the cable delivery system itself may be without power, leaving the consumer without a phone service during a power outage. Therefore, the costs of a backup power system has to be considered before switching to a complete VoIP service. Security and alarm systems may not work with a VoIP service. To counteract this some providers are offering their own proprietary alarm systems. In some occasions facsimile setting have to be changed with VoIP connection. In some occasions if international calls to landlines or mobile phones are made using VoIP, the service providers may add fee for connecting the calls, which adds to costs of the service. ,W¶VQRWFOHDUKRZUHJXODWRUVZLOOGHDOZLWKWHOHPDUNHWHUVZKRVHQG³9R,3VSDP´PHVVDJHVWRVXEVFULEHUV7UDGLWLRQDOSKRQHXVHUV can sign up for “no call” lists to avoid being deluged by telemarketing calls, but there is no such list to restriction with Internet spammers.
803
Impact of Broadband VoIP on Telecoms
broadband as a prerequisite for VoIP initially and then take a closer look at Germany and India.
BROADBAND Broadband as it is commonly known is the shortened version of broadband internet access which refers to high data transmission rate Internet connections, which were rapidly expanded into the consumer market in the early 2000s. The International Telecommunication Union Standardization 6HFWRU ,787 UHFRPPHQGDWLRQ,KDVGH¿QHG broadband as a transmission capacity that is faster than primary rate ISDN, at 1.5 to 2 Mbit/s (ITU, 2006). The popular consumer broadband technologies are via DSL or cable modems, typically capable of transmitting faster than any dial up modem with a real maximum speed of up to 48 Kbit/s. Modern consumer broadband implementations, up to 24 Mbit/s, are several hundred times faster than those available at the time of the popularization of the Internet (such as ISDN and 56 Kbit/s) while costing less than ISDN. However, performance and costs vary widely between countries. The Organization for Economic Cooperation DQG'HYHORSPHQW 2(&' KDVGH¿QHGEURDGEDQGDV 256 Kbit/s in at least one direction (OECD, 2005). This is regarded as the most common baseline around the world. It is clear that the industry as such does not specify bit rates and, therefore, broadband speeds can be lower bit rate transmission. Some
internet service providers (ISPs) use this aspect to their advantage and market the lower bit rate transmission as “the broadband.” The standard broadband technologies are via DSL (Digital Subscriber Line) and cable networks. Table 3 represents WKHPDLQGH¿QLWLRQVLQ'6/DVDGDSWHGIURP'6/ Forum (2007a). According to DSL Forum (2007b), Asymmetric digital subsriber line (ADSL) is a modem technology that transforms ordinary phone lines (also known as “twisted copper pairs”) into high-speed digital lines for ultra-fast Internet access…it enables access to corporate networks for telecommuters, as well as new interactive multimedia applications. 6SHFL¿FDOO\XVLQJGLJLWDOFRGLQJWHFKQLTXHV ADSL modems are able to utilize 99% capacity out of a phone line, without interference with the regular telephone service, thus enabling simultaneous access to normal telephony service as well as multimedia applications such as video and high speed Internet access (DSL Forum, 2007b). The latest development in DSL is the very high speed digital subscriber line (VDSL). According to the DSL Forum (2007c), The newly standardized VDLS2 also enables very high speed Internet access of up to symmetrical 100 Mbps (both up and downstream), EXW DGGLWLRQDOO\ 9'6/ LV VSHFL¿HG WR VXSSRUW
Table 3. Broadband technologies (adapted from DSL Forum, 2007a) DSL GLJLWDOVXEVFULEHUOLQH LVDWHFKQRORJ\WKDWH[SORLWVXQXVHGIUHTXHQFLHVRQFRSSHUWHOHSKRQHOLQHVWRWUDQVPLWWUDI¿F typically at multimegabit speeds. It can allow voice and high-speed data to be sent simultaneously over the same line. The service is “always available”, end-users do not need to dial in or wait for call set-up. It is wired for speed. Asymmetric variations include: ADSL, G.lite ADSL (or simply G.lite), VDSL (ITU G.993.1), and VDSL2 (ITU G.993.2). The standard forms of ADSL (ITU G.992.3, G.992.5, and ANSI T1.413-Issue 2) are all built upon the same technical foundation, discrete multi tone (DMT). The suite of ADSL standards facilitates interoperability between all standard forms of ADSL. VDSL YHU\KLJKELWUDWH'6/ 8SWR0EVRYHUGLVWDQFHVXSWRPHWHUVRQVKRUWORRSVVXFKDVIURP¿EHUWRWKHFXUE 0RVWRIWKHWLPH9'6/OLQHVZLOOEHVHUYHGIURPQHLJKERUKRRGFDELQHWVWKDWOLQNWRDFHQWUDORI¿FHYLDRSWLFDO¿EHU,WLV particularly useful in university campuses or business parks, and it is currently being introduced in market trials to deliver YLGHRVHUYLFHVRYHUH[LVWLQJSKRQHOLQHV,WFDQEHFRQ¿JXUHGLQV\PPHWULFPRGH
804
Impact of Broadband VoIP on Telecoms
DSSOLFDWLRQVVXFKDVPXOWLFKDQQHOKLJKGH¿QLWLRQ TV (HDTV), video on demand, videoconferencing, and VoIP, all using the existing ubiquitous copper telephone line infrastructure. That capability combined with its ATM, ethernet and IP compatibility and the capacity for multimode implementations enabling interoperability with existing ADSL equipment, means that VDSL2 will integrate readily into legacy and next generation telecommunication networks.
many (called T-DSL), making the country one of the major DSL nations. T-DSL has a number of resellers and the major ones are T-Online and 1&1. Many ISPs provide service for it. Alternatively, there are DSL providers in Germany which have their own DSL network and only rent the copper lines such as Arcor, QSC, AliceDSL, and Telefonica. Deutsche Telekom offers T-DSL at the following rates, with a UHTXLUHPHQWRIDWHOHSKRQHFRQWUDFWDW¼DQDORJXH RU¼,6'1 %XQGHVQHW]JHQWXU
The second most common broadband infrastructure is via cable network. Cable broadband RSHUDWHVRYHUWKHSD\WHOHYLVLRQ¿EHURSWLFQHWZRUNV LQVWDOOHGWKURXJKRXWQHLJKERUKRRGV6SHFL¿FDOO\D cable is run from the larger cable running down a street, to a household. This cable is then connected to a special cable modem and then to the household compute. If there is an existing cable television, then the household is ready to connect cable broadband. If not, the cable needs to be run to the residence by a professional installer (Broadbandguide, 2007). In addition, there is also :L0$; ZRUOG interoperability for microwave access), which is expected to enable true broadband speeds over wireless networks at a cost point to enable mass market adoption (Intel, 2006a).
There are two main applications of WiMAX: ¿[HG:L0$;DSSOLFDWLRQVDUHSRLQWWRPXOWLSRLQW enabling broadband access to homes and businesses, whereas mobile WiMAX offers the full mobility of cellular networks at true broadband VSHHGV %RWK ¿[HG DQG PRELOH DSSOLFDWLRQV RI WiMAX are engineered to help deliver ubiquitous, high-throughput broadband wireless services at a low cost. (Intel, 2006a) With the above infrastructure that facilitates high speed Internet, which is the main access point that facilitates VoIP, we now take a closer look at Germany and India.
Broadband in Germany $VRI¿UVWTXDUWHURI'HXWVFKH7HOHNRPKDG approximately 7.5 million DSL customers in Ger-
¼ ¼ ¼ ¼
These prices do not include ISP fees, which DUHW\SLFDOO\DQDGGLWLRQDOÀDWUDWHRI¼¼RQ all speeds) and also vary depending on location (Bundesnetzgentur, 2005). Such ISPs include T-Online, Congster (both Deutsche Telekom), 1&1, and GMX (both United Internet), all of which use/resell the T-Com IP backbone (though 1&1 has been using 7HOHIRQLFD¶VQHWZRUNIRUWKHLU0ELWVVHUYLFH Arcor RIIHUV ¼ ¼ ¼ 7KHVHSULFHVDOUHDG\LQFOXGHWKH ,63ÀDWUDWHEXWUHTXLUHDWHOHSKRQHFRQWUDFWZLWK Arcor. Telefonica has its own DSL infrastructure with 40% coverage based on line sharing and its own backbone, formerly MediaWays. It is used in part by AOL (fourth with 1 million customers) and Freenet, but Telefonica itself does not sell to end-users in Germany. AOL uses T-DSL where Telefonica has no coverage. QSC offers Q-DSL home with 1.5 or 2.5 Mbit/s combined download and upload speed, and the customer chooses how PXFKRILWLVXSORDG3ULFHVDUHDQG¼PRQWK respectively, plus activation fee (Bundesnetzgentur, 2005). DSL accounts for some 95.5% of broadband connections and Deutsche Telekom holds approximately 67% market share (Budde, 2007). Deutsche Telekom/T-Com began the process of building a VDSL network in 2006 and has launched 9'6/LQDQXPEHURI*HUPDQ\¶VODUJHUFLWLHV7KH initial download speed offered is 25Mbit/s and an upload speed of 1.5Mbit/s. A download speed of 50Mbit/s is expected in the future. One of the
805
Impact of Broadband VoIP on Telecoms
PDMRUSHUFHLYHGEHQH¿WVRI9'6/LVWKHSRVVLELOLW\ to transmit TV over the Internet (IPTV). Together with the minimum analog telephone connection UHTXLUHPHQW9'6/KDVDFRPSOHWHFRVWRI¼ per month. The hardware required for VDSL and ,379FRVWVDWRWDORI¼ +'79LVWREHLQFOXGHG in the T-online basic package with in excess of 100 channels. A set-top box from Deutsche Telekom is UHTXLUHGWRUHFHLYHWKH79RIIHULQJ7KH;7 box has an 80GB hard-drive for program recording Moreover, Deutsche Telekom will offer a VOIP ÀDW UDWH RI ¼ SHU PRQWK P\VPH ,Q February 2007, Germany passed a telecom law that made Deutsche Telekom a monopoly for VDSL. However, on February 26, EU launched a lawsuit against over this law which keeps competition off the high-speed Internet networks (CNBC, 2007). The cable TV infrastructure was owned by Deutsche Telekom, which pushed DSL and neglected the cable network. With much political pressure over the past few years, it was sold to Kabel Deutschland, Kabel BW, and so forth (separated geographically), which is slowly investing into XSJUDGLQJWKHFDEOHQHWZRUN¶VEDQGZLGWKFDSDFLW\ and making it two-way-capable. Kabel Deutschland RIIHUV0ELWVIRU¼PRQWK DQGKLJKHUVSHHGV at higher prices) on top of the cable TV fee. In September 2005, it was offering Internet access services with maximum download of 8.2 M/bits (Bundesnetzagentur, 2005). However, cable broadband has accounted for less than 3% of total broadband subscriptions despite extensive cable networks (Budde, 2007). According to latest reports, During 2006 the cable sector was further consolidated through mergers and acquisitions and operators have positioned themselves to compete with Deutsche Telekom through network upgrades and triple play offers. Municipal networks such as NetCologne offer effective alternatives to Deutsche Telekom, resulting in the incumbent being squeezed out of a growing number of communities. (Budde, 2007) According to the same report on Western Europe, the penetration of broadband in Germany,
806
which is the second largest Internet and broadband market in Europe, remains below that of other benchmark countries in the EU (Budde, 2007). As per the Strategy Analytics Research, broadband penetration reached only 28% in 2005 and LVIRUHFDVWWRUHDFKRQO\LQ :L0$; Day, 2007a). Deutsche Telekom still dominates the market through its T-Online ADSL service and the latest right it has obtained to prevent competition access to the VDSL network may slow down broadband penetration, during 2007 (Budde, 2007). Satellite links are another broadband option. TDSL Satellite or skyDSL, which are one-way links based on DVB-S require an uplink via dialup, which often has to be paid by minute. Some offers with two-way satellite connections exist for consumers. Satellite inherently has high latency, and is thus second choice for most people (Bundesnetzagentur, 2005). BusinessCom offers Internet via satellite services for home and individual usage, SOHO, SME, Internet cafe, calling center, and ISP business up to four Mbit/s with SLA quality guarantee, QoS, business-class voice over IP, and videoconferencing support. STARBLAZER offers home or small RI¿FHVHUYLFHVLQ*HUPDQ\ 6DWVLJ Deutsche Breitband Dienste GmbH (DBD), a service provider based in Heidelberg, Germany, KDV EHHQ RIIHULQJ :L0$; EURDGEDQG ,QWHUQHW DFFHVV WKURXJK LWV 0$;;RQDLU EUDQG VLQFH August 2005. The cost-effective, easily deployed :L0$;QHWZRUNKHOSVWKHFRPSDQ\SURYLGHQHZ services that attract both private and businesses users (Intel, 2006b). In November 2006, the Federal Network Agency (Bundesnetzagentur) in Germany announced the receipt of applications from 6 companies for the auction of 3.4-3.6 GHz spectrum licenses in December. Among them, three had applied for national and the other three IRUVSHFL¿FUHJLRQV :L0$;'D\E In Germany, UMTS is being offered at speeds of up to 384 Kbit/s download and 64 Kbit/s upload, particularly since E-Plus/Base offers a true mobile ÀDWUDWHIRUURXJKO\¼PRQWK7KHELJPRELOH providers T-Mobile and Vodafone now offer tariffs ZLWK¿YH*%PRQWKRIGDWDWUDQVIHULQFOXGHGZLWK the advantage of HSDPA (up to 1.8 Mbit/s down-
Impact of Broadband VoIP on Telecoms
load) availability on all UMTS nodes. HSDPA or High-Speed Downlink Packet Access is a new protocol for mobile telephone data transmission. It is also known as a 3.5G technology standard which provides download speeds on a mobile phone equivalent to an ADSL or asymmetric digital subscriber line in a home, removing any limitations placed on the use of a phone by a slow connection. It is an evolution and improvement on W-CDMA or wideband code division multiple access, a 3G protocol (Wisegeek, 2006). Pyramid Research contends that, The average German consumer is not entirely VDWLV¿HGZLWK'6/RIIHULQJVLQ*HUPDQ\ZKLFK comprise the main source of broadband and come mainly from Deutsche Telecom and alternative telcos…Vodafone and O2 attempted to offer 3G services for home use, but users balked at the slow speed and faulty service, and overall 3G has barely made a dent on the market. :L0$; Day, 2007) Clearly, the German market has much potential for high speed broadband, but is very much in transition with emerging technologies.
Broadband in India Broadband in India is expensive and the penetration is disparate across the nation. The main providers are VSNL, MTNL, BSNL (which are older telecom monopolies), Bharti Telecom, and SIFY (Socialtext, 2006). As of 2004, the Telecom Regulatory Authority of India (TRAI, 2006a) has GH¿QHGLQWKHODZWKDWEURDGEDQG,QWHUQHWKDVWR be 256 Kbit/s always on service. However, many SURYLGHUVYLRODWHWKLVGH¿QLWLRQE\RIIHULQJORZHU speeds as broadband services (Shankar, 2005). Unlimited Internet access via broadband, without any limitation on data downloads is available IRUD¿[HGPRQWKO\SD\PHQW7KHIROORZLQJDUH WKHFXUUHQWWDULIIVE\¿YH,63VZKLFKWRJHWKHUKROG 86% of the market share (TRAI, 2006b):
071/%61/ ¼ %61/ ¼
%KDUWL7HOHFRP ¼ 6,)< ¼ 961/ ¼
However, the charges in the category are usually worked out on the assumption that the average usage is 8 hours a day. "The contention ratios are normally very high and uptime suffers due to various infrastructure problems like road digging, water seeping into cables, and overhead cables disrupted due to broken trees during monsoons" (Socialtext, 2006). With telecom liberalization, many private ISPs entered the market with their own local loop and gateway infrastructures, rendering the marketplace LQ¿QLWH DQG ZURXJKW ZLWK ¿HUFH FRPSHWLWLRQ There were 153 ISPs which were operational as of March 31, 2006 (TRAI, 2006b), and nearly 14% of the market is held by private ISPs. For the same reason quality of service remains substandard. In an attempt to restrict market entry and exit, TRAI charges a substantial licensing fee (Socialtext, 2006). Technically, broadband in India includes optiFDO¿EHU'6/FDEOHVDWHOOLWHWHUUHVWULDOZLUHOHVV 2,5G wireless, and 3G wireless. However, the 3G wireless offered in Indian market does not yet qualify as broadband (Business Week, 2005). Educational and research institutions have VSAT or DTH. Using the 1xrtt, Reliance provides a 115 Kbit/s service across India that is comparable to wired two Mbit/s connections, which is reliable, \HWFKHDSZLWK¼SHUPLQXWHDWSHDNWLPHV and half at other times. Cable Internet offers better potential due to localized reach and the popular existing cable television networks (Cable Quest, 2007). In many cases it is the only alternative for dial up consumers. One of the most interesting aspects in the cable network infrastructure is that whether it is relatively unconnected rural/semi urban or metropolitan areas or even slums—the penetration of television is almost pervasive. For example, every village has clusters of households which can easily be connected via cable. In the metropolitan areas, the high income earning families are housed in high rise buildings, which do have cable networks. In addition, the il-
807
Impact of Broadband VoIP on Telecoms
legal tapping into the cable networks, especially in the metropolitan area, cannot be overlooked. Theoretically, these existing cable infrastructures can easily be utilized/upgraded for offering better broadband penetration. Rediff.com (2005) reveals that over 60% of Indian users regularly access the Internet via cyber cafés where the cost of access is approxiPDWHO\ ¼ SHU KRXU 7KLV F\EHU FDIp FRQFHSW was popularized by its predecessor, that is, public telephone booths which made telecommunications a possibility across the nation, although disparate. Better penetration of broadband is dependant on affordability of the PC and the existing subscription costs are such that most households are not able to afford it. There are approximately 65,000 cyber cafes in India, which initially began with dial up facilities and at least 20% have switched to have broadband (Rediff.Com, 2005). It is also interesting to note that VoIP call minutes from cyber cafes grew by 10% in 2005, which resulted in SIFY—a large ISP increasing its franchises and ownerships of such cafes and extending them into 104 cities and towns. In addition to these cafes, VoIP is now enabled by this provider only, through Internet telephone booths ORFDWHGLQKLJKWUDI¿FFRPPHUFLDODQGUHVLGHQWLDO areas throughout the country (Rediff.com, 2005). Conversely, businesses in India tend to use leased lines for broadband access. However, the two major problems are frequent downtimes, especially by DSL providers due to power outages, cable faults, and so forth, and capped usage which is an almost unwritten rule followed by ISPs (Mahajan, 2005).
For example, heavy users are capped at 400 MB by state owned MTNL and BSNL and by private ISPs at 1-2 GB. 2Q'HFHPEHU%61/DQQRXQFHG¿[HG ZLUHOHVV QHWZRUNV YLD FHUWL¿HG :L0$; LQ VL[ Indian cities offering high bandwidth connections to corporate customers (Malik, 2006). It is also expected to cover four rural districts in the State of Haryana. The initial deployments offer speeds upto 2 Mbit/s. According to Malik (2006), “Fixed ZLUHOHVV UHDGFHUWL¿HG:L0$; LVMXVWWKHNLQGRI leapfrog technology that makes sense in emerging telecom markets such as India and China.” 6HYHUDOLQÀXHQFHVIRUWKHJURZWKRI:L0$; as adapted from Wi|0$; DUHSUHVHQWHG in Table 4. +RZHYHU WKH :L0$; GHSOR\PHQW LV VWLOO viewed in India with much skepticism. The adoption RI:L0$;KDVEHHQVORZGXHWRWKHJRYHUQPHQW QRWUHOHDVLQJVXI¿FLHQWZLUHOHVVVSHFWUXP 5HEHLUR 2006). To begin with, the Telecom Regulatory Authority of India (TRAI) who believes that it is the most appropriate technology especially for providing rural and remote area broadband connectivity, recommended 100MHz of spectrum LQ *+]*+] UDQJH IRU :L0$; )XUWKHU it suggested that another 100MHz for broadband wireless applications be made available in the 3.43.6GHz band. According to Lil Mohan, Intel India Managing Director for broadband wireless, Worldwide WiMAX deployments have tended to EHDURXQGWKH*+]*+]*+] frequency bands…If India releases spectrum in
7DEOH,QÀXHQFHVIRU:L0$;JURZWK DGDSWHGIURP:L_0D[ /DUJHUGHSOR\PHQWVZLOOEHJLQ¿UVWTXDUWHURIIRU:L0$;SURYLGHGWKHPXFKQHHGHGVSHFWUXPLVPDGHDYDLODEOH The mobile telecommunication industry envisages offering high-margin data services in urban areas, as it is being faced with decreasing revenue in voice based calls. Mobile TV, IPTV, and other broadband applications are under trial at Reliance, Bharti, and MTNL. The increased level of eCommerce activity–mainly through travel bookings, discount airfares, holiday destination packages, job hunting, and matrimonial services–is creating a huge demand for always-on broadband services. Government-led initiatives with strong technology partners such as Intel, Motorola, and Alcatel will trigger successful applications such as the Railtel cyber-cafe network along the entire rail route of the nation.
808
Impact of Broadband VoIP on Telecoms
these bands, it will enable local service providers to take advantage of standard WiMax silicon and equipment designed for these frequencies. (Rebeiro, 2006) However, some of the frequency bands used ZRUOGZLGHIRU:L0$;LQFOXGLQJWKH*+]LQ India, is used by the Department of Space which LVUHOXFWDQWWRUHOHDVHLW :L0$;9LVLRQ Although TRAI has appealed to the government to UHOHDVHVRPHRIWKHVHEDQGVIRU:L0$;GHSOR\ment, the government is yet to adopt a policy. The ministry of communications initially released approximately 12MHz of spectrum in the 3.3-3.4GHz band to seven ISPs for city-level deployments. To further up the spectrum range, it freed up 50MHz of spectrum in the 5.8GHz range for commercial use (Telecom India, 2007). This move was meant WRDOLJQWKHFRXQWU\ZLWKJOREDO:L0$;VWDQGDUGV DQGSURYLQJDSDWKZD\IRU:L0$;VHUYLFHVLQWKH OLFHQVHH[HPSWEDQG :L0$;9LVLRQ With the above developments, many ISPs are SXVKLQJ :,0$; VHUYLFHV 6,)<²D PDMRU ,63 KDV GHSOR\HG RYHU ¿[HG :L0$; EDVH VWDtions and 3,500 subscriber units for enterprise, residential, and cyber café users in 200 cities, with the equipment operating in the 5.8 GHz frequency band. However, there are warnings in the media that rushing into this frequency could crowd the 50MHz allocation leading to poor service quality. Further, there is also a push towards delicensing RIWKH*+]EDQGDVLQRWKHUFRXQWULHV :L0$; Vision, 2007). ,QWHO¶V([HFXWLYH9LFH3UHVLGHQW6HDQ0DORQH\ FRQWHQGVWKDWRSWLFDO¿EHUQHWZRUNVFDQQRWEHD solution for broadband connectivity in India (Ribeiro, 2007). According to Maloney, (YHQLIWKHFRVWRIRSWLFDO¿EHUWHFKQRORJ\LV GULYHQORZHUOD\LQJ¿EHUIRUEURDGEDQGFRQQHFtions is still expensive. Copper connections for DSL are not widely in place in countries such as India where wired telecom penetration is low, and the emergence of cell phones has removed the requirement to run copper. Clearly, the Indian market has potential for broadband access but needs to overcome many
regulatory glitches and experiment with newer technologies, before it reaches mass market penetration. The most promising technologies DUH FDEOH DQG FHUWL¿HG :L0$; IRU EURDGEDQG deployment. In this section, we have summarized the main developments in Germany and India for broadband, which is the main infrastructure for VoIP deployments. The next section is a cross country analysis, taking into account this infrastructure and new developments that facilitate VoIP.
IMPACT OF VOIP: A CROSS COUNTRY ANALYSIS In modern societies, the role of telecommunications is crucial for progress. Insight research predicts that the telecommunications-industry revenue will UHDFK ¼WULOOLRQE\WKHHQGRI $OIRQVR 2005) placing it just under 3% of the gross world product. It has been long acknowledged that good telecommunications infrastructure is crucial for macro and microeconomic success in the modern world (Lenert, 1998). In this section however, we are launching a discussion on two economies, Germany, a developed economy with strong infrastructure capabilities, and India, a developing economy with disparate and rather low infrastructure capabilities. The infrastructure facilities impact on the deployment of VoIP—which is essentially driven by cost. The technologies that form the backbone infrastructure therefore need to be considered carefully in the cross-country context. Before we launch into a comparative discussion, Table 5 lists the ITU statistics (2005) relevant to the context. From the statistics presented in Table 5, it is evident that Germany as a developed nation has the telecommunication industry at a mature stage, while India as a developing nation is in an emerging and disparate situation. As of December 2005, Germany had 10.7 million broadband subscribers using DSL (OECD, 2005), which translates into almost one-third of all households. In March 2005, India had a total of 815 thousand broadband households of which 595 thousand have DSL lines and 220 thousand users—have cable lines 809
Impact of Broadband VoIP on Telecoms
7DEOH7HOHFRPPXQLFDWLRQVWDWLVWLFV±*HUPDQ\DQG,QGLD Nation
Population
Population
Main
Main
Internet users
Internet
Mobile
Mobile
(total, M)
density (per
telephone
telephone
(000s)
users per 100
subscribers
subscribers
km2)
lines (000s)
lines per 100
inhabitants
(000s)
per 100
inhabitants
inhabitants
Germany
82.69
232
55,046
66.57
37,500
45.35
79,200
95.78
India
1,103.37
348
49,750
4.51
60,000
5.44
90,000
8.16
(Search Agency, 2006). At the end of November 2005, there was a total of 7,5 million subscribers for broadband in India, well below the target of TRAI (2006). These statistics need to be analyzed in the context of VoIP, as broadband is considered the basic infrastructure for its deployment. Ty pically, broadband VoIP telephony eliminates the need for building/maintaining another infrastructure and lowers the cost of telecommunications for both consumers and EXVLQHVV 6SHFL¿FDOO\ LI WKHUH LV DQ HVWDEOLVKHG broadband connection, VoIP can automatically be provisioned as part of the service. The need for the telecommunications provider to add a line rental itemized calling charges and service fee is negated, thus making it a cheaper option. However, before taking a deeper look into the infrastructure for VoIP, we pose the question: how relevant/viable is VoIP in the context of the two nations? To answer WKLVTXHVWLRQZH¿UVWFRQVLGHUHG³ZKDWLVUHJDUGHG as broadband?” in both nations. In Germany, broadband speed is 1 Mbit/s or above, while in India it is 256 Kbit/s. When the bandwidth is narrow, it is less likely that VoIP will gain momentum or acceptance. The German telecom market is deregulated as against the Indian market, which although deregulated and privatized, is still controlled by TRAI. In Germany, the teleservices are subject to the Teleservices Act which does not require QRWL¿FDWLRQRUOLFHQVLQJ *OREDO,3$OOLDQFH The nation is progressing towards a complete switch to VoIP services considering the mass market diffusion of broadband and the popularity of Skype which is threatening the existing ISPs. In India, VoIP telephony is regulated by TRAI
810
(2006), which only permits its usage for making PC-PC based calls within India/overseas, PC to Phone calls overseas and IP based H.323/SIP terminals in India to similar terminals both in India and abroad. VoIP still is not permitted to be used as an alternative to PSTN within India. The restriction of the domestic provisioning of VoIP to only PC-PC calls is protecting the existing telecom providers, who are multiple across varied telecom circles. The low upload speed restricts the nation from switching to a complete broadband VoIP infrastructure regardless of the regulation. 6SHFL¿FDOO\ 9R,3 UHTXLUHV D FHUWDLQ EDQGZLGWK without which it is not a viable solution. There are approximately 80 VoIP providers in Germany (GerVoIP, 2006). Until recently the trend was to offer a VoIP subscription for DSSUR[LPDWHO\¼WRKDYHXQOLPLWHGFDOOVZLWKLQ Germany, with a number provided free with the broadband bundle. Recently, the trend has been H[WHQGHGVXFKWKDWIRUDÀDWIHHRIDSSUR[LPDWHO\ ¼WKH9R,3LVQRZDYDLODEOHIRUWKHZKROHRI Europe. With increasing penetration of broadband and with advanced support of the European Union, VoIP is set to replace the existing PSTN services. Conversely, VoIP Provider Directory (2006) shows approximately 75 VoIP providers within India. Some of the current offerings are such that there is no monthly fee with equipment provided DQGZLWKORQJGLVWDQFHUDWHVRI¼SHUPLQXWH (Webacall, 2006). The reliability of this service is however quite low. Yet others (VoIP.com, RIIHUVXEVFULSWLRQVUDQJLQJIURP¼WR ¼PRQWKO\IHHZLWK ¼IRUORQJGLVWDQFH calls—with a restriction of 600 VoIP call minutes per month. Subscribers are offered unlimited calls
Impact of Broadband VoIP on Telecoms
to US and Canada; and unlimited international calls to anywhere else in the world to the same ISP subscribers. The SIP standard seems to be the most popular within India as it allows for domestic/international VoIP service, free of charge, between the same ISPs. However, it has to be noted that these VoIP subscriptions are mainly held by the high income earning, upper middle class population, which can afford to maintain a broadband infrastructure. Mainly, this population is housed in the urban areas or semi-urban areas with booming businesses. Hindustantimes.com (2006) comments that VoIP could prove to be the most effective and cheapest means of connecting rural India to the rest of the nation. In these areas, the cable infrastructure is already in place, which could easily be upgraded to the provision of offering broadband and VoIP. Capturing this sentiment, there is also a move WRZDUGVXQL¿HGOLFHQVLQJRI9R,3VHUYLFHVE\WKH government. However, there are many issues that need to be considered before VoIP can even become a viable alternative for India. First, the speed that is currently touted as broadband is 256Kbp/s, which is not feasible for a reliable VoIP service. Second, the infrastructure for VoIP itself is the mass market penetration of broadband which still remains a distant reality for India. If there was no telecommunications infrastructure, a complete switch to broadband VoIP would have been feasible as a “leap frogging” technology, eliminating the need to build another infrastructure for the nation. However, the current telecommunications infrastructure, which was a monopoly under strict control of a regulatory authority until a decade ago, had built a disparate infrastructure although it did not connect the whole nation. Conversely, extending the current infrastructure and maintaining is a necessity to keep the communications alive as well as ensuring employability to a populous nation with an almost 50% semiskilled population. When the mass market is considered in the highly developed telecommunications market of Germany, with high broadband diffusion, the existing telecom providers are competing with free VoIP services. The consumer acceptance of new
technologies has been rapid and the market is ready for a switch to VoIP, eliminating the long term need for maintenance of a separate infrastructure for telephones. Conversely, in India, Internet and broadband are far from attaining mass-market penetration. While the costs of having a telephone line itself is prohibitive, most of the mass market consumers do not own a PC at home nor have they a broadband infrastructure per household. Whether it is the young consumer market housed in the urban areas or low income/semiliterate households in rural areas, the main access points are telephone/Internet booths and cyber cafés. The government-owned public booths have been franchised or licensed to varied ISPs depending on the region while the cyber cafés are operated by ISPs. While the cable infrastructure is strong in India, whether it is in rural or urban areas, the need for a broadband/Internet access is low in the “not so FRQQHFWHG´UXUDODUHDV6SHFL¿FDOO\WHOHYLVLRQLV almost all pervasive, which has resulted in the setting up of a cable infrastructure in many rural areas that house illiterate/semiliterate population. However, Internet via broadband may not be in demand at all in these areas. The illiterate population does not perceive the need for Internet access at all, while the semiliterate population access the Internet mainly via Internet booths/cyber cafés, as they still need peer support to browse, navigate, or utilize Internet facilities. Conversely, the reasoning that VoIP can be an alternative to voice is now eliminated by low cost provisioning of mobile services across the nation, including the previously not connected rural areas. For example, as against 2005 statistics reported in Table 3, the latest reports indicate that the number of mobile subscriptions passed the 100 million mark in 2006 (emarketer, 2007). Similarly, Ahmed (2006) reported that there are 50 million Internet users at the end of August 2006, which places India among the fourth largest market in the world. Most of the young generation market is pushing for the growth of full length movies, clips, and 3G mobile services in India. However, in the case of both economies, there seems to be a consistent move towards upgrading
811
Impact of Broadband VoIP on Telecoms
existing facilities for broadband access, to increase penetration. Now, we take a closer look at the promising potential of the emerging broadband access technologies in the relevant economy contexts. VDSL seems to be almost a revolutionary EUHDNWKURXJKIRU*HUPDQ\7KH¿EHURSWLF9'6/ network capable of offering telephony, data services, television, and multimedia facilities seem to hold the major potential for this economy. Having to support one infrastructure seems to be a promising venture for providers. In a developed infrastructure, the upgrade to the VDSL network seems not a distant dream. Once the ISPs are able to work out their market shares and pricing, this infrastructure will eliminate the need for conventional telephony, thus potentially replacing it with VoIP services. &HUWL¿HG:L0$;VHHPLQJO\LVWKHWHFKQRORJ\ that seems to be touted as promising in India. It KDVWREHQRWHGWKDWWKHFHUWL¿HG:L0$;EHLQJ touted/promoted as a new technology for broadband access, holding its promises to offer similar to VDSL infrastructure in Germany including data services, IPTV, and so forth. In the Indian context, VDSL is not a last mile solution. However, even :L0$;FDQRQO\EHDSSOLHGZKHUHUHOHYDQWDQGWR promote access where required. In the rural areas connected with cable infrastructure, it may be more relevant to upgrade this infrastructure. Conversely, the 3G mobile infrastructure may hold even more potential as the move towards making this infrastructure cheaper has resulted in low cost provisioning of Internet, especially via 3G deployments. In this context, VoIP may not be an alternative to normal telephony. If the new 3G telephone instruments are equipped with the SIP standard, capable of VoIP, in the long term this could result in even lower costing VoIP calls, in this developing nation. To sum up, the Indian LQIUDVWUXFWXUHLVLQVXFKDVWDJHWKDWWKHFHUWL¿HG :L0$; DQG * PRELOH EDVHG DFFHVV PD\ EH the suitable broadband infrastructure that will eventually facilitate low cost VoIP. ,W KDV WR EH QRWHG WKDW 9'6/ :L0$; and 3G mobile services facilitate VoIP, but are different in terms of technologies and facilitation. The requirements in each of these categories are 812
relevant in both economies examined, although at varied levels.
OUTLOOK From the brief analysis, both of the nations considered are at different levels of requirement and feasibility as far as VoIP is concerned. Germany has achieved mass market broadband penetration and acceptance of VoIP. The telecommunications providers are now competing with free VoIP providers such as Skype. The nation seems to be ready for a wider switch to broadband and VoIP, eliminating the need for maintaining a separate infrastructure for telecommunication via VDSL network. However, it is yet to be seen whether the nation will completely make the transition, similar to New Zealand which is in the transition to an exclusive VoIP infrastructure (Katsenelson, 2006). In comparison, India is at a stage where the existing infrastructure has to be maintained until broadband reaches higher speed levels that enable better VoIP and mass market penetration. While the rural area infrastructure can be switched easily into broadband per household via the existing cable networks, the demand for Internet access in households is not high enough. Conversely, the low cost mobile communication has long overtaken the ¿[HGOLQHLQIUDVWUXFWXUHIRUYRLFHFRPPXQLFDWLRQ in rural areas, reducing the demand for providing DQRWKHULQIUDVWUXFWXUH6SHFL¿FDOO\WKHQDWXUHRI Internet access is mainly via public booths/cyber cafés in India, rather than through a household broadband connection. Therefore, technically, a VoIP will not be able to replace the need for having a telephone per household. Nevertheless, there is an increasing demand for public accessible Internet booths/cyber cafés with broadband facilitation which may drive up the need for VoIP services, although a complete switch to VoIP may not be a feasible option for this economy. 7KH SURYLVLRQLQJ RI 9R,3 YLD :L0$; KDV potential for both economies. According to Intel D :L0$;UHSUHVHQWVDJOREDOFRQQHFWLYLW\ opportunity in highly developed mobile market segments and in developing countries the
Impact of Broadband VoIP on Telecoms
technology may help provide affordable broadband VHUYLFHV,Q*HUPDQ\:L0$;KDVDOUHDG\SURYHQ to be cost effective for businesses (Intel, 2006b). ,Q,QGLD:L0$;PD\SURYHWREHWKHDIIRUGDEOH broadband technology which will increase the uptake in not well connected areas. Our analysis also reveals that the nature of VoIP is such that it holds potential and promises for the future. The telecommunication providers in both developed and developing economies cannot ignore this technology, as it is bound to gain momentum. Telecommunication providers therefore may eventually need to switch or reinvent themselves into provisioning of new services. While in the highly mature market of Germany this seems apparent, a developing nation with many complexities such as India can not ignore the potential of this technology. The fact that TRAI is regulating the use of VoIP LQ,QGLDLWVHOILVVLJQL¿FDQWSURRIWKDWWKHHFRQRP\ is aware that as the mass market matures—an eventual shift, at least partially, is likely. As the market matures, the telecommunications providers will need to share the marketplace with VoIP providers, or alternatively reinvent themselves by providing competitive services. Maybe the role of telecommunication companies is shifting with the core business being infrastructure providers. Although the political climate may have an impact on the eventual shift in both economies, it is apparent that VoIP is reshaping the telecommunications industry in both cases. As of now, there are economies such as UAE (2006) which have banned VoIP services as against New Zealand which is moving away from PSTN to only VoIP services. Based on the existing evidence, the shift towards VoIP is apparent, although not completely, which is bound to reshape the telecommunications industry in the future.
REFERENCES Ahmed, R. (2006). 50 million Internet users in India: Not a big market? Retrieved July 30, 2007, from http://www.indianraj.com/2006/08/50_million_internet_users_in_i.html
Alfonso, A. (2005). Telecom Industry revenue to reach $1.2 trillion in 2006. VoIP Magazine. Retrieved July 30, 2007, from http://www.voipmagazine.com/content/view/1197 AT&T. (2006a). Voice over IP security, what are the risks and solutions? Point of View. Retrieved July 30, 2007, from http://whitepapers.businessweek. com/detail/RES/1146587317_458.html $7 7 E 7KHEXVLQHVVEHQH¿WVRIVHVVLRQ initiation protocol. Point of View. Retrieved July 30, 2007, from http://whitepapers.techrepublic. com.com/abstract.aspx?docid=166488&promo= 300111 Broadbandguide. (2007). What is cable broadband and how does it work? Broadband Guide. Retrieved July 30, 2007, from http://www.broadbandguide. com.au/Browse/By-Type/Cable.aspx Broadbandreports. (2007). Broadband reports. com, EU sues Germany over broadband laws. Retrieved July 30, 2007, from http://www.dslreports. com/shownews/81886 Budde, P. (2007). 2006-2007 Western European broadband market. Retrieved July 30, 2007, from http://www.marketresearch.com/product/display. asp?productid=1422244&g=1 %XQGHVQHW]JHQWXU 2I¿FLDOGDWD5HWULHYHG July 30, 2007, from www.bundesnetzagentur.de
Cable Quest. (2007). Cable TV. Cable Quest Magazine. Retrieved July 30, 2007, from http://www. cable-quest.in/cable_tv.asp Cain,W.M., & Munch, B. (2005, September 2). Instant messaging interoperability: The battle over standards creates chaos. Retrieved July 30, 2007, from http://www.gartner.com/ DisplayDocument?doc_cd=130603 CNBC. (2007). EU launches lawsuit against Germany. Retrieved July 30, 2007, from http://www. msnbc.msn.com/id/17343675/for/cnbc Continuity Central. (2006). Emerging threats 2006: VoIP. Continuity Central White Paper. Retrieved
813
Impact of Broadband VoIP on Telecoms
July 30, 2007, from http://www.continuitycentral. com/feature0296.htm
www.voipnow.org/2006/01/voip_has_a_future. html
DSL Forum. (2007a). Learning about DSL. Retrieved July 30, 2007, from http://www.dslforum. org/learndsl/glossary.shtml
,QWHO D :L0$; EURDGEDQG ZLUHOHVV technology access. Retrieved July 30, 2007, from http://www.intel.com/netcomms/technologies/ wimax/index.htm
DSL Forum. (2007b). Learning about DSL – about ADSL. Retrieved July 30, 2007, from http://www. dslforum.org/learndsl/adslfaq.shtml DSL Forum. (2007c). Learning about DSL – about VDSL & VDSL2. Retrieved July 30, 2007, from http://www.dslforum.org/learndsl/adslfaq.shtml emarketer. (2007). Indian wireless market booming. Retrieved July 30, 2007, from http://www.emarketer. com/Article.aspx?1004561&src=article_head_ sitesearch )URVW 6XOOLYDQ $SULO 8QL¿HGFRQIHUHQFing and collaboration Market (F430-64). Retrieved July 30, 2007, from http://www.frost.com/prod/ servlet/report-overview.pag?repid=F430-01-0000-00&ctxst=FcmCtx7&ctxht=FcmCtx8&ctxhl =FcmCtx9&ctxixpLink=FcmCtx10&ctxixpLabe l=FcmCtx11 GerVoIP. (2006). Germany VoIP provider listings. Retrieved July 30, 2007, from http://www. myvoipprovider.com/Directory/VoIP_Provider_ Directory/VoIP_Providers/Germany_-_VoIP_ Providers GIPA. (2006). Global VoIP policy status matrix. Retrieved July 30, 2007, from http://www.ipall. org/matrix/#table0 Harkins, J. (2004). Voice over IP: Some accessibility issues. 7KH%OXH%RRN 7',QDWLRQDO directory and resource guide (pp. 21-23). Silver Spring, MD: TDI. Retrieved July 30, 2007, from http://tap.gallaudet.edu/VoIP.htm Herrel, E. (2006, June 8). Enterprise IP telephony plans in 2006. Retrieved July 30, 2007, from http://www.forrester.com/Research/Document/ Excerpt/0,7211,39454,00.html Hindustan Times. (2006, January 2). VoIP has a future in India. Retrieved July 30, 2007, from http://
814
,QWHO E :L0$; KHOSV VHUYLFH SURYLGHUV serve millions of new customers. Retrieved July 30, 2007, from http://www.intel.com/netcomms/ technologies/wimax/IMA93443GermanyBrief. pdf Intertangent. (2006) Advantages and disadvantages of VoIP. Retrieved July 30, 2007, from http:// www.intertangent.com/023346/Articles_and_ News/1409.html Katsenelson, V. (2006). Government goes medieval on half a broadband. Motley Fool Income Investor. Retrieved July 30, 2006, from http://www.fool.com/ news/commentary/2006/commentary06050509. htm Kouroupas, P. (2006). Critical public policy issues impacting the development and delivery of VoIP services. 3DSHU SUHVHQWHG DW 3DFL¿F 7HOHFRPPXQLFDWLRQ&RXQFLO37&¶ Hawaii, USA. Ledford, J.L. (2006). Cut the cord! the consumer’s guide to VoIP. Canada: Thomson Course Technology. Lemon, S. (2006). VOIP presents major security risk. IDG News Service. Retrieved July 30, 2007, from http://www.infoworld.com/archives/emailPrint.jsp?R=printThis&A=/article/06/09/20/HNvoipsecurityrisk_1.html Lenert, E. (1998, December). A communication theory perspective on telecommunications policy. 48(4), 3-23. 0DOLN2 ,Q,QGLD:L0$;WDNHVZLQJV Retrieved July 30, 2007, from http://gigaom. com/2006/12/14/in-india-wimax-takes-wings Marsan, C.D. (2006). IETF taking on 911 problems within VoIP. Retrieved July 30, 2007, from http://
Impact of Broadband VoIP on Telecoms
www.computerworld.com.au/index.php/id;31575 5031;fp;4;fpid;19
Skype. (2006a). Skype. Retrieved July 30, 2007, from http://www.skype.com/helloagain.html
Maslog-Levis, K. (2004). KPMG: Watch out for VoIP risks. ZdNet. Retrieved July 30, 2007, from http://www.silicon.com/research/specialreports/ ip/0,39025013,39125380,00.htm
Skype. (2006b). DUALphone 3088. Retrieved July 30, 2007, from http://accessories.skype.com/ item?SID=c09df3f4dd89b65589763758c5e8b878 3d9:4580&sku=951020
MyNetFone. (2006). MyNetFone plans. Retrieved July 30, 2007, from http://www.mynetfone.com. au/residential/
Socialtext. (2006). Broadband in India. Retrieved July 30, 2007, from http://www.socialtext.net/ broadband/index.cgi?india
MySme. (2006). Internet services in Germany. Retrieved July 30, 2007, from http://www.mysme. de/page36.html?&MMN_position=12:11
Taylor, M., & Coombs, V. (2006, September 1). Technology terms—VoIP. London Computer Technology Law Report (Technology, Media and Telecoms Group Lovells). Retrieved July 30, 2007, from http://www.lovells.com/Lovells/MediaCentre/Articles/Technology+Terms+-+VoIP.htm
OECD. (2005). OECD broadband statistics. Retrieved July 30, 2007, from http://www.oecd.org/ document/39/0,2340,en_2649_201185_36459431_ 1_1_1_1,00.html Olorunda, O., & Olorunda, A. (2006). Bridging the digital divide – the social and cultural impact of VoIP in developing countries: Nigeria as a case study 3DSHU SUHVHQWHG DW 3DFL¿F 7HOHFRPPXQLFDWLRQ&RXQFLO37&¶+DZDLL USA.
Telecom India. (2007). Government frees 5.8 *+]IRU:L0$;5HWULHYHG-XO\IURP http://telecom.broadbandindia.com/2007/02/government-frees-58-ghz-for-wimax.html TRAI. (2006a). Telecom regulatory authority of India. Retrieved July 30, 2007, from http://www. trai.gov.in
PUC. (2006). Customer facts VoIP, public utility commission of Texas. Retrieved July 30, 2007, from www.puc.state.tx.us/ocp/telephone/telefacts/ VoIP.pdf
TRAI. (2006b). Study paper no.2/2006 on analysis of Internet & broadband tariffs in India. Retrieved July 30, 2007, from http://www.trai.gov.in/StudyPapers_list_year.asp
Ribeiro. (2007). Lack of spectrum blamed for WiMax delay in India. Retrieved July 30, 2007, from http://www.infoworld.com/archives/emailPrint.jsp?R=printThis&A=/article/07/01/24/ HNwimixindiadelay_1.html
UAE. (2006). UAE TRA defends VoIP ban. Retrieved July 30, 2007, from http://www.ameinfo. com/98354.html
Richardson, T. (2005, January 20). Skype VoIP threat to Euro telcos. Retrieved July 30, 2007, from http://www.theregister.co.uk/2005/01/10/ skype_euro_telcos Satsig. (2006). Satellite signals. Retrieved July 30, 2007, from http://www.satsig.net/ivsat-europe. htm 6HDUFK$JHQF\ 5HJLRQVSHFL¿FV5HWULHYHG July 30, 2007, from http://www.thesearchagency. com/ResourceLibrary/search_engine_stat.aspx?b FUXP 5HJLRQDO6SHFL¿FV V&DW,G
V&D. (2006). Regulating VoIP for emergency services. Voice and Data Online. Retrieved July 30, 2007, from http://www.voiceanddata.com. au/feature_article/item_022006d.asp VoIP Foro. (2006a). VoIP protocols, H.323. VOIP Forum. Retrieved July 30, 2007, from http://www. en.voipforo.com/H323/H323_objectives.php VoIP Foro. (2006b). H.323. VOIP Forum. Retrieved July 30, 2007, from http://babel.altavista.com/ EDEHO¿VKWUXUOBSDJHFRQWHQW"OS HVBHQ WUXUO KWW p%3a%2f%2fwww.voipforo.com%2fH323%2fH 323objetivo.php
815
Impact of Broadband VoIP on Telecoms
VoIP Foro. (2006c). SIP. VOIP Forum. Retrieved July 30, 2007, from http://babel.altavista.com/ EDEHO¿VKWUXUOBSDJHFRQWHQW"OS HVBHQ WUXUO KWW p%3a%2f%2fwww.voipforo.com%2fSIP%2fSIP arquitectura.php VoIP Provider Directory. (2006). VoIP providers in India. Retrieved July 30, 2007, from http://www. voipproviderslist.com/country/voip-india/voipproviders-india Voip.com. (2006). VoIP plans for India. Retrieved July 30, 2007, from http://www.voipreview.org/service.all2.aspx?Country=99&Area_Code=0&Calli ngArea=0&provider=0&serviceType=1&Adv=0 Wallingford, T. (2005). What is VoIP? White Paper. O’Reilly Network. Retrieved July 30, 2007, from http://www.oreillynet.com/pub/a/network/2005/09/02/what-is-voip.html Webacall. (2006). VoIP plans for India. Retrieved July 30, 2007, from http://www.voipreview.org/service.all2.aspx?Country=99&Area_Code=0&Calli ngArea=0&provider=0&serviceType=1&Adv=0 :L0$; ,QGLDEURDGEDQGDQG:L0$; market analysis. Retrieved July 30, 2007, from http://www.wimax.com/commerce/catalog/ wimax_india_report :L0$;'D\ D :L0$;VSHFWUXPDXFWLRQ stirs debate in Germany. Retrieved July 30, 2007, from http://www.wimaxday.net/site/2006/10/09/ wimax-spectrum-auction-stirs-debate-in-germany :L0$;'D\ E 6L[¿UPVDSSO\IRU:L0$; in Germany, November 10th 2006. Retrieved July 30, 2007, from http://www.wimaxday.net/ site/2006/11/10/six-apply-for-wimax-in-germany :L0$; 9LVLRQ ,QGLD WDNHV IXUWKHU VWHSV WRZDUGV :L0$; :L0$; YLVLRQ 5Htrieved July 30, 2007, from http://www.wimaxvision.com/newt/l/wimaxvision/viewarticle. html?artid=20017407693 Wisegeek. (2006). What is HSDPA. Retrieved July 30, 2007, from http://www.wisegeek.com/what-ishsdpa.htm
816
KEY TERMS 3G/UMTS: Refers to third generation/Universal Mobile Telecommunication Systems which are standardized by European Telecommunications Standards Institute (ETSI) under 3GPP (third generation partnership project) along with other regional standards organizations. 3GPP is the standardization structure producing UMTS VSHFL¿FDWLRQV DQG JOREDO V\VWHPV IRU PRELOH communications (GSM). Broadband: Refers to advanced communications systems capable of providing high-speed transmission of services such as data, voice, and video over the Internet and other networks. Transmission is enabled by a wide range of technologies, including digital subscriber line '6/ ¿EHURSWLFFDEOHZLUHOHVVDQGVDWHOOLWH H3.23: Refers to the oldest of signaling protocols, recommended by the International Telecommunications Union. It was designed to provide teleconferencing with voice, video, and data capacities on packet switching networks. POTS: Refers to plain old telephony systems or traditional analogue based public telecommunication services. PSTN: Refers to public switched telephone network, which is the conventional telephone service that exists today. SIP: Refers to sessions initiation protocol. It is the proposed standard for initiating, modifying, and terminating an interactive user session that involves multimedia elements such as video, voice, instant messaging, online games, and virtual reality. It is an application-layer control protocol, a signaling protocol for Internet telephony. It can establish sessions for features such as audio/ videoconferencing, interactive gaming, and call forwarding to be deployed over IP networks. Skype: Refers to a proprietary VoIP organization which has created the most popular VoIP instant
Impact of Broadband VoIP on Telecoms
messaging tool, known as SKYPE, using P2P technologies. VDSL: Refers to very high speed digital subscriber lines. It is the latest development in DSL, which enables high speed Internet access of up to symmetrical 100 Mbps (both up and downstream) DQG LV VSHFL¿HG WR VXSSRUW DSSOLFDWLRQV VXFK DV PXOWLFKDQQHOKLJKGH¿QLWLRQ79 +'79 YLGHR on demand, videoconferencing, and VoIP, all using the existing ubiquitous copper telephone line infrastructure.
to be used for voice applications, such as telephony, voice instant messaging, teleconferencing, and so forth. WiMAX: Refers to world interoperability for microwave access. It is a technology built to deliver high throughput broadband wireless services at a low cost, to enable mass market adoption. There DUHWZRPDLQDSSOLFDWLRQV¿[HGDQGPRELOH)L[HG :L0$;LVDSRLQWWRPXOWLSRLQWIDFLOLW\WKDWHQables broadband access to homes and businesses. 0RELOH :L0$; RIIHUV IXOO PRELOLW\ RI PRELOH networks at true broadband speeds.
VoIP: Refers to voice over Internet protocol. It is the family of technologies that allow IP networks
817
818
Chapter XLIX
The Adoption of Broadband Internet in Australia and Canada Catherine Middleton Ryerson University, Canada Shanton Chang University of Melbourne, Australia
ABSTRACT %URDGEDQG,QWHUQHWFRQQHFWLYLW\LVVHHQDVDPHDQVWRLQFUHDVHWKHHI¿FLHQF\DQGFRPSHWLWLYHQHVVRIDQ economy. But despite ongoing efforts to promote broadband in Australia, uptake has been much slower than expected. This chapter aims to identify areas that have been holding up the broadband development in Australia. In examining multiple areas for attention (competition, user characteristics and behaviors, applications, network characteristics, and pricing), we refer to the experience of Canada, a leader in broadband deployment, to show the differences in each area. The chapter outlines objectives for the development of a more user-friendly broadband environment in Australia, which would encourage broadband adoption. Although both countries discussed here have their own policy agendas and some unique circumstances related to broadband deployment, the chapter provide valuable insights for policy makers and industry leaders in Australia, and in other countries which are struggling to develop widespread broadband deployment.
INTRODUCTION Broadband has been considered as a key to enhancing competitiveness of an economy and sustaining economic growth (International Telecommunication Union, 2001, 2003c; OECD Directorate for Science Technology and Industry, 2001, 2002). The Commission of the European Communities (2006) states that broadband is “crucial for fostering growth and jobs” (i2010, 2006). Gillett, Lehr, Osorio, and Sirbu (2006) provide some preliminary
evidence to show that broadband access does result LQSRVLWLYHHFRQRPLFEHQH¿WVEXW)UDQVPDQ notes that there is very little evidence to justify the FODLPHGEHQH¿WVRIEURDGEDQGDGRSWLRQ1HYHUWKHless, there is no doubt that governments around the world are committed to extending broadband networks to their citizens (Broadband Advisory *URXS 2I¿FH RI WKH H(QYR\ 7DVN Force on Broadband Communications, 2002). The extent and speed of broadband adoption has varied widely across nations.
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
The Adoption of Broadband Internet in Australia and Canada
The Australian government has been keen to deploy broadband across the country to build a foundation for the information society or knowlHGJH HFRQRP\ DQG WKHUHE\ HQKDQFH $XVWUDOLD¶V national competitiveness (Broadband Advisory Group, 2003). When compared to other countries, however, Australia is far behind in this race of broadband adoption,1 despite strong government support for broadband development over the past decade (Australian Information Economy Advisory Council, 1999; BSEG, 1994; DCITA, 2004). 7KHODWHVW2(&'¿JXUHV E VKRZQLQ7DEOH 1, ranked Australia in 17th place for broadband subscriptions, among the 30 OECD countries. This is an improvement from 21st place in 2004, and a PRYHDERYHWKH2(&'DYHUDJHIRUWKH¿UVWWLPH However, countries like Australia and the U.S. continue to lag behind Korea, Canada, and Scandinavian and European countries (e.g., Iceland, Netherlands, Denmark). It is noted that in 2006, Australia remained in the lower half of the ranks while Canada was superseded by a number of other QDWLRQVLQWKHUDQNLQJV7KH8QLWHG6WDWHV¶SRRU performance has been subject to much scrutiny, with Bleha (2005) suggesting that consumers there have broadband services that are “among the slowest, most expensive, and least reliable in the developed world” (p. 111).
Despite the fact that up to 7 million Australian homes, housing 91% of the population, are ADSL2enabled (meaning that broadband access can be provided over existing telephone lines) (Houghton & Morris, 2001), the one million subscriber milestone was reached only at the end of June 2004 (ACCC, 2004). It is reported that Australia is 2 years behind other developed countries like the U.S. and Canada (Riley, 2004), a comparison that only considers the number of subscribers. When the quality of services (i.e., speed) is considered, the gap is even larger. Whereas access speeds of 1 Mbps (megabits per second) or above are the norm in leading countries (with speeds in excess of 50 Mbps available in countries like Japan and Korea) (International Telecommunication Union, 2006a), standard broadband plans in Australia provide speeds of only 256 Kbps (kilobits per second). Furthermore, there is a strict download cap applied to Australian broadband services which discourages users from becoming active surfers. Although the number of broadband subscribers in Australia is increasing steadily, a 2004 report by IDC predicts that Australia will remain in the broadband “backwater” when compared against other developed countries (IDC, 2004). The purpose of this chapter is to explore the conditions for broadband adoption in Australia to
Table 1. Broadband subscription ranking of OECD countries (excerpt) 2002 2002-2006 June
2003
2004
2005
2006
2002
Rank
2003
Rank
2004
Rank
2005
Rank
2006
Rank
Denmark
6.7
4
11.11
4
17.0
2
21.8
3
29.3
1
Netherlands
3.9
10
9.2
6
15.6
4
22.5
2
28.8
2
Iceland
4.7
7
11.22
3
15.5
5
21.7
4
27.3
3
Korea
19.1
1
23.17
1
24.4
1
25.5
1
26.4
4
Canada
10.2
2
13.27
2
16.7
3
19.2
6
22.4
9
United Kingdom
1.3
19
3.63
18
7.4
15
13.5
13
19.4
10
United States
5.6
6
8.25
10
11.2
11
14.5
12
19.2
12
Australia
1.3
18
2.65
20
5.3
21
10.9
17
17.4
17
OECD average
3.8
6.06
8.6
11.8
13.55
Source: Organization for Economic Co-operation and Development (2006b); Broadband access per LQKDELWDQWVDQQXDO¿JXUHVPHDVXUHGLQ-XQHRIHDFK\HDU
819
The Adoption of Broadband Internet in Australia and Canada
better understand why Australian consumers have not adopted broadband as rapidly as consumers elsewhere in the OECD. The focus on residential broadband users is appropriate because the anWLFLSDWHGVRFLHWDOEHQH¿WVRIEURDGEDQGQHWZRUNV can only be realized if such networks are widely adopted by consumers. The chapter draws on the experiences of broadband deployment in Canada, offering a point of comparison for the Australian situation. On a cultural and geographic basis, Australia and Canada are similar, allowing for valuable insights to be gained from considering broadband adoption patterns in Canada. The chapter outlines the differences between the two nations, illustrating how different contextual factors impact adoption. The chapter provides a starting point for discussion of how to encourage increased broadband uptake by Australians and by those in other countries where broadband adoption has been slower than anticipated. Previous research on national broadband diffusion shows no clear determinants of broadband uptake (e.g., Kim, Bauer, & Wildman, 2003), and highlights the complexities of trying to conceptualize factors that contribute to broadband rollout (Bouwman, Fijnvandraat, & van de Wijngaert, 2006). While various factors are thought to be important, no clear outcomes are drawn when compared on a statistical basis. This chapter takes a less quantitative focus, drawing on lived experiences with broadband in the countries being compared. The chapter begins with a review of literature that addresses the determinants of residential broadband supply and demand. The literature review is followed by a discussion of why studying broadband diffusion in Canada can provide useful insights to help advance broadband adoption in Australia. Broadband development in each country is described, followed by a discussion of differences LQWHUPVRIWKHGHWHUPLQDQWVZHKDYHLGHQWL¿HG LQ WKH OLWHUDWXUH UHYLHZ ,Q WKH ¿QDO VHFWLRQ ZH conclude and present implications for Australia and other countries.
820
LITERATURE REVIEW The uptake of any technology is dependent upon both supply of, and demand for, that technology. Much research has been done in the past few years to better understand the determinants of broadband network supply and demand. As this section VKRZVKRZHYHULWLVGLI¿FXOWWRGUDZFRQFOXVLRQV about the relationships among various supply and demand variables such as price, regulatory policy, and characteristics of the potential adopters, especially when attempting to compare adoption patterns across countries (Kim, Bauer, & Wildman, 2003). Data on broadband supply and demand is frequently proprietary and publicly available data (e.g., International Telecommunication Union, 2006b; Organization for Economic Co-operation DQG'HYHORSPHQWD GRQRWRIIHUVXI¿FLHQW detail to investigate individual supply and demand determinants.
Determinants of Broadband Supply With some exceptions (e.g., LAN-based broadband LQ6ZHGHQ¿EHUWRWKHKRPHDQG¿[HGZLUHOHVV service to remote areas, see International Telecommunication Union, 2003a), most residential consumers who have access to broadband get this access using a telephone line (DSL service) or a cable modem. Theoretically, most urban homes in developed countries can get either DSL or cable modem broadband service (or both), but in practice the actual availability of broadband service is dependent upon a service provider making a decision to offer the service. Supplying consumers with broadband requires substantial infrastructure investment, and while such investment has been ongoing for some time now (Fransman, 2006; Shelanski, 1999), providers still need to consider whether or not they will offer broadband services in a particular area. The availability of broadband LV LQÀXHQFHG E\ WKH QDWXUH RI WKH PDUNHWSODFH GH¿QHGEHORZDV³XVHUFKDUDFWHULVWLFV´ DQGE\ regulatory policy.
The Adoption of Broadband Internet in Australia and Canada
User Characteristics In a study conducted in the U.S. state of Ohio, Grubesic (2003) found that household density, location (rural or urban), education, and income levels were determinants of xDSL supply. He also found that regions with more older residents were less likely to be supplied with broadband. Prieger UHSRUWVVLPLODU¿QGLQJVIRUVXSSO\RI'6/ and cable broadband services in the U.S., noting that income alone was not a reliable indicator of broadband supply. The ITU (2003a) and CavaFerreruela and Alabau-Muñoz (2006) also report that the supply of broadband is related to national income, urbanization, and population density. 7KHVH¿QGLQJVDUHQRWVXUSULVLQJDVWKH\LQGLFDWH that providers consider the potential demand for broadband services when determining whether or not to supply a particular market.
Regulatory Policies: Local Loop Unbundling and Open Access Local loop unbundling requires that the incumbent telephone provider allow competitors access to its existing network infrastructure (OECD Directorate for Science Technology and Industry, 2003a, 2003b), and it is frequently mandated by telecommunications regulatory agencies. Unbundling allows competitors to offer broadband services over existing infrastructure. There is mixed opinion as to the impact of local loop unbundling (LLU) on the supply of broadband (Ure, 2003). OECD research (2001) suggests that the presence of LLU in a market promotes intramodal competition (i.e., FRPSHWLWLRQDPRQJ['6/SURYLGHUV EXW+RZHOO¶V (2002) analysis of the OECD data does not support this contention. More recent OECD research (2003b, p. 6) observes that “arguments that structural separation of the local loop is necessary are inconclusive.” The presence of LLU regulation does not guarantee that potential competitors will actually use the unbundled loop to provide broadband, meaning that the presence of LLU in a broadband market does not necessarily increase the supply of broadband to consumers in that market. One instance where LLU
has increased price competition is Japan, where YahooBB! offers a competitive DSL service by accessing the unbundled local loop (International Telecommunication Union, 2003a). In the cable broadband market, open access or “third party Internet access” rules operate analogously to local loop unbundling, allowing new entrants to purchase broadband capacity from existing cable providers and resell it to consumers. To date there appears to be little evidence of open access to cable networks increasing broadband supply. Rather than the intramodal competition that would result from LLU or open access to cable networks, increased supply more often comes as a result of intermodal, facilities-based competition. Intermodal competition is competition between DSL and cable providers within a single market, and is frequently mentioned as a primary reason for the development of extensive broadband capacity LQ6RXWK.RUHD /HH2¶.HHIH <XQ
Summary In high-income countries around the world, broadband is widely available (International Telecommunication Union, 2003a), although it is noted that the speed of service supplied does vary dramatically (see International Telecommunication Union, 2006a, especially Table 7, on this point). Although many small remote communities in countries with low population densities (e.g., Canada and Australia) are not yet served by commercial broadband providers (Industry Canada, 2002; Sale, 2001), in general, broadband supply is not constrained. As broadband networks are generally accessible to consumers, differences in broadband uptake rates are likely explained by differences in demand for broadband, not supply.
Determinants of Broadband Demand Five major determinants of demand for broadband networks are (i) the nature of competition, (ii) characteristics of the users (on an individual and national basis) and user behaviors, (iii) broadband SULFLQJ LY WKHDYDLODELOLW\RIEURDGEDQGVSHFL¿F
821
The Adoption of Broadband Internet in Australia and Canada
applications, and (v) the characteristics of the broadband network.
Competition Using U.S. data, Aron and Burnstein (2003) studied the impact of intermodal competition on broadband adoption rates, concluding that the presence of intermodal competition increases demand for broadband in a given region. García-Murillo and Gabel (2003) also found that the presence of FRPSHWLWLRQLQDPDUNHWLVDVLJQL¿FDQWSUHGLFWRU of demand for broadband. While it is obvious that broadband supply to a region is a necessary preFXUVRUWRDGRSWLRQ$URQDQG%XUQVWHLQ¶VUHVXOWV indicate that the type of supply (i.e., DSL-only, cable-only, or DSL and cable) has an impact on broadband demand. Howell (2002) also notes that intermodal competition promotes increased broadEDQGXSWDNHD¿QGLQJUHSOLFDWHGLQ'LVWDVR/XSL DQG0DQHQWL¶V VWXG\RIEURDGEDQGSODWIRUP competition in the European Union. Polykalas and Vlachos (2006) argue that bitstream (wholesale) access to incumbent infrastructures can provide competition among DSL providers when interplatform competition is not present. It is noted that when the incumbent telco in a country also owns the cable network, broadband adoption rates are slower than in countries where this is not the case (International Telecommunication Union, 2003a). For incumbent telcos who also own cable companies, there is less incentive to invest in broadband cable infrastructure that would compete with their existing broadband DSL offerings, unless mandated directly by the government or anticompetitive boards, such as in Australia.
User Characteristics and Behaviors Adoption research generally considers individual characteristics of the potential adopter when investigating adoption behaviors (Rogers, 1995). Individual characteristics that are thought to have an impact on broadband adoption are income, education level, age, and family structure. For example, Kridel, Rappoport, and Taylor (2002) found greater uptake of cable modem-based broadband
822
among higher income and better-educated groups. Varian (2002) found that the best predictor of an LQGLYLGXDO¶V ZLOOLQJQHVV WR SD\ IRU EURDGEDQG was his or her occupation, a factor that may be correlated with individual income. Choudrie and 'ZLYHGL¶VZRUN DOVRVXJJHVWVWKDWLQFRPH is an important determinant of broadband demand. ,QDGGLWLRQDOWKRXJK&DPHURQ¶V VWXG\GRHV not directly discuss income, many of the factors she IRXQGWREHVLJQL¿FDQWO\DVVRFLDWHGZLWKEURDGEDQG take-up (e.g., education level, occupation level, home ownership, type of housing, and other ICT equipment in the household) are arguably correlated with income. The impact of the age of the potential adopter upon the decision to adopt a broadband network is less clear. Rappoport, Kridel, and Taylor (2002) found that older people were more likely to adopt broadband than younger people, but commented that this could be an income effect (i.e., older people have more money, and people with more money are more likely to have broadband). In their cable modem adoption study (Kridel et al., 2002), the same authors note that cable modem adoption generally declines with age, but also note that among Internet users, a higher proportion of older users have adopted cable modems. Madden, Savage, and Simpson (1996) found that being over the age of 55 decreased the likelihood of broadband adoption, but their study was based on intention to adopt in a population that did not have access to broadband at the time of the study. Recent work by Choudrie and Dwivedi (2006a) suggests that older people are less likely to use broadband Internet than younger people, arguing that older people do not have the skills or the technology necessary to make use of broadband connectivity. Another factor that may be somewhat correlated with age is household structure. Howell (2002) observes that the large proportion of Koreans aged 18-30 living with their parents may have had an impact on demand for broadband in that country. These Korean youths may have high disposable incomes to spend on broadband connectivity, or may simply encourage their parents to adopt broadband for their households. Madden et al. (1996) found that the presence of children in a home increased
The Adoption of Broadband Internet in Australia and Canada
the likelihood that a family would adopt broadEDQGD¿QGLQJVXSSRUWHGE\.ULGHOHWDO and by Cameron (2004). However, Madden and 6LPSVRQ¶V VWXG\ SUHVHQWHG FRQWUDGLFWRU\ results, reporting that households with more people were less likely to demand broadband access. As with other individual characteristics, the impact of household structure on demand for broadband is still not well understood. In their 2003 study (using the 2001 OECD data on broadband penetration rates), Kim et al. (2003) considered population level determinants of broadband uptake, and concluded that the most VLJQL¿FDQW GHWHUPLQDQWV RI EURDGEDQG SHQHWUDtion were preparedness, the “attitude of a nation towards advanced information technology” (p. 12), and the availability of computers in the country, DQGSRSXODWLRQGHQVLW\7KH,78¶V'LJLWDO$FFHVV Index (2003b) provides a measure of country preparedness. Gardner (2003) also considered user behaviors when investigating “success drivers” for broadband uptake. Using data from 14 countries in North America, Europe, Scandinavia, and the Asia3DFL¿F UHJLRQ *DUGQHU IRXQG WKDW ³KRXUV VSHQW RQOLQHRIIHUHGWKHPRVWGH¿QLWLYHH[SODQDWLRQIRU the rate of broadband take-up” (p. 14). Stern, Gregor, Martin, Goode, and Rolfe (2004) FRQ¿UPHGDOOWKHSUHYLRXV¿QGLQJVZKHQFRPSDUing the relative importance of factors affecting broadband uptake in Australian households. They found that the main factors, in order of importance, were frequency of Internet use, location of households, technophilia, and experience with pay TV. Frequent Internet users were more likely to adopt broadband technology although it is unclear if households use the Internet more because they have broadband, or if they adopt broadband because they are frequent users. Stern et al. (2004) also found that households in metropolitan locations were more likely to adopt broadband technology, D ¿QGLQJ UHSOLFDWHG E\ &DPHURQ ZKR notes that metropolitan areas have easier access to broadband than rural ones. Stern et al. suggested that technophilia is a factor in broadband adoption. Technophilia referred to the extent to which households have a tendency to consume technol-
ogy and may be related to how technologically savvy the households are. This view is supported by Chang, Ahn, and Lee (2006a) in their study of family decision making processes in the purchase of broadband technology in Australia. Chang et al. (2006a) found that the role of the “technology champion,” who is generally the technology savvy family member, is crucial in driving the decision to purchase broadband technology for the home. $GDPV¶V ZRUNVXSSRUWVWKLV¿QGLQJQRWLQJ that a lack of understanding of broadband technolRJ\DQGLWVEHQH¿WVLVDEDUULHUWRDGRSWLRQ
Price Rappoport, Taylor, and Kridel (2003, p. 84) sugJHVW WKDW D ³VLJQL¿FDQW LQFUHDVH LQ EURDGEDQG penetration rates will most likely require large price reductions for access.” While this observation is consistent with economic theory, it is not clear that broadband deployment is always encouraged by lower broadband prices. Kim et al. (2003) found that the price of broadband was QRW D VLJQL¿FDQW SUHGLFWRU RI EURDGEDQG XSWDNH across OECD countries. They also found that higher dial-up prices did not lead to increased broadband uptake, suggesting that broadband is not necessarily a substitute for dial-up. Rappoport and Kridel (2003) note that the price elasticity of broadband access (through DSL or cable modem) is high in instances of low broadband penetration, but lower at higher broadband penetration levels. In the Japanese market, ADSL service was found WREHSULFHLQHODVWLFEXW¿EHUWRWKHKRPHVHUYLFH was price elastic (Ida & Kuroda, 2006). García0XULOORDQG*DEHO GLG¿QGWKDWSULFHZDV VLJQL¿FDQW ZKHQ FRQVLGHULQJ EURDGEDQG XSWDNH rates across the 135 countries included in ITU data (2002). Stern et al. (2004) found that higher income was related to broadband adoption in Australian households. It is likely that cost is an inhibitor to adoption for low income families, a point noted by Choudrie and Dwivedi (2006b). In some countries, the price of broadband is impacted by the presence of download caps that effectively result in a tiered pricing structure. ITU (2003a) research suggests that the presence of download caps does result in lowered demand for broadband networks. 823
The Adoption of Broadband Internet in Australia and Canada
Applications Varian (2002, p. 52) suggests that “the problem with broadband is not access but applications.” This is the “killer application” argument, which suggests that “unless new compelling applications are forthcoming or the price of broadband FRQQHFWLYLW\IDOOVVLJQL¿FDQWO\DVXUJHLQGHPDQG for broadband in the United States should not be expected” (p. 54). Owen (2002, p. 20) supports this argument, commenting that the most obvious reason for low broadband demand is the “lack of any—much less attractive—products and services that require broadband media for their delivery.” :KLOHEURDGEDQGVSHFL¿FDSSOLFDWLRQVOLNHRQOLQH games and viewing time-shifted TV content (Lee & Choudrie, 2002; Lee et al., 2003) are popular in high broadband penetration countries like Korea and Japan, there are other high broadband penetration countries like Canada where a broadband VSHFL¿FNLOOHUDSSOLFDWLRQLVOHVVREYLRXVLILWLV SUHVHQWDWDOO&RXQWHUWR9DULDQ¶VDQG2ZHQ¶VDUJXments, Middleton (2003) argues that there is not a single killer application that will drive broadband demand. However, it is widely noted that broadband connectivity must be perceived to be useful it if is to be adopted by consumers (Adams, 2006; Oh, Ahn, & Kim, 2003).
Network Characteristics The primary feature that differentiates broadband networks from narrowband ones is the network VSHHG:KLOHWKHUHDUHYDULRXVGH¿QLWLRQVRIZKDW constitutes broadband (Sawyer, Allen, & Lee, WKH86)&&¶VVSHFL¿FDWLRQWKDWEURDGEDQG requires a minimum network speed of at least 200 Kbps is widely accepted. It is noted however that “broadband lite” services are offered in some countries, providing speeds of only 128 Kbps. While such services are not fast (only twice as fast as dial-up), they offer the always-on network access that is characteristic of most broadband networks. Crandall and Jackson (2003, p. 163) note that always-on access “substantially increases the value RIFRQQHFWLRQVIURPWKHKRPHRURI¿FHWRWKH,Q-
824
WHUQHW´:KLOHWKHUHLVQRWPXFKVSHFL¿FUHVHDUFK that considers the impact of network speed on demand for bandwidth, it is reasonable to assume that those who frequently use applications that are IDFLOLWDWHGE\KLJKEDQGZLGWK HJ¿OHVKDULQJ will be most likely to adopt higher bandwidth offerings (providing that the price for such offerings LVFRQVLVWHQWZLWKXVHUV¶DVVHVVPHQWRIWKHYDOXH provided by higher bandwidth). Varian (2002) does report that demand for speed was related to user occupation, with only certain groups of users ¿QGLQJVXI¿FLHQWXWLOLW\LQKLJKHUVSHHGDFFHVVWR merit a premium price. Rappoport et al. (2002) note the opportunity costs of low speed Internet access, suggesting that those whose time is more valuable will reduce their opportunity costs by adopting broadband network access. It is also noted that for many users, broadband adoption is driven by factors other than network speed. For instance, only 25% of respondents to a Japanese survey indicated that speed was a reason for choosing their current Internet service, but more than 55% indicated that always-on connectivity was important (Ida & Kuroda, 2006). In an unpublished survey of Canadian university students (Morris & Middleton, 2005), more than 40% of respondents indicated that the most valuable feature of high speed Internet was something other than speed (e.g., always-on connection, free phone line, provision of shared connection within household).
Summary 7KHSUHYLRXVVHFWLRQRXWOLQHV¿YHNH\GHWHUPLQDQWV of broadband demand, as described in the literature. These demand drivers are very similar to the “sucFHVVIDFWRUV´IRUEURDGEDQGGHSOR\PHQWLGHQWL¿HG by the ITU (2003a), which include competition, innovation, applications, procompetitive regulations, pricing, speed, marketing, high ICT usage, and urban demographics (e.g., population density). The ITU also notes that in the absence of success factors, broadband adoption is inhibited. 1HLWKHUWKH,78¶VOLVWRIVXFFHVVIDFWRUVQRUWKH GLVFXVVLRQRIWKH¿YHNH\GHWHUPLQDQWVRIGHPDQG provides an indication of the relative importance
The Adoption of Broadband Internet in Australia and Canada
of various demand determinants. It does appear that pricing is not the most important determinant of demand, and that a simple price-quantity relationship (i.e., lower price leads to higher demand) does not hold with respect to broadband pricing. 7KLVREVHUYDWLRQLVVXSSRUWHGE\*DUGQHU¶V work, which offers a hierarchical ranking of the importance of various demand drivers.
BROADBAND ADOPTION IN AUSTRALIA AND CANADA Rationale for Comparison On a cultural and geographic basis, Australia and Canada are similar, so it is expected that valuable insights can be gained from considering broadband adoption patterns in Canada. Until recently, Canada was close to the top of the OECD ranks, with an adoption rate that was more than triple that of $XVWUDOLDLQDQGFORVHWRGRXEOH$XVWUDOLD¶V LQ7KHQXPEHUVVKRZWKDW$XVWUDOLD¶V DGRSWLRQ UDWH LV FDWFKLQJ XS WR &DQDGD¶V EXW
Australia remains in 17th position among OECD FRXQWULHVFRPSDUHGWR&DQDGD¶V WKSODFHUDQN As shown in Table 2, although Australia is smaller than Canada, the economic indicators of the two countries are similar, as is their multicultural nature. In addition, both countries have a comparable geographic mix of large urban centers (found along the coast in Australia, and along the Canada-U.S. border in Canada) and smaller, geographically isolated communities. Given these similarities between Australia and Canada, it is suggested that studying the Canadian deployment of broadband may provide a useful indication for $XVWUDOLDQ SROLF\ PDNHUV $OWKRXJK $XVWUDOLD¶V EURDGEDQGDGRSWLRQUDWHVWUDLO&DQDGD¶VLWVRYHUDOO Internet adoption rate is higher, as is the mobile phone adoption rate.
Broadband in Canada More than 64% of Canadian households have Internet access. Of these households, 80% have broadband Internet access. In total, more than 50% of Canadian households (6.4 million households)
Table 2. Australia and Canada comparative fact sheetA Australia
Canada
20.4
32.3
7 682 400
9 970 610
3
3
Multicultural
Multicultural
1.28 AUD
1.16 CAD
33 100
34 053
70.4
63.0
7RWDO¿[HGOLQHVXEVFULEHUV
56.84
56.64
Total mobile phone subscribers/100 (2005)
91.39
51.44
Population (millions) 2
Land area (km ) Population Density (per km 2) Culture Currency rate (U.S. $1) (December 2006) GDP per head (USD, purchasing power parity) Internet subscribers per 100 inhabitants (2005)
A
$OO¿JXUHVXQOHVVRWKHUZLVHQRWHG 6RXUFHV 7KH(FRQRPLVW&RXQWU\3UR¿OHV ZZZHFRQRPLVWFRPFRXQWULHV ZZZRDQGDFRPFRQYHUWFODVVLF and International Telecommunication Union (2006a).
825
The Adoption of Broadband Internet in Australia and Canada
have broadband. Market share data collected by the telecommunications regulatory agency, the CRTC, suggest that approximately 42% of Canadian broadband connections are provided by DSL through incumbent telecommunications companies, with 54% using a cable modem conQHFWLRQ,WLVQRWHGWKDWWKHVH¿JXUHVLQFOXGH³OLWH´ broadband services (<256 Kbps) which make up just over 10% of broadband subscriptions in Canada (CRTC, 2006). &RPSHWLWLYHVHUYLFHVLQFOXGHG¿[HGZLUHOHVV or satellite provision (generally found in remote areas), as well as independent Internet service providers who resell DSL or cable access. In addition, wireless Internet services are becoming available in some Canadian municipalities. For example, the city of Fredericton, in New Brunswick, offers free WiFi in the downtown area (www.fred-ezone. com), and Toronto Hydro Telecom is developing a commercial WiFi service that will provide broadband to Toronto residents (www.onezone.ca). To date, these sorts of services are primarily viewed as supplements to existing broadband connections (to allow users access away from the home), rather than as direct competition to services provided by telcos and cablecos. Another new service is marketed as “portable Internet,” and provides a wireless Internet connection at a stated speed of 1.5 Mbps in more than 20 Canadian cities, using “pre-WiMax” technologies (Inukshuk, 2006). Less than 5% of Canadian broadband connections are provided by facilities-based or intramodal competitors (CRTC, 2006), indicating weak competition for the DSL and cable modem services provided by market incumbents. Urban areas in Canada are well-served by broadband providers, meaning that more than 90% of Canadian households have broadband access. However, given the geography of the country, it is still the case that almost 50% of Canadian communities still have no DSL or cable access (CRTC, 2006). Current government initiatives are focusing on extending broadband access to rural and northern communities in Canada (see http://www. broadband.ic.gc.ca/pub/index.html?iin.lang=en), using satellite and wireless technologies. Industry Canada, a federal government department, has
826
actively promoted the development of broadband infrastructure across the country, and supported initiatives to develop CANARIE, the cross-Canada ¿EHUEDFNERQHQHWZRUNWKDWSURYLGHVFRQQHFWLYLW\ to schools, universities, libraries, hospitals, and other organizations (CA.net Institute, 2001). This LQIUDVWUXFWXUHKDVEHQH¿WHGPDQ\&DQDGLDQVEXW it does not provide residential broadband access.
Broadband in Australia In Australia, the picture of broadband take-up is somewhat different. Despite a high proportion of computer users where up to 70% of households had access to a home computer (60% with access to the Internet), only about 28.8% of Australian households have broadband Internet access (Australian Bureau of Statistics, 2006). 7KHODWHVW¿JXUHVIURPWKH$XVWUDOLDQ&RPSHtition & Consumer Commission (2006) show the take-up of broadband services is at 3.5 million (households and nonhouseholds) as at June 2006. 1.4 million services were connected over the previous 12 months (June 2005-2006), representing a 67% increase from same time last year. Of the estimated 3.5 million broadband subscribers in Australia in June 2006, 17% used cable, and some 80% used DSL (ACCC, 2006). Yet, as indicated previously, despite the rapid growth in the past \HDU$XVWUDOLD¶VEURDGEDQGSHQHWUDWLRQLVODJJLQJ behind not only Canada but also behind most other developed countries according to the OECD rankings (17th position). 7KH .30* 5HSRUW RQ $XVWUDOLD¶V broadband future suggested that the three main UHDVRQVIRU$XVWUDOLD¶VODJJLQJSRVLWLRQZHUHWKH lack of competitive infrastructure, the slow roll out of services, and the initial relatively higher cost of broadband in comparison with dial-up services. On the surface, Australia may appear to have encouraged competition, as there are more than 246 Internet service providers, many of whom provide broadband services, with a wide variety of broadband plans and options for subscribers to choose from. (see www.whirlpool.net.au). However, the Australian market is in reality dominated by two main wholesalers, Telstra and Optus. Access
The Adoption of Broadband Internet in Australia and Canada
WR¿[HGOLQHEURDGEDQGWHOHFRPPXQLFDWLRQVLVSURYLGHGPDLQO\E\7HOVWUD¶V$'6/QHWZRUNZKLFKLV available to about 71% of the population (including virtually all population centres with greater than SHRSOH ,QDGGLWLRQWKHDOWHUQDWLYH¿[HGOLQH broadband networks in Australia are the Telstra and Optus cable networks. Whereas there are many retailers of broadband technology and services, the prices they charge and the coverage they can offer are largely determined by Telstra and Optus, and are therefore only marginally cheaper than those of Telstra and Optus. Competitors such as TransACT, Bright Telecommunications, and Neighborhood Cable developed cable modem access services %URDGEDQG $GYLVRU\ *URXS 7HOVWUD¶V BigPond brand also offers wireless broadband “in most places across Australia” using 3G technology (BigPond, 2006), but to date, wireless Internet infrastructure has provided limited competition to DSL or cable services (Budde, 2006). The need for a more responsive and cohesive strategy on broadband connectivity within Australia resulted in the Broadband Advisory *URXS¶V %$* UHSRUWWRWKHJRYHUQPHQW $XVWUDOLD¶V %URDGEDQG &RQQHFWLYLW\ ZKLFK RXWlined key strategies for the government including broadband take-up within key sectors (demand side), key performance indicators (KPI) for the incumbent (supply side competition), and monitoring. Since that report, a number of successive projects have been undertaken by the Australian Government (through the Department of Communications, Information Technology, and the Arts [DCITA]), to improve supply, especially in rural and regional areas, and service quality. The latest program is Broadband Connect, an “$878 million initiative of the Australian Government to support equitable access to high quality, sustainable broadband services across regional and rural Australia” (DCITA, 2006). In addition, through its Community Connectivity program, DCITA is starting to address the demand issue by focusing on sectors of the community where there might still be sociocultural and economic barriers to the take-up of broadband.
COMPARATIVE ANALYSIS: BROADBAND ADOPTION IN AUSTRALIA AND CANADA Framework for Analysis ,QWKLVVHFWLRQWKH¿YHNH\GHWHUPLQDQWVRIEURDGEDQGGHPDQGDUHDVVHVVHGXVLQJ*DUGQHU¶V hierarchical model of demand drivers as a framework for the comparative analysis of broadband markets in Canada and Australia. Accelerating factors include a strong competitive market for broadband, widespread availability of broadbandVSHFL¿FDSSOLFDWLRQVFRQVXPHUSURSHQVLW\WRVSHQG time online, and the availability of high speed network access. Pricing has the potential to stimulate the broadband market but will not necessarily accelerate broadband take-up. Threshold factors such DVWKHQXPEHURIKRXVHKROGVZLWK3&VDQGVXI¿FLHQW population density are necessary preconditions for the deployment of broadband, but on their own are QRWVXI¿FLHQWWRVWLPXODWHRUDFFHOHUDWHLWVXSWDNH by residential consumers. Because the threshold factors for Canada and Australia are relatively similar (Table 2), the focus here is on the roles of accelerators and stimulants in the Canadian and Australian broadband markets (where there are marked differences). Table 4, at the end of this VHFWLRQVXPPDUL]HVWKH¿QGLQJV
Accelerators Competition Healthy competition plays a pivotal role in the deployment of broadband Internet (OECD Directorate for Science Technology and Industry, 2001). As noted earlier, there are two ways to promote competition in the network industry (Christodoulou & Vlahos, 2001; Michalis, 2001). One is competition between infrastructure networks (e.g., cable modem, DSL), also known as facilities-based or intermodal competition. The other is intramodal competition, within each network technology (e.g., among DSL providers). In infrastructure competition, new entrants have to build their own network;
827
The Adoption of Broadband Internet in Australia and Canada
in intramodal competition (enabled by local loop XQEXQGOLQJ QHZHQWUDQWVFDQXVHWKHLQFXPEHQW¶V network and resell capacity on it. In Canada, broadband competition primarily occurs between infrastructure networks, with cable modem access having a larger market share than DSL (54% as compared to 42%). In most urban areas there is only one cable modem provider and one major DSL service provider (the incumbent telco), so most consumers perceive that their choice for broadband service is between cable and DSL. As broadband adoption has become widespread and growth in the market has started to slow, the basis of competition appears to have shifted. Now rather than attempting to convince new users of the superiority of one platform over the other, both cable and DSL providers are focusing on diversifying their product lines, to make broadband more appealing to a wider range of customers. Intense rivalries between providers have subsided, and the industry has settled into an environment where there is little difference in price or service between cable and DSL offerings. Competition now occurs at a more macro level, as both telcos and cablecos try to entice customers with “bundles” of services WKDW LQFOXGH KLJK GH¿QLWLRQ WHOHYLVLRQ PRELOH telephones, and broadband Internet. Efforts are also made to increase broadband penetration by offering “lite” services that are priced almost on par with dial-up services. Some companies have taken advantage of local loop unbundling to establish competing DSL services. This has resulted in some customer “churn” as people move from one provider to another to take advantage of introductory subscriber offers. To date, there is little evidence of competing services being offered via existing infrastructure in the cable sector, so there is effectively no competition within the cable sector. Service providers have divided the market based on geography, and each geographic area has only one cable provider. In summary, the Canadian market looks quite competitive on paper, given the local loop unbundling and open access policies. In reality however, the market is dominated by a few key players who offer very similar services at similar prices, in a situation of moderate rivalry. In 2005, only 3.4%
828
of residential broadband Internet access revenues were earned by companies other than incumbent telcos or cable providers, although 40% of dial-up revenues were earned by competitors to the incumbents (CRTC, 2006). There is healthy intermodal competition, but this has not come about as a result of local loop unbundling. In Australia, it appears that there are many broadband connection providers, particularly in DSL. However, DSL and cable networks are dominated by Telstra and to a lesser extent, by Optus. 0DQ\,63VVHOOLQJ'6/DUHUHVHOOHUVRI7HOVWUD¶V DSL capacity. Local loop unbundling (LLU) is seen as crucial to competition. In Australia, the local ORRS XQEXQGOLQJ ZRXOG HQDEOH WKH LQFXPEHQW¶V 7HOVWUD¶V FRPSHWLWRUVWRLQVWDOOLQIUDVWUXFWXUHLQ local exchanges to provide broadband services such as DSL. However, Telstra initially argued consistently that the simultaneous challenge of rolling out both DSL and LLU was a major impediment to growth, citing Australia as the only place where simultaneous DSL and LLU rollout has been in place, due to pressures from the Australian Competition and Consumer Commission (ACCC). The ACCC is the government body responsible for monitoring competition in industries. Since the near completion of DSL roll out, the delay now is in the debate between the ACCC, Telstra, and its competitors around the issue of the LLU pricing (ACCC, 2006). Until the LLU pricing levels are clear, the ability of competitors to deploy their own infrastructure will be slowed down. LLU is clearly a contentious area for incumbents as can be seen by the New Zealand example where 7HOHFRP 1HZ=HDODQG¶VLQFXPEHQW LVRSSRVHG to LLU, while its Australian subsidiary (AAPT) is keen to take advantage of unbundling in Australia (Watson, 2002). In summary, comparing the impact of competition within the industry on the rate of speed of take-up of broadband services Canada and Australia, it may be surmised that competition is a key determinant of (i) how well and quickly broadband infrastructure may be rolled out, (ii) the level of service quality, and (iii) pricing and choice for consumers. It is important to reiterate that the Canadian broadband market is dominated
The Adoption of Broadband Internet in Australia and Canada
by competition between DSL and cable, and that competition among DSL providers has been less evident. In addition, the growth of cable broadband in Canada was likely facilitated by the widespread uptake of cable television services by Canadians, with 75% of Canadian households subscribing to cable in 2001 (CRTC, 2002). With an existing relationship with both the cable company and the incumbent telco (the DSL provider), Canadians could easily choose between the two broadband platforms.
User Behaviors For Canadian Internet users, popular applications include e-mail, Web browsing, and information searching, as well as general leisure and entertainment activities (e.g., playing games, downloading or listening to music) (Statistics Canada, 2004). More than 55% of Canadians go online for at least 7 hours a week, and more than 60% have been using the Internet for more than 5 years (Zamaria, Caron, & Fletcher, 2005). More recent data show WKDW&DQDGLDQVDUHVWLOOZDU\RIRQOLQH¿QDQFLDO transactions using credit cards (Statistics Canada, 2006), but close to 60% now use the Internet for banking. For many Canadians, especially those in younger age groups, the Internet has become an integral part of their daily lives. 55% of university students surveyed indicated that they strongly agreed with the statement “it would be very hard for me to stop using the Internet,” and 85% of them chatted online with their friends daily. Only 2% of these students had a dial-up connection at home, 97% had either DSL or cable broadband (Morris & Middleton, 2005). For Canadian youth, the Internet plays a large role in maintaining social networks (Shade, Porter, & Sanchez, 2006). In Australia, Chang et al. (2006a) found in their interviews with Australian households that they were still relatively reluctant to use the Internet for recreational and online gaming purposes. In a separate study, Chang, Lee, and Oh (2006b) found that young Australians were not active participants in online discussions and reviews. They mostly preferred to use online interaction as a supplement rather than a substitute for face to face communi-
cation. Additionally, Cameron (2004) also found that people who live in regional Australia were also less likely than their metropolitan counterparts to take up and use broadband technology and applications. In summary, it appears that Canadians have embraced the Internet as part of their daily lives in ways that are not so prevalent in Australia. A possible reason for this difference is that Canadian surfers do not generally face quotas on their Internet use, allowing them to access the Internet without fear of facing restricted bandwidth or being charged for excess usage.
Applications Government policy documents frequently identify WKHEHQH¿WVRIEURDGEDQGVHUYLFHIRUDFFHVVWRH learning, e-government, and e-health (e.g., National %URDGEDQG7DVN)RUFH 2I¿FHRI7HFKQRORJ\ Policy, 2002). But Bauer, Gai, Kim, Muth, and Wildman (2002) argue that the majority of such services can be delivered to consumers using relatively low bandwidth. There are efforts by governments to promote demand for broadband (see http://www.itu. int/osg/spu/ni/promotebroadband/ for activities in this regard), and it is frequently argued that unique and interesting content and applications are needed to create demand for broadband, leading to a search for so-called “killer applications” (International Telecommunication Union, 2001; Smith & Leung, 2002). Middleton (2003) suggests that the search for a content-based killer application overlooks WKHYDOXHEURDGEDQG,QWHUQHWXVHUV¿QGLQEHLQJ able to connect with each other and generate their own content (a point reinforced by the increased popularity of peer-to-peer applications like Flickr, YouTube, MySpace, Facebook, etc.). ,Q &DQDGD LW LV GLI¿FXOW WR LGHQWLI\ VSHFL¿F broadband applications that led users to broadband services. Consumer demand for broadband in Canada seems to have grown independently of government initiatives.3 As noted above, Canadians use broadband for communication (e-mail, messaging), for information gathering (Web searches), to create and share content (e.g., photos, blogs), and for entertainment (e.g., games, downloading
829
The Adoption of Broadband Internet in Australia and Canada
PXVLFDQGYLGHR¿OHV 6RPHXVHUVGRSDUWDNHLQ online learning activities (e.g., accessing course materials), access government information and conduct various transactions (e.g., renewing parking permits), and search for health care information online, but it is not clear that these are their primary activities.
Analysis of Statistics &DQDGD¶V+RXVHKROG,QWHUQHW8VH6XUYH\GDWDRQ broadband adoption shows that more than half of the broadband households in Canada are not high intensity Internet users, meaning that they accessed the Internet less than seven times per week, and used the Internet less than 40 hours a month. Furthermore, as recently as 2003 (the last year the survey was conducted), more than 55% of broadband households did not use the Internet for banking, online purchases, work, or education (Middleton & Ellison, 2006). These data suggest that broadband may be adopted as much for the convenience of always on access as for a need for a high speed connection to support intensive, varied Internet activities, which is consistent with ,GDDQG.XURGD¶V ¿QGLQJVDPRQJ-DSDQHVH broadband users. 6LPLODUO\WKHUHDUHQRDSSOLFDWLRQVLGHQWL¿HG as broadband drivers in Australia. Rather the current discourse asks whether there is sustainable consumer demand for broadband. Awareness and experience of broadband among the public appears to be low in Australia, as highlighted by the BAG (2003) report, which consistently recommends SXEOLF HGXFDWLRQ RQ WKH EHQH¿WV RI EURDGEDQG technologies. Wales (2002) suggested that even some Australians who have been exposed to the use of broadband do not have a full appreciation of its potential, and lack awareness about comparative costs of broadband and dial-up. Adams (2006) found that potential broadband users are confounded by the complexity of the technology and the pricing plans. As noted by Wales (2002), there are few compelling applications for which broadband is essential in Australia. Both Wales (2002) and Sacks (2002) indicated that there might be more interest among Australian users in the
830
attributes of broadband rather than its applications for two main reasons. First, it seems that connectivity (always on) and capacity to network (within households and organizations) are still PRUH DWWUDFWLYH WKDQ VSHFL¿F EURDGEDQGUHODWHG applications or content. This has also been the mainstay of advertizing for broadband services. Second, Wales (2002) and Sacks (2002) both argue that the lack of importance placed on applications in Australia may be due in part to factors such as capped speed and charging by download volume. These constraints mean that for users, applications that are heavily reliant on broadband capacity are still not attractive to Australian users both within households and organizations. However, since those earlier studies, new applications such as YouTube have meant that increased broadband capacity is now becoming more important for households who want to access these newer applications, and may have contributed to the more recent increase in uptake of broadband services in Australian households.
Network Speed 'H¿QLWLRQV RI ³EURDGEDQG´ VSHHGV YDU\ ZLGHO\ around the world. At the highest end of the scale, consumers in Japan and South Korea can get 50 0ESVVHUYLFHRYHU¿EHUQHWZRUNVIRUDERXWSHU month (International Telecommunication Union, 2006a). In contrast, Canadian cable companies now offer “Ultra Lite Broadband,” which is an always-on, 128 Kbps service, for under $20 per month (http://www.shoprogers.com/store/cable/ internetcontent/ultralite_RCI.asp?). Canadian consumers have a choice of four tiers of service. An “ultra” or “pro” package is available that offers upstream bandwidth in excess of 5 Mbps, plus extras like a wireless router. The “standard” package now offers bandwidths of up to 5 Mbps, and the 1 Mbps service (the original “high speed” service in Canada) has been rebranded as basic/lite. As noted above, “ultra lite” offerings provide access speeds in the 128 - 256 Kbps range. In terms of network speed, Australia now DVVXPHVWKHGH¿QLWLRQRIEURDGEDQGDVDQ³DOZD\V on” Internet connection with an access speed equal
The Adoption of Broadband Internet in Australia and Canada
to or greater than 256 Kbps (Australian Bureau of Statistics, 2004). The KPMG (2004) report RQ $XVWUDOLD¶V EURDGEDQG IXWXUH VXJJHVWHG WKDW ³%URDGEDQGLVVXI¿FLHQWGDWDWUDQVPLVVLRQVSHHGWR utilize applications services or content effectively UHODWLYHWRWKHXVHU¶VDFFHVVGHYLFHFDSDELOLWLHV´ There seems to be an implication in Australia that network speed is not a major determinant of broadband take-up. However, it is unclear whether this is due to the fact that Australians do not understand the potential value of faster network speed or if they do not see it as a factor to be concerned with. ,QHLWKHUFDVHLWLVGLI¿FXOWWRDVVHVV$XVWUDOLDQ understanding of faster network speed because of the lack of exposure to the types of speed available in countries such as Canada and Korea.
Stimulants Pricing Pricing is a more complicated element than it would appear, since price can represent the clear cost of broadband access, or the cost as compared to dial-up service. In Canada, there is little price competition for broadband services, but the tiered structure provides access to broadband at several price points. For the standard offering, both DSL and cable providers charge $45 (CAD) per month
($39 USD), with introductory offers that reduce the SULFHIRUWKH¿UVWPRQWKVRIVHUYLFHIUHHSURGXFW giveaways (e.g., 19” computer monitor for customers who sign up for two years) and package deals (bundles) for subscribers who get other services from the provider (e.g., cable television, mobile or landline telephone services). For many early adopters of broadband, this pricing structure was attractive as it allowed Canadians to dispense with the expense of a second phone line that had been acquired for Internet access. Broadband providers are now converting dial-up customers to broadband by offering the lower tiered services at prices comparable to dial-up. Canadian broadband providers have implemented, removed, and now reimplemented bandwidth caps. The bandwidth caps are different for DSL and cable service. For instance, the ultra lite DSL service provided by an incumbent telco is capped at 1 Gb (gigabyte) per month, compared to a 60 Gb per month cap for ultra lite cable access. For the “pro” service, DSL caps are set at 30 Gb, compared to 100Gb for cable service. Both DSL and cable charge a per Gb fee for data transfers over the caps, the DSL fees are higher than the cable fees. Most Canadian Internet users are unaware of the download caps, but with the popularity of bandwidth-intensive applications like video sharing (e.g., using YouTube or BitTorrent) and with
Table 3. Comparison of broadband pricing: Sorted by price per 100 Kbps Country
Speed (Down)B
Subscription/month (US$)
Price per 100
Value Factor C
Kbps (US$)
Japan
51.2 Mbps
$38.19
$0.07
28.13
Korea (Rep.)
51 Mbps
$40.59
$0.08
9.49
USA
4 Mbps
$20.00
$0.49
1.69
Canada
4 Mbps
$41.26
$1.01
3.75
Iceland
6.1 Mbps
$91.39
$1.49
-
Australia
1.5 Mbps
$53.30
$3.45
0.40
B C
(International Telecommunication Union, 2006a) 6DXQGHUV
831
The Adoption of Broadband Internet in Australia and Canada
distribution of television shows over the Internet, the caps will place restrictions on some users. In Australia, the government maintains that $XVWUDOLD¶VEURDGEDQGSULFHVDUHFRPSDUDEOHWRWKH rest of the world (DCITA, 2006), although these contentions are not supported by the data presented in Table 3. In Australia, there has admittedly been increased competition among ISPs which has led to current prices for monthly packages dropping to about $29.95 for 0.2GB of free downloads. As KPMG (2004) notes, the reality remains that Canadians can download much more data on a monthly basis, more cheaply, than Australians can. Yet, perhaps the issue here is not the relative price of Australian broadband services in comparison with that of Canada, but the relative difference between the price of broadband services and dial-up service. In Canada, pricing for lower tier broadband services is now the same as for dial-up access. Australia has among the lowest dial-up Internet prices in the world (slightly lower
WKDQWKRVHLQ&DQDGD 7KHUHIRUHLWLVGLI¿FXOWWR convince consumers to switch away from dial-up to more expensive broadband unless there are JUHDWHUSHUFHLYHGEHQH¿WVWRXVLQJEURDGEDQG7KLV situation is comparable to that in New Zealand, where Howell (2002) found that consumers, despite having access to a well-developed broadband LQIUDVWUXFWXUHGLGQRWSHUFHLYHDQ\EHQH¿WVJUHDW enough to justify the comparatively higher cost than their dial-up services. In addition, as seen in Table 3, the price of bandwidth per 100 Kbps in Australia is still relatively expensive. These arguments are compounded by the fact that in Australia, the download cap is relatively low compared to Canada. The download caps represent a potential obstacle for switching to broadband from dial-up. It negates a main relative advantage of broadband over dial-up services which is the ability to use more advanced applications that require greater download capacity. Therefore, as competition drives ISPs to start introducing plans
Table 4. Summary of comparative analysis Canada
Australia
Competition
Moderate (strong DSL and cable services)
Weak (dominated by Telstra and Optus)
User characteristics and
The Internet is becoming a part of daily
Less understanding of technology capabilities.
behavior
life for many Canadians, especially the
Initial research indicates high use of Internet but less
young who rely on it heavily for interaction
likelihood to interact online.
Accelerators
with friends. Applications
)HZEURDGEDQGVSHFL¿FDSSOLFDWLRQV
)HZEURDGEDQGVSHFL¿FDSSOLFDWLRQVGRZQORDGFDSV constrain adoption of applications
Network characteristics
Fast connectivity available, multiple
Slow connectivity and limited choice
platforms
Stimulants Pricing
Cheap, download caps not a major concern
Expensive, plus download caps
ICT usage
High
High
Urban demographics/
Low density
Low density
Threshold Factors
Population density
832
The Adoption of Broadband Internet in Australia and Canada
with increased free downloadable bytes for higher prices, the trade off for the Australian household is between downloads and price. This is yet another consideration that Australian subscribers have to make when compared to Canadian counterparts who do not generally have this concern and treat EURDGEDQG DV DQ XQOLPLWHG VHUYLFH DW D ¿[HG price. In light of issues already mentioned, Saunders (2005) proposed an alternative way of comparing broadband benchmarks between countries by proposing a “value factor.” The value factor includes considerations for upload and download speeds ZHLJKWHGWRUHÀHFWWKHLPSRUWDQFHRIGRZQORDGV as compared to uploads), and the metered nature of broadband plans (including download caps). In effect, the study provided a value factor that compared the service and price, with the higher factor indicating greater value (see Table 3).
Summary of Comparative Analysis In summary, despite the similarity in threshold factors such as low density population and per capita income between the two nations, there are some key factors in Australia which may explain why the diffusion of broadband access is relatively low. These factors include (i) the lack of competition in the broadband industry, (ii) less likelihood to engaged in prolonged online interaction among Australians, (iii) slower connectivity, and (iv) download caps and pricing.
DISCUSSION As Fransman (2006) demonstrates in Global Broadband Battles: Why the U.S. and Europe Lag while Asia Leads, the answers as to how to achieve increased broadband adoption are complex and require far more consideration than is possible in a single book chapter. But what this chapter does is provide insights into tangible objectives for achieving a more “user-friendly” broadband environment in Australia. These objectives can be summarized as follows:
1.
Reduce constraints on users and usage. It appears that many Australians are somewhat afraid to surf the Internet. They are worried about exceeding download caps and about the cost of their broadband service. This environment is not conducive to experimentation with new types of Internet content (e.g., user generDWHGYLGHRV ¿OHVKDULQJRUXVLQJWKH,QWHUQHW to access music or television content. Most Australians do not experience broadband Internet in the same way that Canadians do, as simply an always-on, fast, convenient, easy to use service, without constraints. Efforts to increase broadband penetration in Australia must address the issue of constrained Internet usage. As progress is made on reducing the constraints on usage, then it is possible to focus on promoting broadband access among current nonusers. $UWLFXODWHWKHEHQH¿WVRIEURDGEDQG,QWHUQHW service. This chapter noted that there are many Canadians who have broadband Internet service but do not consider speed to be its most important attribute. Australian consumers VKRXOGXQGHUVWDQGWKDWEURDGEDQG¶V³YDOXH proposition” goes beyond faster Internet, providing ease of access through always-on connectivity, enabling access to new services (e.g., voice over Internet protocol), allowing multiple users to share a connection within a residence (using wired or wireless networking), and freeing up the telephone line. 3. Improve the price/speed equation. In Canada, the lower tiers of broadband service are available at about the same price as dial-up service. *LYHQWKHEHQH¿WVQRWHGDERYHLWLVDYHU\ simple proposition to convince people of the value of abandoning dial-up for broadband. In Australia, the cost of entry level broadband is higher than in Canada, and given the constraints on usage, people are not so willing to pay a premium for a limited service. 4. Consider alternatives to local loop unbundling to increase market competition. It is frequently argued that local loop unbundling is necessary to improve the broadband environment in Australia. But in the Canadian
833
The Adoption of Broadband Internet in Australia and Canada
context, even with local loop unbundling there is little competition among DSL providers, and cable and DSL connections have fairly equal market shares. Local loop unbundling may result in more competitive services available to Australian consumers, but it will not increase cross-platform competition. The market for cable broadband in Australia remains limited because cable television networks only serve approximately 20% of the population (Breznick, 2005). Because the prospects for higher uptake of cable broadband are not strong, competition among DSL providers seems attractive. But it is noted that wireless Internet access infrastructure can provide robust connectivity to Australians in urban and rural regions (see International Telecommunication Union, 2006a, and OECD Directorate for Science Technology and Industry, 2006, for a discusVLRQRIWKHSRVVLELOLWLHVDIIRUGHGE\:L0$; and satellite services), and may prove to be a better means of increasing competition in the Internet access market than increasing the number of DSL providers in the market.
CONCLUSION In closing the chapter, we suggest paying attention to the following two facts. First, Canada became a leader in broadband in a short time frame (3 to 4 years). Second, 91% of Australian households are within DSL reach, but among these seven million homes there is only a 30% take-up. Fast penetration is feasible and possible in Australia if and when momentum is created. We have identi¿HG¿YHDUHDVIRUDWWHQWLRQLQWKHGHYHORSPHQWRI broadband Internet: competition, user behaviors, application availability, network characteristics, and pricing. We maintain that deploying broadband access is not a matter of building networks. Rather, it has more to do with the context in which such networks are deployed and whether the environment is friendly to consumers or not. The chapter LGHQWL¿HVHOHPHQWVRIWKHXVHUIULHQGO\EURDGEDQG environment in Canada, and we encourage further
834
work to determine appropriate social, governmental, and regulatory initiatives to create a more user friendly broadband climate in Australia. Increased broadband deployment should be possible in Australia, if policies encouraging supporting the provision of knowledge and the persuasion of consumers are put in place, and if accelerating market drivers, as opposed to merely threshold drivers, are the focus of policy. To date in Australia, government efforts to encourage broadband uptake have centered on suppliers and providers of broadband services. The focus needs now to switch from the supply side to the demand side as it becomes increasingly evident that despite available infrastructure, the demand for broadband in Australia continues to be “sluggish” (Firth, Longstaff, & Mellor, 2002). As the DCITA (2006) correctly pointed out, it is important to understand the context of country when analyzing broadband diffusion. This chapter draws on the Canadian context to illustrate a way forward for broadband development in Australia.
ACKNOWLEDGMENT The authors wish to thank Alison Powell for editorial assistance on this chapter and Daniel Barkho for research assistance. The research was supported by the Faculty of Business at Ryerson University, and the Social Sciences and Humanities Research Council of Canada. An earlier version of this chapter, written with colleague Heejin Lee, was presented at the 2003 International Telecommunications Society Asia-Australasian Regional Conference in Perth, Australia.
REFERENCES ACCC. (2004). Snapshot of broadband deployment as at 30 September 2004. Retrieved July 31, 2007, from http://www.accc.gov.au/content/item. SKWPO"LWHP,G QRGH,G ¿OHFFIHG e&fn=Broadband%20snapshot%2030%20Septem ber%202004.pdf
The Adoption of Broadband Internet in Australia and Canada
ACCC. (2006). Snapshot of broadband deployment. Retrieved November, 2006 from http://www.accc. gov.au/content/index.phtml/itemId/693170 ACCC. (2006). Strategic review of the regulaWLRQRI¿[HGQHWZRUNVHUYLFH5HWULHYHG-XO\ 2007, from http://www.accc.gov.au/content/index. phtml/itemId/719844 Adams, P. (2006). The complexity of the purchase decision – a barrier to broadband adoption for mainstream consumers. Paper presented at the Association of Internet Researchers Annual Conference, Brisbane. Aron, D.J., & Burnstein, D.E. (2003). Broadband adoption in the United States: An empirical analysis. Paper presented at the Telecommunication Policy Research Conference, Arlington, VA. Australian Bureau of Statistics. (2004). Measures of a knowledge-based economy and society, Australia: Information and communications technology indicators, number of household ISP subscribers. Retrieved July 31, 2007, from http://www.abs.gov. au/Ausstats/[email protected]/0/FA4AE6295749722FCA2 56D97002C8645?Open Australian Bureau of Statistics. (2006). Household use of information technology, 2005-2006. Retrieved July 31, 2007, from http://www.abs.gov. au/Ausstats/[email protected]/0/acc2d18cc958bc7bca2568 a9001393ae?OpenDocument Australian Information Economy Advisory Council. (1999). National bandwidth inquiry. Canberra, Australia: Department of Communications, Information Technology and the Arts (DCITA). Bauer, J.M., Gai, J., Kim, T.A., Muth, S., & Wildman, S. (2002). Whither broadband policy. Paper presented at the Telecommunication Policy Research Conference, Alexandria, VA. Bigpond. (2006). Wireless broadband. Retrieved July 31, 2007, from http://my.bigpond.com/internetplans/broadband/wireless/about/default.jsp Bleha, T. (2005). Down to the wire. Foreign Affairs, 84(3), 111-124.
Bouwman, H., Fijnvandraat, M., & Van De Wijngaert, L. (2006). White spots and black holes: Developing a conceptual model for broadband rollout. Info - The Journal of Policy, Regulation and Strategy for Telecommunications, 8(1), 72-90. Breznick, A. (2005). Australian digital cable. Retrieved July 31, 2007, from http://www.cablelabs. com/news/newsletter/SPECS/NovDec_2005/ story6.html Broadband Advisory Group. (2002). Report on broadband advisory group consultative forums. Canberra, Australia: NOIE Access Branch. Broadband Advisory Group. (2003). Australia¶s broadband connectivity. Canberra, Australia: 1DWLRQDO2I¿FHIRUWKH,QIRUPDWLRQ(FRQRP\ BSEG. (1994). Networking Australia¶s future: 7KH¿QDOUHSRUWRIWKHEURDGEDQGVHUYLFHVH[SHUW group. Canberra, Australia: AGPS. Budde, P. (2006). Australia - wireless broadband - infrastructure overview and developments. Bucketty, NSW: BuddeComm. CA*Net Institute. (2001). A nation goes online. Ottawa, Canada: CANARIE. Cameron, A. (2004). Factors affecting broadband adoption and demand. Telecommunications JourQDORI$XVWUDOLD(2), 53-60. Cava-Ferreruela, I., & Alabau-Muñoz, A. (2006). Broadband policy assessment: A cross-national empirical analysis. Telecommunications Policy, 30(8-9), 445-463. Chang, S., Ahn, S., & Lee, H. (2006a). The family purchasing process of broadband Internet in Australia. International Journal of Technology Marketing, 1(3), 283-300. Chang, S., Lee, H., & Oh, S. (2006b). Australians as prosumers; How active are we? Paper presented at the Digital Natives and Australia and Korea, Melbourne. Choudrie, J., & Dwivedi, Y.K. (2005). The demographics of broadband residential consumers in a British local community: The London borough of
835
The Adoption of Broadband Internet in Australia and Canada
Hillingdon. The Journal of Computer Information 6\VWHPV(4), 93-101. Choudrie, J., & Dwivedi, Y.K. (2006a). A comparative study to examine the socio-economic characteristics of broadband adopters and non-adopters. Electronic Government, An International Journal, 3(3), 272-288. Choudrie, J., & Dwivedi, Y.K. (2006b). InvestigatLQJIDFWRUVLQÀXHQFLQJDGRSWLRQRIEURDGEDQGLQWKH household. The Journal of Computer Information Systems, 46(4), 25-34. Christodoulou, K., & Vlahos, K. (2001). Implications of regulation for entry and investment in the local loop. 7HOHFRPPXQLFDWLRQV3ROLF\(10-11), 743-757. Commission of the European Communities. (2006). Communication from the commission to the council, the European Parliament, the European economic and social committee and the committee of the regions: Bridging the broadband gap. Brussels, Belgium: European Union. Crandall, R.W., & Jackson, C.L. (2003). The $500 ELOOLRQRSSRUWXQLW\ 7KHSRWHQWLDOHFRQRPLFEHQH¿W of widespread diffusion of broadband Internet access. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications technologies (pp. 155-193). New York: Nova Science Publishers, Inc. CRTC. (2002). Broadcasting policy monitoring report 2002. Gatineau, Canada: CRTC. CRTC. (2006). CRTC telecommunications monitoring report: Status of competition in Canadian telecommunications markets—deployment/accessibility of advanced telecommunications infrastructure and services. Gatineau, Canada: CRTC. DCITA. (2004). Australia¶s national broadband strategy. Canberra, Australia: NOIE. DCITA. (2006). Broadband blueprint. Canberra, Australia: DCITA. Distaso, W., Lupi, P., & Manenti, F.M. (2006). Platform competition and broadband uptake:
836
Theory and empirical evidence from the European union. Information Economics and Policy, 18(1), 87-106. Firth, L., Longstaff, P.H., & Mellor, D. (2002). Convergence driving regulatory change: Regulation driving convergence. Paper presented at the 14th Biennial Conference of the International Telecommunications Society, Seoul, Korea. Fransman, M. (2006). Global broadband battles: Why the U.S. and Europe lag while Asia leads. Stanford, CA: Stanford Business Books. García-Murillo, M., & Gabel, D. (2003). International broadband deployment: The impact of unbundling. Paper presented at the Telecommunication Policy Research Conference, Arlington, VA. Gardner, K. (2003). Broadband: The great Australasian divide. Paper presented at the International Telecommunications Society Asia-Australasian Regional Conference, Perth, Australia. Gillett, S.E., Lehr, W.H., Osorio, C.A., & Sirbu, M.A. (2006). Measuring broadband’s economic impact: Final report. Washington, DC: U.S. Department of Commerce, Economic Development Administration. Grubesic, T.H. (2003). Inequities in the broadband revolution. The Annals of Regional Science, 37, 263-289. Houghton, J., & Morris, P. (2001). Case study: Broadband the case of Australia. Geneva: International Telecommunication Union. Howell, B. (2002). Infrastructure regulation and the demand for broadband services: Evidence from OECD countries. Communications & Strategies, 47, 33-62. i2010. (2006). Broadband for all: Commission mobilizes all its policy instruments to bridge the broadband gap. Retrieved July 31, 2007, from http://europa.eu.int/rapid/pressReleasesAction. do?reference=IP/06/ 340&format=HTML&aged =0&language=EN&guiLanguage=en
The Adoption of Broadband Internet in Australia and Canada
Ida, T., & Kuroda, T. (2006). Discrete choice analysis of demand for broadband in Japan. Journal of Regulatory Economics, 29(1), 5-22. IDC. (2004). Press release: Australian broadband market trailing with 2.8 million subscribers by 2008, predicts IDC. Retrieved July 31, 2007, from http://www.idc.com.au/press/detail. asp?releaseid=88 Industry Canada. (2002). Broadband for rural and northern development. Ottawa, Canada. International Telecommunication Union. (2001). A broadband future: Reconciling opportunities and uncertainties. ITU News, 6. International Telecommunication Union. (2002). Trends in telecommunications reform 2002. Geneva: ITU. International Telecommunication Union. (2003a). Birth of broadband. Geneva: ITU. International Telecommunication Union. (2003b). Digital access index. Retrieved July 31, 2007, from http://www.itu.int/ITU-D/ict/dai/ International Telecommunication Union. (2003c). Promoting broadband: Background paper. Retrieved July 31, 2007, from www.itu.int/osg/spu/ni/ promotebroadband/PB03-PromotingBroadband. doc International Telecommunication Union. (2006a). Digital.Life: Internet report 2006. Geneva: ITU. International Telecommunication Union. (2006b). Statistics and analysis: International telecom statistics and analysis. Retrieved July 31, 2007, from http://www.itu.int/osg/spu/statistics/ Inukshuk. (2006). Inukshuk – who are we? Retrieved July 31, 2007, from http://www.inukshuk. ca/anglais/offre.html Kim, J.H., Bauer, J.M., & Wildman, S.S. (2003). Broadband uptake in OECD countries: Policy lessons from comparative statistical analysis. Paper presented at the Telecommunication Policy Research Conference, Arlington, VA.
.30* /HDGHUVRUODJJDUGV"$XVWUDOLD¶V broadband future. Retrieved July 31, 2007, from http://www.kpmg.com.au/Default.aspx?tabid=70 &kpmgarticleitemid=207 Kridel, D., Rappoport, P., & Taylor, L. (2002). The demand for high-speed access to the Internet: The case of cable modems. In D.G. Loomis & L.D. Taylor (Eds.), Forecasting the Internet: Understanding the explosive growth of data communications (pp. 11-22). Norwell, MA: Kluwer Academic Publishers. Lee, H., & Choudrie, J. (2002). Investigating broadband technology deployment in South Korea: Brunel-DTI International Technology Services Mission to South Korea. /HH+2¶.HHIH50 <XQ. 7KH growth of broadband and electronic commerce in South Korea: Contributing factors. The Information Society, 19(1), 81-93. Madden, G., Savage, S., & Simpson, M. (1996). Information inequality and broadband network access: An analysis of Australian household data. ,QGXVWULDO DQG &RUSRUDWH &KDQJH (4), 10491066. Madden, G., & Simpson, M. (1997). Residential broadband subscription demand: An econometric analysis of Australian choice experiment data. Applied Economics, 29, 1073-1078. Michalis, M. (2001). Local competition and the UROH RI UHJXODWLRQ 7KH (8 GHEDWH DQG %ULWDLQ¶V experience. 7HOHFRPPXQLFDWLRQV 3ROLF\ (1011), 759-776. Middleton, C.A. (2003). What if there is no killer application? An exploration of a user-centric perspective on broadband. Journal of Information Technology, 18(4), 231-246. Middleton, C.A., & Ellison, J. (2006). All broadband households are not the same: Why scope and intensity of use matter. Paper presented at the Statistics Canada Socio-Economic Conference, Ottawa.
837
The Adoption of Broadband Internet in Australia and Canada
Morris, J., & Middleton, C.A. (2005). The net generation? Exploring the complexities of innovation adoption among youth. In Association of Internet Researchers Annual Conference, Chicago, IL. National Broadband Task Force. (2001). The new national dream: Networking the nation for broadband access. Ottawa, Canada: Industry Canada. OECD Directorate for Science Technology and Industry. (2001). The development of broadband access in OECD countries. Paris: OECD. OECD Directorate for Science Technology and Industry. (2002). Broadband access for business. Paris: OECD. OECD Directorate for Science Technology and Industry. (2003a). Developments in local loop unbundling. Paris: OECD. OECD Directorate for Science Technology and Industry. (2003b). 7KHEHQH¿WVDQGFRVWVRIVWUXFWXUDO separation of the local loop. Paris: OECD. OECD Directorate for Science Technology and Industry. (2006). The implications of WiMAX for competition and regulation. Paris: OECD. 2I¿FHRI7HFKQRORJ\3ROLF\ Understanding broadband demand: A review of critical issues. Washington, DC: U.S. Department of Commerce. 2I¿FH RI WKH ((QYR\ 8. RQOLQH 7KH broadband future. An action plan to facilitate rollout of higher bandwidth and broadband services. $YDLODEOH IURP KWWS DUFKLYHFDELQHWRI¿FHJRY XNHHQYR\UHSRUWVEURDGEDQG¿OHLQGH[KWP Oh, S., Ahn, J., & Kim, B. (2003). Adoption of broadband Internet in Korea: The role of experience in building attitudes. Journal of Information Technology, 18(4), 267-280. Organization for Economic Co-Operation and Development. (2006a). OECD broadband statistics. Retrieved July 31, 2007, from www.oecd. org/sti/ict/broadband Organization for Economic Co-Operation and Development. (2006b). OECD broadband statistics,
838
June 2006. Retrieved July 31, 2007, from http:// www.oecd.org/dataoecd/50/49/37530046.xls Owen, B. (2002). Broadband mysteries. In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 9-38). Washington, DC: AEI-Brookings Joint Center for Regulatory Studies. Polykalas, S.E., & Vlachos, K.G. (2006). Broadband penetration and broadband competition: Evidence and analysis in the EU market. Info - The Journal of Policy, Regulation and Strategy for Telecommunications, 8(6), 15-30. Prieger, J.E. (2003). The supply side of the digital divide: Is there equal availability in the broadband Internet access market? Economic Inquiry, 41(2), 346-363. Rappoport, P.N., & Kridel, D.J. (2003). Residential demand for access to the Internet. In G. Madden (Ed.), The international handbook of telecommunications economics volume II: Emerging telecommunications networks (pp. 55-72). Cheltenham, UK: Edward Elgar. Rappoport, P.N., Kridel, D.J., & Taylor, L.D. (2002). The demand for broadband: Access, content, and the value of time. In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 57-82). Washington, DC: AEIBrookings Joint Center for Regulatory Studies. Rappoport, P., Taylor, L.D., & Kridel, D.J. (2003). Willingness-to-pay and the demand for broadband service. In A.L. Shampine (Ed.), Down to the wire: Studies in the diffusion and regulation of telecommunications tchnologies (pp. 75-86). New York: Nova Science Publishers, Inc. Riley, J. (2004, August 31). Our broadband two years behind. Australian IT. Rogers, E.M. (1995). Diffusion of innovations (4th ed.). New York: The Free Press. Sacks. (2002). Determinants of broadband adoption for NGOs in the healthcare industry. Melbourne, Australia: University of Melbourne, Honors Thesis.
The Adoption of Broadband Internet in Australia and Canada
Sale, A. (2001). Broadband Internet access in regional Australia. JRPIT, 33(4), 346-355. Saunders, R. (2005). Benchmarking broadband. 7HOHFRPPXQLFDWLRQV-RXUQDORI$XVWUDOLD(2), 20-24. Sawyer, S., Allen, J.P., & Lee, H. (2003). Broadband and mobile opportunities: A socio-technical perspective. Journal of Information Technology, 18, 121-136. Shade, L.R., Porter, N., & Sanchez, W. (2006). You can see anything on the Internet, you can do anything on the Internet! Young Canadians talk about the Internet. Canadian Journal of Communication, 30(4), 503-526. Shelanski, H.A. (1999). The speed gap: Broadband infrastructure and electronic commerce (the legal and policy framework for global electronic commerce). Berkeley Technology Law Journal, 14(2), 721. Smith, M.W., & Leung, H.-T. (2002). Finding the killer application - the role of the broadband applications laboratory. BT Technology Journal, 20(1), 11-21. Statistics Canada. (2004). Microdata user guide household Internet use survey 2003 reference year. Ottawa, Canada: Statistics Canada. Statistics Canada. (2006). Canadian Internet use survey. Retrieved July 31, 2007, from http://www. statcan.ca/Daily/English/060815/d060815b.htm Stern, S., Gregor, S., Martin, M., Goode, S., & Rolfe, J. (2004). $FODVVL¿FDWLRQWUHHDQDO\VLVRI broadband adoption in Australian households. Paper presented at the Sixth International Conference on Electronic Commerce, Delft, Holland. Task Force on Broadband Communications. (2002). Task force on broadband communications report. Italy: Ministry of Communications, Minister for Innovation and Technologies. Ure, J. (2003). Competition in the local loop: Unbundling or unbungling? Info - The Journal of Policy, Regulation and Strategy for TelecomPXQLFDWLRQV(5), 38-46.
Varian, H.R. (2002). The demand for bandwidth: Evidence from the index project. In R.W. Crandall & J.H. Alleman (Eds.), Broadband: Should we regulate high-speed Internet access? (pp. 39-56). Washington, DC: AEI-Brookings Joint Center for Regulatory Studies. Wales, C. (2002). In search of the elusive “killer application” that will drive broadband uptake in Australian households. Melbourne, Australia: University of Melbourne, Honours Thesis. Watson, D. (2002, December 21). Unbundling to be investigated. Computerworld. Zamaria, C., Caron, A.H., & Fletcher, F. (2005). Canada online! A comparative analysis of Internet users and non-users in Canada and the world: Behavior, attitudes and trends 2004. Toronto, Canada: Canadian Internet Project.
KEY TERMS Local Loop Unbundling: The term “local loop” refers to telephone infrastructure used to provide DSL (digital subscriber line) broadband Internet access to consumers in their homes. In many countries, the local loop was built and operated by a monopolist telephone company, referred to as the “incumbent” carrier. In order to increase competitive provisioning of DSL service, “local loop unbundling” policies have been developed. These policies force the owners of the local loop infrastructure to make their infrastructure available to other providers (for a fee), so that competitors can provide DSL service without building new infrastructure. There is no consensus as to whether such policies have been successful in promoting broadband uptake. Open Access: “Open access” policies are similar to local loop unbundling policies, but apply to the cable industry. Also known as “third party Internet access” rules, existing cable operators are required to make their network capacity available to competitors in an open access environment.
839
The Adoption of Broadband Internet in Australia and Canada
Facilities-Based Competition: This term, also known as “intermodal competition” refers to competition between DSL and cable providers. There is high intermodal competition in the U.S. and Canadian consumer broadband markets, and it is believed that intense intermodal competition stimulates demand for broadband access. In countries where the cable company was/is owned by the telephone company, intermodal competition KDVEHHQOHVV¿HUFH Download Caps: Many broadband service providers offer tiered pricing schemes for broadband connectivity. Common to such pricing schemes are restrictions on the volume of data that can be downloaded within a given period. For instance, in Australia, some broadband packages allow 0.2 Gb of “free” downloads monthly, compared to packages in Canada that offer 30-60 Gb of downloads monthly. When download caps are low, they can reduce demand for broadband connectivity. Depending on the provider and the service package chosen, consumers may have to pay for excess downloads, or may have their connection speeds reduced to dial-up speeds. Broadband Success Drivers: Gardner (2003) argues that success factors for broadband growth (which include favorable pricing, competitive market environment, and population readiness, among others) have differential impacts on growth LQ GHPDQG IRU EURDGEDQG VHUYLFHV *DUGQHU¶V UHVHDUFKLGHQWL¿HV DFFHOHUDWRUVIRUEURDGEDQG demand, (2) stimulants for broadband demand, and (3) threshold factors for broadband demand. Recognition of this hierarchy of success factors is important in developing policies and taking actions to increase broadband adoption. Accelerators: 7KH IDFWRUV WKDW PRVW VLJQL¿cantly increase demand for broadband services, and include the establishment of a competitive
840
marketplace, user comfort levels with the technology, and network characteristics (e.g., speed, multiple platforms). Stimulants: Have a moderate impact on increasing demand for broadband services. Stimulants include pricing and the sophistication of marketing broadband services. Threshold Factors: Necessary, but not suf¿FLHQW WR HQFRXUDJH EURDGEDQG GHPDQG 7KH\ include penetration of PCs in households, narrowband Internet adoption, and adequate development of broadband infrastructure.
ENDNOTES 1
2
3
Note that there are multiple sources of data on broadband adoption worldwide. Frequently cited statistics come from the OECD (www. oecd.org/sti/telecom) and the ITU (www.itu. int/osg/spu/statistics). Not all sources contain information on all countries. Some statistics measure broadband adoption by numbers of subscribers (e.g., subscribers per 100 inhabitants), others by numbers of households (e.g., percentage of national totals), and these numbers are not necessarily comparable. DSL is an abbreviation for digital subscriber line. The acronym xDSL is used to represent generic DSL service, for example asymmetrical DSL is ADSL. Throughout the paper, references to DSL service in Canada and Australia describe ADSL service, but the simpler DSL acronym is used. For instance, less than 15% of Canadian students agreed with the suggestion that they used the internet because the government promoted internet usage (Morris & Middleton, 2005).
841
Chapter L
The Evolution of Broadband Industry in the Developing World: A Comparison of China and India Nir Kshetri University of North Carolina at Greensboro, USA Nikhilesh Dholakia University of Rhode Island, USA
ABSTRACT Telecommunications networks of India and the People’s Republic of China are among the largest in the world. The two economies have a number of areas for broadband use ripe for exploration. Broadband networks in some regions in these two economies are even more developed than in some parts of the industrialized world. There are, however, a number of reasons to believe that these two countries may exhibit distinct and varied patterns of broadband diffusion. This chapter compares and contrasts the diffusion patterns of broadband technology in the two economies. We examine factors driving broadband diffusion in the two economies in three major categories: demand and cost conditions, industry structure, and export conditions.
INTRODUCTION Telecommunications networks of India and the 3HRSOH¶V 5HSXEOLF RI &KLQD KHQFHIRUWK &KLQD are among the largest in the world. Both have a number of areas for broadband use ripe for exploration. One estimate suggests that thanks mainly to rapid broadband diffusion in China and India, EURDGEDQGUHYHQXHLQWKH$VLD3DFL¿FUHJLRQZLOO
cross $55 billion by 2011 compared to $20 billion in 2005 (Indiantelevision.com, 2006). As of 2005, in terms of the numbers of the Internet users as well as broadband users, China ranked the second in the world—only after the U.S. One study suggested that by the early 2006, a Chinese Internet user was more likely to be on broadband connections than U.S. counterparts (Koprowski, 2006). By the early 2006, over half of Chinese Internet-users had broadband compared
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
The Evolution of Broadband in the Developing World
to only 6.6 percent at the end of 2002 (Special report, 2006).1 By the early 2004, China had 10.95 million DSL (a form of broadband) users, the highest in the world.2 0RUHRYHU &KLQD¶V EURDGEDQG network growth rate is among the fastest in the world (Country growth rates, 2004). By the end of 2006, China is expected to have more Internet users and broadband lines than any other country in the world (China.net, 2004).
Currently, India is far behind China in terms of broadband and other related indicators (Table 1). A report released by Forrester Research in May 2006 indicated that the top three socioeconomic classes in urban India had only a 3% adoption rate of the broadband technology (zdnetindia.com, 2006). Nonetheless, India is rapidly catching up LQWKHEURDGEDQGUDFH,QGLD¶V0LQLVWU\RI&RPmunications and Information Technology hopes
Table 1. Indicators related to broadband development in China and India Indicator
China
India
3HUVRQDOFRPSXWHUV 3&V LQXVH µ
62241.13
12627.19
Online households (million) (2004)
20.42
5.85
PC households online (% of PC households) (2004)
70.97
53.17
1XPEHURI,QWHUQHWXVHUV µ
162097.46
54159.55
,6'1VXEVFULEHUV µ
1615.74
37.53
Broadband users as a percentage of Internet users (2005) d
45
6
Number of broadband households (2005-end)b
30 million
550,000
Composition of broadband access modes (mid-2006)
70% via DSL, 18% YLD¿EHUWRWKH premises (FTTP), 12% via cable or other means (e.g., ¿[HGZLUHOHVV:L)L or satellite)e
Average monthly broadband charges ($, 2004)a
16
20
Digital main lines % of telephone main lines (2004)
100
100
Capital investment in telecommunications (million, $) (2004)
26033.78
5036.52
7HOHSKRQHOLQHVLQXVH µ
282524
47188
National telephone calls (million minutes) (2004)
66339.08
36.52
International outgoing telephone calls (million minutes) (2004)
1263.16
680.81
Mobile telecom revenues (% of telecom revenue) (2004)
57.8
17.95
0RELOHWHOHSKRQHXVHUV µ
334082
40323.45
Mobile SMS messages sent, million (2004)
93156.85
NA
Number of television sets (2005) c
400 million
Number of cable TV subscribers (2005) c
150 million
Sources: a Mishra (2004); b%XUURZVHWDO cBasu (2006); d1HHGWRUHYROXWLRQL]H e Lindstrom (2006). All other data are from Euromonitor International’s Global Market Information Database.
842
The Evolution of Broadband in the Developing World
that broadband will reach 12 million homes in the country by 2010 (Burrows, Kripalani, & Einhorn, 2005). Broadband networks in some regions in these two economies such as Shanghai and other coastal towns of China (Mallaby, 2005) and Bangalore and Hyderabad in India are more developed than in some parts of the industrialized world (COMMWEB, April 26). Sperling (2004) notes that “there is better broadband in Bangalore than [in] Buffalo.” /LNHZLVHFLWLQJDQH[SHULHQFHRID86EDVHG¿UP Wilson (2005) has observed that it is easier to get a broadband connection in some cities in India and China “than to get one to [a] factory 70 miles away in Batesville, [Mississippi (USA)].” Thanks to broadband and other information and communications technology (ICT), skilled
manpower in the two economies have been able to actively participate in the design and development of innovative new products thus breaking traditional geographical barriers (Skills gap survey, 2006). For instance, broadband has been among the PDMRUGULYHUVRI,QGLD¶VUDSLGO\JURZLQJRXWVRXUFing industry. Some analysts argue that the theory RIFRPSDUDWLYHDGYDQWDJHPD\QHHGPRGL¿FDWLRQ when developing countries such as China and India start competing with industrialized countries with high skilled workers that have access to broadband and the Internet (How America can meet, 2004) There are a number of reasons to believe that these two countries may exhibit distinct and varied patterns of broadband diffusion. First, the two economies differ widely in terms of the potential of possible broadband applications such as multi-
Table 2. Sources of leadership in broadband technology: China vs. India Dimension
China
India
Demand and cost
•
Higher income
•
conditions
0RELOHDQG¿[HGOLQHSHQHWUDWLRQUDWHVKLJKHU
connected to cable
•
Better infrastructure than India
•
•
About 75% of mobile phones based on GSM
and BP outsourcing.
standard, requirement to build new 3G networks from
Higher proportion of TV household Position as a global capital for IT
•
Sophisticated rail network.
Ahead of China in telecom
scratch •
Higher International bandwidth
•
Produces most IT products (including broadband
equipment) domestically
Industry structure
•
Lower monthly subscription rate
•
Wider and deeper adoption of the Internet
•
Reengineered the telecom sector in the 1990s
•
•
Opening the telecom market for competition.
liberalization •
Some telecom equipment
manufacturers have programs focused on India Export conditions
•
•
Heavy exporter of software
exports much higher
Export of hi-tech products as a proportion of total
•
Taking several measures in recent
years
+HDY\LQÀRZRI)',LQWHFKQRORJ\VHFWRU
facilitated export via inward internationalization •
Higher market size
Sources: Beise (2001), SinoCast (2004, March 9), Lehrer et al. (2002), Lawcommerce.com (2001), Pyramid Research (2001), Business (2002), and authors’ research.
843
The Evolution of Broadband in the Developing World
media and animation, business-to-business (B2B), business-to-consumer (B2C) e-commerce, and IT enabled services. Second, they differ widely in terms of the level and composition of base technologies for the development of different varieties of broadband (Table 1). For instance, broadband access in the forms of ISDN/DSL, cable modems and wireless GHSHQGVRQWKHSHQHWUDWLRQUDWHVRI¿[HGSKRQHV cable TV, and wireless phones respectively. China is ahead of India in terms of the penetration rates of all these technologies (Table 1). Third, China and India differ in terms of the degree and type of liberalization of telecom sectors. While a stronger economy is driving the demand of broadband related applications in China, competition for broadband and traditional telecom services is considered to be “healthier” in India than in China (Pyramid Research, 2001). Fourth, other supporting infrastructures required for broadband development are not equally developed in the two economies. For instance, the Chinese initiative to launch its version of national information infrastructure (NII), known as the “Golden Projects” in response to similar initiatives in the developed countries (Tan, Meuller, & Foster, 1997), has put in place a very developed backbone telecom infrastructure in China than in India (Pyramid Research, 2001). What are the sources of similarity and differences in broadband diffusion pattern of India and China? In this chapter, we examine in depth the factors that could shape the patterns of diffusion of broadband telecommunications technologies in the two countries.
FACTORS IMPACTING THE BROADBAND POTENTIAL Following Beise (2001), Lehrer (2004), and Lehrer, Dholakia, and Kshetri (2002), we divide factors LQÀXHQFLQJWKHGLIIXVLRQRIEURDGEDQGLQWKHWZR economies into three groups (Table 2): demand and cost conditions (Linder 1961; Vernon 1966), industry structure (Porter, 1990), and export conditions (Beise, 2001; Tilton, 1971) (Figure 1).
844
Demand and Cost Conditions Demand and cost conditions include factors such as consumer preferences, income level, input costs, and prior national experience with related or previous generations of technology. China and India have favorable environments in terms of some of these factors. For instance, governments in both countries subsidize technology and innovation (Newman, Hook, & Moothart, 2006). Next, consider technical workers needed for the broadband industry. China and India graduated over 600,000 and 350,000 engineers in 2004 compared to 70,000 in the U.S. (Broad, 2005). According to a survey, the annual cost to employ a chip design engineer in 2002 was $28,000 in Shanghai, $24,000 in Suzhou (China) and $30,000 in India compared to $300,000 in the U.S. (Ernst, 2005). The Bangalore unit of the Intel team has planned to launch microprocessor chips for high-speed broadband wireless technology in 2006 (Mehra & Padmanabhan, 2006). Regarding the quality of technology workforce, Intel CEO Craig Barrett argues that the Chinese are “capable of doing any engineering, any software job, any managerial job that people in the United States are capable of” (Segal, 2004). Indeed, Chinese telecom equipment vendors Huawei and ZTE have supplied most of the network equipment needed for deploying broadband and related applications such as ,379LQ&KLQD/LNHZLVH,QGLDDQG&KLQDKDYHQ¶W invested heavily in the legacy systems that most industrialized countries have (Vance, 2005). Nonetheless, there are some factors that are likely to hamper the growth of the broadband network. First, shared broadband access is relatively more common in China and India than in other industrialized countries, which is likely to hamper the demand for broadband (SinoCast China IT Watch, 2006). Second, limited authorized content on the Internet may act as a roadblock to a faster diffusion of this technology (Clendenin, 2005). %RWK FRXQWULHV¶ ODZV SURKLELW FRQWHQWV WKDW DUH politically and/or culturally objectionable. So far, broadband has penetrated only a tiny fraction of the population in both economies. Broadband subscription is beyond the reach of the majority of Chinese and Indians (Chotrani,
The Evolution of Broadband in the Developing World
Figure 1. A framework to explain broadband diffusion in developing countries Consumer characteristics impacting broadband demand o Income o Innovativeness o Adoption of related technologies o Related infrastructure o Existence of domestic standards o Input characteristics o Quality of IT workforce o Cost of hiring
%URDGEDQG diffusion x Diffusion level x Diffusion pattern
Industry structure o Competitive rivalry o Horizontal collaboration
Related infrastructure x Export orientation of industry x Market size
2002). Demand and cost conditions, however, favor China compared to India. First, China has a higher income level than India and the income is growing more rapidly. For instance, the Chinese economy grew by 7% annually in the 1980s, 8.5% annually during 1990-20033 and the material consumption per capita increased by over 100% during 1978 6NODLU 7KHVHQXPEHUVDUHUHÀHFWHG LQ&KLQD¶VPDUNHWVL]HIRU,7SURGXFWVLQJHQHUDO and broadband in particular. For instance, China EHFDPH WKH $VLD¶V ODUJHVW ,7 PDUNHW LQ ,Q&KLQDEHFDPHWKHZRUOG¶VELJJHVW,&7 H[SRUWHU,QWKH&KLQD¶V,6'1PDUNHWZDV DERXWWLPHVELJJHUWKDQ,QGLD¶V 7DEOH Second, “technological innovativeness” or the propensity to adopt modern technologies such as broadband is higher in China than in India.
For instance, in the mid-1980s, the penetration rates of consumer durables in China were about the same as South Korea, Japan, and then USSR (Sklair, 1994). Thanks to such “technological innovativeness” of Chinese consumers, in terms of the technology achievement index recently FRQVWUXFWHGE\WKH81'3 &KLQD¶VUDQNRI 45 (out of 72 economies considered) puts it in the group of “dynamic adopters” of new technologies and ahead of other developing countries with higher per capita GDP such as Bolivia, Colombia, Peru, Paraguay, Jamaica, and Tunisia. The same study ranked India 63rd out of the 72 economies. Chinese also have a wider and deeper adoption of the Internet compared to Indians (Kshetri, 2002), which is likely to boost broadband demand. For instance, estimates suggested that the numbers of
845
The Evolution of Broadband in the Developing World
blogs in China were in the range of 1-2 million in 2005 and growing fast (French, 2005). Third, China has made much heavier investment in telecom sector (Table 1). For instance, in the mid-1980s, annual investment in the telecom infrastructure averaged $300 million (30% of revenue) which resulted in an annual network growth of 14% (Kelly 2000). During the early VLQÀDWLRQLQWKH&KLQHVHHFRQRP\WKHJRYernment cut off investments in other sectors, but not in the telecom sector. During 1992 and 1993, investment was again doubled which resulted in telecom growth rates of 36 and 48% respectively (Kelly, 2000). The telecom investment experienced similar level of growth rate during and after the $VLDQ¿QDQFLDOFULVLV,WVWHOHFRPLQYHVWPHQWLQ 1999 was about eight times as much as that of India (International Marketing Data and Statistics, 6LPLODUO\&KLQD¶VFDSLWDOLQYHVWPHQWLQWKH WHOHFRPVHFWRULQZDVRYHU¿YHWLPHVDVELJ DV,QGLD¶V 7DEOH Fourth, a high proportion of the telecom investment in China went to the “most modern” available infrastructure because the government aspired for “nothing but the best.” For instance, consider Shanghai, the most populous Chinese city. In the late 1980s and early 1990s, Shanghai 7HOHFRPXQGHUWRRNSURMHFWVWROD\XQGHUWKHFLW\¶V VWUHHWVRQHRIWKHZRUOG¶VODUJHVW¿EHURSWLFFDEOH networks that contained much more bandwidth than needed for simple telephone services (McGill, 2001). The huge surplus in bandwidth accelerated the diffusion of broadband technology. The investment in the “most modern” infrastructure is a part of Chinese national initiatives to develop telecom infrastructure and high-speed data networks. China launched its version of national information infrastructure (NII), known as the “Golden Projects” in response to similar initiatives in the developed countries (Tan et al., 1997). There are six network organizations that operate international telecommunication circuits and network facilities: China Public Computer Network (CHINANET), China *ROGHQ%ULGJH1HWZRUN &+,1$*%1 8QLFRP¶V UNINET, China Netcom (CNCNET), Chinese Science and Technology Network (CSTNET), and China Education Network (CERNET) (Anderson
846
2001). China received better marks than India for infrastructure (Pyramid Research, 2001). Fifth, China has developed its own version of a 3G cellular standard known as time division-synchronous code division multiple access (TD-SCDMA), which has been approved by the International Telecommunications Union (ITU). The existence of a home-grown standard is likely to accelerate the growth of cellular broadband in China. 6L[WK DOWKRXJK KLJKHU SHUFHQWDJH RI ,QGLD¶V television is cable ready, the absolute number is much higher in China. Thus, the potential of cable broadband is also higher in China. Seventh, government agencies in China are among the early adopters of most IT applications LQFOXGLQJ EURDGEDQG &KLQD¶V HJRYHUQPHQW performance, in fact, is rated ahead of global IT leaders such as Switzerland, UK, Singapore, and Germany (West, 2002). Federal and local government agencies are rapidly adopting a wide range of broadband applications. For instance, Motorola is working with the Chinese Ministry of Public Security (MPS) in an ambitious project to deploy YLGHRVXUYHLOODQFHV\VWHPVFRQQHFWHGYLD¿EHULQ major cities in the country (Lindstrom, 2006). While China fares better than India in several dimensions related to demand and cost conditions, ,QGLDKDVVRPHUHODWLYHDGYDQWDJHVDVZHOO,QGLD¶V position as a major supplier of global IT and business process services is likely to trigger the demand of broadband services. An estimate suggests that global IT services will employ 1.1 million Indian white-collar workers generating nearly $19 billion in annual revenues by 2008 (Dhume, 1999). Thanks to improved telecommunications and the Internet, multinationals such as GE Capital, British Airways, Swiss Air, and American Express DUHXWLOL]LQJ,QGLD¶VYDVWSRRORI(QJOLVKVSHDNLQJ and computer-literate workers to perform whitecollar tasks remotely from India (Dhume, 1999; Ebusinessforum.com, 2001). Similarly, to relocate its accounting division from Washington D.C., World Bank evaluated several countries in terms of the availability of IT manpower at reasonable cost, and chose Chennai, India (Goonewardene, %\ ,QGLD¶V ,7UHODWHG VRIWZDUH DQG
The Evolution of Broadband in the Developing World
service industries revenue had touched $9 billion and was expanding at a steady clip of 30-40% annually. Another estimate suggests that IT outsourcing brought India more than $7 billion a year in revenue in 2003, which is expected to double by 2007 and quadruple by 2012 (Donlan, 2005). ,QWKH¿VFDO\HDUWKH,QGLDQRIIVKRUH outsourcing industry generated an estimated $17.3 billion and employed 695,000 people (Fairell, Kaka, & Stürze, 2005).4 According to a NASSCOM/KPMG study, in 2004, the Indian business process outsourcing (BPO) industry grew by 40% to reach $5.8 billion and is expected to reach $64 billion in 2012, employing 3 million people (Hamm, 2005). What is more, companies based in industrialized countries are outsourcing more cutting edge works to India such as medical outsourcing (e.g., CT scans and other images transmitted via broadband) (Waldman, 2004) and R&D. The following example illustrates how broadband has facilitated the outsourcing of sophisticated works to India: Narayana Hrudalaya, a Bangalore heart KRVSLWDOKDVOLQNVWRDI¿OLDWHVWKURXJKRXW,QGLD in Kuala Lumpur and Mauritius, and a medical school in Hannover, Germany. The Indian doctors consult with patients over broadband connections, on which they can see X-rays and angiogram results. Future plans include robotic surgery by remote control, perhaps the ultimate in outsourcing. (Singh, 2004) ,QGLD¶VZHOOGHYHORSHGUDLOQHWZRUNLVH[SHFWHG to be another major driver of broadband. India is using its rail network to bring broadband access economically and quickly to the rural population (International Telecommunication Union, 2003).
Industry Structure Industry structure includes the market openness for competition in broadband and traditional telecom sectors. China and India are taking several measures to increase competition in the telecom market. A study conducted by Pyramid Research (2001) indicated, however, that the competition for broadband and traditional services is better in India than in China.
According to the recommendation provided by a high-powered IT task force, the Indian government privatized the national long-distance market, permitted ISPs to set up their own submarine cable landing stations and share bandwidth with other ISP, and allowed the use of Ku-band in both Indian and foreign satellites among other measures to liberalize telecom sector. These factors are bringing a bandwidth boom in India. During mid 2000 to mid 2002, an estimated $5 billion investment LVH[SHFWHGLQQHZ¿EHURSWLFV\VWHPV (ULFNVRQ 2000). MNCs such as Lucent Technologies and $LUWHODUHOD\LQJ¿EHURSWLFOLQHV,Q2FWREHU Singapore Telecom announced an eight-terabit cable connection between Singapore and India (Lynch, 2001). Likewise, VSNL, the state owned telephone company, which listed on the New York Stock Exchange in the mid-2000, has proposed to increase its capacity to 13 gigabits by 2004 (Erickson, 2000). In mid 2004, the government issued a report which stressed the need for facilitating the expansion of broadband and emphasized especially the role of wireless broadband.5 In recent years, investment in programming and digital distribution has increased, which is likely to grow competition in broadband (BBC Monitoring International Reports, 2006). The Chinese government, on the other hand, designed a series of programs to accelerate telecom development in the 1990s which included extensive re-engineering of and intense competition in this sector. Telecom companies were also required to adapt to the rigorous disclosure requirements of WKH1<6(1$6'$4DQG+RQJ.RQJ¶V*URZWK Enterprise Market (McDaniels & Waterman, 2000). Especially, ADSL rollout from China Telecom and China Netcom triggered the growth of broadband network in the country (BBC Monitoring International Reports, 2006). In India, horizontal collaboration, which entails FRPELQLQJVHOHFWHGHOHPHQWVIURPHDFKEXVLQHVV¶V supply chain, is also facilitating broadband growth. An alliance formed by Microsoft, HCL (a manuIDFWXUHURIFRPSXWHUVDQGRI¿FHHTXLSPHQW DQG BSNL offers a case in point. The alliance makes a PC loaded with Microsoft software ready for %61/¶VEURDGEDQGVHUYLFHIRUDPRQWKO\SD\PHQW
847
The Evolution of Broadband in the Developing World
of less than $11 (Goyal, 2005). Thanks to such initiatives, new PC sales in India are also growing at 25% annually (Goyal, 2005).
Export Conditions Export conditions include factors such as trade policy, the export orientation of industry, strategic regulation, and market size. China fares better than India in terms of several dimensions of transfer and export conditions related to broadband technology. First, the ICT market size is higher in China than in India. Because of its higher market size, China is ahead of India in terms of several indicators related to broadband such the number of hosts, number of users, and total e-commerce transactions (Kshetri, 6HFRQG&KLQD¶VKLJKWHFKQRORJ\H[SRUWV as a proportion of manufactured exports is much higher than in India. More to the point, Chinese telecom equipment vendors Huawei and ZTE have DOUHDG\VROG¿[HGDQGPRELOHEURDGEDQGHTXLSPHQW in a number of developing and developed countries DFURVVDOO¿YHFRQWLQHQWV &KLQD¶V LQZDUG LQWHUQDWLRQDOL]DWLRQ WR VRPH extent, facilitated the export of high technology SURGXFWV&KLQD¶VSRVW0DRUHIRUPFRPSRXQGHG E\LWVELOOLRQSOXVPDUNHWDWWUDFWHGKXJHLQÀRZRI foreign direct investment (FDI) in the technology sector (inward internationalization). The annual )',LQÀRZLQ&KLQDDYHUDJHGELOOLRQGXULQJ 1985-1995 and $42 billion during 1996-1999, which FRPSDUHZLWKFRUUHVSRQGLQJ¿JXUHVIRU,QGLD billion and 2.7 billion respectively (UNCTAD, 2000). FDI in China averaged 13.9% of gross ¿[HGFDSLWDOIRUPDWLRQGXULQJ ZKLFKLV VLJQL¿FDQWO\KLJKHUWKDQWKDWRIWKHZRUOG and developing countries average (10.1%) during that period (UNCTAD, 2000). Several factors restricted the internationalL]DWLRQ RI ,QGLD¶V ,&7 LQGXVWU\ SDUWLFXODUO\ WKH hardware sector, in the 1980s and 1990s. The government protected the hardware industry from external competition. Its distorted tariff structure further hampered the internationalization of the hardware sector (Gopalan, 2002). Reduced access to foreign exchange and distorted tariff structure increased the transaction cost for exporting
848
hardware which is more import-intensive (64% of sales volume was imports in the early 1990s) than software (40% import-intensity) (Tandon, 1991, p. 65) and encouraged even hardware companies to consider software exports to earn foreign exchange (McDowell, 1995). In recent years, liberalization in Indian economy has led to a decline of such barriers. Moreover, competitive threats from China and other neighboring countries have forced India to develop a strong hardware industry to support software and services sectors. Thanks to measures taken in recent years, the export of the information and communications technology (ICT) industry is, however, growing DWDSKHQRPHQDOUDWHLQ,QGLD,QWKH¿VFDO\HDU ,QGLD¶VH[SRUWRIKDUGZDUHVRIWZDUH and information technology (IT) enabled services amounted $1 billion (Elcina.com 2002), $6.2 billion, and $896 million (NASSCOM, 2002) respectively. By 2008, the Indian ICT industry is estimated to earn export revenues of $10 billion from hardware (Elcina.com, 2002), $50 billion from software and $17 billion from IT enabled services (NASSCOM, 2002) and provide employment to 7 million people.6
DISCUSSION AND CONCLUSION In this chapter, we examined several factors that DUH OLNHO\ WR LQÀXHQFH WKH XQIROGLQJ EURDGEDQG potential in the two Asian giants. Our analysis indicates, in terms of the demand and cost conditions affecting the potential of broadband, many factors favor China. These factors include higher income, higher propensity of Chinese consumers to adopt new technologies, higher investment in WHOHFRPVHFWRUDQGDVLJQL¿FDQWSURSRUWLRQRILW going to the most modern technology, much higher mobile phone and cable penetration. On the other KDQG,QGLD¶VSRVLWLRQLQWKHJOREDO,7PDSDVD major provider of IT services is likely to trigger the demand for broadband. The competition levels in the broadband and traditional telecom sectors are roughly comparable in the two economies. India, however, fares slightly better in terms of opening up the telecom sector to competition. As a result,
The Evolution of Broadband in the Developing World
the broadband subscription costs are declining rapidly in both economies. Such cost declines are likely to further drive the demand for broadband technology. Although China fares better in terms of several dimensions of transfer and export conditions, India has taken several measures in recent years to improve the situation. Our analysis in this chapter should be of value to broadband manufacturers and service providers as they devise appropriate strategies for these two giant Asian markets as well as for other emerging markets. For instance, this chapter provides insights into the nature of business and consumer demands and performances current business models offered in the two economies. Furthermore, the insights provided by this study would help policymakers and industry lobby groups to craft appropriate policies to accelerate the diffusion of broadband technology in emerging and transition economies. For instance, the Indian case suggested that a properly formed horizontal collaboration could be an important trigger for driving broadband diffusion. It would thus be in the best interest of the government to encourage such collaborations.
REFERENCES Basu, I. (2006, January 4). China excited for IPTV. UPI. Calcutta, India: UPI. http://www. accessmylibrary.com/coms2/summary_028612226682_ITM BBC monitoring international reports. (2006, March 2). India, Japan forecast to lead Asian SD\79 PDUNHW E\ (Study). Reading, UK: BBC Monitoring. Beise, M. (2001). /HDGPDUNHWV &RXQWU\VSHFL¿F success factors of the global diffusion of innovations. Heidelberg, Germany: Physica-Verlag. Broad, W.J. (2005, October 13). Top advisory panel warns of an erosion of the U.S. competitive edge in science. New York Times, p. A22. Burrows, P., Kripalani, M., & Einhorn, B. (2005, November 28). Cisco: Sold on India. Business Week, 3961, 50.
Business: 3G by any other name, mobile telecoms. (2002, January 12). The Economist, 72. China.Net; China will soon be no. 1 in Web users. That will unleash a world of opportunity. (2004, March 15). Business Week, 22. Chotrani, R. (2002). Keep it cheap. CIO. Retrieved August 1, 2007, from http://www.cio-asia.com/ pcio.nsf/unidlookup/FEE795C1C754FDC44825 6BEC00124804?OpenDocument Clendenin, M. (2005, December 19). Internet TV gets running start in large Asia markets – Hong Kong service thrives; China, India face roadblocks. Electronic Engineering Times, 1. COMMWEB. (2006, April 26). World’s digital divide is narrowing. Australia: Commonwealth Bank of Australia. Country growth rates in broadband compared. (2004, May 3). Oil and Gas Investor’s A&D Watch, 2(17), 1. Dhume, S. (1999, September 2). Catching the bus. Far Eastern Economic Review, 10-11. Donlan, T.G. (2005, February 14). Catch a tiger by the tail. %DUURQ¶V(7), 39. Ebusinessforum.com. (2001, May 3). India – back RI¿FHWRWKHZRUOGThe Economist. London. Elcina.com. (2002). Budget 2002-3: Observations and implications. Retrieved from: http://www. elcina.com/sg_future.html Erickson, J. (2000, September 6). The net: Bandwidth bottleneck. India sorely needs more telecom capacity. Asiaweek. Ernst, D. (2005). Complexity and internationalization of innovation - why is chip design moving to Asia? (Working Paper). East West Center. Retrieved August 1, 2007, from http://www.eastwestcenter. org/stored/misc/Complexity_and_Internationalisation_Chip_design_1.3.05.pdf Euromonitor. (2001). International Marketing Data and Statistics. London: Euromonitor.
849
The Evolution of Broadband in the Developing World
Fairell, D., Kaka, N., & Stürze, S. (2005). EnsurLQJ ,QGLD¶V RIIVKRULQJ IXWXUH >6SHFLDO HGLWLRQ@ McKinsey Quarterly, 74-83. French, H.W. (2005, November 24). A party girl OHDGV&KLQD¶VRQOLQHUHYROXWLRQNew York Times, p. A1. Goonewardene, D. (2001). World bank moves some departments to India. The Sunday Times. Retrieved August 1, 2007, from http://www.is.lk/ times/010701/busm.html
Retrieved August 1, 2007, from http://www.technewsworld.com/story/48219.html Kshetri, N. (2002). What determines Internet diffusion loci in developing countries: Evidence from China and India. 3DFL¿F7HOHFRPPXQLFDWLRQV Review, 23(3), 25-34. Lawcommerce.com. (2001, August 13). The PeoSOH¶V5HSXEOLFRI&KLQDSRLVHGWRMRLQWKHZRUOG trade organization: Implications for foreign telecommunications and media companies. Retrieved August 1, 2007, from http://www.lawcommerce. com/newsletters/art_LW_telecomm010813.asp
Gopalan, R. (2002). A study of the performance of the Indian IT sector. Prepared for the Nautilus Institute for Security and Sustainable Development, San Francisco, CA. http://www.nautilus. org/archives/cap/reports/IndiaChapters1-3.pdf
Lehrer, M. (2004). National lead markets and the design competition for 3G network applications. -RXUQDORI%XVLQHVV5HVHDUFK 1397-1401.
Goyal, M. (2005, October 17). High speed, low cost; video chats, e-commerce, online gaming DQGVWRFNWUDGLQJQRERG\GRXEWHGWKHEHQH¿WVRI broadband. Now better availability at a low price is turning the virtual dream into a reality. India Today, 50.
Lehrer, M., Dholakia, N., & Kshetri, N. (2002, August 8-11). National sources of leadership in 3G M-business applications: A framework and evidence from three global regions. In Proceedings of the Americas’ Conference on Information Systems (AMCIS), Dallas, Texas.
Hamm, S. (2005, August 22). Taking a page from 7R\RWD¶VSOD\ERRNBusiness Week, 3948, 69-72.
Linder, S.B. (1961). An essay on trade and transformation. Uppsala: Almqvist & Wiksells.
Indiantelevision.com. (2006). India, China to proSHO$VLD3DFL¿FEURDGEDQGUHYHQXHWRELOOLRQ by 2011 (Study). Retrieved August 1, 2007, from http://www.indiantelevision.com/headlines/y2k6/ may/may208.htm
Lindstrom, A. (2006). Citizens of China and India are hoping to reap the economic rewards of increased broadband wireless access in cities and remote areas. Retrieved August 1, 2007, from http://www.connectwithcanopy.com/index. cfm?canopy=spotlight.story&aid=393
International Telecommunication Union. (2003, September). ITU Internet reports: Birth of broadband. Retrieved August 1, 2007, from http://www. itu.int/osg/spu/publications/sales/birthof broadband/exec_summary.html Internet use deepens; broadband overtakes dial-up. (2006). The China Business Review, 33(1), 37. Kelly, Tim. (2000). If the telecommunications industry is so successful, why is it unable to reduce the waiting list for telephone service? Geneva: International Telecommunications Union. Koprowski, G.J. (2006, January 13). Report: China Internet use catching up with US. TechNewsWorld.
850
Lynch, G. (2001, February 1). The coming bandwidth bubble burst. America’s Network. Mallaby, S. (2005, July 11). In Beijing, a growing problem [Final edition]. The Washington Post, A15. McDaniels, I.K. & Waterman, J. (2000). WTO: A done deal? The China Business Review, 27(6), 22-27. McDowell, S.D. (1995). The decline of the license Raj: Indian software export policies. Journal of &RPPXQLFDWLRQ(2), 25.
The Evolution of Broadband in the Developing World
McGill, D.C. (2001). Shanghai WebTV is banking on broadband. Retrieved August 1, 2007, from KWWS ZZZYLUWXDOFKLQDFRPDUFKLYH¿QDQFHVWLUfry/032300-stirfry-dcm.html
Sklair, L. (1994). The culture-ideology of consumHULVPLQXUEDQ&KLQD 6RPH¿QGLQJVIURPDVXUYH\ in Shanghai. Research in Consumer Behavior, 7, 259-292.
Mehra, P. & Padmanabhan, A. (2006, March 6). Business as the bait; business between the two countries has surged, but things can be much better with more active support from governments on either side. India Today, 44.
Special report: The party, the people and the power of cyber-talk - China and the Internet (2006, April 29). The Economist, 379(8475), 28.
Mishra, P. (2004, June 28). India to drive broadband usage a la Korea. Asia Computer Weekly, 1. NASSCOM. (2002). Export destination. National Association for Software and Service Companies. Retrieved August 1, 2007, from http://www.nasscom.org/it_industry/export_destinations.asp Need to revolutionize IT sector. (2005, December 11). The Nation. Newman, R.J., Hook, C.S., & Moothart, A. (2006, March 27). Can America keep up? Why so many smart folks fear that the United States is falling behind in the race for global economic leadership. U.S. News & World Report, 140(11), 48-56. Pyramid Research. (2001). Broadband in Asia: The next big thing. Retrieved August 1, 2007, from http://www.pyramidresearch.com/research/ APBrdbnd01.asp Porter, M. (1999). The competitive advantage of nations. London: MacMillan. Segal, A. (2004, Nov/Dec). Is America losing its edge? Foreign Affairs, 83(6), 2. Singh, S. (2004, October 18). Doctor visits. Newsweek, 144(16), E20. SinoCast. (2004, March 9). China becomes top DSL country. SinoCast China Business Daily News, 1. SinoCast China IT watch. (2006, January 19). Latest report on China’s Internet. China: SinoCast LLC. Skills gap survey: Begin now to attract, train and develop manufacturing workers. (2006). Plant Engineering, 60(1), 23-24.
Sperling, G. (2004, March 1). A new consensus on free trade. The Washington Post, p. A19. 7DQ=0HXOOHU0 )RVWHU: &KLQD¶V new regulations. Communications of the ACM, 40(12), 11-16. Tandon, A. (1991). Export or prish? Dataquest, 9, 63-65. Tilton, J.E. (1971). International diffusion of technology: The case of semiconductors. Washington, DC: The Brookings Institution. UNCTAD. (2000). World investment report 2000: Cross-border mergers and acquisitions and development. Geneva. UNDP. (2001). Human development report 2001. New York: United Nations Development Program. Retrieved August 1, 2007, from http://www.undp. org/hdr2001/completenew.pdf Vance, A. (2005, April 26). India, China poised to feast on US IT complacency. Retrieved August 1, 2007, from http://www.theregister. co.uk/2005/04/26/sun_schwartz_china/ Vernon, R. (1966). International investment and international trade in the product cycle. Quarterly Journal of Economics, 80, 190-207. Waldman, A. (2004, July 24). Indians go home, EXWGRQ¶WOHDYH86EHKLQGTimes, p. A1. West, D.M. (2002). Global e-government, 2002. Center for Public Policy, Brown University. Retrieved August 1, 2007, from http://www.insidepolitics.org/egovt02int.PDF Wilson, C. (2005, April 25). Municipal networks gain ground. Telephony, 246(8), 6-7.
851
The Evolution of Broadband in the Developing World
World IT Report. (2004, June 24). WebSky to set up JV for operation of wireless broadband system in India. World IT Report, 1. ]GQHWLQGLDFRP 0D\ ,QGLD¶VEURDGEDQG penetration extremely low. Retrieved August 1, 2007, from http://www.zdnetindia.com/news/communication/stories/147843.html
KEY TERMS Digital Subscriber Line (DSL): A family of technologies providing digital data transmission over the wires of a local telephone network. Gross Domestic Product (GDP): The sum of the total value of consumption expenditure, total value of investment expenditure, and government purchases of goods and services. Information and Communications Technologies (ICTs): Technologies that facilitate the capturing, processing, storage, and transfer of information. Time Division - Synchronous Code Division Multiple Access (TD-SCDMA): A 3G mobile telecommunications standard being pursued in China. Internet: The “global information system that (i) is logically linked together by a globally unique address space based on the Internet protocol (IP) or its subsequent extensions/follow-ons; (ii) is able to support communications using the transmission control protocol/Internet protocol (TCP/IP) suite or its subsequent extensions/follow-ons, and/or other
852
IP-compatible protocols; and (iii) provides, uses, or makes accessible, either publicly or privately, high level services layered on the communications and related infrastructure described herein” (The )HGHUDO1HWZRUNLQJ&RXQFLOGH¿QLWLRQ Offshore Outsourcing: Subcontracting a FRPSDQ\¶VQRQFRUHEXVLQHVVSURFHVVHVWRDQH[ternal service provider in a foreign country. Internet Protocol Television (IPTV): A system in which a digital TV service is delivered using the Internet protocol. E-Commerce: Any transaction in which at least one of the following activities– production, distribution, marketing, sale, or delivery–takes place by electronic means.
ENDNOTES 1
2
3
4
5
6
Another estimate suggested that about 60% of Internet users in 2005 went online via broadband (Internet use deepens, 2006). See “China Becomes Top DSL Country,” SinoCast China Business Daily News, March 9, 2004, p. 1. See http://hdr.undp.org/reports/global/2005/ pdf/HDR05_HDI.pdf The same estimate suggests that by 2007-08, WKLVLQGXVWU\ZLOODFFRXQWIRURI,QGLD¶V GDP employing 1.4-1.5 million people. See “WebSky to set up JV for operation of wireless broadband Internet system in India,” World IT Report, June 24, 2004, p. 1. See http://www.commonwealthknowledge. net/documents/Kummar/_Toc500136352
