Innovation in Low-Tech Firms and Industries (Industrial Dynamics, Entrepreneuship and Innovation Series)

Innovation in Low-Tech Firms and Industries

INDUSTRIAL DYNAMICS, ENTREPRENEURSHIP AND INNOVATION Series Editors: David B. Audretsch, Max Planck Institute of Economics, Jena, Germany and Ameritech Chair of Economic Development, Indiana University, Bloomington, USA, Dirk Fornahl, Institute for Economic Policy Research – Section System Dynamics and Innovation, Universität Karlsruhe (TH), Germany and Hariolf Grupp, Deputy Director, Fraunhofer Institute for Systems and Innovation Research (FhG-ISI), Karlsruhe, Germany and Professor, Faculty of Economics and Management, Universität Karlsruhe (TH), Germany This series aims to discover important new insights in the relationship between the three cornerstones of economic development: industrial dynamics, entrepreneurship and innovation. In particular, the series will focus on the critical linkages between these three foundations. For example, the entry and exit of firms with differentiated growth processes can influence industrial development, but at the same time can also reflect the current industrial context shaping the entrepreneurial activities of single firms or individuals. A similar interaction linking industrial dynamics to entrepreneurship and innovation can also be identified. For instance, the particular technological regimes of industries may influence innovative activities, but the technological trajectory and type of innovative activity can, in turn, have a positive or negative influence on industry development. Innovation and entrepreneurship are also closely linked, since many types of entrepreneurial activity are barely distinguishable from similar innovative endeavors. Hence, the series addresses the linkages among the three fields in order to gain new findings concerning the nature of economic change. Theoretical, empirical as well as policy-oriented contributions are welcome.

Innovation in Low-Tech Firms and Industries Edited by

Hartmut Hirsch-Kreinsen Professor of Economic and Industrial Sociology, Dortmund University of Technology, Germany and

David Jacobson Professor of Economics, Dublin City University, Ireland

INDUSTRIAL DYNAMICS, ENTREPRENEURSHIP AND INNOVATION

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Hartmut Hirsch-Kreinsen and David Jacobson 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Control Number: 2008932880

ISBN 978 1 84720 823 1 Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents List of figures List of tables List of contributors

vii viii x

INTRODUCTION 1.

The low-tech issue Hartmut Hirsch-Kreinsen, Katrin Hahn and David Jacobson

3

PART I: INNOVATION IN LMT: CONDITIONS AND REQUIREMENTS 2.

3.

4. 5.

How to grasp innovativeness of organizations: outline of a conceptual tool Gerd Bender Standard-setting competition and open innovation in non-HT industries: mechanical engineering and machinery Alexander Gerybadze and André Slowak The moral economy of technology indicators Benoît Godin Critical comments on the ‘moral economy of technology indicators’ Hariolf Grupp

25

43 64

85

PART II: TECHNOLOGICAL DIFFUSION AND INTERRELATIONSHIPS BETWEEN SECTORS 6.

7.

Distributed knowledge bases in low- and medium-technology industries Paul L. Robertson and Keith Smith LMT innovations in a high-tech environment: human-factor ‘tools’ for the airline industry David Jacobson and Bernard Musyck

v

93

118

vi

8.

9. 10.

Contents

Technology fusion and organizational structures in low- and medium-tech companies Daniela Freddi Industrial innovations in relation to service sectors Marja Toivonen The relevance of services for high-, medium- and low-tech firms – an empirical analysis in German industry Eva Kirner, Gunter Lay and Steffen Kinkel

140 160

175

PART III: LOCAL VERSUS GLOBAL PERSPECTIVES IN INNOVATION 11.

Innovation activities versus competitiveness in low- and medium-technology-based economies: the case of Poland Anna Wzia˛tek-Kubiak 12. Low-tech industries between traded and untraded interdependencies: a dynamic concept of industrial complementarities Martin Heidenreich 13. High-tech innovation in catching-up countries: conditions and perspectives Staffan Laestadius, Linda Gustavsson and Vicky Long 14. Worshipping at the shrine of the knowledge-based society? James Wickham Index

197

221

245 267

285

Figures 3.1 3.2 3.3 3.4 5.1 6.1 6.2 9.1 9.2 10.1 10.2 10.3 10.4 10.5

10.6 12.1

13.1 13.2

Comparison of R&D investment patterns in EU-based corporations and non-EU firms The transformation of a medium-tech industry: the case of automobiles Layered model of innovation and standard-setting in collaborative partnership in the PROFIBUS/PROFINET case Organizational form and the transnational network for a standard-setting partnership Survey/overview of important innovation stages and their typology Innovation networks Layers of innovation networks The service model of Edvardsson et al., redrawn A model of service strategies in manufacturing Low-, medium- and high-tech firms by sector and firm size Capital and labour productivity of low-, medium- and high-tech firms by firm size Type of product-related services offered by low-, mediumand high-tech firms Share of product and service innovations among low-, medium- and high-tech firms Turnover with sales of product-related services (introduced within the last three years versus offered for longer than three years) by firm size and R&D intensity Share of turnover with product-related services (directly and indirectly accounted) by firm size and R&D intensity Dynamic complementarities between low- and high-tech industries (as a function of traded and untraded interdependencies) Total R&D investment of the top 100 Chinese ICT firms Total sales revenue of the top 100 Chinese ICT firms over 19 years

vii

45 48 55 57 86 105 106 163 167 179 181 183 185

187 189

226 251 252

Tables 3.1 4.1 4.2 6.1 6.2 11.1

11.2

11.3

11.4 11.5 11.6 11.7 12.1 12.2

12.3 12.4

Comparison of open innovation, closed innovation and semi-open innovation processes OECD technology intensity levels (1986) Terms used for modern technology Changing technological competencies of 500 large firms: 1981 to 2000 (percentage shares). Changing technological competencies of large firms in food: 1981 to 2000 Changes in the structure of Polish manufacturing production and exports by technology level (based on OECD classification 1997) from 1998–2004 (in percentage) Changes in the relative unit labour costs (RULC) and relative unit export values (RUEV) of Polish manufacturing from 1996–2003 Average changes in wages, productivity, turnover and employment of Polish and EU manufacturing in 1998–2003 (at current prices) Dynamics of innovation expenditure in Polish manufacturing and their structure in 2000–4 (in percentage) Structure of expenditure on innovation by firm size in 2002–4 (in percentage) Structure of expenditure on innovation in firms by ownership type in 2002–4 (in percentage) Structure of expenditure on innovation in terms of technology level in 2002–4 (in percentage) Type of innovation activities in the EU (2002–4) Highly important sources of information for innovation, as a percentage of innovative enterprises, 20 EU member states Innovation activity and cooperation during 2002–4 (in percentage of all innovation enterprises) Territorial dimension of innovation cooperation (2002–4; in percentage of all innovation enterprises)

viii

56 70 72 107 109

200

201

202 205 209 211 214 227

228 230 232

Tables

12.5 12.6

Employment in high- and low-technology sectors (2006; in percentage of total employment) Innovation activity and cooperation in Poland and the EU (2002–4; in percentage of all innovative enterprises)

ix

234 240

Contributors Gerd Bender is Professor of Sociology at the HdBA – School of Labour Market Studies, Mannheim, Germany. He worked as a teacher and a researcher in various universities such as Goethe University, Frankfurt, Vienna University of Technology and Dortmund University, and in both public and private institutes. Innovation in traditional sectors in particular is one of his main research interests. His recent publications include: ‘Peculiarities and relevance of non-research-intensive industries in the knowledge-based economy’, final project report, Dortmund 2006 (www.pilot-project.org/publications/publications.html); Technologieentwicklung als Institutionalisierungsprozess, Berlin, 2006; ‘Non-researchintensive industries in the knowledge economy’, Perspectives on Economic Political and Social Integration, (editor with David Jacobson and Paul L. Robertson) XI, no. 1–2, special edition, Lublin, 2005. Daniela Freddi is a PhD student in industrial economics. Her main research interests are industrial economics, learning and innovation in low- and medium-tech sectors, and regional development policies. Her recent publications include: ‘From industrial districts to company network’, (with A. Bardi), in H. Hirsch-Kreinsen, D. Jacobson, S. Laestadius (eds), Low-Tech Innovation in the Knowledge Economy, 2005; ‘The integration of old and new technological paradigms in LMT sectors: the case of mechatronics’, Research Policy, special issue on technological change in low- and mediumtechnology industries, forthcoming. Alexander Gerybadze is Professor of International Management, Hohenheim University, Stuttgart, Germany. His studies have been in economics, mathematics and business administration in Heidelberg (1973–8) and Stanford (1979–80). He earned his PhD on Evolutionary Models of Technical Change at Heidelberg University in 1980. Other assignments have been 1981–3 at the VDI Technology Center Berlin, 1983–90 at the Arthur D. Little International in Wiesbaden and Habilitation on Managing Networks and Strategic Alliances in 1991. He was Professor of Technology Management at St Gallen Business School, Switzerland 1991–5, and since 1996 has been Director, Center for International Management and Innovation, Hohenheim University. He is a member of the Executive Board, Center for Innovation and Services and an Honorary Research x

Contributors

xi

Fellow at the University of Manchester. His recent work includes research on innovation and knowledge management in multinational firms, R&D internationalization and offshoring, knowledge transfer in distributed teams, management and organization of innovation clusters, and on standard-setting consortia. Benoît Godin is professor at Institut National de la Recherche Scientifique in Montreal, QC, Canada. He holds a DPhil in science policy from Sussex University, UK. He has written extensively on science policy and statistics. He is currently involved in a project on the history of science and technology statistics from which two books have recently been published: Measurement and Statistics on S&T: 1920 to the Present, London: Routledge, 2005, and La science sous observation: cent ans de measures sur les scientifiques, 1906–2006, Québec: Presses de l’Université Laval, 2005. He has recently started a largescale project on the intellectual history of innovation as a category, from the Middle-Ages to the present day. Hariolf Grupp studied physics and mathematics at Heidelberg University and received his doctorate in 1978. Post-doctoral studies were in Jerusalem, Grenoble, Tbilisi, and Cambridge, MA. Since 1985 he has been at the Fraunhofer ISI and was Deputy Director of the institute from 1996 to 2005, and Director from 2005 to 31 March 2007. He is a recipient of the Fraunhofer Prize. On 1 January 2001 he was appointed Professor of the Chair of System Dynamics and Innovation at the Institute for Economic Policy Research (IWW) at Karlsruhe University. His present research topics include: the economics of technical change, science and technology indicators, industrial R&D management and research policy, science and innovation studies, technology assessment and foresight. Grupp is editor of the series of books Industrial Dynamics, Entrepreneurship and Innovation, and has written many journal articles and made book contributions. Linda Gustavsson is a PhD candidate at the Royal Institute of Technology (KTH), Stockholm. Her research interest is in knowledge formation processes in globalized industrial systems. She has previously participated in the EU-project PILOT (Policy and Innovation in Low-Tech) and in a project studying Swedish regional innovation policy. Katrin Hahn, a PhD student, works at the Chair of Economic and Industrial Sociology at Dortmund University of Technology. Her main research interests are innovation and diffusion in industrial sectors and European innovation policy. Her recent publications include ‘Der Lissabon-Prozess: Das Innovationskonzept und die Auswirkungen auf die Politikgestaltung’, sociological working paper no. 20/2008, edited by Hartmut Hirsch-Kreinsen and Johannes Weyer, TU Dortmund.

xii

Contributors

Martin Heidenreich is involved in the studies of sociology and business administration in Bielefeld, Bologna and Paris, as Professor of Sociology with special attention to Social Stratification. His research interests are regional and national patterns of work, management and innovation and the Europeanization of national societies. He has published nearly 100 books and articles on regional innovation systems and regional experimentalism and the organization of work, technologies, management and innovation in knowledge societies. Among them are: Regional Innovation Systems (co-editor, with Hans-Joachim Braczyk and Philip Cooke), London, 2002; ‘Regional inequalities in the enlarged EU’, Journal of European Social Policy, 2003; ‘The renewal of regional capabilities’, Research Policy, 2005; and ‘Innovation in European low- and mediumtechnology industries’, Research Policy, 2008. Hartmut Hirsch-Kreinsen is Professor of Economic and Industrial Sociology at the Dortmund University of Technology, Germany. He has done research at the Institute for Social Research in Munich (ISF), at the University of Wisconsin in Madison, WI, and at the Arbetsmiljöinstitutet in Stockholm. His working fields include the internationalization of companies and company networks, the change of sectoral structures and development of work, and innovation studies and development processes of new technologies. His recent publications include: Wirtschafts- und Industriesoziologie, Weinheim 2005; Low-tech Innovation in the Knowledge Economy (editor with David Jacobson and Staffan Laestadius) Frankfurt 2005; and ‘ “Low-tech” innovations’, Industry & Innovation, 15, no. 1 February, 2008, 19–43. David Jacobson is Professor of Economics at Dublin City University in Ireland. He completed his first degree at Hebrew University of Jerusalem, his Masters at Sussex University and his PhD at Trinity College, Dublin. In addition to spending most of his working life in Ireland, he has taught in the USA, France, Israel and Cyprus. His research interests include subsectoral industry studies, systems of innovation, economic geography, MNEs and European integration. His recent articles have been published in such journals as Economic Geography, Journal of Economic and Social Geography, Prometheus and European Planning Studies. His recent books include Industrial Economics and Organization: A European Perspective (with Bernadette Andreosso). With Paul Robertson and Richard Langlois he is a contributor to the Handbook of Industrial Districts. Steffen Kinkel, PhD, is head of the competence center Industrial and Service Innovations at Fraunhofer Institute for Systems and Innovation Research (ISI), Karlsruhe, Germany. The main focus of his research is international

Contributors

xiii

production and strategic controlling. Several of his numerous publications examine production relocations, R&D offshoring and backsourcing activities as well as instruments for strategic location assessment and holistic innovation management of companies. They include: Patterns of Organisational Change in European Industry (PORCH). Ways to Strengthen the Empirical Basis of Research and Policy, (with H. Armbruster et al.) Innovation Papers no. 46, Brussels, Luxembourg: European Commission, 2007, and ‘Development, motives and employment effects of manufacturing offshoring of German SMEs’, (with G. Lay and S. Maloca) International Journal of Entrepreneurship and Small Business, 2007, 4, 256–276. Eva Kirner, PhD, is a researcher at Fraunhofer Institute for Systems and Innovation Research in Karlsruhe, Germany. She obtained her degree in business administration at the University of Mannheim with her main focus on Management Studies and Organization. She received her PhD from the University of Dortmund. Her research mainly focuses on analysis of innovative organization and work concepts, especially future challenges in this field, competence management and innovation measurement. Her recent publications include: Patterns of Organisational Change in European Industry (PORCH). Ways to Strengthen the Empirical Basis of Research and Policy, (with H. Armbruster et al.) Innovation Papers no. 46, Brussels, Luxembourg: European Commission, 2007. Staffan Laestadius, PhD, is Professor of Industrial Dynamics at the Royal Institute of Technology, Stockholm, Sweden. His research interests include knowledge formation processes, in particular non-R&D-based innovations, globalization and industrial and technical implications of climate change. Gunter Lay, PhD, studied business administration at the University of Mannheim. In 1978 he started work at the Fraunhofer ISI, where he set up the Innovations in Production group. The results of his research activities have been published in a great number of professional journals and compilations. He has been invited to sit on a number of panels of experts and advisory boards on account of his experience. Parallel to his work at Fraunhofer ISI, he was offered lecturing assignments by the Universities of Strasbourg, Hanover and Kassel as well as the Technical College Schmalkalden, Germany. In 1996 he was awarded a visiting professorship at the University of Grenoble, France. The main focus of his present work in Fraunhofer ISI is the coordination of the institute’s survey on innovations in production. Vicky Long is a PhD candidate at the Royal Institute of Technology Stockholm. Her current research projects include ‘Catching up and from

xiv

Contributors

above – the case of the Chinese ICT international expansion’, and ‘The mechanisms of wireless industrial and technological transformation’. Bernard Musyck is Associate Professor at the School of Economic Sciences and Administration, Frederick University, Nicosia, Cyprus. His main research interests include innovation policies, localized learning and endogenous industrialization with special emphasis on small- and medium-sized enterprises. Since June 2005 he has been working on the economics of innovation in the aviation industry in Europe within the context of a European Commission-funded Integrated Project aimed at transforming the aviation industry and improving flight safety through the integration of human factors knowledge into all aviation-related activities. Paul L. Robertson holds an adjunct professorship in the Faculty of Business of the University of Tasmania, Australia. He is the author, co-author or editor of more than 70 books and articles including Firms, Markets and Economic Change: A Dynamic Theory of Business Institutions (written with Richard N. Langlois). He is also the co-author of articles on the role of lowand medium-technology sectors in modern economies which have been published in Research Policy, Industry and Innovation and Prometheus. André Slowak is Research Associate at the Hohenheim University’s Department for International Management in Stuttgart, Germany. He is also a PhD candidate at Hohenheim’s Center for International Management and Innovation. He holds a Diplom-Kaufmann as well as a Diplom-Ökonom from the Technical University of Dortmund, Faculty of Economics and Social Sciences. Meanwhile, in his studies, he has been Research Assistant and Student Management Consultant in several projects at the Fraunhofer Institute for Material Flow and Logistics, Dortmund. Keith Smith, Professor, is Executive Director of the Australian Innovation Research Center (AIRC), University of Tasmania, Australia. After training as an economist at Cambridge and Sussex universities, he worked extensively on issues related to the economics of innovation. His recent publications include: ‘Globalization and innovation policies: the continuing role of national innovation systems’, in S. Kuhlmann and R. Smits (eds), Innovation systems and Innovation Policy, Oxford, forthcoming; ‘Does the EU do too little corporate R&D? Comparing the EU and non-EU worlds’ (with Pietro Moncada, Constantin Cuipagea, Mike Tubbs, and Alexander Tübke), Research Policy, forthcoming; and The Knowledge Economy in the Australian Context, a report to Australian Business Foundation, May 2006. Marja Toivonen, PhD, is Research Director at the Innovation Management Institute at Helsinki University of Technology, Finland. Toivonen leads a

Contributors

xv

research group focusing on services innovation. She has written several articles on this topic and has been an invited speaker in conferences organized in this emerging field. Earlier, Toivonen worked as foresight project manager and head of the research unit at the Employment and Economic Development Centre for the Helsinki region. In this work, she participated in several international groups developing foresight, for example the High Level Expert Group of the EU project ‘Blueprints for foresight actions in the regions’. Her recent publications include ‘The role of KIBS in the IC development of regional clusters’, Journal of Intellectual Capital, (with A. Smedlund), 2007, 8, 159–70; ‘Innovation process interlinked with the process of service delivery – a management challenge in KIBS’, (with T. Tuominen and S. Brax) Economies et Sociétés, 3, 2007, 355–84. James Wickham, PhD, is Director of the Employment Research Centre, and Jean Monnet Professor of European Labour Market Studies in the Department of Sociology at Trinity College, Dublin, Ireland. He has researched and published on Irish industrialization and labour market issues, especially in the electronics industry. Main interests include hightechnology industry and high-skill labour markets, equal opportunities, employment and transport, and European employment policy. His publications include: ‘The development of policy relevance in European social research’ (with L. Greco, P. Landri and M. Tomassini) in M. Kuhn and O. Remoe (eds), Building the European Research Area – Socio-economic Research in Practice, New York and Berlin, 2005, pp. 177–237; ‘Training cubs for the Celtic Tiger: the volume production of technical graduates in the Irish educational system’ (with G. Boucher); Journal of Education and Work, 2004, 17, no. 4, 377–95; ‘Something new in old Europe? Innovations in EU-funded social research’, Innovation, European Journal of Social Science Research, 2004, 17, no. 3, 187–204. Anna Wzia˛tek-Kubiak is Professor of Economics and Head of Department at the Institute of Economics, Polish Academy of Sciences, Lazarsky School of Commerce and Law, Warsaw, Poland, and a consultant at the CASE Foundation in Washington, DC. She has been involved in research on industrial competitiveness, enterprise restructuring, innovation and international trade. In addition to chapters in books and journal articles, her previous publications include: Changes in Competitiveness of Polish Manufacturing Industries and Structural Change and Exchange Rate Dynamics in the Context of the EU Enlargement (edited with Paul J. Welfens), Springer, 2005; Industrial Competitiveness and Restructuring in Enlarged Europe. How Accession Countries Catch Up and Integrate in the European Union (edited with Iraj Hoshi and Paul J. Welfens), Palgrave Macmillan, 2007.

Introduction

1.

The low-tech issue Hartmut Hirsch-Kreinsen, Katrin Hahn and David Jacobson

INTRODUCTION As the European Union evolves into a knowledge society, the ability to generate, use, diffuse and absorb new knowledge is increasingly viewed as critical to economic success and societal development. While the attention of policy makers, scholars and the public at large has been concentrated disproportionately on the 3 to 10 per cent of modern economies conventionally classified as ‘high-tech’, the importance of innovation activities in the established sectors that comprise the bulk of economic activity has tended to be overlooked. This is not to say that these sectors do not attract any attention or that they are regarded as totally unimportant – recent policy steps taken in the EU and the US to protect their textile industries prove that low- and medium-technology (LMT) industries have considerable political presence. But these non-research-intensive industries are not well understood in terms of their specific innovative capabilities, their role within the economy, current technologies or their probable future development. Moreover this lack of understanding has broad implications because it reflects inaccurate – or at least inadequately nuanced – views about the role of research and development (R&D) and technological upgrading in modern economies as a whole. In this scenario LMT industries are deemed to offer severely limited prospects for future growth in comparison to hightech ones, and as a result, receive less explicit policy attention and support. Recent research1 (Bender et al., 2005; Hirsch-Kreinsen et al., 2005, 2006) has convincingly shown that many popular notions of the role of technology in economic growth are misleading. In particular it is counterproductive to categorize activities as high- or low-technology, as is the norm in the OECD, among other authorities (Hatzichronoglou, 1997). Because of the narrowness of the indicators used, such classifications often mislead in terms of how technology is generated, used and diffused in individual firms and in economies as a whole. For example the OECD schema discussed by Hatzichronoglou (1997) puts sectors into narrow categories on the basis of 3

4

Introduction

the proportions of their revenues devoted to research and development, and as a result perpetuates a number of errors concerning the role of technology. Sectors classed as ‘LMT’ are by the schema’s definition relatively stagnant due to their low levels of investment in R&D. However our studies show that they are nevertheless quite dynamic technologically. In fact there is a great deal of evidence that LMT sectors ● ● ● ●

achieve respectable growth in productivity; draw heavily on high-technology sectors (often but not always process technology); generate substantial innovation themselves (though these activities may not be captured in R&D statistics); and are an important element in the innovativeness and effectiveness of regional and transnational industrial value chains.

Because it can be assumed that there is a broad range of issues concerning innovation and the use of technology in LMT sectors that needs to be covered in greater depth than has been done to date2 the present volume is our attempt to help remedy this need.

CHALLENGES AND CONTEXT OF RESEARCH ON LOW-TECH INDUSTRIES The main research starting point in relation to innovation in low-tech industries in the advanced industrialized societies of the European Union is a fundamental criticism of the high-tech focus of both policy makers and academics. This criticism necessitates first of all a re-examination of the relevance of LMT sectors. To a large extent the focus on high-tech reflects the idea that much change in modern societies can be characterized as typical of the emerging ‘Knowledge Society’ (Drucker, 1994; Stehr, 1994; Willke, 1998; David and Foray, 2003) or ‘Learning Economy’ (Lundvall and Borrás, 1997). These writers and others share the idea that modern organizations and societies are undergoing a fundamental process of change based on the enhanced significance of knowledge as a productive force and asset. In this view, continual innovation accompanied by a restructuring of work processes and organization is a decisive determinant of economic and social development, while the generation, diffusion and utilization of knowledge are core characteristics of firms and of economic activity as a whole. To be sure, these discourses on the emerging knowledge society do describe important tendencies in economic and social development. We share the view that knowledge is an increasingly important resource, but we

The low-tech issue

5

dispute much of the conventional wisdom about how the knowledge economy is structured and the implications for economic trends and hence policy measures. However, although referring to the economy as a whole, the knowledge economy is actually identified with a very small number of research-based or science-based activities, especially information and communication technologies (ICT), biotechnology and nanotechnology. A key issue is the policy consequences of the argument that, as a consequence of increased knowledge intensity, the economies of industrialized countries in Europe and elsewhere are currently going through at least two great changes (Carson, 1998): ●

●

A significant part of industrial production is relocating from its traditional sites to developing countries. The classic example is the exodus of textile firms from the rich world over the past three decades. This applies particularly to labour-intensive ‘mature’ industries: quite soon, it is claimed, many big Western firms in such industries will have more employees and even customers in developing countries than in developed ones. In many industrialized countries the balance of economic activity is swinging from manufacturing to services. Even in Germany and Japan, which rebuilt so many factories after 1945, manufacturing’s general share of jobs in relation to the whole economy is declining rapidly in favour of high-tech manufacturing and services.

Particularly in Western countries, authors focusing on these trends have continued the debate originating in the 1970s (Bell, 1973/1999; Fröbel et al., 1977) over an ongoing process of ‘de-industrialization’. By the end of the 1980s many American and European experts had come to believe that their countries’ industries were being ‘hollowed out’ as many basic production activities relocated to other areas (Uchitelle, 2006). The policy consequence drawn from this development is the well-known objective of making the EU the world’s most competitive knowledge-based economy. How this objective can be reached has been widely debated, with policy-makers focusing – in the Lisbon Strategy for example – on an important target indicator selected to reflect this goal, namely that the EU should achieve an R&D to GDP ratio of 3 per cent. This political and economic objective has strongly promoted high-R&D industries. These arguments are related to the aforementioned indicator measuring the ratio of R&D expenditure to turnover for a company or business sector. The OECD classifies industrial sectors as follows: High-technology sectors (‘high tech’) are those with an R&D intensity above 5 per cent, as well as those with complex technology (‘medium-high tech’) with an R&D

6

Introduction

intensity of 3 to 5 per cent. Industries which are not research-intensive (‘medium-low tech’ and ‘low tech’) have an R&D intensity below 3 per cent. Pharmaceutical, electronics, motor vehicle, aerospace and mechanical engineering industries for instance are categorized as high or medium-high tech. By contrast the LMT category includes ‘more mature’ industries such as the household appliance manufacture, food processing, paper, printing and publishing, wood and furniture, metal (for example foundry) and plastic products industries. In this debate it is frequently ignored that in all industrialized countries a large proportion of manufacturing is in LMT industries, and that these industries (whatever their vintage) provide goods and services absolutely vital to the function of modern societies. In spite of growing global competition, particularly in traditional and mature industries, this continues to hold true for the industrialized countries of Western Europe and the transition economies of Middle and Eastern Europe. Further evidence for the importance of the LMT sector is provided by a number of empirical findings which emphasize the innovative ability of the low-tech sector particularly in high-tech countries.3 Thus The Economist (1998) has referred to ‘the strange life’ of low-tech industries in high-tech California. In Europe too: ●

●

On the one hand there is the clear trend towards manufacturing’s rapidly decreasing share of total employment and the service sector’s equally if not more rapidly increasing share. On the other hand a close examination of industry reveals that LMT sectors are surprisingly significant.

All data show4 that LMT industries play a very important role in employment in all industrialized countries, accounting for nearly 60 per cent of employment in manufacturing. There has been a tendency for the low-tech industries’ proportion of manufacturing to decline since the 1980s, while that of high-tech industries has been stable. A similar trend can be observed regarding the share of value added of the different sectors in manufacturing. In the long run, starting from a low level, high-tech sectors show a rising share of the value added in manufacturing while the share of the LMT sectors is declining. However, these declines are not marked, and LMT industries still add by far the largest part of the value added in manufacturing in OECD economies. It is debatable whether there is a real structural change – from LMT to high tech – in the period examined here. In fact the low-tech sectors continue to evince remarkable stability and a large share of employment. In addition, there is no clear connection between high-tech intensity and national growth rates. The question is whether

The low-tech issue

7

countries with more high-tech sectors have better overall growth records. On the basis of the statistical data no positive correlation can be found between the high-tech share of manufacturing value added and the rate of growth of GDP per inhabitant (Kaloudis et al., 2005). Furthermore it is wrong to conclude that only high-tech countries are also high-growth countries. Rather, it is clear that even in the EU many predominantly low-tech countries have impressive growth rates. The findings lead again to the question: What are the reasons for this remarkable stability of LMT industries and firms? The answer requires a discussion of the mode of innovation in non-science-based industries.

LMT MODE OF INNOVATION Company and Intercompany Relations The starting point of an answer to this question is an elaboration of the specific knowledge base of LMT firms. Because these enterprises pursue virtually no R&D activities, it stands to reason that formalised processes of knowledge generation and use play an insignificant role and that instead innovation activities proceed in the form of ‘practical and pragmatic ways by doing and using’ (Von Tunzelmann and Acha, 2005: 417). Knowledge that is relevant to these enterprises should therefore in general be regarded as application-oriented, practical knowledge. Unlike scientifically and theoretically generated knowledge that rests on criteria such as theoretical relevance and universality, practical knowledge is generated in application contexts of new technologies and obeys validity criteria such as practicability, functionality, efficiency and failure-free use of a specific technology. The two types of knowledge can be difficult to distinguish however. Simplifying matters, theoretical and scientific knowledge in enterprises, for instance in the form of systematically acquired engineering knowledge, can primarily be associated with research, development and generation processes, while practical knowledge arises in the context of ongoing operating processes. The term ‘practical knowledge’ stands for a complex bundle of different knowledge elements that comprises explicit, codified and formalized elements such as design drawing and requirement specifications for new products, as well as above all implicit elements such as accumulated experience and well-established, proven routines for solving technical problems. The latter are closely connected with everyday experience and processes of ‘learning by doing’ and ‘learning by using’ which constitute a typical individual but also collective form of acquisition of practical knowledge.5 The acquisition and generation of innovative knowledge by no means

8

Introduction

takes place only within companies. The research on LMT mentioned above shows that external knowledge sources are also relevant. For many LMT companies the knowledge held by other firms and organizations as well as the systematic use of that knowledge in pursuit of innovation plays a decisive role. This is true for both practical knowledge and especially also for scientifically generated knowledge in various forms. Examples of external sources include the experience of long-time customers concerning new market and demand trends, the expertise of consultants and other suppliers, and information about foreseeable market trends gained during visits to exhibitions and fairs. Suppliers may provide both external practical and theoretical or scientific knowledge. In relation to practical knowledge, the fashion-oriented design of products such as chairs by external design agencies plays a far from marginal role in successful sales strategies. In relation to theoretical/ scientific knowledge, machine manufacturers and suppliers provide that knowledge embodied in production technologies and components. On the whole the knowledge base of non-research-intensive enterprises can be characterized as a ‘distributed knowledge base’ (Smith, 2003) comprising different forms of knowledge from sources, including LMT firms, independent of each other and often in different sectors and technology fields. Management and Organization of Knowledge The specific knowledge base can be regarded as the central condition for the innovation strategies of LMT enterprises. It largely determines the direction and the scope of the innovation strategies, insofar as it defines the framework for action of the companies. The knowledge base sets the technological trajectory that can only be deviated from with difficulty. Accumulated knowledge bases naturally cause a high degree of inertia (path dependency) in the sense of the types of activities and the direction of innovation. Even if the sources of change are external to the existing knowledge base, the extent to which it can be adopted or absorbed is determined by the knowledge base. This is because the knowledge base largely determines the orientations and expectations prevailing in companies concerning possible and viable innovation perspectives, thus influencing the company’s ability to assess, adopt and integrate new, external knowledge into its knowledge base.6 As the research findings show, of decisive importance for the innovation strategies of companies is how they effectively make use of their internally available as well as externally accessible knowledge. The organizational structures and the internal processes within companies can be regarded as among the determining factors in this respect. This connection can be

The low-tech issue

9

specified by reverting to the concept of ‘dynamic capability’.7 This is the ability of companies to cultivate and develop their knowledge base strategically, mobilize it and, in doing so, to combine the individual knowledge elements in specific ways in order ultimately to generate technological innovations. In the case of the non-research-intensive sector, this can be qualified as follows: ●

●

●

First, it involves the ability to use and advance the knowledge in principle already available in the context of product and process innovations; that is to say, to continually transform this knowledge. Second, it concerns the ability continually to recombine available knowledge and technology elements to enhance products and process structures.8 Third, this denotes the ability to integrate new knowledge. This relates to the fact that a number of the companies examined more or less continually take up new, generally externally generated knowledge. This can vary from practical experience of the sales personnel in changing marketing conditions, to research results from engineering science concerning new machining procedures or potential product materials. These companies can integrate this new knowledge into their existing knowledge base to develop new products and processes.

This ability to utilize knowledge is to a large extent dependent on the routines and structures of the company organization, for instance on the mode of the division of labour, the prevailing forms of communication and co-operation, and the related qualification and personnel structures (see Henderson and Clark, 1990: 15; Cohen and Levinthal, 1990: 131). In many LMT companies these practices are embedded in a company and work organization form that is centralized and based on a marked division of labour (see Schmierl and Köhler, 2005). One can observe a concentration of knowledge in the hands of a small group of managers and technical experts while the more or less qualified production workforce is only responsible for carrying out tasks. In many cases one can speak of the dominance of Tayloristic forms of work organization which are however often accompanied by flexibilized staff deployment as regards working hours and the workplace. Only in some of the enterprises, especially those using highly automated process technologies, can one observe qualification- and holistically-oriented forms of work organization which allow a regular workforce of skilled technical staff much leeway for decisions and manoeuvre. Due to the great importance of external actors and their specialist knowledge, the ability to effectively coordinate network relations across company borders, especially with other companies within the value chain,

10

Introduction

is a central precondition to successful LMT innovation strategies. This ability incorporates also determining the appropriate nature of the relationships with other actors, such as how close or how formal they should be. An essential requirement is a company organizational structure geared to the demands of cross-company cooperation and providing adequate channels of communication, gateways and personnel responsibilities conducive to cooperation. A further important aspect is the professionalism of management. It has to be able to harmonize and regulate the specific competencies and associated interests of many different partners so that the transfer of the required knowledge is assured. As network research findings (Semlinger, 2003) also show, management’s ability to communicate intensively regarding both everyday matters and strategic aspects of cooperation is of great importance in this connection. This lays the foundation for the development of cooperative relations, overcoming restraints and barriers and creating the necessary reliability.

SOCIETAL AND INSTITUTIONAL CONDITIONS The knowledge base and the enterprises’ ability to make strategic use of it are always also embedded in socio-structural conditions. By taking a closer look at these conditions, the specifics of a LMT innovation mode can be stated more precisely. It can be positioned on a broad spectrum of couplings of varying intensity between the innovation strategies and the societal institutions. Loose Coupling of Innovation Policy and Vocational Education Only a small number of LMT enterprises regard the labour market and institutions of vocational education and training as relevant to their innovation ability. These enterprises employ the most modern production technologies and therefore need sufficiently experienced and competent manpower for the continuous operation of the often complex robotized lines. For such enterprises, bottlenecks and constraints because of difficulties in recruiting employees with specific qualifications are significant and as a result, for them the labour market and vocation training systems are important. Above all, the enterprises refer to a shortage of so-called ‘hybrid qualifications’ that encompass both traditional technical and professional competencies as well as skills relating to new technologies and organization forms. Qualifications described with keywords such as ‘communication skills’ and ‘the ability to work in teams’ are frequently in short supply on the labour market. This issue of qualification deficits is raised by

The low-tech issue

11

many enterprises. They emphasize the fact that the contents of vocational training do not always correspond to the actual requirements of the new technologies and new organization structures, and that therefore additional and costly training is often necessary. For most LMT enterprises however the institutions of the labour market and system of vocational education have proved to be of little significance for their innovation ability. For many non-research-intensive enterprises their internal knowledge is concentrated in a few managers and experts while the majority of the employees are more or less semi-skilled workers. In these circumstances there are few recruitment or qualification problems and these companies therefore have low levels of dependence on systems of vocational training and qualification. In some cases enterprises’ representatives express their interest in unskilled but quickly trainable and motivated manpower, in particular, migrants. Migrant workers often fit this profile and the LMT companies’ interests are in reducing restrictions on employment of such migrants. Where a relatively high proportion of employees are unskilled, which is the case in many LMT companies, the regulated structures of a professional or even an internal labour market are irrelevant. There is a similarly loose relationship between LMT companies’ innovation strategies and the policy-regulatory context; there is even a low level of awareness among the LMT companies of how to engage with some of the policies (see Jacobson and Heanue, 2005). Regardless of the type of strategy, LMT companies emphasize negative factors such as high costs, particularly labour costs and taxes, and inflexible and restrictive state bureaucracies. This emphasis is not surprising given the intensive cost and competitive pressures that LMT enterprises face. There is an awareness among them of existing public promotion of technological innovations and state-aided extension or even start-up of factories as innovation-promoting conditions, but they frequently do not see these as helpful or relevant. Indeed these promotion measures often either fail to specify target sectors or aim specifically at R&D and high tech. In general the specific concerns of low-tech enterprises are not included in public innovation support programmes. As a result non-researchintensive enterprises often refer to economic and innovation policy actors as ‘lacking awareness’ of their needs. At best most innovation support programmes and measures promote the innovation ability of LMT companies only indirectly since they generally aim to improve the technological and economic conditions of industrial production. Close Coupling of Economic and Industrial Structures Given the importance of the distributed knowledge base for the innovation ability of LMT enterprises, it is not surprising that their embeddedness in

12

Introduction

the economic and industrial structures around them in many cases proves important to their ability to innovate. This involves networking with ‘neighbouring’ and ‘supporting’ companies and organizations which, as the need arises, provide new technologies and knowledge (Porter, 1998: 166). It can be argued that this is particularly facilitated by the advanced industrialisation of Western Europe for a number of reasons. First, the concentration of suppliers in many Western European countries allows manufacturers – depending on their innovation and production requirements – to change suppliers more easily than in less industrialized countries. The value chain is thus more flexible and can be adapted quickly to new requirements. Second and somewhat paradoxically,9 close relations with the developers and manufacturers of production technologies are crucial for many companies. This is so particularly if technical equipment is custom-designed, or if at least certain components and functions are adapted to particular user needs (or have high asset specificity). Naturally this presupposes relatively close coordination, communication and learning processes between the partners concerned. These findings fit well with the general results of the study based on statistical data by Nascia and Perani (2002) on innovation in Europe. They show that the scientific and technological environment in which an enterprise is located affects the enterprise’s ability to use existing and available knowledge. Third, service providers with specialized knowledge occasionally play an important role in the innovation strategies of LMT companies. Design companies in some cases may assume responsibility for parts of the product design; in other cases firms or institutes may provide special competencies and facilities for quality tests, for answering special technical development questions, or for undertaking market research. Other examples of service provision closely related to innovation strategy include assigning specialized research institutes such product development tasks as materials tests or calculations, and process innovation tasks such as installation design. These furnish the engineering knowledge necessary for low-tech innovations. The LMT companies also occasionally draw on consultants, for example to solve problems of process development and optimization. Altogether the forms of exchange between the different actors in the distributed knowledge base can be very diverse, ranging from relatively anonymous, marketregulated exchange to well established, intensive collaborative relations. We have thus far considered the close links between sectors in terms of the innovative contributions of high-tech to low-tech. But low tech also contributes to high tech. It is often overlooked for example that LMT firms are the main market for high-tech products. Through their purchases they contribute to the profits of research-intensive enterprises and are thus vital to the amortization of the latter’s R&D and continuing investments. A

The low-tech issue

13

determining factor in this respect is the rate of diffusion of new technologies, which is in turn very strongly influenced by the dynamic capabilities of the non-research-intensive companies (Robertson and Patel, 2005). Another way in which LMT companies influence high-tech innovation is through their technical and economic specifications of application requirements for new technologies. LMT companies with a strategy of process specialization play a decisive role here, very often influencing the directions of development of new technologies, especially when a number of them have similar requirements. From the manufacturer’s point of view a broad application field for complex products is opened up. Examples here are the process technologies in woodworking and paper manufacture. Initiated by individual users, these became industry-wide, providing intersectoral marketing and application opportunities.10 In general, LMT enterprises can be regarded both as important ‘recipients’ of new technologies and also as ‘carriers’ of their further development. They constitute a sector that links different industrial sectors and is essential to industrial, regional and national innovation ability as a whole (Robertson and Patel, 2005; Von Tunzelmann and Acha, 2005). Erosion of Regional Ties There is generally a close coupling of LMT and the economic and industrial environment. This does not mean that the regional embeddedness of LMT enterprises is necessarily always relevant to their ability to innovate, and there are indeed examples of innovations driven by international linkages. This suggests the ‘paradox of territories’ in the debate over socioscientific regionalization and globalization (Crouch et al., 2001: 21). On the one hand, many LMT companies are under pressure to spatially extend their cooperation with suppliers and expand their customer base in order to secure and improve their positions. Suppliers to the automotive industry for example often have to relocate production sites when cost conditions and/or the location of their large customers change (Garibaldo, 2006). As a result industrial agglomerations such as the well-known industrial districts of Emilia Romagna (and elsewhere) are under threat. The changing patterns of spatial proximity less and less frequently feature the tightly networked structures and coordination forms of a comparatively closed regional innovation and production system. Market regulated and contractually formalized exchange relations that are increasingly far removed from the region and are characterized by increasing cost competition are gaining in importance (Garibaldo and Jacobson, 2005). On the other hand, ‘going global’ implies a growing importance of social and cultural proximity – and therefore usually also spatial proximity – for the

14

Introduction

general strategic ability of enterprises. Social and cultural proximity provides enterprises with differentiated ownership specific advantages such as in supplier and customer relations, logistics, information and communication. Another important aspect of regional location is the relationship between companies and their local occupational training and continuing education institutions. As the training provided by these institutions is often geared to the needs of the regional LMT companies, relatively trouble-free and rapid recruitment of appropriately qualified staff is guaranteed. Finally, for some enterprises regionally established and relevantly specialized scientific organizations, technology liaison offices, political institutions, industry associations, chambers of commerce and industry, and regionally focused support programmes contribute significantly to their innovation ability. Such actors and activities often provide knowledge that initiates learning processes leading to concrete innovation measures in companies.

DEVELOPMENT PERSPECTIVES OF LMT In spite of the difficult economic situation of LMT industries and the challenges of globalization, prospects for many LMT sectors and companies are reasonably promising even in countries with advanced economies. This is true for a number of reasons: ●

●

First, the specific knowledge base and capability of many low-tech companies are deeply embedded in the company social system and local environment and therefore, being fairly inaccessible to potential competitors, cannot easily be copied or transferred. This – paradoxically – applies to standardized products usually considered easy to imitate. Even though standardized, such products are often designintensive and have major potentials for technological upgrading via complex knowledge inputs. Second, the geographical and social proximity to sales markets and specific customer groups, as well as the capabilities of many LMT companies to use and influence these advantages in a flexible manner, are a further important reason for the relatively favourable development perspectives of such companies. (This argument is somewhat offset however by the factor discussed above of the declining importance of local networks.) For low-cost competitors from other countries it is often a time-consuming, difficult task to establish the necessary contacts and gain the information required to enter localized markets.

The low-tech issue ●

15

Third, a considerable number of low-tech companies are obviously capable of employing high-tech process technologies systematically and efficiently. Their specific process skills, and frequent, wellestablished contacts with the manufacturers of such technologies form the basis for this capability. The high-tech environment is a central requirement for the development perspectives of low-tech enterprises in this case.

These considerations should lead to a new understanding of the restructuring of the European economic landscape in the first years of the twentyfirst century. The economy does not appear to be undergoing a wholesale structural replacement of ‘old’ sectors with ‘new’ ones, or a substitution of ‘old’ technologies for ‘new’. In fact this process of change is evolving as a restructuring of sectoral and technological systems, a transformation more from within than without, not dominated by industrial activities in which competitive advantage, capabilities and economic growth are generated by front-line technological knowledge. In this context it must again be emphasized that industrial innovations are not for the most part based on newly-created scientific knowledge. Even where technical change is based on scientific activities, it is not necessarily recent; innovations stemming from the stock of knowledge and solutions to practical problems of various types may be more important than the creation of new knowledge. The relationship may, in addition, work the other way around, that is, when technology creates the foundation for scientific knowledge. LMT industries are well-placed to play a decisive role in innovation because the contribution of LMT companies is frequently an important precondition both to the innovativeness of value chains – or production systems – and the design, fabrication and use of a range of high-tech products. As is convincingly shown by Robertson and Patel (2005), the relationships between high-tech and non-high-tech sectors in developed economies are highly symbiotic and the well-being of high-tech firms and industries depends heavily on their ability to sell their outputs to other sectors in developed economies. Collaboration and networking between companies in different industries at regional, national, and transnational levels are increasingly important determinants of the innovativeness and competitiveness of individual companies. These value chains, filières or clusters include low-tech companies not just as third-tier participants in supply chains. The low-tech companies are frequently active participants, through user specification, in the development of technologically advanced machinery and equipment. Furthermore the dynamics and efficiency of value chains may crucially depend on the reliability, effectiveness, capabilities and specific knowledge

16

Introduction

of their low-tech partners, and on their integration into innovation processes of other firms in the cluster, whether low- or high-tech. This focus on the contribution of low-tech industries to the innovativeness of industry as a whole is extremely important from a policy perspective at both national and regional levels, and indispensable to assessing the overall growth and performance possibilities of the European economy. To follow the above line of argument, the high-tech prospects of many economies are based on the presence of and dynamic interaction with reliable low-tech functions and processes. The significance of low-tech companies as regards innovation policy must ultimately also be seen against the background of the strong and probably increasing international competitive pressure on providers of complex technologies and products. Their market position can by no means be regarded as permanently stable and promising. High technologies and the corresponding know-how can, in the context of global economic integration, diffuse rapidly, and the crucial point is that they are also quickly utilizable for innovations, so that the window for realizing innovation profits in this sector is in many cases quite small. One instructive example is a developing country like China which in a few years will be one of the largest developers and producers of such high-tech products as mobile telephones. Another example is the situation of the medium-high-tech automotive industry in countries like Germany. The dependence of German manufacturing on the auto industry provides specialization advantages but also increases the risk of severe damage from competition as highly sophisticated cars are increasingly being produced more cheaply in newly industrialized countries (albeit often by German firms). The policy conclusion to be drawn is that it is necessary to focus on the industrial innovation chain as a whole, concentrate more intensely on inter-sectoral connections, and identify the potentials of low-tech industries. Most notably, research shows that there are favourable development potentials for low-tech industries, not least in the high-tech-oriented countries of the European Union.

THE STRUCTURE OF THIS VOLUME The contributions in this volume pursue these topics and questions. They modify and criticize the theses outlined at the beginning and thus all contribute to deepening the understanding of the particular innovative ability and related development perspectives of LMT enterprises and sectors. In Part I, ‘Innovation in LMT: conditions and requirements’, the topic of the organizational and cognitive preconditions for (and drawbacks of) innovativeness on the organizational level of an enterprise is broached. First Gerd Bender, drawing on the ‘dynamic capabilities’ literature, introduces the

The low-tech issue

17

concept of ‘innovation-enabling capabilities’. Then Alexander Gerybadze and André Slowak show that successful innovation is critically dependent on the ability to influence standards and dominant designs. Their contribution explores the relationship between R&D, innovation strategy and the participation of firms in standard-setting groups and consortia, and the relevance of this process particularly to LMT companies. In the next chapter Benoît Godin thematicizes the concept of innovation by taking up the well-known debate on innovation indicators. First documenting how indicators for high technology came to be constructed, and identifying the moral economy on which they are based, he then contrasts high-tech with the recent literature on innovation, which he interprets as a move away from technological innovation and its moral superiority. In a provocative conclusion he suggests going still further: forgetting technology for a moment and going back to the very definition of innovation as novelty of any kind. Finally, Hariolf Grupp in turn criticizes this position, arguing that the concepts of innovation indicators are well developed in the scientific field and are, at least to some extent, capable of expressing the special characteristics of LMT innovations. He concludes that although adequate innovation indicators are available, they are rarely employed at the level of innovation policy. In Part II, contributions addressing the increasingly important networking of companies from different economic sectors are subsumed under the heading of ‘Technological diffusion and interrelationships between sectors’. As mentioned above, the innovative ability of LMT enterprises in particular is based on a close coupling to companies in other sectors. The chapters in Part II deal with these interrelations from various perspectives. Paul Robertson and Keith Smith show in their chapter that new knowledge which is vital for competitive reasons derives from unexpected directions and requires considerable reworking before it is of value. This places severe burdens on firms that must learn to manage their distributed knowledge bases in order to analyse, transform, combine and apply disparate items of technological intelligence in ways appropriate to their own environments. They examine how LMT firms manage channels for new knowledge by looking into the formal and informal ways they can capture and make sense of information from unfamiliar as well as from familiar sources. They show that the operationalization of new knowledge goes well beyond absorptive capacity and requires the development of a range of other capabilities. These interrelations are subsequently concretized by David Jacobson and Bernard Musyck. They present a case study of the interrelationship between LMT and high-tech in parts of the airline industry, adding new evidence from a case area not studied before in the context of research on innovation and LMT. It also focuses on primarily service-based activities rather than on the manufacturing focus of PILOT and other LMT research. Daniela

18

Introduction

Freddi discusses the issue of ‘Technology fusion and organizational structures in low- and medium-tech companies’. This chapter summarizes the possible evolution of LMT sectors thanks to the fusion of traditional and radical new technologies. In particular she focuses on the shift from traditional mechanical engineering towards mechatronics and tries to assess the consequences of this change for organizational structures. In particular she analyses the relationship between complexity and enterprise boundaries, exploring to what extent complex problems can be solved and complex products realized within or between companies. The growing importance of the integration of industrial production and service sectors is taken up in the two following contributions. Marja Toivonen in her chapter ‘Industrial innovations in relation to service sectors’ discusses the possibilities and challenges linked to the combination of services and manufacturing. She focuses mainly on the production of services in manufacturing, but also briefly on new features in the purchase of services by manufacturers. Her contribution starts with the introduction of a service model that illuminates the special nature of services and the ways in which they differ from goods. Thereafter she outlines different ways goods and services can be combined into valueproducing packages. Throughout her analysis the perspective is that of innovation: she examines which elements are capable of renewal and how the renewals can be effected. In the next chapter Eva Kirner, Gunter Lay and Steffen Kinkel turn to the relevance of product-related services in the light of low-, medium- and high-tech industries. Their research focus is the question whether an alternative innovation approach exists based on productrelated services – instead of an R&D-based innovation strategy – and whether such an approach might be an explanation for the survival of lowand medium-tech firms under the competitive conditions in high-wage countries. Based on the data from the German Manufacturing Survey 2006, the level of their analysis is not the industrial sector but the single firm. In their examination of different innovation and economic strategies related to different levels of R&D intensity, the authors deem the micro-level analysis to be preferable to a sectoral analysis because industrial sectors are composed of a variety of low-, medium- and high-tech firms. Part III, entitled ‘Local versus global perspectives in innovation’, discusses the development perspectives of LMT sectors and enterprises in the context of global market competition and the changing international division of labour. These interactions are examined by Anna Wzia˛tek-Kubiak on the basis of the economic transformation process in Poland. Though the new EU member states have made good progress in institutional reforms and adjustment to market-economy institutions, the major emerging issue is how to sustain further growth. In this context the author discusses the importance of traditional versus knowledge-based determinants of change for the

The low-tech issue

19

competitiveness of Polish manufacturing in LMT industries. The shift in factors that determine the competitiveness of ‘catch-up’ economies is at the heart of the discussion raised in the chapter. This relates to the innovativeness of LMT as compared to that of high tech, the specifics of innovation sources, forms and outputs, and the role they play in economic growth. The contribution by Martin Heidenreich follows up on this topic. Beginning with the theoretical debate on different patterns of interdependencies and on the dynamic balance between the advantages of spatial proximity of low- and high-tech industries and outsourcing, he analyses the traded and untraded interdependencies existing between low-, medium- and high-tech companies in Europe. These territorial dynamics of LMT industries are described on the basis of regional EU data. The attempts to face the risks and limitations of mostly traded interdependencies of LMT and MHT (medium- and hightech) companies are analysed by examining the newly designed Polish regional policies. Finally, Staffan Laestadius, Linda Gustavsson and Vicky Long take up the topic of ‘High-tech innovations in catching-up countries’. They argue that the present rapid catching-up processes in countries such as the ‘Asian Tigers’ challenges conventional theories on growth, stages of growth and product-cycle theories. By means of examples from the ICT sector the chapter shows that the tendencies towards leapfrogging are accelerated by the fast technological change in that sector. It argues at the outset that the development path now visible in China is characterised by high-tech innovations in combination with classical growth, and that this new situation is a challenge to the traditional understanding of complexity and of knowledge-transfer mechanisms. The concluding chapter to this volume, offered by James Wickham, makes a critical summary of the scientific as well as public debate on the emerging ‘Knowledge-Based Society’ (KBS). Wickham begins by examining key texts by which social scientists found an audience and influence beyond the academy, each of which posits a particular relationship between ‘knowledge’ and social structure. He then suggests that a key political use of the term today is to argue for enhanced expenditure on R&D and for the ‘reform’ of European higher education along explicitly American lines, while pointing to some problems with the ‘excellence’ of US education which Europeans tend to ignore in their attempts to emulate that system. Another crucial part of the KBS thesis discussed is the notion of the ‘learning organization’. The author suggests that this is at variance with some developments in vocational education which paradoxically amount to an undermining of the achievements of some European vocational educational systems. All these approaches accept the basic idea of the KBS, namely the equation of ‘knowledge’ with knowledge that is used at work. As the summary makes clear, this is an extremely inadequate basis upon which to build innovation theories and policies.

20

Introduction

NOTES Cf. in particular the findings of the European Commission project on ‘Policy and innovation in low-tech: knowledge formation, employment and growth contributions of the “Old Economy” industries in Europe’ – PILOT, which ran from the beginning of December 2002 to the end of 2005. 2. See also Schmierl (ed.) (2000); Palmberg (2001); Von Tunzelmann and Acha (2005). 3. Maskell (1998): ‘Learning in the village economy of Denmark: the role of institutions and policy in sustaining competitiveness’, in H.-J. Braczyk, P. Cooke and M. Heidenreich (eds) (1998) Regional Innovation Systems, London, pp. 190–213; Palmberg, C., op. cit.; and Von Tunzelmann, N.; Acha, op. cit. 4. For more details see Kaloudis, (2005) et al. 5. Similar correlations are indicated by Nonaka and Takeuchi’s (1997, pp. 70) category ‘operational knowledge’, which describes the process of integration (‘internalization’) of explicit and codified knowledge into ongoing operating processes that are strongly characterized by tacit knowledge. 6. This is related to the phenomenon that is discussed in innovation research with the concept of the ‘absorptive capacity’ of companies (Cohen and Levinthal, 1990). According to this concept, the prior knowledge within companies to a large extent determines the company’s ability to assess new knowledge and to use it systematically for innovations. 7. Putting it simply, the central argument of this resource-oriented analysis concept (e.g. Dosi et al., 2000) stemming from management research is that enterprises are characterised by a specific combination of special and rare resources, especially of knowledge of different kinds and that they have to possess a specific ability, designated as ‘dynamic capability’ to be able to use these resources for their strategic objectives (Bender and Laestadius, 2005; Laestadius, 2005). 8. Following Kogut and Zander (1992), who speak of a ‘combinative capability’. 9. The paradox is that on the one hand the industrial system facilitates flexibility in the sense of easily finding alternative suppliers and on the other it facilitates close – and long-lasting – relations with key suppliers. 10. In innovation research, this phenomenon of the generalisation of certain technologies is understood as a process of ‘technological convergence’ between different companies and sectors of industry (Rosenberg, 1963). 1.

REFERENCES Bell, Daniel (1973/1999), The Coming of the Post-Industrial Society: A Venture in Social Forecasting, New York: Basic Books. Bender, G. and S. Laestadius (2005), ‘Non-science based innovativeness: on capabilities relevant to generate profitable novelty’, in G. Bender, D. Jacobson and P. L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, 123–70. Bender, Gerd, David Jacobson and Paul L. Robertson (eds) (2005), Non-ResearchIntensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue. Carson, I. (1998), ‘Meet the global factory’, The Economist, survey of manufacturing, 20 June, 1–22. Cohen, W.M. and D.A. Levinthal (1990), ‘Absorptive capacity: a new perspective on learning and innovation’, Administrative Science Quarterly, 35 (1), 128–52.

The low-tech issue

21

Crouch, Colin, Patrick le Galès, Carlo Trigilia and Helmut Voelzkow (2001), Local Production Systems in Europe, Oxford: Oxford University Press. David, P.A. and D. Foray (2003), ‘Economic fundamentals of the knowledge society. Policy futures in education’, An e-Journal, 1(1) special issue, January. Dosi, Giovanni, Richard R. Nelson and Sidney G. Winter (2000), ‘Introduction: the nature and dynamics of organizational capabilities’, in Giovanni Dosi, Richard R. Nelson, Sidney G. Winter (eds), The Nature and Dynamics of Organizational Capabilities, Oxford: Oxford University Press, pp. 1–22. Drucker, Peter F. (1994), Post-Capitalist Society, New York: Harper Business. Fröbel, Folker, Jürgen Heinrichs and Otto Kreye (1977), Die neue internationale Arbeitsteilung, Reinbek: Rowohlt. Garibaldo, Francesco (2006), News from the Land of Districts and Flexible Manufacturing, IpL Bologna. Garibaldo, F. and D. Jacobson (2005), ‘The role of company and social networks in low-tech industries’, in G. Bender, D. Jacobson, P.L. Robertson (eds), NonResearch-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, pp. 233–69. Hatzichronoglou, T. (1997), ‘Revision of the high technology sector and product classification’, STI working papers 2, Paris: OECD. Henderson, R.M. and K.B. Clark (1990), ‘Architectural innovation: the reconfiguration of existing product technologies and the failure of established firms’, Administrative Science Quarterly, 35, 9–30. Hirsch-Kreinsen, Hartmut, David Jacobson and Staffan Laestadius (eds) (2005), Low-tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang. Hirsch-Kreinsen, H., D. Jacobson and P.L. Robertson (2006), ‘ “Low-tech” industries: innovativeness and development perspectives – a summary of a European research project’, Prometheus, 24 (1), 4–21. Jacobson, D. and K. Heanue (2005), ‘Policy conclusions and recommendations’, in G. Bender, D. Jacobson and P.L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, 359–416. Kaloudis, A., T. Sandven and K. Smith (2005), ‘Structural change, growth and innovation: the roles of medium and low-tech industries 1980–2000’, in G. Bender, D. Jacobson and P. L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, 49–73. Kogut, B. and U. Zander (1992), ‘Knowledge of the firm, combinative capabilities, and the replication of technology’, Organization Science, 3 (3), 383–97. Laestadius, Staffan (2005), ‘Innovation – on the development of a concept and its relevance in the knowledge economy’, in Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), Low-tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang, pp. 99–122. Lundvall, B.-Å. and S. Borrás (1997), The Globalising Learning Economy, Luxembourg: European Communities. Maskell, Peter (1998), ‘Learning in the village economy of Denmark: the role of institutions and policy in sustaining competitiveness’, in Hans-Joachim Braczyk, P. Cooke and Martin Heidenreich (eds), Regional Innovation Systems, London, pp. 190–213. Nascia, L. and G. Perani (2002), ‘Diversity of innovation in Europe’, International Review of Applied Economics, 16 (3), 277–93.

22

Introduction

Nonaka, Ikujiro and Hirotaka Takeuchi (1997), Die Organisation des Wissens, Frankfurt am Main: Campus. Palmberg, C. (2001), Sectoral Patterns of Innovation and Competence Requirements. A Closer Look at Low-tech Industries, Sitra report Series, 8, Helsinki: Sitra. Porter, Michael E. (1998), On Competition, Boston, MA: Harvard Business School Press. Robertson, P.L. and P.R. Patel (2005), ‘New wine in old bottles – technological diffusion in developed economies’, in G. Bender, D. Jacobson, P.L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, 271–304. Rosenberg, N. (1963), ‘Technological change in the machine tool industry, 1840–1910’, Journal of Economic History, 23, 414–46. Schmierl, K. (ed.) (2000), Intelligente Produktion einfacher Produkte am Standort, Frankfurt and New York: Campus. Schmierl, K. and H.-D. Köhler (2005), ‘Organisational learning: knowledge management and training in low-tech and medium low-tech companies’, in G. Bender, D. Jacobson and P. L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, published in Perspectives on Economic Political and Social Integration, XI, (1–2), special issue, 171–221. Semlinger, Klaus (2003), ‘Vertrauen als Kooperationshemmnis – Kooperationsprobleme von kleinen und mittleren Unternehmen und Auswege aus der Vertrauensfalle’, in Hartmut Hirsch-Kreinsen and Manfred Wannöffel (eds), Netzwerke kleiner Unternehmen, Berlin: Edition Sigma, pp. 61–88. Smith, Keith (2003), ‘What is the knowledge economy? Knowledge-intensive industries and distributed knowledge bases’, paper presented at the PILOT Workshop on Concepts, Theory, Taxonomies and Data, Department of Industrial Economics and Management, Royal Institute of Technology, Stockholm, 26–27 September. Stehr, Nico (1994), Knowledge Societies, London: Sage Publications. The Economist (1998), ‘The strange life of low-tech America’, 17 October, 85–6. Uchitelle, L. (2006), ‘Good-bye, production (and maybe innovation)’, The New York Times, 24 December. Von Tunzelmann, G. Nick and Virginia Acha (2005), ‘Innovation in “low tech” industries’, in Jan Fagerberg, David Mowery and Richard R. Nelson (eds), The Oxford Handbook of Innovation, Oxford and New York: Oxford University Press, pp. 407–32. Willke, Helmut (1998), Systemisches Wissensmanagement, Stuttgart: Lucius and Lucius.

PART I

Innovation in LMT: conditions and requirements

2.

How to grasp innovativeness of organizations: outline of a conceptual tool1 Gerd Bender

INTRODUCTION Sometimes persistent errors can be fruitful in a way. Take the linear model of innovation as an example. It is something of a conceptual zombie. Though reputed to be dead for at least two decades it still inspires some innovation research and the bulk of innovation policies. At the core of this model is the understanding that there is a sequence from scientific research via experimental development of new technology to innovative, marketable products. If this was true, technological innovations in nonresearch-intensive industries – low-tech industries according to the conventional classification (Hatzichronoglou, 1997) – would by definition be derivational phenomena. Innovators in these sectors would only use what others produce, that is to say, live on the pool of knowledge fed from – in the last instance – basic research. This is of course not true. Innovations are not necessarily based on scientific research or even on scientific knowledge; apparently most of them are non-science-based innovations. But there is nevertheless something fruitful in this false conception: it calls our attention to processes of interchange and transformations. Innovation is usually a distributed process. Some actors take up knowledge and other building blocks produced by other actors and transform both according to their own needs, aims and imaginations. The problem with the linear model is that it paints a far too simple picture of this complex entanglement of diverse players in space and time. Technological progress and innovation – taking up a picture used by Gibbons et al. (1994) – resemble a football match rather than a relay race. Different actors with different roles interact and none of these roles is in a somehow naturally primal or privileged position. Though some of the players are more successful in producing novelty than others, the degree of innovativeness does not seem to have anything to do with an actor’s 25

26

Innovation in LMT

proximity to the realm of science (see Nowotny et al., 2001 and, for the PILOT2 experience, Bender, 2006). An actor shall be called innovative when that actor is able to meet changing requirements and to recognize and seize arising new chances in novel ways. When this actor is an individual we tend to ascribe sophisticated powers of comprehension and creativity to her or him. How does this translate into features of an organization? One may assume that an organization will never be creative when not at least some of its members are. But this is obviously not enough; there is a plethora of examples of organizational rules and routines squelching individual creativity. What is it that makes an organization innovative? And what makes some organizations more innovative than others? It is not – this is the basic assumption of this chapter – R&D capacity. Most innovations are not a result of systematic scientific research or latest scientific and technological knowledge – thus, research and development is by no means the most important activity for the creation of innovation. Individual creativity is obviously important – but ‘to innovate’ is usually a collective action and, thus, organizations and organizational capabilities play a vital role. This chapter’s main intention is to outline a tool that helps to identify some preconditions for innovativeness on the level of the organization. In the next section the concept of innovation enabling capabilities will be introduced in a critical examination of related literature. In the subsequent paragraph it will be illustrated drawing on case-study work conducted as part of a research project on Policy and Innovation in Low-tech Industries (PILOT).

THE CONCEPT OF INNOVATION ENABLING CAPABILITIES One difference between economists and other social scientists is that the former, due to their research interests, tend to underestimate differences between firms (Nelson, 1991). Penrose (1959) was among the first in her profession who explicitly stated that firms differ due to internal mechanisms, not only as a consequence of the competitive environment in which they are embedded: A ‘firm is more than an administrative unit; it is also a collection of productive resources the disposal of which between different uses and over time is determined by administrative decision . . . The fact that most resources can provide a variety of different services is of great importance for the productive opportunity of a firm. It is the heterogeneity, and not the homogeneity, of the productive services available or potentially available

How to grasp innovativeness of organizations

27

from its resources that gives each firm its unique character’ (Penrose, 1959: 75–6). Heterogeneity and adequate organizational responses to varying demands from different environments had been in the focus of various important studies later on in both industrial and organizational sociology (among others Burns and Stalker, 1961; Woodward, 1965, 1970) and economics (Cyert and March, 1963; Chandler, 1966). Over the past two decades this discourse has developed into what is today labelled as the resourcebased view of the firm (Priem and Butler, 2001) or, a more recent offspring, the dynamic capabilities approach (Dosi et al., 2000). The core argument in this tradition is that differences between firms or smaller organizational subunits of such may be analysed in terms of capabilities orchestrating and mobilizing resources as regards knowledge formation and productive activities (for example Foss and Robertson, 2000; Zollo and Winter, 2002). Drawing on this we introduce the concept of innovation enabling capabilities (IEC). Such capabilities develop in timeconsuming learning processes, they become operative in firm-specific knowledge – both codified and tacit – and routines and thus also constitute path dependency. And as they are part of a firm’s ‘deep structure’ they cannot simply be copied by another organization. There are many firms that develop significant parts or even most of these capabilities in the R&D department. With our approach, however, the relative importance of the R&D units becomes an empirical question rather than something postulated a priori. The coupling of novelty-formation to capabilities rather than to R&D can also rely on findings in the field of Science and Technology Studies (STS) that clearly evidence that knowledge production and innovation are multi-actor and multi-level processes that involve a wide range of different types of knowledge, actors and interests (for example Callon, 1986; Bijker and Law, 1992; Disco and van der Meulen, 1998; Sørensen and Williams, 2002; Borup et al., 2006). Using this literature as a further conceptual pillar our capabilities approach leaves us with a much wider concept for describing and understanding the performance of innovative organizations and its preconditions. Though it was developed for the analysis of firms that do not have high levels of R&D the STS literature in particular may support expectations that it can be developed further for use in high-tech environments also. The use of the capabilities concept in the literature is in no way uniform. Dosi et al. (2000: 3) perhaps only slightly exaggerate when they note that ‘the term “capabilities” floats in the literature like an iceberg in the Arctic sea . . . not easily recognized as different from several icebergs nearby’; neither is the concept in itself unambiguously distinctive. This can be illustrated by two fairly recent, and prominent, definitions of the term. For

28

Innovation in LMT

Teece et al. (2000: 339) ‘dynamic capabilities are the ability to reconfigure, redirect, transform, and appropriately shape and integrate existing core competences with external resources and . . . assets to meet the challenges of a time-pressured, rapidly changing Schumpeterian world of competition and innovation’. In contrast, for Zollo and Winter (2002: 340) ‘a dynamic capability is a learned and stable pattern of collective activity through which the organization systematically generates and modifies its operation routines in pursuit of improved effectiveness’ (italics in both quotations added). For our purpose here the more abstract understanding advocated by Teece and his colleagues seems to be more appropriate. That is to say, we understand capabilities not as a pattern of activities but rather use the term to address specific preconditions for specific activities: a particular configuration of enabling cognitive, financial and material (machinery and so on) resources which characterizes an organization and which constitutes potentialities for this organization. Following Dosi et al. (2000) we include intentionality and conscious decision making as well as routines as building blocks of the concept. The notion capability is related to a recognizable purpose expressed in terms of significant outcomes it is supposed to enable. This also means that capability building can be a strategic aim (cf. Tidd et al., 2001) in which both the actual process of capability building and the definition of specific aims are affected by the capabilities already present at any point in time. The latter direction of impact has been labelled with the term absorptive capacity (Cohen and Levinthal, 1990) or absorptive capabilities (Laestadius, 1995). Here we are particularly interested in the dialectical interplay between outcomes and preconditions. We may identify two analytical dimensions of innovation enabling capabilities, transformational and configurational capabilities, which are tightly interwoven empirically.3 Transformational capabilities constitute the enduring ability of an organization to transform available general knowledge into plant, firm or task specific knowledge and competence. This is a core competence particularly in low-tech industries; general knowledge on traditional industrial techniques like welding, for instance, is spread all over the world. The ability to transform it into specialized and economically competitive ‘high quality zero defect’ competence separates the profitable firms from the rest. One may describe the mechanism underlying this as a shift between levels which has to be mastered by an organization: globally available knowledge is being accommodated and transformed locally for local use. Here again, we may link our reasoning to a long-standing discussion in the STS field. Rip (1997) for instance analysed the relevance of the distinction between global and local. He argues that global knowledge is in principle generally

How to grasp innovativeness of organizations

29

available whereas local knowledge refers to, and is embedded in, a certain confined situation. The two types of knowledge differ as regards the claimed validity – universality in the one case versus adequacy in the other. And they differ in form as well. Global knowledge is always codified as it refers to a paradigm4 whereas local knowledge, though having codified elements (instructions, handbooks, formal organizational rules, technical process protocol and so on), is usually characterized by some degree of tacitness. This difference has considerable practical consequences. The change from the global level to the local is not just a transposition of the same but always involves transformations. Transformation of global knowledge in local settings proceeds as contextualization, that is, the global knowledge is not simply replicated locally (Zollo and Winter, 2002) but it has to be translated according to local conditions which may include both its (re-)codification and its practical adoption ‘by using’. And this in turn may necessitate some reorganization of competencies and resources available in the firms’ capital – human as well as material. The phrase ‘application of generally available knowledge’ (that is global knowledge) tends to shroud both the complex processes of transformation and adaptation and their individual and organizational preconditions. The ability to render global, for instance technological knowledge useful in and for specific local circumstances always presupposes not only professional ‘literacy’ as for the respective technological discipline but also contextual experience and practical knowledge – that is, knowledge and know-how concerning the local (cultural, technological, financial, and so on) possibilities and needs. Both together are fundamental for an organization’s transformational capabilities. Configurational capabilities constitute the enduring ability to synthesize novelty by creating new configurations of knowledge, artefacts and actors. There are two facets of configurational capabilities. 1.

Integration of dispersed knowledge. The generation of novelty is to a large extent based on the ‘synthesizing competence’ (Bender, 2005) of innovators, that is, on their ability to tap distributed knowledge and know-how from sometimes very different areas and to recombine them creatively (see ‘combinative capability’, Kogut and Zander, 1992). This may include knowledge embodied in hard- and software, it may be scientific knowledge, design competence, or expertise in logistics; it may be codified knowledge or tacit knowledge incorporated in individuals or teams. Although most of the technologies used in industry have been well known in general as well as in most of their details for years, what

30

2.

Innovation in LMT

counts is often the precision and speed of a new architecture perhaps only marginally different – although by definition unique in its important elements – from another one. It may be argued that to a large extent this synthesizing activity across different technologies and knowledge fields is what qualified engineering is essentially about. The other dimension of configurational capabilities – empirically tightly interwoven with the first – is an organizational one: the enduring ability not only to combine pieces of knowledge and technology but also to link actors together that possess relevant knowledge, technology and competence. That is, configurational capabilities include an organization’s aptitude to efficiently provide for access to and use of distributed sources of relevant knowledge and competence; this may in many cases involve the ability to cooperate with external R&D facilities or design houses. And it also embraces an organization’s competence to manage logistics in a timely and flexible manner.

Though developed for analysis on the level of the organization our notion of configurational capabilities can easily be linked to the fundamental STS argument that new technology and innovation can entail a change in networks of social relations even on the societal level. There is a variety of concepts that address this co-evolutionary process: Hughes (1986) analysed the development of power supply systems in a ‘seamless web’ of technology and political and economic institutions. Callon (1986) argued that the stability of technological solutions is dependent on the successful networking of human and non-human elements; a process Law (1987) called ‘heterogeneous engineering’. Bender (1999) analysed the development of new technology as a process of aligning actors, artefacts, interests and technological concepts that brings new ‘sociotechnical configurations’ into the world. All these studies document that innovation is not merely a scientific or technological process with economic consequences (and presuppositions) but normally implies changes of ‘heterogeneous’ configurations and social relations. We may conclude this discussion with a pointed reformulation of a basic practical problem for innovators. The major task is not necessarily to develop or apply the latest technological knowledge; but innovation always entails the creation and management of sustainable new configurations of various types of knowledge, actors and artefacts. And a crucial organizational precondition for this is the creation and reproduction of appropriate innovation enabling capabilities in the sense just explicated. From this vantage point the problems analysed by Henderson and Clark (1990) in their work on architectural innovation turn out to be not specific for this certain type of novel products. They show how certain ways of

How to grasp innovativeness of organizations

31

doing things tend to crystallize into organizational structures, information filters, and communication channels which, on the whole, shape an organization’s capabilities. They also highlight the problems incumbent firms face when they are confronted with architectural innovations. We would argue that what the two authors describe as a specific threat in this specific situation seems not so specific at all when you take into account that innovation is always about reconfiguring existing knowledge, components and actors. That is to say, the organizational processes and mechanisms analysed by Henderson and Clark play at all times – admittedly to a greater or lesser degree – a role, no matter which type of innovation we have in mind. In the next section we give some empirical illustrations for innovation enabling capabilities. We focus on the two dimensions elaborated above: transformational capabilities, being able to transform different kinds of globally available knowledge into local knowledge and competence; and configurational capabilities, being able to configure dispersed knowledge and know-how as well as actors and other ‘repositories’ of such knowledge and know-how.

INNOVATION ENABLING CAPABILITIES OF A FEW LOW-TECH FIRMS The following discussion draws on approximately 40 company case studies in 11 European countries conducted as part of the PILOT project. The selection was not a representative sample; to qualify as a case a company had to be innovative (regarding products and/or processes), economically successful and of a critical minimum size. And because innovation in lowtech has not been investigated very well so far, the case studies’ purpose within the overall project was rather to state problems more precisely than to answer questions (see Schmierl and Kämpf, 2004 for a general overview). A standardized questionnaire was used to collect basic data on the respective company, its production process and its relations to suppliers, clients and partners. This survey was complemented by about half a dozen semistructured extensive interviews for each case study with firm and workforce representatives (based on a master guideline common for all national project teams), by site inspections and by an analysis of available documents of the firms (catalogues, product specifications, website and so on). One intention of the case studies was to identify relevant aspects of innovation enabling capabilities of low-tech companies, that is, cognitive, organizational and material abilities a non-research-intensive firm has to develop to be able to generate profitable innovations.

32

Innovation in LMT

Transformational Capabilities With respect to transformational capabilities, we have argued above that high performance in innovation does not necessarily imply creation or use of latest scientific insights but rather the ability to identify potentially relevant knowledge and transform it according to the local conditions with which a firm is confronted. This may or may not include new scientific and technological knowledge; but in low-tech industries we can assume that this will – almost by definition – not be the main ‘raw material’ for successful innovations. A few examples illustrate the point. One of the companies in the sample was a German producer of large (up to more than 4 metres diameter) longitudinally welded steel pipes used mainly in the construction industry or for pipelines. A major problem for the transportation of aggressive fluids or gases in pipelines is that the tubes are subject to rigorous and sometimes conflicting demands. They must be highly resistant to detrimental chemical effects of the conveyed substance and at the same time very robust to endure physical impacts such as extreme heat or, for instance when they are laid in the deep sea, high pressure. A major technical difficulty is that steel qualities that provide good protection against corrosion are usually susceptible to fatigue and vice versa. The solution to this problem is so-called clad tubes the outer shell of which is made of a very hard, pressure resistant steel while stainless steel and alloys based on nickel are used for the interior of the tube, that is, material which is dimensionally less stable but unsusceptible to corrosion. Such pipes combine high strength with distinct surface properties in a single product. The principle of this ‘coated tube’ is well known, but the principle alone is not a usable – let alone saleable – solution. Different pipes have to be designed for different applications and the most delicate point in terms of knowledge and know-how is the welding of the respective compound material. That is to say, innovation in this field does not mean developing fundamentally new principles or artefacts but developing modified products and related manufacturing procedures. ‘A clad tube’ is nothing new. ‘The clad tube’ for a discerning customer very often is. Dimensions need to be established and tested, welding material and processes need to be adjusted. To develop such a product can take up to two years and may cost 100 000 euro. But this work is neither ‘R’ nor ‘D’ in the sense defined in the Oslo or Frascati Manuals. In the firm we have studied the novel products are developed by small teams of engineers and technicians alongside their everyday jobs in the plant. If needed the teams get in contact with external partners; among them are scientific institutes and test labs. This firm is obviously innovative and it has to be because otherwise it would not stay in business much longer; they are too small to compete with

How to grasp innovativeness of organizations

33

the very large producers of standard pipes. Instead, they serve niche markets and hence innovation is their core business. But the company’s innovativeness is in no way determined by excellence in R&D in the usual understanding. Instead, what gives them a competitive edge are the capabilities to adopt what is well known in principle and to transform it into solutions for particular needs. Transformation of relevant knowledge in the sense brought forward here is not limited to expertise stored in books, databases or brains but may also include technical equipment. Here too we may use the company just mentioned for an apt illustration. In the late 1980s they bought a huge plate press which had been used in a shipyard heretofore. With this machine an enormous sword hit a steel plate which was moved laterally in defined intervals and in so doing it curved flanks of ships. The purchase of this machine was not just an ‘import’ of ready-made equipment from the shipbuilding industry to tube manufacturing. The buyer had to reconstruct it extensively. In particular it was necessary to develop a new control programme to meet the very much tighter tolerances characteristic for their business which they did in cooperation with an engineering company. This new software also added flexibility to the whole business process. For once it allowed for processing tubes of very different sizes in best time. But not only in manufacturing did it make a difference; it also opened new possibilities for the design of tubes.5 The whole system was tested, modified and re-tested. The staff developed skills and routines in using the new equipment. In fact, the tube company has absorbed machinery and know-how developed, approved and used elsewhere and transformed it according to their specific needs and possibilities. This resulted in innovative technical equipment and novel process technologies that enabled the firm to meet new demands and to design and manufacture new products. Another case study, focused on a Finnish engineering firm, illustrates the point in a slightly different way. This company is one of the world’s leading providers of propulsion systems for ships with high manoeuvrability demands (such as ice breakers). Their proficiency is based on accumulated experience in propeller design as well as casting and grinding techniques. The design competency, in its turn, is based on acquisition and creative transformation of available technology which the firm has learned to master and to integrate with academic knowledge in hydromechanics acquired through collaboration with technical research institutions in Finland. Thus they gather globally available technologies and blend them with their local casting and grinding competence to meet the highly specialized needs of propeller manufacturing; this includes the metallurgical composition of the castings as well as the processing and quality management. In doing so, the firm steadily develops new expertise in engineering

34

Innovation in LMT

of propulsion systems which, again, strengthens its transformational capabilities as for relevant new hydromechanical knowledge. An Irish engineering firm can illustrate another aspect of transformational capabilities and of their commercial relevance. The company designs, installs and validates stainless steel process systems and manufactures atmospheric pressure rated vessels to applicable design codes. One recent innovation project of the firm was Cleaning in Place (CIP), which means that once a product has been processed through any particular piece of machinery, that apparatus has to be cleaned before either a different product or another batch of the same material is processed so that no cross-contamination can occur. The most effective way to implement CIP technology is to design it into a process and this is what the Irish firm does. It involves the addition of spray systems, tank cleaners, nozzles, and seals in order to automate the cleaning process. CIP technology was always relatively extensively used in the dairy industry and now the volume of its use is increasing dramatically in the pharmaceutical industry. The company already had significant experience with CIP for the food industry and was able to develop it further to meet the requirements of applications in the pharmaceutical industry. One key to the company’s success is exactly this dynamic character of its proficiency; they are a competitive designer and producer of vessels and process systems because they have capabilities to transform expertise in one field into knowledge adjusted to requirements and conditions in another environment. That is, they possess capabilities to redesign and creatively reproduce their own knowledge bases and, thus, capabilities. All of the cases outlined here show that innovation enabling capabilities in general and transformational capabilities in particular trigger a sort of Matthew effect. The more sophisticated they are the better can an organization integrate new knowledge and other resources and convert them into novel ideas and eventually products. Furthermore the cases support two main arguments brought forward in much of the literature discussed above. New technological knowledge does not necessarily emerge from R&D in the usual sense. And elaborate new technological knowledge is not inevitably a precondition for technical innovation. Instead new knowledge and novel technologies may be the result of the creative use of well-established knowledge which usually implies its transformation – and its (re)configuration. Configurational Capabilities To be successful, low-tech companies very often must develop capabilities to creatively combine distributed knowledge (which may include its prior transformation). The cognitive side of this (know-what) is to some extent congruent with what has been addressed with the term absorptive capacity

How to grasp innovativeness of organizations

35

(Cohen and Levinthal, 1990) – the ability to evaluate and utilize outside knowledge. But it also has an organizational side and that is the ability to both identify actors and other ‘repositories’ of potentially relevant knowledge (know-where) and organize some kind of interchange with them. In fact one could talk about a double-helix of know-what and know-where. That is to say, what we label configurational capabilities is an important aspect of an organization’s proficiency in networking or in other forms of trans-organizational interaction. (Re-)configure distributed knowledge There are many examples of creative configuring of distributed knowledge and competence in the PILOT case studies. One of them is an Austrian manufacturer of railroad tracks. The company’s capabilities to join forces with external expertise in order to foster development of innovative own product solutions are a critical success factor. This does not just mean that they are able to organize external support for continuous improvements of their core products; the competitive edge results from the ability to be more creative when needed. The company is confronted with a general trend in the railway industries. Due in part to structural changes within many railway transportation companies customers tend to ask for integrated system solutions rather than simply for tracks. Hence, the supplier has to be able to functionally augment its core products without giving up the advantages of specialization. And it is: the firm is well known for being able to produce the worldwide longest head hardened non-welded pieces of rail (120 metres). When this product was introduced in the 1990s it was in itself an innovation. But the firm also offers a novel process to lay these bulky pieces. The appropriate handling system was developed in collaboration with a German manufacturer of railway equipment and machinery; its design embodies the merged expertise of both partners.6 The integration of distributed knowledge and competence in this case took the form of a joint product of independent partners. The German tube manufacturer, already mentioned, provides quite another example for what we call configuring distributed knowledge. To enhance product qualities some kinds of tubes have to be heated to more than 1000ºC after welding and straightening. If you do this using an oven or any other kind of external heat supply it is very likely that the heavy tubes will deform and in the worst case they simply collapse because heating softens the steel. To prevent this, engineers of the firm implemented an innovative annealing process applying electromagnetic induction which allows for extremely high temperatures without making any alterations in the dimensional tolerances of the pipes necessary. Neither the idea nor the general process was something new heretofore. It was even used for tubes before, if just for

36

Innovation in LMT

much smaller ones. But this firm constructed, in collaboration with an engineering consultant, the machinery to accommodate this ‘in principle available technology’ to their products and local needs. This was much more than simple scaling up in size. In fact, they creatively configured knowledge on fundamental physical principles and technical solutions based on these principles with own expertise in handling huge steel tubes. And thus they synthesized a new technological system and related new knowledge. The novelty results from the ingenious combination of more or less well-known technological and cognitive ‘components’. Capabilities to configure distributed knowledge may also have a temporal dimension: the ability to anticipate future customer needs. The importance can be illustrated by the experience of a Spanish manufacturer of rails. As engineers of this company are in permanent contact with colleagues in their main customer’s quality and purchase department they are fairly well able to foresee what the client may need even before any requirements have been expressed. This shows that acquaintance with local conditions and knowledge on the side of partners (downstream in this particular case but also upstream) may be a constituent of innovation enabling capabilities. We are aware that there are many examples of the opposite in the literature. One may draw the conclusion that whether familiarity leads to lock-ins or to innovation can in itself depend on the innovation enabling capabilities of the actors involved of which familiarity with the circumstances of utilization of one’s own product is but one element. Nevertheless we would argue that knowledge about the context of application of your own products and the ability to integrate this knowledge with ‘domestic’ technological expertise is an important aspect of a supplier’s innovation enabling capabilities in general. Though it is basically impossible for a producer to know everything needed to know in advance on how to keep customers satisfied, the better the context of application is known the quicker a producer may come up with appropriate new designs or prototypes. This is particularly true when the supplier produces components for larger systems. In most of the respective companies in the sample, firm representatives experienced it as a serious shortcoming that they often know only the interface between the customer’s system and their own subsystem. This is perhaps a rather mundane problem but the consequences are far from trivial. Some of the firms invest considerable sums to accumulate such knowledge mainly by sending personnel to appropriate courses;7 others – the small ones in particular – would have liked to but cannot afford it. (Re-)configure actors and other ‘repositories’ of knowledge and know-how Configuring distributed knowledge and configuring actors are very often two sides of the same coin. This holds for smaller low-tech companies in

How to grasp innovativeness of organizations

37

particular because due to lack of resources they are often not able to incorporate knowledge sources by hiring appropriate experts or paying steadily for specialized service providers. Hence, they have to organize enduringly distributed ‘repositories’ of relevant knowledge. What this means can be easily exemplified. A small manufacturer of heating elements in the sample pursues a second-starter innovation strategy. To develop new product or process technologies itself is not an option because it is considered too risky and too expensive. Hence, decision makers in the company very carefully watch new developments presented at trade fairs and on the market and not until they come to the conclusion that a novelty may be important for their own business do they start looking for a partner who can supply them with the appropriate technology. It is obvious that this strategy’s success depends very much on the firm’s configurational capabilities, that is, on both the available know-what and know-where and the accessibility of financial and other resources to keep this knowledge constantly up-to-date. The latter tends to be not in the centre of the capabilities discussion. However, the case at hand illustrates that it is important to conceptualize financial and other material resources as constitutive elements of innovation enabling capabilities. This particular firm’s cash-cow is customized micro-coil heating elements used for nozzles of plastic casting machines. The main component of these elements is a tiny heating spiral with a diameter of about 1 millimetre. The state-of-the-art insulation technology for them is with ceramic powder but this solution has reached its limits. When you miniaturize the coils further you need another insulation material – and an alternative process to apply it – to ensure safe and reliable functioning. Though there is currently no demand for still smaller micro-coil heating elements this may change quickly. Hence, the firm’s engineers have tried out alternative substances but they have not been able to do this very systematically. That they have not is not caused by limited ‘absorptive capacity’ (Cohen and Levinthal, 1990). Due to lack of time and resources the company simply cannot afford proper technological experiments or methodical search for suppliers of alternative technologies just to be prepared for uncertain future demands. That is, there is a potentially critical weakness in the firm’s configurational capabilities. Pooling knowledge by configuring actors is an issue for other cases in the sample too, such as, again, the tube producer. Very basic features of their products are actually determined elsewhere, namely in the blast furnace of the steel mill the company cooperates with and in the relationship with the suppliers of welding material. In the standard business case, that is, when the tube specifications can be met with standardized material the company can treat suppliers as a black box; the company orders a well-defined commodity and

38

Innovation in LMT

it is supplied. There is no necessity to make the relevant knowledge and competence the provider possesses explicit; it is embodied in the steel or in the welding wire and so on. But when the company needs non-standard steel qualities, communication is inevitable and the fundamentally distributed character of the tube manufacturer’s knowledge bases becomes apparent. This organization does not know everything that must be known to do the job; hence they draw on outside repositories of knowledge on a regular basis. This is just another way of saying that the core company8 functions as an intermediary between steel mill, supplier of welding technology and end customer. They know the client’s requirements and at least in principle they also know what the steel mill among others can do. Part of their engineers’ competence is to be able to organize trade-offs between the various functional features of a product a client demands. In a way this implies an integration of knowledge distributed along the value chain. It also frequently implies the generation of new knowledge; the definition of product specifications quite often involves the definition of objectives for a collaborative development project of the tube producer and its steel supplier. Critical for the success of the firm even in everyday business is the ability to handle this distributedness. In other words, parts of their configurational capabilities accrue from their employees’ skills in communicating across the boundaries of their immediate field of expertise. For some of the cases in our sample a major competence and one key to success is the competence to timely and flexibly manage logistics. A striking example is a Swedish producer of high-end furniture for offices and public spaces. The components in their products are up to 90 per cent manufactured by suppliers, most of them are based on standard components which are upgraded and they are delivered just in time. The logistics dimension in this production system is crucial. This small firm cooperates with up to 300 different suppliers; most of the supply chain is located within 200 km of the factory leaving only the most standardized of the components used in most products to be purchased from global suppliers. The firm’s output are bespoke products but to a large extent based on 3000 modules of components and fittings. To be able to cope with this complexity the firm has had to develop capabilities to configure and manage a wide spread network of actors with complementary knowledge and expertise. They are successful because – and as long as – they are able to configure suppliers along the value chain and orchestrate these actors’ activities. The quintessential point seems to be the firm’s configurational competence, that is, the capabilities to capitalize on a system of coordinated links between actors. Whether the focus is on the cognitive or on the organizational dimension of configurational capabilities, the examples presented support our thesis that the economic success of low-tech companies very often depends on the

How to grasp innovativeness of organizations

39

ability to bridge the gap between different knowledge domains creatively and to create new bonds between different actors and/or processes and/or components. And they also provide evidence that this may mean rather different practical demands depending on the specific technological and economic environment a firm is embedded in.

CONCLUSION This chapter’s main intention has been to outline a tool that helps identify fundamental conditions for innovativeness in organizations. The suggested answer to the question, what makes an organization innovative, is innovation enabling capabilities. Two dimensions of a related concept – transformational and configurational capabilities – have been unfolded. It is crucial to note that the discrimination between the two is strictly analytical; it intends to reduce complexity but is not an empirical description. That is, we do not say that some firms need or possess more transformational capabilities and others mainly configurational ones. In order to configure distributed knowledge in novel ways it is usually also necessary to transform it because otherwise it may not match the new configuration. Thus, transformational and configurational capabilities are two sides of the same coin; to be innovative an organization has to develop both. But the two are – therefore the distinction – not the same. To build up transformational capabilities requires investment in other things as compared with configurational capabilities. Transformational capabilities are about cognition and learning; they are based to a large extent on what Cohen and Levinthal (1990) have called absorptive capacity. But while their interest is on organizational abilities and constraints to identify knowledge created elsewhere as relevant and to absorb it, our focus is on transformations. More often than not will absorption of knowledge need to be adapted to local needs and possibilities and this requires specific capabilities. Whether the knowledge in question is the latest scientific or any other type of knowledge is not a matter of principle but an empirical issue. And the complementary term configurational capabilities gets us one step further because it helps to identify the complexity of the learning conditions in a given case. While configuring actors and equipment in novel ways may always entail learning processes, we have provided evidence that it often requires not only cognitive and other intellectual abilities. The approach to innovation we introduced here may have advantages for both our understanding of innovation and its preconditions as well as for innovation policy. It is based on the understanding that proper R&D is only one form of knowledge creation among others that are relevant for

40

Innovation in LMT

technical innovations. The innovation enabling capacities concept was designed to cope with this diversity. Though it was developed with a focus on experiences within low-tech firms, it may thus be useful for an analysis of science-based innovation too – for the very reason that it aims at organizational and cognitive preconditions for the generation of novelty in general. The particular features of an organization that we describe as having innovation enabling capabilities, like any other capabilities, cannot be copied easily. But their development can be influenced strategically. And this is where the policy relevance of our conceptual exercise comes into play. Notwithstanding the undeniable importance of public support for R&D and academic education, to promote low-tech companies’ innovativeness – their potential capacity to both utilize and generate technological novelty – should be a major element of sustainable innovation policies. That is to say that innovation policies in the European Union and its member states should focus on all innovative companies in the economy – not only on the R&D-intensive small segment we tend to label high-tech. And the concept of innovation enabling capabilities may help to determine more precisely where to apply the lever.

NOTES 1. This chapter is based on previous collaboration with Staffan Laestadius (see Bender and Laestadius, 2005). 2. European Commission project ‘Policy and innovation in low-tech industries: knowledge formation, employment and growth contributions of the “Old Economy” industries in Europe – PILOT’ which ran from the beginning of December 2002 to the end of 2005. 3. This bears some resemblance to Teece et al.’s (2000: 345) discussion of transformation and reconfiguration as part of a firm’s organizational and managerial process aiming ‘to reconfigure the firm’s asset structure, and to accomplish the necessary internal and external transformation’. But they focus on transformation and reconfiguration of a firm’s capabilities. With our concept we want to underline that the acquirement of, for example, distributed knowledge may entail not only transformations (and re/configuration) on the side of the acquiring organization but also of the acquired knowledge. 4. This is what makes global knowledge – different from local knowledge – easily transferable in principle; global knowledge is by definition mobile. 5. With the traditional bending roll machines it is not feasible to machine thick walled tubes wider than 4 metres. 6. The Austrian ‘rail smithy’ holds the exclusive sales rights globally for this device produced by the German partner. For the Austrians this means not only diversification of their product range, the availability of the dedicated handling system also supports marketing of their innovative ultra-long rails. 7. Individual knowledge acquired is not yet a capability of the organization but this way of incorporating knowledge about the application environment can be a means to build up its innovation enabling capabilities. 8. ‘Core’ is meant here in a strictly topographical sense, that is, without any allusion to power or dependencies.

How to grasp innovativeness of organizations

41

REFERENCES Bender, G. (1999), ‘Shaping technology as a means of transforming society. The case of the GSM standard for mobile telecommunication’, Science Studies, 12 (2), 64–82. Bender, Gerd (2005), ‘Innovation in low-tech companies – towards a conceptualisation of non-science-based innovation’, in Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), Low-tech Innovation in the Knowledge Economy, Frankfurt: Peter Lang, pp. 85–98. Bender, Gerd (2006), ‘Peculiarities and relevance of non-research-intensive industries in the knowledge-based economy, final project report’, http://pilotproject.org/publications/publications.html, accessed 15 October 2007. Bender, Gerd and Staffan Laestadius (2005), ‘Non-science-based innovativeness – on capabilities relevant to generate profitable novelty’, in Gerd Bender, David Jacobson and Paul L. Robertson (eds), Non-Research-Intensive Industries in the Knowledge Economy, Perspectives on Economic Political and Social Integration, XI (1–2), Special Issue, 123–70. Bijker, Wiebe and John Law (eds) (1992), Shaping Technology-Building Society: Studies in Sociotechnical Change, Cambridge, MA: MIT Press. Borup, M., N. Brown, K. Konrad and H. van Lente (2006), The sociology of expectations in science and technology, Technology Analysis & Strategic Management, 18 (3–4) (special issue). Burns, Tom and G.M. Stalker (1961), The Management of Innovation, London: Tavistock. Callon, Michel (1986), ‘The sociology of an actor-network: the case of the electric vehicle’, in Michel Callon, John Law and Arie Rip (eds), Mapping the Dynamics of Science and Technology. Sociology of Science in the Real World, London: Macmillian, pp. 19–34. Chandler, Alfred D. (1966), Strategy and Structure, New York: Anchor Books. Cohen, W.M. and D.A. Levinthal (1990), ‘Absorptive capacity: a new perspective on learning and innovation’, Administrative Science Quarterly, 35 (1), 128–52. Cyert, Richard M. and James G. March (1963), A Behavioural Theory of the Firm, Englewood Cliffs, NJ: Prentice-Hall. Disco, Cornelis and Barend van der Meulen (1998), ‘Getting case studies together. Conclusions on the coordination of sociotechnical order’, in Cornelis Disco and Barend van der Meulen (eds), Getting New Technologies Together. Studies in Making Sociotechnical Order, Berlin and New York: de Gruyter. Dosi, Giovanni, Richard Nelson and Sidney Winter (2000), The Nature and Dynamics of Organizational Capabilities, Oxford: Oxford University Press. Foss, Nicolai J. and Paul L. Robertson (eds) (2000), Resources, Technology and Strategy, London: Routledge. Gibbons, Michael, Camille Limoges, Helga Nowotny, Simon Schwartzman, Peter Scott, and Martin Trow (1994), The New Production of Knowledge. The Dynamics of Science and Research in Contemporary Societies, London: Sage. Hatzichronoglou, T. (1997), ‘Revision of the high technology sector and product classification’, STI Working Papers 2, Paris: OECD. Henderson, R.M. and K.B. Clark (1990), ‘Architectural innovation: the reconfiguration of existing product technologies and the failure of established firms’, Administrative Science Quarterly, 35 (1), 9–30.

42

Innovation in LMT

Hughes, T.P. (1986), ‘The seamless web: technology, science, etcetera, etcetera’, Social Studies of Science, 16 (2), 281–92. Kogut, B. and U. Zander (1992), ‘Knowledge of the firm, combinative capabilities, and the replication of technology’, Organization Science, 3 (3), 383–97. Laestadius, Staffan (1995), ‘Tacit knowledge in a low tech firm’, European Journal of Vocational Training, 6, 27–33. Law, John (1987), ‘Technology and heterogeneous engineering’, in Wiebe Bijker, Thomas P. Hughes and Trevor J. Pinch (eds), The Social Construction of Technological Systems, Cambridge, MA: MIT Press, pp. 111–34. Nelson, Richard (1991), ‘Why do firms differ, and how does it matter?’, Strategic Management Journal, 12, 61–74. Nowotny, Helga, Peter Scott and Michael Gibbons (2001), Re-Thinking Science: Knowledge and the Public in an Age of Uncertainty, Cambridge: Polity Press. Penrose, Edith (1959), The Theory of the Growth of the Firm, London: Blackwell. Priem, R.L. and J.E. Butler (2001), ‘Is the resource-based “View” a useful perspective for strategic management research?’, Academy of Management Review, 26 (1), 22–40. Rip, Arie (1997), ‘A cognitive approach to relevance of science’, Social Science Information, 36 (4), 615–40. Schmierl, Klaus and Tobias Kämpf (2004), ‘ “Low-tech companies” – some preliminary remarks on production, knowledge and innovation in low-tech industries’, PILOT Newsletter 2, 19–29, www.pilot-project.org/newsletter/newsletter. html, accessed 15 October 2007. Sørensen, Knut and Robin Williams (eds) (2002), Shaping Technology, Guiding Policy: Concepts, Spaces and Tools, Cheltenham, UK and Northampton, MA, US: Edward Elgar. Teece, David, G. Pisano and A. Shuen (2000), ‘Dynamic capabilities and strategic management’, in Giovanni Dosi, Richard R. Nelson and Sidney G. Winter (eds), The Nature and Dynamics of Organizational Capabilities, Oxford: Oxford University Press, pp. 334–62. Tidd, Joe, John Bessant and Keith Pavitt (2001), Managing Innovation, Chichester: Wiley. Woodward, Joan (1965), Industrial Organization: Theory and Practice, Oxford: Oxford University Press. Woodward, Joan (1970), Industrial Organization: Behaviour and Control, Oxford: Oxford University Press. Zollo, Maurizio and Sidney Winter (2002), ‘Deliberate learning and the evolution of dynamic capabilities’, Organization Science, 13 (3), 339–51.

3.

Standard-setting competition and open innovation in non-HT industries: mechanical engineering and machinery Alexander Gerybadze and André Slowak

THE NEW DYNAMICS OF OPEN INNOVATION IN NON-HIGH-TECH INDUSTRIES This chapter describes the new dynamics of innovation and the complex interrelationship between R&D strategies, inter-firm alliances and standard-setting agreements. In most high-tech as well as in low- and medium-tech industries, new products and services increasingly build on a complex configuration of technologies and complementary assets. Even the largest corporations need to set up cooperative agreements with suppliers, complementary service providers, research centres, regulators and the like in order to manage complex innovations successfully. There is a growing stream of research on open innovation and new forms of alliances. Most of these new studies address case studies and critical issues in high-tech industries, with a strong focus on information technology and the new role of the Internet. In a recent survey on the state-of-the-art in open innovation, Chesbrough (2006b) has emphasized the need to extend this new research programme to the study of non-high-tech industries and to include more case studies on innovation in different European countries. The European research group on the study of innovation in low-tech industries thus provides an ideal platform for the study of new forms of open innovation which are typical of more traditional industries in European countries. Our research focuses on industries that can be considered medium-tech and play a strong role in Europe: mechanical engineering, metal-working, machine-tools, measurement and process control. Products and systems are highly complex and innovation depends on strong linkages between industrial users, suppliers, complementary service providers, education and research. The mechanical engineering and metal cluster1 is very strong in Germany (particularly in the southern part), Switzerland, Italy as well as in 43

44

Innovation in LMT

the Nordic countries (Denmark, Finland, Sweden). Over the years, traditional manufacturers have been able to absorb new technologies that were originally generated in other industries: in information technology (IT), electronics, instrumentation as well as advanced materials. The new dynamics of innovation – the way more traditional industries have absorbed new technologies, but also how they need to develop new business relationships as well as the new strategic value of technical and business standards – may serve as a model for many other important sectors of the European economy.

THE GAME BETWEEN INCUMBENTS AND CHALLENGERS National innovation systems in Europe have not often been very successful in their attempts to create radically new high-tech industries. Other countries – the US as well as some in Asia – have been much more effective in building and sustaining new technology-based industries. European corporations have often followed the strategy of absorbing new technologies and transforming business concepts within well-established industries (as in metal-working, mechanical engineering, transportation, food and beverages and others). Modernizing traditional industries can be described as a game between incumbents and challengers over the restructuring of a ‘market field’. Using the idea of markets as fields requires one to specify what a market is, who the players are, what is meant to be an incumbent and a challenger, and how the social relationships and cultural understandings that come into play create stable fields by solving the main problems of competition and controlling uncertainty. . . . A stable ‘market as field’ means that the main players in a given market are able to reproduce their firms. . . . Incumbent firms are those that dominate a particular market by creating stable relationships with other producers, important suppliers, customers, and the government. They exploit their position by reacting to what other dominant firms are doing. Challenger firms fit into the dominant logic of a stable market, either by finding a spot in the market (i.e. a niche) or imitating dominant firms. (Fligstein, 2001: 17)2

Market fields in Europe seem to operate differently, tending to work more in favor of incumbent firms than in the US where radical innovation is often pursued by challengers and high-tech start-up firms.3 As a result the innovation system in Europe is more ‘stable’ in traditional industries often characterized as low- or medium-tech rather than in new industries. The creation of new, high-tech and research-intensive industries is a pattern more typical of the US innovation system and certain countries in Asia. The most recent EU Industrial R&D Investment Scoreboard has compared investment patterns of the world’s leading 1400 technology corporations.

45

Standard-setting competition and open innovation Pharmaceuticals and biotechnology Automobiles and parts Software and computer servies Aerospace and defence Other EU Non-EU

Per cent 0

17%

11%

20%

10

23%

21%

20

Technology hardware and equipment Electronic and electrical equipment Chemicals Leisure goods

30

7% 3% 6%

14%

40

50

8%

8% 2%

23%

9% 4% 3% 5%

60

70

80

16%

90

100

Percentages express R&D shares of sectors over total R&D. Source:

JRC-IPTS (2007, 17)/European Commission, JRC/DG RTD.

Figure 3.1 Comparison of R&D investment patterns in EU-based corporations and non-EU firms The top ten R&D-intensive industries of the world account for €315 million in 2006, representing 86 per cent of industrial R&D spending. The most dynamic of these top-ten industries are: ● ● ● ●

pharmaceuticals and biotechnology; IT-hardware and equipment; software and computer services; and consumer electronics/electronic leisure goods.

In these four very dynamic high-tech industries, which account for 56 per cent of global corporate R&D spending, very few European-based corporations are really in the forefront of world competition. The innovation race and major R&D investment programmes in these dynamic fields are dominated by North American and Asian corporations. European firms by contrast have remained strong in certain industries in which innovation has more of an incremental and structure-enhancing nature. As Figure 3.1 shows, in comparison to non-EU firms a much greater share of the R&D of European corporations is devoted to: ● ● ●

automobiles and parts (23 in European versus 14 per cent in nonEuropean firms); chemicals excluding life-sciences (6 per cent in European versus 4 per cent in non-European firms); as well as in ‘other sectors’ (23 EU versus 16 per cent non-EU).

46

Innovation in LMT

This group of ‘other sectors’ includes a wide array of low- and mediumtech industries such as metal-working, machinery, plastic and rubber, food and beverages and others. In Europe there is a relatively clear pattern of specialization on industrial activities for which R&D intensities are in the range of 1–4 per cent and which do not account for dynamic growth in R&D spending.4 European incumbents have often concentrated on defending their territory in existing industries which have had to continuously adapt to technological change. This is the characteristic pattern for innovation in automobiles, chemicals, industrial engineering and many more ‘traditional’ sectors (metal/machinery, food, textiles and the like). Major new technologies in many of these traditional industries were often generated in other high-tech and science-driven industries such as computers and software, semiconductors, biotech and new materials. Most challenger firms generating new technology in these dynamic fields were primarily active outside of Europe. As a result, innovation in Europe has often been a game between European incumbent firms in low- and medium-tech industries and their more dynamic challengers in the US and Asia who are at the forefront among high-tech industries. Adapting new technology developed abroad and building on spillovers from high-tech industries can be a successful formula as long as incumbents are able to control critical knowledge downstream and existing strengths in application-engineering and user-support are used to advantage. European firms in industries like machinery, automobiles, chemicals and other traditional sectors have remained strong where they have been able to absorb high-tech breakthroughs from foreign sources into their existing industry and user environment without changing the status hierarchy and the power structure in established business relations.5 As long as the status hierarchy of incumbents has been reproduced European firms have been able to effectively follow ‘successful user’ or adaptor strategies. In many other cases however challenger firms from the US or Asia have had an incentive to become active downstream and to aggressively penetrate markets in more traditional European user industries. Typical examples are digital technology and semiconductors which have radically changed the audio and video industry. European firms which have been traditionally active in these fields have lost out to foreign challengers, as have established pharmaceutical firms that have not been able to adapt fast enough to biotechnology and genetic engineering. Three factors are of critical importance for restructuring traditional industries and supporting the role of incumbents in European end-user markets: 1.

Corporations need to be continuously innovative and be able to absorb key technologies, whether generated in Europe or elsewhere.

Standard-setting competition and open innovation

2.

3.

47

If technologies are generated and controlled by foreign high-tech challengers, European firms need to build on complementary strengths in engineering, logistics, user support and services. A third, very important factor is influence and control of standardsetting activities. Standards tend to stabilize fields and it is critical whether standards are pushed by foreign challengers or whether European incumbents take an active, leading part in the standardsetting process.6

The first two issues have often been addressed in comparative innovation studies. In this chapter we want to emphasize the dynamic relationship between innovation strategy and standard-setting strategies. We focus on a particular combination of industries and their supplier-user relationships, which represent a traditional strength of European firms: industrial engineering and machinery as well as metal-processing. During the last ten years firms in more traditional mechanical-engineering sectors have had to continuously adapt to digital electronics and process control. The process of how incumbent firms in Europe have been able to adapt to new technologies generated by foreign challengers is described in the following pages.

THE ROLE OF STANDARD-SETTING IN LOW- AND MEDIUM-TECH INDUSTRIES Innovation in low- and medium-tech industries is critically dependent on the absorption of new products and technological spillovers from other industries. Absorptive capabilities need to be combined with other related forms of downstream innovation (design, customer integration, complementary services, changing business models). Innovation becomes highly complex and business transactions need to be standardized. Two different types of standard-setting activities have to be followed simultaneously. First, the selection of technology standards for the most advanced and pacing technologies that drive the industry. Second, corporations need to adapt business system standards for their specific market and user environment. While European firms are often successful in controlling their industry and user environment, they often encounter problems in adapting to major new technologies that are pushed upon them by non-European challenger firms. Companies in the machinery industry need to absorb digital controls and software, and these technologies are often developed and controlled by foreign firms that are not interested in highly-fragmented machinery markets. Automobile manufacturers need to absorb semiconductor and software technology, for which the most innovative firms

48

Innovation in LMT Software

High-tech

(15% R&D intensity)

8–10% R&D intensity

Semiconductors (12% R&D intensity)

Automobile electronics and vehicle IT

5–7% R&D intensity

Low-tech

3–4% R&D intensity

Automobile-OEM (Final assembly and system integration)

Automobile suppliers (sub-systems and components)

European strengths

US dominated

Figure 3.2 The transformation of a medium-tech industry: the case of automobiles are often located in the US or Japan (see Figure 3.2). European chemical manufacturers need to absorb new technologies like high throughput experimentation (HTE) or molecular modelling, and these again are often developed by foreign high-tech firms. The critical issue in these cases is: do users of new technologies get the best and most reliable information on the most advanced technologies? Will they be appropriately supported in their downstream innovation activities if they are dependent on foreign suppliers? Will they be able to bet on the ‘right’ standard that is being set by non-European firms? Furthermore, foreign suppliers of critical technologies, components and sub-systems may be inclined to move downstream in order to penetrate European end-user markets. By controlling technology standards, they may see opportunities for influencing the business systems standards downstream. As an example, an ever greater percentage of innovation in the automobile industry is influenced by advances in automobile electronics, vehicle-IT and advanced manufacturing. Technology standards are strongly dependent on foreign semiconductor suppliers, design firms and software companies. Will companies like Microsoft, Oracle, Intel or NEC also try to get a stronger influence on business system standards in the automobile and supplier industry? Or will European incumbents remain strong enough in shaping the standard-setting process for automotive electronics?7

Standard-setting competition and open innovation

49

Figure 3.2 illustrates this transformation process and the new dynamics of innovation and standard-setting in the European automobile industry. Similar conflicts between European incumbents and non-European challengers can be observed in other low-tech industries (for instance food and beverages, paper products and printing) as well as in medium-tech industries (such as machinery and the more traditional electromechanical and chemical products). Business system standards evolve through a mediated process of social and technical engineering and through institution building. The innovation system in Germany during the first half of the twentieth century was shaped by its research universities, by the growth in industrial R&D and by the emergence of the patent system. Advanced R&D and strong patent protection are typical characteristics of high-tech activities. These need to be complemented by what we call high-norm activities, which include: ● ● ●

the development of standard-setting institutions and regulatory bodies; technical societies, engineering communities and industry associations; and innovation marketing and opinion-building through trade fairs, international exhibitions and conferences.

This system of complementary technical and market institutions supporting R&D activities at the public and private level has traditionally been a cornerstone of Germany’s industrial success in mechanical engineering, chemicals and transportation. In most of these areas standard-setting institutions such as the DIN, technical and engineering societies (VDI, VDE and others), as well as leading international trade fairs (such as Achema, the Hanover industrial fair and others) have played a strong role in shaping both technical standards and dominant business standards in their respective industries. This system of technical and social engineering supported the evolution of transportation, mechanical engineering and chemicals and the status hierarchy of German incumbents well into the latter half of the twentieth century. In Neil Fligstein’s words: Differing conditions of market stability produce different kinds of politics. A stable market is defined as a situation in which the identities and status hierarchy of producer firms (i.e. who are the incumbents and who are the challengers) is well known, and a conception of control that guides actors who lead firms is shared. Firms resemble one another in tactics and organizational structure. Politics reproduce the position of the advantaged groups. (Fligstein 2001: 76).

The status hierarchy became undermined when German firms and research institutions were no longer in the ‘driver’s seat’ of standardization.

50

Innovation in LMT

During the 1980s and 1990s the German innovation system had to adapt to a new wave of high-technology in which leading-edge developments no longer came from German research institutions. Investment programmes of German corporations as well as government R&D support programmes in computers, semiconductors and biotechnology often had to be considered failures. These difficulties may be explained not so much by limited R&D capabilities, but more by shortcomings and inappropriate judgments in the standard-setting process. German corporations encountered great difficulties in innovation projects which were both high-tech and highnorm. As high-norm we define products and activities whose big commercial success depends on the ability to influence a common world standard. If the world standard for a new product is influenced and defined by foreign corporations, institutions and opinion leaders with just peripheral network relationships to agents in Germany, it will be very difficult to participate successfully, even with strong research capabilities. R&D activities and standard-setting activities must be effectively synchronized in order to be successful in the global innovation race. More recently German firms in low- and medium-tech industries seem to have learned from failures observed during the 1980s and 1990s. Particularly in industries in which German firms had a traditionally strong position, and business system standards are shaped through the engineering and user system, we have observed successful examples of sustainable innovation at the interface between high-tech and low- and medium-tech: ●

● ● ●

Innovation in machinery, mechatronics and the adaptation of field bus systems to mechanical engineering markets (– a very interesting example presented in detail below). New combinations of automobile technology, electronics and embedded software (case examples studied at our research centre). Chemical engineering in combination with paint, coatings, surface treatments, process automation and advanced materials. Finally, the combination of low- and high-tech in the creation of business-system norms is very effective in housing and construction, environmental technology and new energy systems.

CASE STUDY: EUROPEAN STANDARDS IN THE MACHINERY SECTOR AND THE CREATION OF PROFIBUS European firms traditionally have strength in mechanical engineering and machine-tools, and these have been increasingly affected by the development

Standard-setting competition and open innovation

51

of digital controls/control systems and data communication. Standardization processes for data communication lines in and between factories, the so-called field bus standards, became increasingly important during the 1990s. However, manufacturers in different European countries followed different solutions and this resulted in the European ‘field bus war’. Until 1995 there was head-on competition between the German concept PROFIBUS and the French counterpart field bus FIP. These ‘standards battles’ weakened the European influence on International Electrotechnical Commission (IEC) standardsetting attempts in industrial automation in favour of the United States, respectively ISA/ANSI.8 Challenged by their American counterparts, the dominant European field bus manufacturers agreed on a common European norm at the European body CENELEC in 1995–8. Instead of finding a compromise, all national standards were compiled (see EN 50170, 50254, 50295 and 50325), so that the related EN norms and subsequent IEC norm 61158 are just a summary of the different field buses used in the market. Finally in 1999 the Fieldbus Foundation, Fisher Rosemount, ControlNet International, Rockwell Automation, the PROFIBUS User Organization (PROFIBUS Nutzerorganisation e.V., Karlsruhe, Germany), and Siemens signed a ‘Memorandum of Understanding’ within the IEC to terminate the field bus wars. Since then field bus associations have been searching for more diplomatic, indirect ways to cultivate their community around their field bus technology and to set international standards by market penetration instead of blocking each other in the standard-setting committees (Felser, 2002a, 2002b). Mechanical engineering, industrial automation and electronics are increasingly integrated into complex mechatronic systems. As the cell level9 of manufacturing and process plants gains new functionalities, the traditionally different segments of automation controls, industrial communication, and motion control and drives need to become more integrated. Innovations in each of the once distinct industry segments may now enhance technologies in the other segments. Traditional R&D approaches hence do not allow for the new variety and complexity of automation products and advanced automation concepts.10 Firms and research institutions need to build up alliances in order to manage technological integration processes and agree on common standards and protocols. New forms of standard-setting partnerships emerge which try to win a global community of supporters and technology developers out of which will hopefully result one dominant global standard. Essential characteristics of the standard-setting dynamics and the innovation process in our case study are: ●

First, the diverse array of capabilities needed (engineering, factory construction and operation, maintenance and troubleshooting, IT

52

Innovation in LMT

●

●

and software, electronics, industrial services) builds on global communities of expertise that need to be included in the partnerships. Second, we observe a shift from conventional field buses to Ethernet buses, and this increases market complexity and opens a new standard-setting race. Asian vendors are participating in the IEC Ethernet real time specification process (see Sauter and Felser, 2004) and ASIC manufacturers (like NEC for PROFINET11) are gaining stronger influence on future factory automation technology. As a third factor, global industrial customers need to implement and procure field bus solutions on a global scale and thus demand a global technology certification (preferably through IEC) for their products and plant-related services. If relying on proprietary strategies, single firms cannot have the necessary influence on the IEC that a field bus association is able to have.

These three factors will lead to a substitution of national or regionally centred standards by emerging global standards. The key issue is: Who will influence the formation of the global standard? Can national or European coalitions of firms, research and standardization bodies be strong enough to influence global standards? The following section concentrates on the evolution of the PROFIBUS standard, which emerged from a joint industry project in 1987–90 supported by the German Federal Ministry of Research and Technology. The project was accomplished by German firms and universities with strong engineering and IT capabilities, in particular Siemens, Bosch, Klöckner & Moeller, the RWTH Aachen and the Karlsruhe Research Center for Information Technologies (FZI). Fourteen companies and five research institutes were involved (Hoffmann, 1988; Felser, 2000). In 1989 the PROFIBUS user organization (PNO) was founded in order to augment and to maintain the technology designed in the above research project. PROFIBUS International (PI) was set up in 1995 and has since had a continuously growing number of members. Recently PROFIBUS International has emerged as one of the largest field bus organizations worldwide with several regional chapters (so-called RPAs) in major countries in the Triad regions. Today, PROFIBUS/PROFINET is represented by PROFIBUS and PROFINET International (PI, responsible for user support and technology-marketing), and the PROFIBUS User Organization (PNO, responsible for enhancing technology and influencing coherent international standards). PROFIBUS’ market position is predominant in factory automation. In recent years it has also won a significant world market share in process automation. Here it is used in chemical engineering, refineries and other

Standard-setting competition and open innovation

53

process industries. In November 2004 the German Automobile Manufacturers’ Initiative (AIDA) – AUDI, BMW, Daimler-Chrysler and Volkswagen – decided to implement PROFINET in their assembly plants. Other non-factoring applications were also penetrated.12 These developments indicate that PROFIBUS/PROFINET in combination has a great chance of becoming a world standard. Furthermore the PNOmember pool covers the full automation value chain. Several PNO members are either leading industry firms, highly specialized companies, or well-known engineering institutions, for example. Together they represent a kind of ecology of firms from specialized industry domains as follows: ●

● ● ● ●

●

● ●

Solution providers offering automation systems and integrated automation solutions (factory automation: Siemens; process automation: in particular ABB, but also Endress  Hauser Process Solutions, Yokogawa). Strong industry users like BASF or Shell and DEA Oil. ASIC manufacturers, necessary to running PROFINET in real time (NEC, HMS, Hilscher, profichip). Field bus competitors (Mitsubishi Electric, OMRON, Phoenix Contact, Schneider Electric). Prestigious German applied-research institutions from the field of engineering and IT as well as university research in industrial automation. Technical specialists, in particular in the fields of motion control and drives (Festo, Bosch Rexroth and others); components; robotics (FANUC, KUKA); safety in motion control and embedded systems, as well as industrial software. Test, diagnosis and plant control (such as Softing, Bürkert Fluid Control). Firms focusing on new automation approaches like Linux/Open Source-based automation; wireless identification systems (such as RFID), or the implementation of industrial Ethernet.

The PNO is registered as an association according to German law (eingetragener Verein) and thus can admit both legal entities and natural persons as members. The association statutes defines members as hardware, software and system suppliers; industrial solution providers; industrial users; research institutes as well as industry associations. Members or competitors who purposely impede PNO’s technical or standard-setting progress can be excluded. The firms’ member fees are differentiated according to member status.13

54

Innovation in LMT

OPEN INNOVATION, SEMI-OPEN INNOVATION AND NEW FORMS OF LAYERED ORGANIZATIONS In order to influence and enhance the de facto standard PROFIBUS/ PROFINET, the PNO organizes around 50 working groups specialized in different technical domains (‘Technical Committees’). The organization is governed by the Board of Directors (BoD) and the Advisory Board (AB). PI members outside the PNO (the PNO at the same time constitutes the German Regional Association, RPA) are informed and trained about new PNO specifications via their regional association. The PNO Business Office processes the financial controlling of WG projects and assists delegated members participating in standard-setting bodies, in particular the IEC. It furthermore manages the global review and feedback process. PNO proposals have to undergo approval by PROFIBUS and PROFINET International, which also executes the PNO ‘call for experts’ to establish new working groups. In addition, it provides the necessary infrastructure for the working group and for marketing activities. The Technical Committees coordinate specification work done in Working Groups (WGs, JWGs and ad hoc WGs). Disclosure and licensing terms are regulated by the PNO IPR Policy (2006). In general, members are obliged to disclose ‘relevant’14 intellectual property rights (IPR) and to license them free of royalty fees within PROFIBUS/PROFINET specifications. Hence although disclosure does not imply that a member passes on IPR, relevant IPR turns into an association good. In exceptional cases however a member may ask the AB for permission to license relevant IPR to the PNO under RAND-terms. Furthermore, PNO rights to use a member’s relevant IPR on a royalty-free basis do not terminate if the member resigns from the PNO or sells this IPR. This novel form of international alliance represents an intermediary or hybrid form between open and closed innovation. So far the literature distinguishes between extreme forms of ‘open’ and ‘closed innovation’ (Chesbrough, 2003, Chesbrough et al, 2006). We prefer to call it ‘semiopen’, in the sense that some areas of expertise are proprietary and exclusive, while others are accessible to different groups of members. The PNO organization is a very interesting example of such a novel form of semiopen standard-setting partnership. Figure 3.3 outlines the major characteristics of semi-open innovation processes in the lower section. These may be differentiated from closed innovation processes which can be observed in pharmaceuticals or automotive manufacturing. IPRs (in particular patent rights) are formally arranged, and intellectual property regimes in these industries are very

Standard-setting competition and open innovation

55

PI International • No innovation activities for the PNO • Thus no influence on innovation agenda and PNO strategy • Standard-setting involvement limited to feedback to PNO specification drafts

Complementary, proprietary member assets PROFIBUS and PROFINET International PNO members PNO core

PNO core • Members of executive or advisory board, committee chairs • Mapping out PNO strategy • Unrestricted access to all PROFIBUS/PROFINET relevant intellectual property

Other PNO members • Control relevant sub-task or component • Access to such intellectual property relevant to the particular working group • No direct involvement in mapping out and auditing PNO strategy

Figure 3.3 Layered model of innovation and standard-setting in collaborative partnership in the PROFIBUS/PROFINET case

56

Innovation in LMT

Table 3.1 Comparison of open innovation, closed innovation and semi-open innovation processes Innovation paradigm ●

Open innovation processes

Characteristics ● ● ● ●

●

Closed innovation processes

● ● ● ●

●

Semi-open innovation processes

●

●

●

Know-how as a public good Activities coram publico Open access/participation Full open source Proprietary IPR policies In-house controlling rights Strict access Closed innovation Business and technical domains broken down into activity domains Particular activities are open to a broad circle of members Others are accessible to few core members only

strong. By contrast open innovation processes are characterized by highly interdependent competence fields and overlapping industries. The firms involved must draw on technologies and know-how owned by other companies and research institutions. Sometimes there are also public spheres of relevant know-how (‘public knowledge’). IPRs are either unspecified or the use of know-how by third parties is subject to the granting of licenses (in extremis royalty-free licenses like the ‘Open-Source General Public License’). Today, open innovation in particular is encouraged by the ‘OpenSource’ movement. It has been well described for the software industry, but can also be observed in various other industries.15 Many recent joint industry projects and standard-setting partnerships display mixed structures of open and closed innovation in which particular domains are open but others closed by strict proprietary policies. Companies establish partnerships to break down business and technical domains into activity domains and knowledge fields. Particular technologies and know-how domains are open for a more extensive group of firms, others are accessible for few core members only. Hence we differentiate between different layers of participation, whereas entry rules, IPRs and rights of disposal are specified for each layer. Innovation targets within the PNO are set and driven by the Board of Directors. It decides about the appointment of Working Groups and WG leaders, the creation of projects with external partners (so-called ‘Joint

57

Standard-setting competition and open innovation

Standardsetting authority

Regional/countryspecific chapters

Global standard-setting community

Multinational firms

Administrating entity

Application-specific user communities

Figure 3.4 Organizational form and the transnational network for a standard-setting partnership Working Groups’) and the appointment of temporary projects (‘ad hoc WGs’). The BoD also proposes the appointment of Technical Committees (TCs) and TC leaders, as well as the closure of a TC or WG. TC proposals need the approval of the Advisory Board (AB) on a caseby-case basis (PNO TC/WG Guidelines 2006). The TC/WG Guidelines furthermore specify how WG work is to be coordinated. Financing is provided on a voluntary basis by the WG forming a project consortium. Within the PNO/PI organization the PNO integrates and develops new technologies for the PI organization as well as initiates PROFIBUS/PROFINET standardsetting activities within the IEC. PNO members develop additional proprietary functionalities and services to enrich the PROFIBUS standard in their business solutions (as Siemens does within SIMATIC solutions). The different PNO/PI layers of power or decision rights are illustrated in Figure 3.4. Different Working-Group members write the PNO specification wording chapter by chapter on a voluntary basis. Generally acceptable phrases have previously been discussed, but not yet been formally documented. Text modules are finalized via vote; they can be accepted, neglected or alternated. The WG further proposes the specification parts to be taken into the international norm processes. All standard-setting activities are coordinated by TC5 (recent chair: Siemens) – WG3, Standardization Strategy (chair: PNO Business Office). The most influential members can be grouped as those defining and auditing the overall PNO/PI strategy. They can discipline members, and they set the innovation agenda. The PNO core16 thus consists of the PNO Board of Directors (BoD), the Advisory Board (AB), the Business Office and the PNO TC chairmen. BoD members are Siemens (BoD chairman); Endress +

58

Innovation in LMT

Hauser Process Solutions; and the TU Munich, Institute for Information Technology (ITM). Elected AB members actually are strong engineering firms from the segment motion control and drives (SICK, FESTO); components and sensors (WAGO Kontakttechnik, Pepperl  Fuchs); and field bus competitors (Mitsubishi Electric, Phoenix Contact).17 SIEMENS, Endress  Hauser Process Solutions and ABB belong to the group of the TC chair companies. In contrast to the PNO, WGs do not set the PROFIBUS/PROFINET policy, although their work in progress shapes the PNO strategy. Nonetheless PNO members are involved in PROFIBUS/PROFINET innovation processes and standard-setting at the IEC. They thus gain a temporary leadtime advantage vis-à-vis PI members and can influence WG outcomes to the advantage of their product and service portfolio. The PNO facilitates and stimulates coordinated collective action of crucial German industry players which is necessary for giving PROFIBUS global momentum. Such German success stories are also observed within automotive electronics (in particular CAN and FlexRay)18 and depend on the ability of strong industry players to form tailored organization structures or a standard-setting and innovation partnership around a technology, as well as on astute pooling and configuration of intellectual property.19 Furthermore, the PNO has established a global network to diffuse its technology: PROFIBUS and PROFINET International (PI) members are informed about new IEC, PROFIBUS/PROFINET-relevant standards (via PI marketing documents like newsletters) and pre-final WG documents (via the review process). They can seek support and training from the PNO. Apart from that, they can build networks via regional PI associations, market themselves as PI competence centers, and certificate their products for PROFIBUS/PROFINET. However PI members from outside the PNO (and thus, non-German firms) are usually not involved in the innovation process. Additionally, proprietary single-firm activities provide such product-embedded technical functionalities, but also services which either bundle the different components for integrated automation solutions, or which establish niche markets (such as plant trouble-shooters). These assets are thus complementary to the goods produced and diffused within the PNO-PI organization.

CONCLUSIONS: SUCCESS FACTORS IN IMPLEMENTING STANDARD-SETTING PARTNERSHIPS In recent years European firms have taken a more active role in managing innovation and influencing the standard-setting process. We observe a new

Standard-setting competition and open innovation

59

form of partnership which goes beyond the traditional form of precompetitive, collaborative R&D. Firms work together in joint R&D projects, but their common purpose is not primarily R&D work: they are driven by their desire to have substantial influence on a strong standard. They work together for a new technology and want to develop products or subsystems that work smoothly together and are compatible with the dominant design in a particular market. A major part of collaborative work is directed towards a business system standard that has a high likelihood of becoming a strong world standard. Often participating firms mention that the objective of agreeing/achieving a common standard is more important than the purpose of shared R&D work. If this is the case, we should call this type of cooperation a standard-setting partnership or standard-setting community. Our research centre has studied standard-setting partnership agreements in a number of industries including: ●

●

● ●

standard-formation in mechanical engineering and machinery including PROFIBUS, the application of industrial Ethernet and PROFINET (which has just been described in the preceding sections); standard-setting in housing and construction, including the development of standards for electronics in buildings and for new sanitary equipment; standard-setting in industrial coatings and paints and the application of new surface-treatment technologies to traditional sectors; and finally, we have studied standard-setting agreements in automotive electronics such as CAN, FlexRay, MOST and AUTOSAR.

In most of these cases European firms – often from a low- or mediumtech environment – have become quite influential in defining application standards and elaborating new ‘business-system standards’ later adopted by other firms worldwide. The study of standard-setting activities in different industries has displayed certain patterns and ‘success formulas’. The following organizational factors may be considered important elements of a coherent standard-setting strategy. Often some firms are more active and better informed than other participating firms. These core members of a standard-setting partnership pursue a consistent strategy emphasizing the following ‘building blocks’. 1.

In a first step, core partners explore a basic concept that is most likely to result in an accepted standard. This step may be initiated by a joint study group.

60

2.

3.

4.

5.

Innovation in LMT

The second and most critical step requires involving a lead customer who signals strong commitment. These signals induce other firms and downstream ‘complementors’ to invest in widespread application. This triggers demand-side network externalities and fosters the emergence of a common standard. As a result of growing market opportunities, suppliers and upstream complementors, in a third step, invest in complementary components, infrastructure, equipment, tools, services and the like. A powerful business system emerges and this further strengthens the emerging common standard. A most critical part in this standard-setting exercise which is often underestimated involves stringent intellectual property management. Knowledge and intellectual property ownership (including patents, trademarks, copyrights and trade secrets) is distributed among several partner firms and therefore it is very important to find a workable agreement on knowledge management among them.20 In a final step, members in a standard-setting partnership have to find the appropriate form of organization. There may be different phases of the partnership and the organization must be flexible enough to serve as a framework. In each phase partners have to agree on membership criteria and on the roles, responsibilities and access rules for its members.

In the case of PROFIBUS-Organization, the following type of organization was selected. A global standard-setting community has its headquarters in Karlsruhe, Germany and serves as an administrating entity and controls the major rules for the PROFIBUS standard. Its members are multinational firms that have agreed to comply with this standard. In addition, there are application-specific user communities which are primarily interested in adapting the standard to a particular user environment (such as for woodworking machinery). The international network is organized into a series of regional and country-specific chapters (such as the PROFIBUS user organization in Japan). These country-specific and application-specific subgroups organize workshops, learning groups, conferences and other events. Together this serves as a powerful transnational organization to make sure that the business system standard is supported worldwide.

NOTES 1.

An excellent description of this metals and machinery cluster and its particular strengths in different regions of Europe can be found in OECD (2001: 283).

Standard-setting competition and open innovation 2. 3. 4.

5.

6. 7. 8.

9. 10.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

61

Neil Fligstein’s theory of global markets and the persistent game between incumbents and challengers is a fascinating modern version of Schumpeter’s theory of creative destruction. See Christensen (1997), who emphasizes the role of the challenger firm, on the innovator’s dilemma. By ‘dynamic growth in R&D’ spending we mean average annual increases in sectoral R&D spending considerably above GDP growth. In pharmaceuticals and biotech as well as in IT hardware and software we encounter annual growth rates of 13–15 per cent in R&D spending. The concept of a ‘status hierarchy’ is based on Podolny’s (1993, 2005) sociological studies of dynamic market competition. The game between incumbents and challengers, and the question whether status hierarchies are replicated or overturned, is addressed in Fligstein (2001). See Fligstein (2001, 35ff) on the role of standardization in global competition and on the influence of standardization on the stability of markets. See Gerybadze (2007) for a more recent analysis of standard-setting partnerships in automotive electronics. The International Electrotechnical Commission (IEC) and the European Committee for Electrotechnical Standardization (CENELEC) are formal standard-setting bodies in the areas of electrical engineering, electronics and partly in information technologies. The IEC prepares global standards, whereas CENELEC represents the European Economic Community and EFTA countries. IEC is a specialized affiliate of the ISO, and CENELEC, of CEN. The American National Standards Institute (ANSI), a full member, represents the United States at the IEC. Within the field of automation technologies, the Instrumentation, Systems, and Automation Society (ISA) and ANSI often prepare joint standard drafts for IEC procedures. Engineering disciplines separate an automation system into the plant, cell and field level. A cell is a group, for example, of robots and machines. The field level then contents these single devices, logic controllers, robots and machines. For example, see PROFINET ‘CBA’, where all functionalities shall be stored in a kind of automation plant library, so that various machines and cells in the entire field bus network can access and apply them. Approaches with such a high degree of integration are much too complex to be developed by a single company’s proprietary field bus solution. Strong communities are needed to provide a rich technology platform and complementary assets – thus big firms as well as small entrepreneurs. PROFINET is the PNO Ethernet standard developed from PROFIBUS. The Ethernet approach combines Internet standards with field bus specific standards and real time hardware (ASICs). As an example, PNO works on the implementation of PROFINET in passenger trains. Yearly fees range from €1250 (member category G) to €15 000 (member category A). Industrial users and nonprofit research institutes, however many employees they have, pay annual fees of €1250 and €625 respectively. ‘Relevant’ according to PNO IPR Policy (2006) are IPRs which are necessary to implement a PNO specification within a product or process. Examples are Linux-based automation and embedded systems, or digital commons in music and art. The above layer definitions and the term ‘PNO core’ are the authors’ empirical, casestudy-based conceptualizations and not PNO/PI terminology. Furthermore, TC chairs are automatic members of the AB. For case examples see also Gerybadze (2007). The latter covers the disclosure of IP to the core members, access to IP and rights to its use, but also clear rules concerning the ownership and use of IP collectively generated within WG projects. This is also a core element in European R&D consortia, and there is a special regulation on IP-management for the EU-Framework programmes.

62

Innovation in LMT

REFERENCES Blind, K. (2006), ‘Explanatory factors for participation in formal standardisation process: empirical evidence at firm level’, Journal of Economic Innovation and New Technology, 15 (2), 157–70. Chesbrough, Henry (2003), Open Innovation: The New Imperative for Creating and Profiting from Technology, Boston, MA: Harvard Business School Press. Chesbrough, Henry (2006a), Open Business Models: How to Thrive in the New Innovation Landscape, Boston, MA: Harvard Business School Press. Chesbrough, Henry (2006b), ‘Open innovation: a new paradigm for understanding industrial innovation’, in Henry Chesbrough, Wim Vanhaverbeke and Joel West (eds), Open Innovation: Researching a New Paradigm, Oxford and New York: Oxford University Press, pp. 1–12. Chesbrough, Henry, Wim Vanhaverbeke and Joel West (eds) (2006), Open Innovation: Researching a New Paradigm, Oxford and New York: Oxford University Press. Christenson, C.M. (1997), The Innovator’s Dilemma. When New Technologies Cause Great Firms to Fail, Boston, MA: Harvard Business School Press. Felser, M. (2000), ‘PROFIBUS international: status und geschichte’, Neue Technik, 4/2000, Mellingen: Diagonal Verlag, accessed at http://felser.ch/download/FETR-0002.pdf. Felser, M. (2002a), ‘The fieldbus standard: history and structure’, Technology Leadership Day 2002, MICROSWISS Network, HTA, Lucerne, Switzerland, 10 October, accessed at http://felser.ch/download/FE-TR-0205.pdf. Felser, M. (2002b), ‘Vom Feldbus-Krieg zur Feldbus-Koexistenz’, Bulletin des Schweizerischen Elektrotechnischen Vereins (SEV), Fehraltorf, Switzerland, 9/2002, pp. 24–6, accessed at http://felser.ch/download/FE-TR-0202.pdf. Fligstein, Neil (2001), The Architecture of Markets. An Economic Sociology of Twenty-First Century Capitalist Societies, Princeton, NJ and Oxford: Princeton University Press. Gerybadze, Alexander (2007), ‘Innovationspartnerschaften, Patentpools und Standardsetzungsgemeinschaften: Verteilung und Zuteilung der Rechte und neue Organisationsformen’, in Hoffmann-Riem, Wolfgang and Martin Eifert (eds), Geistiges Eigentum und Innovation, Baden-Baden: Nomos. Grindley, Peter C. (1995), Standards, Strategy and Policy. Cases and Stories, Oxford and New York: Oxford University Press. Hoffmann, H. (1988), Das BMFT-Verbundprojekt ‘Feldbus’, Automatisierungstechnische Praxis, 5, 212–16. JRC-IPTS (2007), ‘EU R&D investment scoreboard’, Seville, Spain, accessed at http://iri.jrc.ec.europa.eu/research/docs/2007/sb_2007.pdf. PNO IPR Policy (2006), V 1.0, 8 December. PNO TC/WG Guidelines (2006), V 3.0, 8 December. OECD (2001), Innovation Clusters, Paris: Organization for Cooperation and Developments. Podolny, J.M. (1993), ‘A status-based model of market competition’, American Journal of Sociology, 98 (4), 829–72. Podolny Joel M. (2005), Status Signals. A Sociological Study of Market Competition, Princeton, NJ and Oxford: Princeton University Press. Sauter, T. and M. Felser (2004), ‘Standardization of industrial ethernet: the next battlefield?’, 5th IEEE International Workshop on Factory Communications

Standard-setting competition and open innovation

63

Systems (WFCS2004), 22–4 September, Vienna, Austria, pp. 413–21, accessed at http://felser.ch/download/FE-TR-0403.pdf. Shapira, C. and H.R. Varian (1999), ‘The art of standards war’, California Management Review, 41 (2), Winter 1999, 8–32. Simcoe, Tim S. (2006), ‘Open standards and intellectual property rights’, in Henry Chesbrough, Wim Vanhaverbeke and Joel West (eds), Open Innovation Researching a New Paradigm, Oxford and New York: Oxford University Press, pp. 161–83.

4.

The moral economy of technology indicators Benoît Godin

INTRODUCTION In 1939 the British left-wing scientist J.D. Bernal suggested a type of measurement that became the main indicator of science and technology: the research budget as a percentage of the national income (Bernal, 1939). He compared the UK’s performance with that of the United States and the USSR, and suggested that Great Britain should devote between 0.5 per cent and 1 per cent of its national income to research. The number was arrived at by comparing expenditures in other countries, among them the United States, which invested 0.6 per cent, and the Soviet Union which invested 0.8 per cent, while Great Britain spent only 0.1 per cent. This certainly seems a very low percentage and at least it could be said that any increase up to tenfold of the expenditure on science would not notably interfere with the immediate consumption of the community; as it is it represents only 3 per cent of what is spent on tobacco, 2 per cent of what is spent on drink, and 1 percent of what is spent on gambling in the country. (Bernal, 1939: 64)1 The scale of expenditure on science is probably less than one-tenth of what would be reasonable and desirable in any civilized country. (Bernal, 1939: 65)

Bernal was soon followed by many other organizations, among them the US President’s Scientific Research Board in a survey on government R&D. In 1947, the Board introduced into science policy the indicator first suggested by Bernal, which is still used by governments today: R&D expenditures as a percentage of national income (US PSRB 1947). Unlike Bernal however, the Board did not explain how it arrived at a 1 per cent goal for 1957. Nevertheless, President Truman subsequently incorporated this objective into his address to the American Association for the Advancement of Science (AAAS) in 1948. The origin of such ratios is probably a very early calculation made by British economist L. Levi (1869). Using data from a circular sent to British 64

The moral economy of technology indicators

65

scientific societies, Levi computed a ratio of incomes of scientific societies to national income of 0.04 per cent. Another such calculation before Bernal was that of E.B. Rosa, chief scientist at the US Bureau of Standards. In 1920, Rosa compiled, for the first time in American history, a government budget for ‘research-education-development’ (Rosa, 1921). Rosa estimated that the government’s expenditures on research amounted to 1 per cent of the federal budget. In the following year, J.M. Cattell, editor of Science, would use the ratio (1 per cent) in his crusade for the advancement of science (Cattell, 1922). In the next decades, variants of the ratio took on names like research intensity, then technology intensity, then high technology. This chapter documents the construction of the high-technology concept and its statistics. The first part summarizes the history of the concept, as given in Godin (2004). To this history, the chapter adds the story of the origin of the classification of the technology indicator into three levels: high, medium and low. The chapter then discusses the moral economy behind the concept, or its ‘aperspectival’ qualities. It identifies six virtues the indicator possesses, and which make it a popular indicator. In light of the recent literature on innovation and its new moral economy, the author concludes by predicting the death of the indicator in the near future.

A VERY BASIC RATIO The simplest indicator of high technology is constructed by dividing industrial R&D expenditures by production (that is value added, turnover or sales) and then classifying industries according to this ratio. As R.N. Anthony, author of an influential survey on industrial R&D, once wrote: ‘Use of this ratio implies that there is some relationship between research spending and sales; to the extent that sales is a measure of the size of the company, this implication is in general warranted’ (Anthony and Day, 1952: 295). The indicator has precursors that go back to the 1930s. For decades managers have constructed ‘return on investment’ (ROI) ratios in order to evaluate their investments. Very early on, these ratios came to be applied to R&D activities (Chandler, 1977; Johnson, 1978; Johnson and Kaplan, 1987; Hounshell and Smith, 1988). By the 1950s most companies calculated ratios like R&D as a percentage of earnings, as a percentage of sales, or as a percentage of value-added (Olsen, 1948; Abrams, 1951; Anthony and Day, 1952: 286–300; Quinn, 1960), and a whole ‘industry’ developed around studying the ‘effectiveness’ of research in such terms (Hogan, 1950; Pelz, 1956; Quinn, 1959; Kaplan, 1960; Institution of Chemical Engineers, 1963; Lipetz, 1965; Seiler, 1965; Yovits et al., 1966; Pelz and Andrews, 1966; Dean, 1968). Very few administrative decisions actually did rely automatically on

66

Innovation in LMT

metrics (NSF, 1956a; Rubenstein, 1957; Seeber, 1964), but it was not long before the ratios came to be applied to aggregated statistics on national R&D expenditures (Bernal, 1939; Ewell, 1955; NSF, 1956a) and industrial R&D (Sherman, 1941; Compton, 1941; US National Association of Manufacturers, 1949; US Bureau of Labor Statistics, 1953a, 1953b; Dearborn et al., 1953; NSF, 1956b; Federation of British Industries, 1952; DSIR, 1958; Canadian Dominion Bureau of Statistics, 1956). The US National Research Council (NRC) conducted the first such industrial analysis among the industrialized countries. In 1933, its Division of Engineering and Industrial Research tried to assess the effect of the Great Depression on industrial laboratories (Holland and Spraragen, 1933). The report classified companies according to whether they spent over 10 per cent of sales revenue on R&D, 5–10 per cent of sales, 1–5 per cent, or under 1 per cent. With the data in hand, the authors concluded: ‘it appears that those companies the products of which more nearly approach the classification of raw materials spent a smaller percentage of their sales income for research than the companies in which products are of a highly manufactured character’. Similar ratios in the United States were produced by or with the National Association of Manufacturers (NRPB, 1941: 124; NAM, 1949: 3), the US Bureau of Labor Statistics (1953b: 12–13); US Bureau of Labor Statistics (1953a: 12–13, 1953b: 26–9); Dearborn et al., (1953: 29s), and the National Science Foundation (NSF) who related R&D expenditures to sales – and tentatively to assets (NSF, 1956a, 1953: 33s), then to value added. From an analytical point of view, the statistic served to assess and compare the relative efforts of industries in terms of R&D, and to look at the impact of R&D on industries’ economic performances. The message was to influence policies supporting R&D, particularly in big firms that invest more than others. A totally different ‘rhetoric’ accompanied the indicator for high technology.

VARIATIONS ON A THEME What characterized the construction of the high-technology indicator was that a specific ‘rhetoric’ came to be associated with the statistic. First, labels were associated with the ratio of R&D expenditures to sales: research intensity, technology intensity, high technology. Second, the indicator was developed and increasingly used in the context of debates on the competitiveness of countries. The United States was at the origin of the rhetoric, and the OECD was at the heart of the indicator’s worldwide dissemination.

The moral economy of technology indicators

67

Research Intensity In 1958, E. Hoffmeyer coined the term ‘research intensity’ to talk about the performances of industries in terms of R&D effort (Hoffmeyer, 1958). The balance of payments deficit in the United States,2 but also in Europe,3 was the context in which Hoffmeyer published his analysis of US foreign trade over the twentieth century. Using the NSF ratios of R&D expenditures to sales, Hoffmeyer looked at the structure of US foreign trade: 11 industries were classified into four groups according to their research effort or intensity.4 With the data, Hoffmeyer showed that the United States had a competitive advantage in the research-intensive industries. To the best of my knowledge, Hoffmeyer was the first to use this term ‘research intensity’, as well as the first to use it in the context of an analysis of international trade.5 He would soon be imitated worldwide, first of all at the OECD. In 1963, the OECD published a study it had (later) also presented to the first ministerial meeting on science. The study, written by C. Freeman, R. Poignant and I. Svennilson, was the result of the OECD’s early research programme on the economics of science. Using available statistics, the authors looked at industrial R&D, and constructed three industry groups classified according to the ratio of R&D expenditures to sales (OECD, 1963: 81).6 The first group (Group A) was called research-intensive industries and was composed of aircraft, vehicles, electronics, other electrical, machinery, instruments, and chemicals. The study determined that ‘all the industrial countries considered show over two-thirds (the United States and the United Kingdom over nine-tenths) of their industrial R&D expenditure in Group A which comprises the research-intensive industries’ (p. 30). To the authors, research-intensive industries had several characteristics that made them valuable from the point of view of policies: (a) they were generally the fastest-growing industries (p. 29); (b) their share of world trade was growing (p. 32); and (c) they had the highest balance of technological payments (p. 33). The use of such a term was part of the OECD campaign for science policies. In fact, in the early 1960s, the OECD was campaigning to convince governments to develop science policies and to set up ministries to this end (Godin, 2005). Thus, research-intensive industries were a phenomenon that policies should work on, but also a symbol with ‘rhetorical’ overtones that precisely fitted the OECD’s efforts to convince officials to bring their countries into the modern economy. The R&D intensity ratio would be calculated regularly in the following decades, particularly in OECD International Statistical Year studies and analyses on trends in R&D (Godin, 2005).

68

Innovation in LMT

A few years after the 1963 report, the OECD took part in a second campaign, this one calling for closure of the technological gaps between European countries and the United States. It confirmed earlier findings of the organization. Research-intensive groups accounted for a large share of manufacturing exports, and the Americans had the lead over Europe, followed by the medium-sized countries and then the smaller industrialized OECD-member countries: ‘there is a high concentration of United States exports in those product groups where there is a high concentration of R&D effort’ (p. 255). However, the OECD calculated that the share of the United States in OECD exports had declined from 23.7 per cent in 1962 to 21.3 per cent in 1966 as a result of catching-up by other OECD countries. Technology Intensity It was precisely in the context of issues on international competitiveness that a new term – ‘technology intensity’ – appeared to describe the same phenomenon. In the course of the debate on the technological gaps between Europe and the United States, the US government set up an interdepartmental committee to study the issue (Godin, 2002). At the request of the (Hornig) committee, the Department of Commerce conducted one of the first surveys of American investments and operations in Europe. To the best of my knowledge, it is this survey that coined the term technologyintensive industries to document the structure of US direct investment in Western Europe (US Interdepartmental Committee on the Technological Gaps, 1967). According to the committee, 80 per cent of all US direct investments in manufacturing in Western Europe was in technology-intensive industries, and Americans controlled large segments of the market in such technology-intensive products as computers. Then, M.T. Boretsky, director of the Technological Gap Study Program (1967–9) at the Department of Commerce, introduced a variant to the indicator: technology-intensive products (although he used industries as the units, not products) (Boretsky, 1971, 1973, 1975; Science, 1971). Boretsky showed that the United States was in danger of losing its preeminence in advanced technologies, particularly those that are important in world trade. American exports of technology-intensive manufactured products were leveling off. This was so mainly because of the narrowing of the gap with other OECD countries, and because of faster growth rates in these countries. The Department of Commerce continued to develop and improve the indicator in the following years (Kelly, 1976, 1977; Davis, 1982, 1988), and used the data to document America’s competitiveness (US Department of Commerce, 1983; Hatter, 1985). In fact, the 1980s was a time when the US

The moral economy of technology indicators

69

government became obsessed with (Japan and) international competitiveness. Soon, other US organizations started developing their own classifications, among them the Department of Labor (Aho and Rosen, 1980), the Bureau of Census (Worden, 1986; Abbott et al., 1989; Abbott, 1991; McGuckin et al., 1992; Doms and McGuckin, 1992), and the NSF.7 It was in this context that concepts like critical technologies, core technologies, basic technologies, advanced technologies, new technologies, strategic technologies and emerging technologies came onto the scene (US National Research Council, 1983; US Department of Commerce, 1987; Giersch, 1982; Science Council of Canada, 1986; OECD, 1985, 1988b). From these efforts, the use of the indicator soon spread to other countries (Legler, 1987; OECD, 1988a; Butchart, 1987; MSST, 1978) and international organizations like the European Commission (1994) and the OECD.

HIGH TECHNOLOGY In the mid-1980s, the term ‘high technology’ began to be used concurrently or in place of technology intensity, as evidenced in the Department of Commerce reports.8 Nothing had really changed with regard to the statistic, however, but a valued and prestigious label (high) was now assigned to it. The OECD was an important catalyst in the dissemination of this term. The real impetus to work on high technology came from the OECD Council of Ministers, which asked the Secretariat in 1982 to examine the problems that could arise from the trade of high-technology products. The first international statistics were published in 1986 in the second issue of the OECD’s Science and Technology Indicators (OECD, 1986). The organization improved over previous works, in two senses. First, it began systematically using a new label – ‘high technology’. Second, it broke down the statistics into subclasses. Up until then, there was generally only one class of industries or products, classified according to technology intensity. Other classes were simply forgotten or called ‘non-technology-intensive’. With the OECD work, three categories of technology intensity were now constructed: ‘high’, ‘medium’, and ‘low’ (Table 4.1). The OECD improved on its methodology in the following years. It developed a new classification based on a broader sample of countries (OECD, 1984), on products rather than industries, and taking into account embodied technology. Unlike other OECD science and technology indicators, the work of the organization on high technology never led to a methodological manual. Several times, among them during the fourth revision of the Frascati manual, a manual devoted to high technology was envisioned

70

Innovation in LMT

Table 4.1 OECD technology intensity levels (1986) High

Medium

Low

Aerospace Office machines, computers Electronics and components Drugs

Automobiles Chemicals Other manufacturing Non-electrical machinery Rubber, plastics Non-ferrous metals

Stone, clay, glass Food, beverage, tobacco Shipbuilding Petroleum refineries

Instruments Electrical machinery

Ferrous metals Fabricated metal products Paper, printing Wood, cork, furniture Textiles, footwear, leather

(OECD, 1978, 1988a, 1991a, 1991b) but never written. Nevertheless indicators on international trade (export/import ratios) in high-technology industries were published regularly in Main Science and Technology Indicators (MSTI) starting in 1988. The source of the high-tech classification is most probably the economist R.W. Maclaurin of MIT. Maclaurin is an author totally forgotten today. One finds neither his biography nor anything on his role in the literature on innovation, except for old citations (Nelson, 1959; Mansfield, 1968). Beginning in the early 1940s he developed the first programme of research on the economics of technological change (Godin, 2008). He used Schumpeter’s ideas, analysing innovation as a process composed of several stages or steps, and proposed a theory of innovation, later called the ‘linear model’ of innovation. From this research he developed the first full-length discussion of the model, and suggested the first list of indicators for measuring innovation (Maclaurin, 1953). Then, in 1954, Maclaurin suggested a measure of ‘technological progress’ based on a three-stage classification (high, medium and low) and classified 13 industries according to volume of R&D expenditures, number of patents issued and scientists employed (Maclaurin, 1954): Highest rate of progress Chemical Photographic Airplane Oil High rate of Progress Radio and television Electric light

The moral economy of technology indicators

71

Medium rate of Progress Automobile Paper Steel Lower rate of progress Food processing Cotton textile Coal mining House assembling in tracts

This classification was influential. In the following years similar classifications appeared. In the late 1950s C.F. Carter and B.R. Williams, respectively from Belfast and Keele universities, carried out a series of studies on innovation for the Science and Industry Committee of the British Association for the Advancement of Science (Carter and Williams, 1957, 1958, 1959a). One of these studies looked at the characteristics of firms that make them ‘technically progressive’, or innovative, defined as using science and technology, and capable of producing or adopting new products and processes (Carter and Williams, 1957: 108–11 and 177–88; 1959b). The suggested classification for over 150 firms in their sample population was: progressive, moderately progressive and non-progressive. From their calculations, Carter and Williams measured a relation between progressiveness and firms’ performance as measured by figures such as profits.

THE MORAL ECONOMY OF HIGH-TECHNOLOGY Why did high technology get such a central place in science and technology statistics? Why did such an indicator, which rests on a rather simple and century-old ratio, occupy so many statisticians for decades? The answers to these questions lies in the moral economy of statistics. To a certain extent, every human activity relies on trust. Science and technology are no exception. Several authors have recently documented how individuals attach social values to scientific truth, and how these values stand for epistemic truth.9 Similarly, machines, instruments and technology carry these values to most people, since they are automated rather than manual, and possess values like regularity and accuracy (Gooday, 2004; Wise, 1994), although tacit knowledge is often necessary to make them work properly (Collins, 1985). Following other authors, I call these ‘aperspectival’ dimensions: moral economy. Concepts like high technology have their own moral economy. High technology is a category produced in a specific social context, namely debates

72

Innovation in LMT

Table 4.2 Terms used for modern technology Advanced Core (or Basic) Generic Strategic Critical Key Progressive

Major Revolutionary Paradigmatic Radical Systemic Emerging High

on international competitiveness between countries during the 1960s. This context induced moral forces that were translated into virtues and qualities carried by the category or attributed to it. The following three elements gave strength to the category and contributed to its diffusion and use. Naming and Classifying In the 1980s, a series of terms or labels came on the scene to describe those technologies that were said to exercise influence in many economic sectors (Table 4.2). Such terms are qualitative and superlative, and they therefore carry a moral overtone. They are terms that suggest modernity. They also create hierarchies and dichotomize reality. Science and technology is full of such dichotomies: natural sciences and engineering versus social sciences and humanities; manufacturing versus services; science versus technology; basic research versus applied research and development; R&D versus related scientific activities; and high technology versus low technology. In fact, oppositions, polarities, dichotomies or dualisms (let’s call these dyads) are at the very heart of language and thinking, ever since Protagoras’ time,10 Greek mythology (Mason, 1988) and the Bible’s Genesis11 at the very latest (Tarde, 1897; Hertz, 1909; Durkheim, 1914; Perelman and Obrechts-Tyleca, 1958; Douglas, 1966; Lloyd, 1966; Needham, 1973). They often correspond to tensions that people experience and oppose. But why do human beings cut up reality and the mental space, and classify things into dyads or triads? A triad is a dyad where the two opposites are changed into two extremes with a middle term or third dimension (Coirault, 1935), like Aristotle’s Nicomachean Ethics: episteme (knowledge of unchanging things), praxis (action), and techne (craft production). High technology started as a one-level classification (high), then became a dyad or two-level classification (high versus the residual, then high/low), and then shifted to a triad or three-level classification

The moral economy of technology indicators

73

(high/medium/low). It thus got more and more ‘objective’ by covering a whole spectrum of technological types rather than the highest level only. The reasons for classifying are many (Durkheim and Mauss, 1901–2; Needham, 1979; Douglas and Hull, 1992). Classification is a cognitive action (simplifying reality) to accomplish social actions; it institutionalizes facts and practices and loads them with moral content (symbolic, political or ideological) (Edwards, 1991). Classification puts order into the world for regulating or controlling, distinguishing and separating things, people and groups, and suggesting moral values and actions. In a dyad, for example, one term is devalued, and the other privileged. The history of the high/low dichotomy is a good example of such classification, and a very old one indeed. Ginzburg goes as far back as St Paul, who wrote ‘be not high-minded, but fear’ (Ginzburg, 1976). By this, St Paul meant to be wise. Then, in the sixteenth and seventeenth centuries, this moral meaning shifted to an intellectual meaning: do not seek to know high things. By high things, people meant cosmology (sky, secrets of Nature), religion (God), and politics (power). The aim of the lesson was to maintain the social and political hierarchy of the time. More recently, the high/low dichotomy has continued to be used, in cultural matters for example. To Malcolm Arnold (1822–1888), a Victorian poet, lifelong educator, literary critic, government official (Inspector of Schools) and public figure, culture is the best which has been thought and said in the world. High culture came to mean the greatest artistic and literary achievement of a society, while low (or popular) culture meant the cultural themes characterized by production for and consumption by the masses (Gans, 1974; Bourdieu, 1979, 1984; Levine, 1988). High culture came to be opposed/contrasted to low culture by the upper social classes in order to maintain their privileged position in society. It gave them a sense of having things in common (identity), and imposing their cultural ideals on the rest of society enhanced their social status through the claim that their culture was superior to that of the rest of society. The prefix ‘high’ has been added to the term ‘technology’ in the second elevation of the status of technology in history. The first occurred at the end of the nineteenth century when the mechanical or useful arts, with their image of soiled hands, finally achieved a social and professional status more or less equal to that of high or fine (creative, imaginative) arts (Marx, 1997). The second elevation of status was the creation of a specific class of technologies, those named high technology. One sees here how naming or labelling and classifying has a powerful impact because it automatically suggests one way of looking at things (and not other ways). It puts things in good terms and draws attention towards them. ‘High’ speaks about advancement, progress and modernity.

74

Innovation in LMT

Quantifying As a rule, the more a concept is supported by statistics, the more it crystallizes and has long life. Of all the above terms or labels (Table 4.2), only one remains widely used and relatively precise, and it is the only one that is measured systematically: high technology. No other term has a ratio, formula or statistics as criterion or definition. If the success of an indicator lies in its statistics, the success of the statistics lies in objectivity and impartiality. Objectivity and impartiality refer here to two characteristics. First, the high technology indicator is based on statistics that have a long history. The same very old statistics (R&D/sales or R&D/value added) have served to quantify, partly or wholly, any of the three concepts (or names) used over history for high tech: ● ● ●

Research intensity Technical intensity High technology

In fact, among official statistics, R&D expenditures is one of the oldest statistics on science and technology. It has statistical series going back to the 1940s in the case of several countries. No one disputes the numbers. The statistics on R&D are reputed to be ‘solid’, and are among the most cherished indicators. Second, the statistics come from an authoritative source. In the early 1960s member countries of the OECD conventionalized the statistics on R&D in the form of a methodological manual, the Frascati manual. Since then, R&D has been a consensual statistics supported by an international organization, the OECD, and a network of officials (the OECD NESTI Group). Over the past decade, the high technology concept has been the object of continuous work on improving the numbers. As we have seen, the statistics were first applied at the level of industries, then extended to that of products. A second development concerned whether technology is disembodied or embodied technology. Finally, the list of high-tech products was updated. What the indicator did not get, unlike many other indicators, was a methodological manual. Finally, the indicator was calculated and published regularly by the OECD. For every member country, an export/import ratio was calculated for five groups of industries, and the results were published twice a year in Main Science and Technology Indicators (MSTI). The groups were aerospace, electronics, office machinery and computer industry, drugs, and other manufacturing industries, plus total manufacturing. Every country could compare itself to others, and economists could develop models linking economic performances to (high) technology.

The moral economy of technology indicators

75

Storytelling The OECD is a producer of stories on the theme of science and technology. Conceptual frameworks like the Knowledge-Based Economy or the Information Economy incorporate stories that tell what the issue of the day is, outline its characteristics, and suggest a course of action. High technology would probably not be so popular had it not been integrated into such stories, in the present case, histories on/about trade and competitiveness. Stories are the OECD’s way to sell its concepts and convince governments to devote resources to science and technology. A story gives meaning to science and technology and helps put the issue on the political agenda. It carries/transmits three elements. First, a story develops an imaginary/ imagery in which people can believe. Second, it identifies the impacts/ impact/effects associated with the phenomenon or event under study. Third, it puts forward a policy-relevant discourse for policy-makers. Generally, a story goes like this: 1. 2. 3. 4. 5. 6. 7. 8. 9.

Premise: science and technology are good for you and society. Something new is happening in the economy. This something is different from the past. Let’s call it NEW NAME (high technology). This new phenomenon or event will bring big rewards, as well as the possibility of leadership to those in the front line. It is therefore necessary to know more about it. Let’s collect STATISTICS. It is also essential that policies be developed. Let’s imagine a CONCEPTUAL FRAMEWORK to this end.

CONCLUSION No statistic has been so central in the history of science and technology policy as R&D expenditures. High-technology indicators rely precisely on R&D, as did early indicators on innovation. For example, until recently innovation was defined as formal innovation, namely restricted to innovation arising from R&D activities. Such a restriction comes from statistics: measuring activities that are easily identifiable for accounting. A privileged significance is attached only to that which is easily measurable: ‘that which cannot be easily measured is first disregarded, then treated as unimportant, and then as if it did not really exist at all’ (Gooday, 2004: xvii). Such a philosophy was in fact at the heart of the very first measurement of science and technology, as suggested in the OECD Frascati Manual in

76

Innovation in LMT

1962: an activity must be conducted on a systematic basis in order to be measured. The 1993 edition of the manual further mentioned (OECD, 1994: 105–6): R&D has two elements. R&D carried out in formal R&D departments and R&D of an informal nature carried out in units for which it is not the main activity. In theory, surveys should identify and measure all financial and personnel resources devoted to all R&D activities. It is recognized that in practice it may not be possible to survey all R&D activities and that it may be necessary to make a distinction between ‘significant’ R&D activities which are surveyed regularly and ‘marginal’ ones which are too small and/or dispersed to be included in R&D surveys. . . . This is mainly a problem in the business enterprise sector where it may be difficult or costly to break out all the ad hoc R&D of small companies.

This meant that a large part of activities was not counted in the statistics: activities conducted on an ad hoc basis, and activities which do not rely wholly on R&D, among them those of small- and medium-sized enterprises (SMEs) (Laestadius, 2006; Bogers and Lhuillery, 2006). In contrast, informal innovation is ‘innovation that is not explicitly planned and budgeted and therefore remains largely hidden in (aggregate) innovation data’ (Bogers and Lhuillery, 2006). It often takes place without formal R&D, as part of a process of experimentation or problem-solving: learning-byusing, learning-by-doing. Some recent studies have estimated that around half of the innovative firms develop innovations without any R&D, and that over one-third of the innovative sales and production-cost reductions can be attributed to informal innovation. I would like to suggest that the recent literature on low-tech is part of a ‘movement’ away from the past hierarchy and moral economy of science and technology (Walsh, 1996; Gertler and Vinodrai, 2006; Lambert, 2006; Hippel, 2006; Bogers and Lhuillery, 2006; Laestadius, 2006; Pedersen and Sandven, 2006). Essentially, the new literature suggests refocusing analyses of science and technology on: ● ● ● ● ●

development and design rather than research; innovation rather than R&D; non-technological innovation (organizational, marketing, aesthetic) rather than technological innovation; processes rather than output; firms (including services) rather than technologies.

This new ‘moral economy’ is concerned with the less visible, the intangible, and the minor or continuous development. What is still missing in this new

The moral economy of technology indicators

77

innovation framework is society. Innovation studies and surveys still focus on firms and marketed innovation. This is an economically-oriented approach that has driven the whole science and technology field, particularly the science policy and statistical subfields, for decades (Godin, 2006). However, long before economists appropriated innovation as a term and defined it as the commercialization of inventions, the category was used and studied by many social scientists beginning in the nineteenth century and applied to diverse dimensions of society. In reality, innovation means novelty (of any kind). I know of no discipline in history that had not studied invention and innovation, whatever the name: Philosophy: progress Literature: originality, imagination, creation Arts and engineering: invention Biology: evolution Anthropology: culture change Sociology: social change Psychology: creativity History and philosophy of science: discovery, scientific change Economics: technological change Politics and management: organizational change. However, over time, we have lost the richness of/in the term ‘innovation’. The dominant interpretation – the economic one – has reduced innovation to very few things. Going back to basics would require not making technology: (a) the definition of innovation (or novelty); and (b) the focus of policy-oriented studies (commercialization of inventions).

NOTES 1.

2.

3.

In 1914 J.M. Cattell, editor of Science, offered a similar rationale (Cattell, 1914: 161–2): ‘Over a billion dollars a year are spent in the United States on the drinking of alcohol and its consequences, a comparable amount on prostitution and its ensuing diseases. We devote twice as much money to each of these destructive agencies as to our entire educational work. Pleasure automobiles or moving-picture shows cost each year more than the support of the teachers in all our schools. The national wealth is ample to double the salary of every teacher.’ In the late 1950s the balance of payments came to be an important economic issue (for the second time in a decade), and the competitiveness of countries was measured by it. In the 1960s the debate resumed again. See: US Joint Economic Committee (1962); Johnson (1963). Britain, for example, seconded by the OEEC’s (OECD predecessor) numbers, increased its laments of ‘economic decline’, particularly in light of its balance-of-payments deficit. France launched a campaign against foreign investment that led to the well-known debate over technological gaps.

78 4. 5. 6. 7. 8. 9. 10. 11.

Innovation in LMT Complementary goods, standard commodities (not research-intensive), standard commodities (research-intensive), other commodities. In his early studies on the structural (technological) basis of the American economy, W. Leontief coined the term ‘capital-intensive goods’ for those commodities that require for their manufacture large quantities of capital. See Leontief (1953). OECD (1963, p. 81). Such a grouping comes from a study published by C. Freeman in 1962, except that no label was used with regard to research intensity. See Freeman (1962). See US National Science Board (1974) and subsequent editions. Starting with the 1993 edition, new indicators were constructed by researchers from the Georgia Institute of Technology (A.L. Porter and J.D. Roessner). An early use of the term appears in Cooper (1971, p. 9). Fairness, fidelity, honesty, detachment, impartiality, disinterestedness, self-effacement, independence, integrity, truthfulness, credibility. See Porter (1995), Shapin (1994), Daston (1992). There are always two sides to every issue. Dividing the world: light vs darkness, waters versus heaven.

REFERENCES Abbott, T.A. (1991), ‘Measuring high technology trade: contrasting international trade administration and bureau of census methodologies and results’, Journal of Economic and Social Measurement, 17, 17–44. Abbott, Thomas A., R. McGuckin, P. Herrick and L. Norfolk (1989), Measuring the Trade Balance in Advanced Technology Products, Washington, DC: Center for Economic Studies, US Bureau of Census, Washington, DC. Abrams, A. (1951), ‘Contribution to the session on measuring the returns from research’, in Engineering Research Institute, Proceedings of the Fourth Annual Conference on the Administration of Research, at the University of Michigan, 11–13 September 1950, pp. 22–4. Aho, C.M. and H.F. Rosen (1980), Trends in Technology-Intensive Trade with Special Reference to US Competitiveness, Washington, DC: Office of Foreign Economic Research: US Department of Labor. Anthony, R.N. and J.S. Day (1952), Management Controls in Industrial Research Organizations, Boston, MA: Harvard University Press. Bernal, John D. (1939) [1973], The Social Function of Science, Cambridge, MA: MIT Press. Bogers, Marcel and Stéphane Lhuillery (2006), ‘Measuring Informal Innovation: From Non-R&D to On-line Knowledge Production’, École polytechnique fédérale de Lausanne, College of Management of Technology, CEMI-REPORT2006-009. Boretsky, Michael (1971), Concerns About the Present American Position in International Trade, Washington, DC: National Academy of Engineering, pp. 18–66. Boretsky, Michael (1973), US Technology: Trends and Policy Issues, revised version of a paper presented at a seminar sponsored by the Graduate Program in Science, Technology and Public Policy of the George Washington University, Washington, DC. Boretsky, M. (1975), ‘Trends in US technology: a political economist’s view’, American Scientist, 63, 70–82.

The moral economy of technology indicators

79

Bourdieu, Pierre (1979), La distinction: critique sociale du jugement, Paris: Les Éditions de Minuit. Bourdieu, Pierre (1984), Homo Academicus, Paris: Les Éditions de Minuit. Butchart, R.L. (1987), ‘A new UK definition of the high-technology industries’, Economic Trends, 400, 82–8. Canadian Dominion Bureau of Statistics (1956), ‘Industrial R&D expenditures in Canada 1955’, reference paper no. 75, Ottawa. Carter, Charles F. and Bruce R. Williams (1957), Industry and Technical Progress: Factors Governing the Speed of Application of Science, London: Oxford University Press. Carter, Charles F. and Bruce R. Williams (1958), Investment in Innovation, London: Oxford University Press. Carter, Charles F. and Bruce R. Williams (1959a), Science in Industry: Policy for Progress, London: Oxford University Press. Carter, C.F. and B.R. Williams (1959b), ‘The characteristics of technically progressive firms’, Journal of Industrial Economics, 7 (2), 87–104. Cattell, J.M. (1914), ‘Science, education and democracy’, Science, 39 (996), 30 January, 154–64. Cattell, J.M. (1922), ‘The organization of scientific men’, The Scientific Monthly, June, 568–78. Chandler, Alfred D. (1977), The Visible Hand: The Managerial Revolution in American Business, Cambridge, MA: Belknap Press. Coirault, P. (1935), ‘Sur les dyades et triades dans la pensée et l’expression’, Journal de psychologie normale et pathologique, 32, 83–90. Collins, Harry M. (1985), Changing Order: Replication and Induction in Scientific Practice, London: Sage. Compton, K.T. (1941), ‘Industrial research expenditures’, Science, XXX, 124–5. Cooper, R.N. (1971), ‘Technology and US trade: a historical review’, in National Academy of Engineering, Technology and International Trade, proceedings of a symposium held 14 and 15 October 1970, Washington, DC: NAE. Daston, L. (1992), ‘Objectivity and the escape from perspective’, Social Studies of Science, 22 (4), 597–631. Davis, L. (1982), Technology Intensity of US Output and Trade, Department of Commerce, International Trade Administration, Washington, DC: USGPO. Davis, L.A. (1988), Technology Intensity of US, Canadian and Japanese Manufacturers’ Output and Exports, New York: Office of Trade and Investment Analysis, Department of Commerce. Dean, B.V. (1968), Evaluating, Selecting, and Controlling R&D Projects, American Management Association. Dearborn, D.C., R.W. Kneznek and R.N. Anthony (1953), Spending for Industrial Research, 1951–1952, Division of Research, Graduate School of Business Administration, Harvard University. Doms, M.E. and R.H. McGuckin (1992), ‘Trade in high technology products’, Science and Public Policy, 19 (6), 343–6. Douglas, M. (1966), Purity and Danger: An Analysis of Concepts of Pollution and Taboo, London: Routledge. Douglas, Mary and David Hull (eds) (1992), How Classification Works: Nelson Goodman Among the Social Sciences, Edinburgh: Edinburgh University Press.

80

Innovation in LMT

DSIR (1958), Estimates of Resources Devoted to Scientific and Engineering R&D in British Manufacturing Industry, 1955, London. Durkheim, Emile (1914), ‘The dualism of human nature and its social conditions’, in R.N. Bellah (ed.), Emile Durkheim on Morality and Society: Selected Writings, Chicago, IL: University of Chicago Press, pp. 149–63. Durkheim, E. and M. Mauss (1901–1902), ‘De quelques formes primitives de classification: contribution à l’étude des représentations collectives’, L’Année sociologique, 1–72. Edwards, D. (1991), ‘Categories are for talking: on the cognitive and discursive bases of categorization’, Theory and Psychology, 1 (4), 515–42. European Commission (1994), ‘First European report on science and technology indicators’, Brussels. Ewell, R.H. (1955), ‘Role of research in economic growth’, Chemical and Engineering News, 18 July, 2980–5. Federation of British Industries (1952), R&D in British Industry, London: FBI. Freeman, C. (1962), ‘Research and development: a comparison between British and American industry’, National Institute Economic Review, 20 (May), 21–39. Gans, H.J. (1974), Popular Culture and High Culture: An Analysis and Evaluation of Taste, New York: Basic Books. Gertler, Meric S. and Tara Vinodrai (2006), Better by Design? Capturing the Role of Design in Innovation, presentation to Blue Sky II: What Indicators for Science, Technology and Innovation Policies in the 21st Century?, OECD, Ottawa, Canada, 25–27 September. Giersch, Herbert (ed.) (1982), Emerging Technologies: Consequences for Economic Growth, Structural Change and Employment, Tübingen: Mohr. Ginzburg, C. (1976), ‘High and low: the theme of forbidden knowledge in the sixteenth and seventeenth centuries’, Past and Present, 73, 28–40. Godin, B. (2002), ‘Technological gaps: an important episode in the construction of S&T’, Technology in Society, 24, 387–413. Godin, B. (2004), ‘The obsession for competitiveness and its impact on statistics: the construction of high-technology indicators’, Research Policy, 33 (8), 1217–29. Godin, B. (2005), Measurement and Statistics in Science and Technology, London: Routledge. Godin, Benoît (2006), Statistics and STI Policy: How to Get Relevant Indicators, communication presented at the OECD Blue Sky Conference, ‘What Indicators for Science, Technology and Innovation Policies in the 21st Century?’, Ottawa, Canada, 25–27 September accessed 11 January 2007 at www.oecd.org/ document/60/0,2340,en_2649_34409_37083516_1_1_1_1,00.html#Wednesday. Godin, Benoît (2008), In the Shadow of Schumpeter: W. Rupert Maclaurin and the Study of Technological Innovation, Minerva Press. Gooday, Graeme J. N. (2004), The Morals of Measurement: Accuracy, Irony, and Trust in Late Victorian Electrical Practice, Cambridge: Cambridge University Press. Hatter, V.L. (1985), US High Technology Trade and Competitiveness, Department of Commerce, International Trade Administration, Washington, DC: USGPO. Hertz, R. (1909), ‘La prééminence de la main droite: étude sur la polarité religieuse’, Revue philosophique, 68, 553–80. Hoffmeyer, Eric (1958), Dollar Shortage and the Structure of US Foreign Trade, Amsterdam: North-Holland.

The moral economy of technology indicators

81

Hogan, R.M. (1950), ‘Productivity in research and development’, Science, 112 (2917), 24 November, 613–16. Holland, M. and W. Spraragen (1933), Research in Hard Times, Washington, DC: Division of Engineering and Industrial Research, National Research Council. Hounshell, David A. and John K. Smith (1988), Science and Corporate Strategy: DuPont R&D, 1902–1980, Cambridge: Cambridge University Press. Institution of Chemical Engineers (1963), ‘Productivity in research’, proceedings of a symposium held in London 11–12 December 1963, London. Johnson, H.G. (1963), ‘The International Competitive Position of the United States and the Balance of Payments Prospect for 1968’, Review of Economics and Statistics, 66 (February), 14–32. Johnson, H.T. (1978), ‘Management accounting in an early multidivisional organization: General Motors in the 1920s’, Business History Review, 52 (4), 490–517. Johnson, H. Thomas and Robert S. Kaplan (1987), Relevance Lost: The Rise and Fall of Management Accounting, Boston, MA: Harvard Business School Press. Kaplan, N. (1960), ‘Some organizational factors affecting creativity’, IEEE Transactions of Engineering Management, 30, 24–30. Kelly, R.K. (1976), Alternative Measurements of Technology-Intensive Trade, Washington, DC: Office of International Economic Research, Department of Commerce. Kelly, R.K. (1977), The Impact of Technology Innovation on International Trade Patterns, Washington, DC: Department of Commerce. Laestadius, Staffan (2006), Beyond the High-Tech/Low-Tech Divide: Towards a new Taxonomy and New Indicators to Guide the Transformation to a Knowledge Society, Stockholm: Department of Industrial Economics and Management, Royal Institute of Technology. Laestadius, Staffan, Trond E. Pedersen and Tore Sandven (2006), Towards a New Understanding of Innovativeness – and of Innovation Based Indicators, Brussels: PILOT Project. Lambert, Ray (2006), Design as a Source and Enabler of Innovation: New and Improved Indicators, presentation to Blue Sky II: What Indicators for Science, Technology and Innovation Policies in the 21st Century, OECD, Ottawa, Canada, 25–27 September. Legler, Harald (1987), ‘West German competitiveness of technology intensive products’, in Hariolf Grupp (ed.), Problems of Measuring Technological Change, Cologne: Verlag TÜV Rheinland GmbH. Leontief, W. (1953), ‘Domestic production and foreign trade: the American capital position reexamined’, Proceedings of the American Philosophical Society, 97 (4), reprinted in Input-Output Economics, Oxford: Oxford University Press, 1986, 65–93. Levi, L. (1869), ‘On the progress of learned societies, illustrative of the advancement of science in the United Kingdom during the last thirty years’, in Report of the 38th Meeting of the British Association for the Advancement of Science (1868), London: John Murray, pp. 169–73. Levine, Lawrence W. (1988), Highbrow/Lowbrow: The Emergence of Cultural Hierarchy in America, Cambridge, MA: Harvard University Press. Lipetz, B.-A. (1965), The Measurement of Efficiency of Scientific Research, Carlisle: Intermedia. Lloyd, G.E.R. (1966), Polarity and Analogy: Two Types of Argumentation in Early Greek Thought, Cambridge: Cambridge University Press.

82

Innovation in LMT

Maclaurin, W.R. (1953), ‘The sequence from invention to innovation and its relation to economic growth’, Quarterly Journal of Economics, 67 (1), 97–111. Maclaurin, W.R. (1954), ‘Technological progress in some American industries’, American Economic Review, 44 (2), 178–200. Mansfield, Edwin (1968), The Economics of Technological Change, New York: Norton. Mason, J.H. (1988), ‘The character of creativity: two traditions’, History of European Ideas, 9 (6), 697–715. Marx, L. (1997), ‘Technology: the emergence of a hazardous concept’, Social Research, 64 (3), 965–88. McGuckin, R.H., T.A. Abbott, P. Herrick and L. Norfolk (1992), ‘Measuring advanced technology products trade: a new approach’, Journal of Official Statistics, 8 (2), 223–33. Ministry of State, Science and Technology (1978), Canadian Trade in TechnologyIntensive Manufactures, 1964–1976, Ottawa. National Association of Manufacturers (1949), Trends in Industrial Research and Patent Practices, Ithaca, NY: Cornell University, pp. 77–79. Needham, Rodney (1979), Symbolic Classification, Santa Monica, CA: Goodyear Publishing. Needham, Rodney (ed.) (1973), Right and Left, Chicago, IL: University of Chicago Press. Nelson, R.R. (1959), ‘The simple economics of basic scientific research’, Journal of Political Economy, 67, 297–306. National Science Foundation (NSF) (1953), ‘Federal funds for scientific R&D at nonprofit institutions, 1950–1951 and 1951–1952’, Washington, DC. NSF (1956a), ‘Expenditures for R&D in the United States: 1953’, Reviews of Data on R&D, 1, 56–28, Washington, DC. NSF (1956b), Science and Engineering in American Industry: Final Report on a 1953–1954 Survey, NSF 56-16, Washington, DC: Bureau of Labor Statistics. NSF (1960), Funds for R&D in Industry: 1957, NSF 60–49, Washington, DC. OECD (1963), Science, Economic Growth and Government Policy, Paris: OECD. OECD (1970), Gaps in Technology, Paris: OECD. OECD (1978), Problems of Establishing the R&D Intensities of Industries, DSTI/SPR/78.44, Paris: OECD. OECD (1984), Specialization and Competitiveness in High, Medium and Low R&DIntensity Manufacturing Industries: General Trends, DSTI/SPR/84.49, Paris: OECD. OECD (1985), Analytical Report of the Ad Hoc Group on Science, Technology and Competitiveness, SPT (84) 26, Paris: OECD. OECD (1986), Science and Technology Indicators: RD, invention et compétitivité, Paris: OECD. OECD (1988a), La mesure de la haute technologie: méthodes existantes et améliorations possibles, DSTI/IP/88.43, Paris: OECD. OECD (1988b), New Technologies in the 1990s: A Socio-Economic Strategy, Paris: OECD. OECD (1991a), High Technology Products: Background Document, DSTI/STII (91) 35, Paris: OECD. OECD (1991b), Future Work on High Technology, DSTI/STII/IND/WP9 (91) 7, Paris: OECD. OECD (1992), High Technology Industry and Products Indicators: Preparation of a Manual, DSTI/STII/IND/WP9 (92) 6, Paris: OECD.

The moral economy of technology indicators

83

OECD (1993), Seminar on High Technology Industry and Products Indicators: Preparation of a Manual, DSTI/EAS/IND/STP (93) 2, Paris: OECD. OECD (1994), The Measurement of Scientific and Technical Activities: Proposed Standard Practice for Surveys of Research and Experimental Development, Paris: OECD. Olsen, F. (1948), ‘Evaluating the results of research’, in C. C. Furnas (ed.), Research in Industry: Its Organization and Management, Princeton, NJ: D. Van Nostrand, pp. 402–15. Pedersen, E. and T. Sandven (2006), ‘Toward a new understanding of innovativeness – and of innovation based indicators’, PILOT Project, Brussels. Pelz, D.C. (1956), ‘Some social factors related to performance in a research organization’, Administrative Science Quarterly, 1, 310–25. Pelz, Donald C. and Frank M. Andrews (1966), Scientists in Organizations: Productive Climate for Research and Development, New York: John Wiley. Perelman, C. and L. Obrechts-Tyleca (1958), Traité de l’argumentation, Brussels: Université libre de Bruxelles, pp. 550–609. Porter, Theodore M. (1995), Trust in Numbers: The Pursuit of Objectivity in Science and Public Life, Chicago, IL: University of Chicago Press. Quinn, James B. (1959), Yardsticks for Industrial Research: The Evaluation of Research and Development Output, New York: Ronald Press. Quinn, J.B. (1960), ‘How to evaluate research output’, Harvard Business Review, March–April, pp. 69–80. Rosa, E.B. (1920), ‘Scientific research: the economic importance of the scientific work of the government’, Journal of the Washington Academy of Science, 10 (12), 341–82. Rosa, E.B. (1921), ‘Expenditures and revenues of the federal government’, Annals of the American Academy of Political and Social Sciences, 95 (May), 26–33. Rubenstein, A.H. (1957), ‘Setting Criteria for R&D’, Harvard Business Review, January-February, 95–104. Science (1971), ‘Technology and world trade: is there cause for alarm?’, Science, 172 (3978), 37–41. Science Council of Canada (1986), A National Consultation on Emerging Technologies, Ottawa: Science Council of Canada. Seeber, N.C. (1964), ‘Decision-making on R&D in the business firm’, Reviews of Data on R&D, 44, February, NSF 64-6, Washington, DC: NSF. Seiler, R.E. (1965), Improving the Effectiveness of Research and Development, New York: McGraw Hill. Shapin, Steven (1994), A Social History of Truth: Civility and Science in 17thCentury England, Chicago, IL: University of Chicago Press. Sherman, J.V. (1941), ‘Research as a growth factor in industry’, in National Resources Planning Board (ed.), Research: A National Resource (II): Industrial Research, Washington, DC: USGPO, pp. 120–3. Tarde, G. (1897), L’opposition universelle: essai sur une théorie des contraires, Paris: Félix Alcan. US Bureau of Labor Statistics (1953a), Industrial R&D: A Preliminary Report, Washington, DC: Department of Labor and Department of Defense. US Bureau of Labor Statistics (1953b), Scientific R&D in American Industry: A Study of Manpower and Costs, bulletin no. 1148, Washington, DC. US Department of Commerce (1983), An Assessment of US Competitiveness in High Technology Industries, Washington, DC: International Trade Administration.

84

Innovation in LMT

US Department of Commerce (1987), The Status of Emerging Technologies: An Economic/Technological Assessment to the Year 2000, NBSIR 87-3671, Washington, DC: Department of Commerce. US Interdepartmental Committee on the Technological Gaps (1967), report submitted to the President, 22 December, White House. US Joint Economic Committee (1962), Factors Affecting the United States Balance of Payments, findings of the subcommittee on International Exchange and Payments, Congress of the United States, Washington: USGPO. US National Association of Manufacturers (1949), Trends in Industrial Research and Patent Practices, NAM. US National Research Council (1941), Research: A National Resource (II): Industrial Research, Washington, DC: National Research Planning Board. US National Research Council (1983), International Competition in Advanced Technology: Decisions for America, Washington, DC: National Academy Press. US National Science Board (1974), Science Indicators, Washington, DC: National Science Foundation. US President’s Scientific Research Board (1947), Science and Public Policy, President’s Scientific Research Board, Washington, DC: USGPO. von Hippel, Eric (2006), ‘Democratizing innovation: the evolving phenomenon of user innovation’, presentation to Blue Sky II: What Indicators for Science, Technology and Innovation Policies in the 21st Century, OECD, Ottawa, Canada, 25–27 September. Walsh, V. (1996), ‘Design, innovation and the boundaries of the firm’, Research Policy, 25, 509–29. Wise, Norton (ed.) (1994), The Values of Precision, Princeton: Princeton University Press. Worden, Gaylord (1986), Problems in Defining High-Technology Industries, Washington, DC: Bureau of Census. Yovits, M.C., D.M. Gilford, R.H. Wilcox, E. Staveley and H. D. Lemer (eds) (1966), Research Program Effectiveness, New York: Gordon and Breach.

5.

Critical comments on the ‘moral economy of technology indicators’ Hariolf Grupp

It is always good practice to look at definitions and conventions of terms which have become common today from a historical perspective. Godin in this volume points out several misconceptions of technology indicators which have emerged in the course of time. From Schumpeter’s basic idea of ‘quasi-rents’ up to today’s convention in the OECD countries, the meaning of product, process, marketing and organizational innovation has been developed in various ways, but basically along Schumpeter’s lines of thought (Schumpeter, [1964] 1911). To be more specific, those innovative products or services that need a great amount of research and development (R&D), the so-called ‘R&D-intensive’ products or services, are often described as ‘high technology’, which is meant to be synonymous. In Chapter 4 Godin criticizes the use of the term ‘high technology’ on the basis of his observation of historical misuse. The underlying concept uses R&D intensity as the sole criterion for innovative activities, which is narrower than Schumpeter’s definition, and is outdated in modern models and statistical descriptions. Nevertheless, this dated perception that ‘high’ means ‘better’ is still very powerful in the policy arena. That is reason enough to tackle it, as Godin does. In his contribution, Godin reviews this history of ‘high technology’ which he has published already in several earlier papers on similar issues. Nevertheless, his section on the moral dimension of what ‘high’ means is new, but comparably simple and short. This critical comment starts from modern heuristic models of innovation which discern several stages of innovation (see Figure 5.1). From theoretical references it becomes clear how important it is to differentiate between R&D activities and innovation stages (see Grupp, 1998, Chapter 1, or more recently, Grupp, 2007: 503 onwards). Input to innovation accounts for resources in general, not for R&D expenditures alone (see Figure 5.1). But for any definition of ‘superior’ or ‘better’, innovation output has to be considered, although the conceptual definitions of it in the literature are far from being consistent. However, if one defines technical progress as the 85

86

Innovation in LMT

Science

Knowledge stock

Type of R&D:

Technology Fundamental Applied Experimental research research development

Standard ization Innovation stages

Idea, theory, discovery

Possible functional interplay

Technical design Product design, innovation Imitation, improvement, diffusion, exploitation, disposal

Source: Reproduced from Grupp, 1998: 19.

Figure 5.1

Survey/overview of important innovation stages and their typology

‘creation’ of new products and as the ‘transition’ to new production processes, as in the neo-Schumpeterian tradition, then the emphasis is more on the procedural aspects than on output. What the literature calls ‘byput’ or ‘throughput’ (Freeman, 1982: p. 8) are ‘attendant’ or ‘partial’ effects of technical progress, and thus are not always a prerequisite for innovation and progress. Firms with no R&D activities at all and thus zero R&Dintensity can arrive at successful innovations, in particular in the service sectors. It is important to note that the innovation stages derived from a survey of the literature in Figure 5.1 do not mean that innovation is a linear, that is, sequential process as the caricature of the ‘pipeline’ model from the 1950s suggested: put some R&D money in, wait a while and then innovation will come out. On the contrary: the stages of innovation show strong mutual dependence. R&D must not be considered one activity preceeding innovation, but a sequence of activities that can be called upon at any time if required. Interactions between R&D activities and innovation processes should be considered in a functional way. Here, ‘functional’ means that

Critical comments on ‘moral economy of technology indicators’

87

social or economic phenomena contributing to the total system are organized in a defined manner to serve a specific function, that is, knowledge creation, intellectual property acquisition, resource financing and so on. In this view, R&D serves as a tool for solving problems at each innovation stage which cannot be solved by drawing from the existing stock of knowledge alone. But innovation is possible without R&D if the knowledge stock in the firm and in the published science is sufficient. Against this modern background of theoretical and heuristic understanding of innovation, Godin’s angle of R&D intensity seems to be quite narrow. We have to admit however that Schumpeter’s understanding of innovation was not followed in the statistical demarcation of the OECD countries for many decades. Data were collected only for product and process innovations and for the monetary as well as human resources devoted to R&D. This narrowing of the broader understanding of innovation was suggested by neoclassical theories of all kinds (competition-oriented, decision-making or game-theory) that describe process innovations as a sudden replacement of the production function from ‘before’ to ‘after’, or as a race to be the first firm coming up with a new product. But several years ago, the OECD countries in their joint Oslo Manual brought Schumpeter’s organizational and marketing innovations back onto the statistical agenda (OECD, 2005). Not surprisingly, in the first surveys, these forms of innovation are considered to be quite important in the manufacturing sector (EFI, 2008, data from Germany for 2005, unpublished, p. 88), while organizational innovations dominate in science-intensive and other services. From these first empirical results it is clear that ‘high technology’ as criticized by Godin does not reflect innovation quality or quantity by R&D input alone. In most practical cases there are rigid lists of high technology by some product-classification scheme, meaning that one includes all products even when there is disparity within the product category. High technology must be separated from low technology by a cut-off rule which is always ambiguous. In many cases the average R&D intensity is taken as the threshold, but we know now that a number of innovative products, and in particular, services, come about with zero R&D input. How relevant is low technology? By diffusion we mean the spread of innovation and knowledge over branches of the economy, geographical areas or technical disciplines. It means adapting and using novel technology with little or no R&D, that is, by definition, with low technology. In theory, unfortunately, innovation and diffusion were strictly separated. This again goes back to Schumpeter who expressed the opinion that innovation is done by heroes while diffusion is unimaginative. But since then, on the basis of the functional model, it has been recognized that incremental innovation takes place during diffusion. The feedback processes occurring in the adoption of

88

Innovation in LMT

new technology lead to learning processes. Most macroeconomic discussions over growth and the reduction of unemployment by innovation still ignore diffusion and focus, for example, on invention and technology transfer or human capital input. But as we are now confronted with new questions, we need new measurement concepts for new indicators and new empirical tests, otherwise the theoretical discussions will remain undecided. Therefore we have to ask the question: Can low technology explain growth? Or: to what extent does economic growth depend on innovation in general and not on high technology in particular? It has long been known that about half of all growth may be attributed to capital investments and the labour force. In industrialized countries the other half depends on technical change. But if this latter factor is divided into domestic invention, licensing of knowledge from abroad for ‘imported’ technology, and diffusion of new technology, it is found (for Germany between 1961 and 2005; see Jungmittag et al., 1999) that diffusion and thus the use of low technology act in considerable proportions as economic motors as well. These were, admittedly, larger in the postwar period in Germany (the ‘economic miracle’), but have never ceased to exist. Therefore, that high technology is economically ‘better’ than ‘low’ technology – the ‘moral value’ criticized by Godin – is just not true. We further have to consider another moral value: the impact of innovation on employment, or better, that it is thought to be a way to remedy unemployment. Innovation is frequently proclaimed a universal remedy for reducing unemployment and this is a hot issue these days. But the matter has not always been seen like this: rationalization was much discussed during the 1970s by trade unions, with the catchword of robots being ‘workers of steel’. Successful process innovation leads to the release of human work force to a large extent and thus was considered negative (in moral terms). Hence the debate over the effects of innovation on employment has a long tradition in economic theory, but there is still no generally accepted causal relation. The main reason for this is again that the argument does not differentiate between product, process, marketing or organizational innovation. Product innovation by itself has net positive effects on the labour market, but process innovation (rationalization) and organizational innovations have unknown effects on employment. One always has to take compensation effects into consideration. New products create new jobs, but if the product replaces an older product, the jobs in making that older product are lost. If process innovation is labour-saving but leads to cost reductions which are handed on to the customers in terms of price reductions, then customers have more money to spend on other products, which increases employment there. So the question of the relation of innovation to employment is a

Critical comments on ‘moral economy of technology indicators’

89

question of how great the substitution effects are, and whether compensation outweighs the direct effects. Add to this international competition: modernization is important for remaining internationally competitive. Waiving process innovation may imply a worse trade balance and thus destroy jobs as well. To conclude: Godin’s critique of the use of the term ‘high technology’ is based on historical observations of misuse. The underlying concept uses R&D intensity as the sole criterion for successful innovation which is outdated in modern models and statistical descriptions. Nevertheless, this dated perception is still very powerful in the policy arena. That is reason enough to debunk it. High and low technology, including indigenous invention and diffusion, both contribute to economic growth and employment, however in a complex way. The structure of Godin’s arguments is to build up a ‘bugaboo’ (bugbear) first and then tear it down. Let me suggest replacing the term ‘high technology’ by another one with less moral implications. Already in 1990 it was suggested calling R&Dintensive products ‘Schumpeter products’ in order to honour the early work of Schumpeter for our understanding of innovation (Klotd 1990: 61; the author refers to Giersch who introduced the term). It is like calling the unit of electric power ‘the Watt’ in honour of James Watt, the inventor of the steam engine. Schumpeter goods may be physical products or knowledgeintensive services that are specific to other products and services in that they need above-average R&D inputs in terms of money or human capital. Thus the allegation of a misuse of the concept of ‘high technology’ is of little relevance to theoretical and empirical research on innovation processes today, and, if at all, merely the term itself can be considered obsolete, but not the concept of technology indicators behind it.

REFERENCES Expertenkommission Forschung und Innovation (EFI) (ed.) (2008), Gutachten zu Forschung, Innovation und Technologischer Leistungsfähigkeit 2008, Berlin: EFI. Freeman, Christopher (1982), The Economics of Industrial Innovation, 2nd edn, London: Pinter. Grupp, Hariolf (1998), Foundations of the Economics of Innovation: Theory, Measurement and Practice, Cheltenham, UK and Lyme, NH, USA: Edward Elgar. Grupp, Hariolf (2007), ‘Typology of science and technology indicators’, in Horst Hanusch and Andreas Pyka (eds), Elgar Companion to Neo-Schumpeterian Economics, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 503–24. Jungmittag, A., K. Blind and H. Grupp (1999), ‘Innovation, standardization and the long-term production function’, Zeitschrift für Wirtschafts- und Sozialwissenschaften – German Economic Review, 119 (2), 205–22.

90

Innovation in LMT

Klodt, H. (1990), ‘Technologietransfer und internationale Wettbewerbsfähigkeit’, Außenwirtschaft, 45 (I), 57–79. Organisation for Economic Co-operation and Development (OECD) (ed.) (2005), The Measurement of Scientific and Technological Activities – Oslo Manual, 3rd edn, Paris: OECD. Schumpeter, Joseph A. (1964), Theorie der wirtschaftlichen Entwicklung, originally published 1911 in Munich und Leipzig, quoted from the 6th edn, published in Berlin.

PART II

Technological diffusion and interrelationships between sectors

6.

Distributed knowledge bases in lowand medium-technology industries Paul L. Robertson and Keith Smith

INTRODUCTION The study of innovation has always focused on learning, just as public policies for science, technology and innovation have always been aimed primarily at creating and diffusing knowledge. In recent years however, the focus of learning and knowledge generation has become broader as firms have embraced an increasingly wide range of sources when undertaking innovation. Although this tendency – which is sometimes associated with ‘open innovation’ – is not, in fact, new (Chesbrough, 2003; Christensen, 2006), it does present a more balanced view of the creation and use of knowledge than has been provided by older concepts according to which, through research and development activities, new science eventually inspires new technology which in turn helps to foster new products and processes, without any provision for significant feedback (Kline and Rosenberg, 1986). Our emphasis in this chapter is on the generation and use of knowledge in established industries which constitute by far the largest share of the manufacturing and service sectors in most developed economies. Although relatively old, even mature in some cases, these industries are on balance reasonably innovative. They engage in frequent changes in both product and process technologies which, although perhaps less spectacular (in a literal sense) than some of the innovations in newer industries, contribute substantially to their own productivity and competitiveness and to better macroeconomic performance. Through innovation, established industries not only benefit themselves but, in their role as consumers of new products and new ideas, they are also significant contributors to the growth of hightechnology industries. The ability of established industries to engage in frequent technological upgrading is an important determinant of prosperity in economies at all levels of development and should be a major preoccupation of both managers and policy makers (Robertson et al., 2003; Robertson and Patel, 2007). 93

94

Technological diffusion and interrelationships

In following innovative paths, firms in most established industries – sectors that have been offering variations on essentially the same product for many years and have gone through long periods of evolutionary change – do not engage extensively in formal research and development (R&D) activities as generally defined (Hirsch–Kreinsen et al., 2006).1 As a result, when discussing innovative activity at the level of firms, industries and public policy, it is necessary to reach beyond a focus on science-based activities, high technology, and internally-generated knowledge to investigate the broader range of innovative activities that drive change in practice – to the diverse channels, both formal and informal, through which firms obtain information from many sources which they then actively combine through internal adjustments. At present, however, the diverse sources of knowledge that firms rely on for their innovative activities are rarely charted. In order to increase the understanding that managers, policy makers and scholars have of the processes by which knowledge is collected and used, we propose that the distribution of sources of innovative knowledge should be mapped systematically to increase our appreciation of the richness of the innovation process. In this way it will be possible to better guide future improvements in the use of knowledge. The problems associated with the concept of a ‘knowledge economy’ are outlined in the next part of the chapter, followed by discussions of distributed knowledge sources and of ways of mapping them. In the final part, before the conclusion, we illustrate our point with an informal mapping of several sectors in food production and processing.

A KNOWLEDGE ECONOMY? The economics of innovation has always focused on learning, just as public policies for science, technology and innovation have always been aimed primarily at creating and diffusing knowledge. In recent years, however, learning and knowledge have attracted increasing attention as a result of claims that knowledge-intensive industries are now at the core of growth, and that we are now entering a new type of knowledge-driven economy or even a completely new form of ‘knowledge society’. But what does it mean to speak of a ‘knowledge-intensive’ industry or a ‘knowledge-based’ economy? Policy initiatives and public and analytical discussion of innovation issues have taken a very narrow view of this question, identifying the knowledge economy with a highly restricted group of economic activities. These activities tend to be characterized either as those directly involving the creation and transmission of information, or as those associated with high levels of direct R&D, high rates of patenting, and

Distributed knowledge bases

95

direct links to extensive scientific publishing. This approach to the knowledge economy has two basic limitations. First, it takes a narrow view of the cognitive characteristics of knowledge, focusing on the use of knowledge that is formally created via investment in R&D, thus obscuring the fact that other knowledge forms may exist with different but economically important characteristics. Second, the focus on direct creation of knowledge tends to obscure the complexity of economic processes which in practice involve interdependence and the flow of knowledge between activities. Our argument is that a ‘distributed knowledge base’ compensates very adequately in many low- and medium-technology (LMT) sectors for their low levels of expenditure on R&D.

THE MEANING OF KNOWLEDGE Concern about the role of knowledge in the economy is hardly new. Karl Marx for instance argued that a distinguishing feature of mid-nineteenthcentury capitalism was ‘the conscious application of science’, and he explicitly treated separation of the conception and execution of tasks (that is, of a knowledge function) as central to mechanisation. In a fascinating appendix to Capital Vol I, drafted in the early 1860s, Marx (1976 [1867]: 1024) wrote: The social productive forces of labour . . . come into being through cooperation, division of labour within the workshop, the use of machinery, and in general the transformation of production by the conscious use of the sciences, of mechanics, chemistry, etc. for specific ends, technology, etc. and similarly, through the enormous increase of scale corresponding to such developments (for it is only socialized labour that is capable of applying the general products of human development, such as mathematics, to the immediate process of production; and, conversely, progress in these sciences presupposes a certain level of material production).

In discussing the knowledge economy we are limited by the absence of a coherent definition, let alone theoretical concept, covering this term: it is at best a widely-used metaphor, rather than a clear concept. The OECD (1995: 7) has spoken of knowledge-based economies in very general terms, as meaning ‘those which are directly based on the production, distribution and use of knowledge and information’. This definition is a good example of the problems of the term, for it seems to cover everything and nothing: all economies are in some way based on knowledge, but it is hard to think that any are directly based on knowledge, if that means the production and distribution of knowledge and information products.

96

Technological diffusion and interrelationships

The weakness, or even complete absence, of definition is actually pervasive in the literature. For example, Powell and Snellman (2004: 201), define the knowledge economy as production and services based on knowledgeintensive activities that contribute to an accelerated pace of technological and scientific advance as well as equally rapid obsolescence. The key components of a knowledge economy include a greater reliance on intellectual capabilities than on physical inputs or natural resources, combined with efforts to integrate improvements in every stage of the production process, from the R&D lab to the factory floor to the interface with customers. These changes are reflected in the increasing relative share of the gross domestic product that is attributable to ‘intangible’ capital.

As the relative (not to mention the absolute) reliance on intellectual capabilities is extremely difficult to determine and intangible capital comprises a diverse set of characteristics that are also hard to measure, it is not easy to pinpoint the explicit qualities of a knowledge economy in this way. The definitional problems often seem to follow from reluctance to consider what knowledge is in epistemological or cognitive terms. Almost the only way in which this matter is addressed in the literature is via the concepts of codified and tacit knowledge. However these are themselves hazy (as well as not necessarily distinguishable) concepts and they do not say much about the cognitive content of knowledge. These issues go far beyond the scope of this chapter, but it is important to point out that ‘knowledge’ is in most forms of discourse a highly differentiated and to some extent hierarchical concept. It normally has to do with understanding, with the resolution of perplexity or uncertainty. But this may take many different forms. It may involve explicit theoretical concepts or principles, data generation procedures, canons of evidence and so on, all linked into some kind of explanatory structure. It is this type of knowledge that raises major questions concerning truth content, and that has been the domain of the philosophy of science. At another point on the spectrum, knowledge may involve simply the transmission of data in the context of comprehensible practical guidelines for use. These differences correspond to psychological or cognitive differences in those who ‘know’. At one extreme knowledge requires a transformative internalizing of some new principle, and at the other it simply involves accessing an intelligible account of how to do something. Such differences – and of course much finer categories could be pointed to – are important in determining what we are talking about with respect to knowledge, but they are often ignored within the literature. Related to this is the matter of institutions. At whatever level we think about the nature of knowledge, institutions are required as generative frameworks and as a kind of social memory (the latter being a precondition

Distributed knowledge bases

97

for transmission), and these too are of very different forms. The reason for making these very preliminary distinctions is that, because the literature rarely makes any attempt to grapple with such dimensions of knowledge itself, it is often able to slide between very different implicit notions of knowledge, and this is one of the many imprecisions that make the notion of ‘knowledge economy’ so rhetorical rather than analytically useful.

KNOWLEDGE-INTENSIVE INDUSTRIES? Before moving to a discussion of knowledge in industry, it is necessary to make a diversion via the concept of ‘high-technology’. In much policy analysis it is common to use the terms ‘high-technology’ or ‘knowledgeintensive industries’ in a somewhat loose way, as though in fact they were both meaningful and interchangeable terms. But we ought to remember that the term ‘high technology’ is a rather recent invention, and that its meaning is far from clear. The standard approach in this area rests on a classification developed by the OECD in the mid-1980s (OECD, 1984). The OECD distinguished between industries in terms of R&D intensities, with those (such as ICT or pharmaceuticals) spending more than 4 per cent of turnover being classified as high-technology, those spending between 1 and 4 per cent of turnover (such as vehicles or chemicals) being classified as medium-tech, and those spending less than 1 per cent (such as textiles or food) as ‘low-tech’. In fact the OECD discussion of this classification was rather careful, and offered many qualifications. Chief among these is the point that direct R&D is but one indicator of knowledge content, and that technology intensity is not mapped solely by R&D. Unfortunately the qualifications were forgotten in practice, and this classification has taken on a life of its own; it is widely used both in policy circles and in the press as a basis for talking about knowledgeintensive as opposed to traditional or non-knowledge-intensive industries. This is a serious problem, since the OECD classification as it is used rests on only one indicator, namely intramural R&D. This is open to two important objections. First, it is by no means the only measure of knowledgecreating activities. Second, it ignores the fact that the knowledge that is relevant to an industry may be distributed across many sectors or agents: thus a low-R&D industry may well be a major user of knowledge generated elsewhere. This issue will be discussed in a more empirical manner below. Even so, it is not clear that this classification helps us, even in a limited analysis of trends. One great problem is that the high-tech sector thus defined is small, and there are therefore some difficulties in arguing that it is driving the growth process. In the OECD the high-tech sector is no more

98

Technological diffusion and interrelationships

than 3 per cent of GDP even in the US (Robertson and Patel, 2007).2 It is hard to see how the direct impact of such a small component of output could, on its own, have a significant effect on overall economic growth. Most discussions of the role of high-tech are conducted in terms of share analyses. This can easily confuse matters. In virtually all of the OECD economies the share of high-tech in total manufacturing has risen in the longer term, and this is widely used as an argument for the claim that such industries are central to growth. However, this is complicated by the fact that the share of manufacturing in total output has been in long-term decline.3 It is not uncommon to see quite sweeping claims made for the high-tech sector which are not supported by readily available evidence. For example, the OECD’s The Knowledge Based Economy (1995: 9) claims that ‘Output and employment are expanding fastest in high-technology industries, such as computers, electronics and aerospace’. But the OECD’s own ‘Scoreboard of Indicators’ actually shows long-term negative growth rates of employment in high-tech manufacturing in 11 of 15 OECD countries for which data are presented (including the US, where high-tech employment declined at a faster rate than manufacturing employment generally).4 Despite these basic problems, the high-medium-low-tech approach has been extended, to divide the medium-tech category into medium-high and medium-low technology industries (Hatzichronoglou, 1997). Such classificatory manoeuvres cannot, however, alter the fundamental limitations of the category, and ought to cause us to question the identification of knowledge-intensive and high-tech industries.

FIRMS AND INDUSTRY EXPENDITURES ON KNOWLEDGE CREATION Although much analysis of knowledge creation rests on R&D data, particularly intramural R&D carried out by firms, modern innovation theory sees knowledge creation in a much more diffuse way. First, innovation rests not on discovery but on learning. Learning need not necessarily imply discovery of new technical or scientific principles, and can equally be based on activities which recombine or adapt existing forms of knowledge (Schulz, 2001). This in turn implies that activities such as design and trial production (which is a form of engineering experimentation) can be knowledgegenerating activities. A second key emphasis in modern innovation analysis is on the external environment of the firm. Firms interact with other institutions in a range of ways, including purchase of intermediate or capital goods embodying knowledge. Then there is the purchase of licences to use protected

Distributed knowledge bases

99

knowledge. Finally, firms seek to explore their markets. Given that innovations are economic implementations of new ideas, then the exploration and understanding of markets, and the use of market information to shape the creation of new products, are central to innovation. These points imply a more complex view of innovation in which ideas concerning the properties of markets are a framework for the recombination and creation of knowledge via a range of activities. In this framework R&D is important, but tends to be seen as a problem-solving activity in the context of innovation processes, rather than an initiating act of discovery. Some of these points are illustrated by data taken from the third round (1998–2000) of the European Union’s ‘Community Innovation Survey’ (CIS) (OECD, 2005: 38–9). The CIS revealed that around a third of smalland medium-size enterprises5 (SMEs) – the great majority of which were low- and medium-technology firms in established sectors – developed innovations in-house, but this did not necessarily involve formal R&D activities.6 They also engaged even more extensively in ‘non-technological’ innovative activities such as new product design and ‘advanced management techniques’. As these figures do not include innovations embodied in new machinery and other inputs, however, the full scope of innovative activity undertaken by SMEs would have been considerably greater.

THE ROLE OF KNOWLEDGE AND LEARNING IN INNOVATION ACROSS INDUSTRIES How do capital investment, intermediate good-acquisition and non-R&D expenditures relate to the structure of knowledge in an industry? Some innovation theorists have explored such aspects of learning as cumulativeness, tacitness, and interactivity, or such issues as the institutional structure of knowledge creation across economies. Others such as Lundvall and Johnson (1994) have highlighted the components of knowledge and firmlevel competence – distinguishing between specific factual information, knowledge of basic scientific principles, specific and selective social knowledge and practical skills and capabilities. But these approaches do not focus on the actual content of the knowledge base of a firm or industry, or on how it is organized institutionally. How then can the knowledge content of an industry be understood and described? We can distinguish between three areas of productionrelevant knowledge, namely firm-specific knowledge, sector or productfield-specific knowledge, and generally applicable knowledge.7 At the firm level, the knowledge bases of particular firms may be highly localized and specific to very specialized product characteristics in firms with one or a few

100

Technological diffusion and interrelationships

technologies which they understand well and which form the basis of their competitive position, or they may reside in more broadly based multitechnology firms or firms with complex products (Granstrand et al., 1997; Patel and Pavitt, 1998). Second there are knowledge bases at the level of the industry or product-field. At this level, modern innovation analysis emphasizes the fact that industries often share particular scientific and technological parameters; there are shared intellectual understandings concerning the technical functions, performance characteristics, use of materials and so on of products.8 This part of the industrial knowledge base is public (not in the sense that it is produced by the public sector, but public in the sense that it is accessible knowledge which is in principle available to all firms): it is a body of knowledge and practice which shapes the performance of all firms in an industry. Of course this knowledge base does not exist in a vacuum. It is developed, maintained and disseminated by institutions of various kinds, and it requires resources (often on a large scale). Finally, there are widely applicable knowledge bases, of which the most important technically is the general scientific knowledge base. This is itself highly differentiated internally and of widely varying relevance for industrial production; but some fields – such as molecular biology, solid-state physics, genetics or inorganic chemistry – have close connections with major industrial sectors.

DISTRIBUTED KNOWLEDGE BASES The creation, diffusion and use of knowledge have been modelled in several ways that are not always compatible. As we have shown, much analysis of knowledge creation rests on R&D data, particularly intramural R&D carried out by firms. This is often measured indirectly, as Powell and Snellman (2004) have done, through the use of patent data. But while patents are a valuable indicator when used carefully (Patel and Pavitt, 1995; Griliches, 1990), it is a mistake to over-identify knowledge creation with intramural R&D and patenting because of the knowledge-generating properties of design and trial production and similar activities. In practice, the relevant knowledge base for many industries is not solely, or even principally, internal to the industry, but is distributed across a range of technologies, actors and industries.9 A ‘distributed knowledge base’ is a set of knowledges/knowledge sources maintained across an economically and/or socially integrated set of agents and institutions. In general, enterprises do not depend on a single technology or on single sources of technological knowledge. They must blend knowledge that is distributed among various knowledge bases according to such factors as industrial source, geographical location, intellectual (scientific or technical) location,

Distributed knowledge bases

101

social location and chronology. Although the relative importance of these may vary from enterprise to enterprise and sector to sector, innovation management in a dynamic environment consists largely of finding efficient ways of detecting, comprehending and mixing and integrating distributed knowledge to achieve outcomes that are economically efficient and lead to acceptable social outcomes both within enterprises and in broader societal and political contexts. Although some authors (Coombs and Metcalfe, 2000; Coombs et al., 2003) concentrate on distributed activity that is formally structured – through joint ventures, strategic alliances, conscious outsourcing, and other well-defined organizational forms – we contend that distributed knowledge bases are often inchoate in important ways. As we discuss below, because of uncertainty and uneven distributions of knowledge, it is often difficult to know where to look for appropriate knowledge, if indeed there is any reason to suppose that such knowledge currently exists. The chains through which knowledge is conveyed may have several links, and not all chains are interconnected. Even when knowledge is ‘in the air’, a particular firm may not be breathing in the right spot to inhale it.

DISTRIBUTED KNOWLEDGE SOURCES OF LMT INDUSTRIES The management of distributed knowledge in established low- and medium-technology industries presents distinct challenges, but it is vital. Because the sources of knowledge may be widespread, however, a first stage towards managing them is to trace where firms in individual industries gain their knowledge, weigh the importance of the various sources, and map them to determine which sources are most important and under which circumstances. Theorizing and modelling in isolation are likely to generate empty boxes if they are not grounded in detailed empirical findings. As established LMT industries comprise 97 per cent or more of GDP and are similarly dominant in employment and investment (Robertson and Patel, 2007), the general economic importance of high technology is primarily exerted through its influence on productivity in LMT industries rather than directly. Because research and development activities as traditionally defined are only minor contributors to change in many established industries, technological upgrading often involves incorporating new technologies based on knowledge that originates outside the enterprise10 into complex existing frameworks (Hirsch-Kreinsen et al., 2006; Kodama, 1992). Firms in established LMT industries, even those among the very oldest, need to manage distributed knowledge bases effectively in order to

102

Technological diffusion and interrelationships

maintain domestic and international competitiveness. Unless they innovate, these firms risk being overtaken by rivals that implement new product or process technologies or manage change in other parts of their supply chains more successfully. This is often a complicated task, however, that requires commitments of substantial financial and intellectual resources. Differentiation of knowledge and its communication within organizations is well-known (Schulz, 2001). Porter’s diamond also discusses the value to firms (and to national economies) of having widespread sources of external knowledge through ‘related and supporting industries’ that allow relatively inexpensive and quick access to new knowledge (Porter, 1990). Porter’s main emphasis is on clusters or industrial districts in which highly specialized firms in close physical proximity are able to develop and diffuse information efficiently using mechanisms similar to the horizontal and vertical channels discussed by Schulz (2001). Although there is no doubt that industrial districts may be effective in promoting innovation in some cases (and quite ineffective in others) (Robertson et al., forthcoming), for many firms much information comes from further afield.

EMBODIED AND DISEMBODIED FLOWS OF TECHNOLOGY Inter-agent or inter-industry flows conventionally take two basic forms, ‘embodied’ and ‘disembodied’. Embodied flows involve knowledge incorporated in machinery and equipment. Disembodied flows involve the use of knowledge, transmitted through scientific and technical literature, consultancy, education systems, movement of personnel and so on. Pavitt (1984) indicates that embodied transfers of knowledge are especially characteristic of LMT sectors in which there is little intramural R&D. The basis of embodied flows is the fact that most research-intensive industries (such as the advanced materials sector, the chemicals sector, or the ICT complex) develop products that are used within other industries. Such products enter as capital or intermediate inputs into the production processes of other firms and industries, that is, as machines and equipment, or as components and materials. When this happens, performance improvements generated in one firm or industry therefore show up as productivity or quality improvements in another. Thus technological competition leads directly to the inter-industry diffusion of technologies, and therefore to the inter-industry use of the knowledge which is ‘embodied’ in these technologies. The receiving industry must in turn develop the skills and competences/competencies/ capability-ties/capacity-ties to use these advanced knowledge-based

Distributed knowledge bases

103

technologies. Competitiveness within ‘receiving’ industries depends heavily on a firm’s ability to access and use such technologies. This may entail problems in trying to mix different vintages of technology (Kodama, 1992). Embodied knowledge cannot necessarily be slotted into an existing framework on a turn-key basis. On the contrary, for both product and process technologies, embodied knowledge may lead users to make substantial adjustments that involve significant development (and sometimes scientific research) activities on their part. Both products and processes tend to evolve over time as incremental improvements are fitted into existing patterns and procedures. Where there is conscious modularity and design rules have been laid down (Sanchez and Mahoney, 1996; Baldwin and Clark, 2000) some changes may be effected through an easy substitution of new components for older ones, but such seamlessness is not always possible. This forces managers to rethink existing practices in order to make the best use of new developments. Knock-on effects from small changes may be significant as a result. Moreover when potential improvements of several kinds become available almost simultaneously but have been developed in different environments that are not subject to the same sets of design rules, managers may be forced to choose among them because of incompatibility. In addition the uncertainties imposed by widely distributed knowledge generate opportunities for strategic initiatives as managers look for niches in which some types of innovation are especially sought after while others offer less advantage. Specialization in knowledge acquisition is therefore possible, but carries a risk of generating technological inflexibility if neglected areas turn out subsequently to be competitively important. In other cases, however, firms may not be able to outsource knowledge acquisition because outsiders do not understand their problems and opportunities as well as the firms themselves. As a result, in dynamic environments firms need to develop absorptive capacity to access knowledge directly (as well as to increase their ‘receptive capacity’ (Robertson et al., 2003) by acquiring a range of other capabilities needed for successful implementation of change). As new knowledge may come from widely distributed sources, this is a difficult problem to manage because, in order to contain costs, firms are forced to gamble on which sources will turn out to be most profitable.

MAPPING DISTRIBUTED KNOWLEDGE SOURCES The production of detailed maps – tracing as many of the technological influences operating in a sector as possible and assessing their relative

104

Technological diffusion and interrelationships

importance – is essential for good firm management as well as for formulating sound public policies because, in the course of studying complex knowledge flows, maps are useful for identifying strengths and weaknesses. But knowledge bases are often diverse and it may be hard to define their characteristics across several dimensions. Which of the contributing sciences and technologies are the most important, and what criteria are used in making a determination? Is a specific category of knowledge (new or old) needed in-house and, if so, how much expertise is really needed? Does a specific type of new knowledge come to a firm in an embodied or disembodied form? What sorts of organizational relationships are needed to deal successfully with outside sources of knowledge? These and similar questions are empirical issues and are likely to vary across knowledge categories and from firm to firm and sector to sector. In a complex environment, both managers and policy-makers must be able to answer these questions accurately for specific cases rather than relying on broad a priori models. A number of different types of maps can be generated. For example to capture knowledge transmission through formal channels, maps of alliances and networks or modular chains are useful. Informal networks such as communities of practice may also be important in charting distributed knowledge flows. However, any single map is unlikely to capture all of the important dimensions in a distributed knowledge base. Restricting the investigation to formal channels of knowledge transmission could miss important flows travelling through informal channels. Regarding knowledge simply as ideas and concepts overlooks embodied flows. Using only regional or sectoral frameworks is likewise inadequate – or at least it may be, since the relative importance of each type of channel can differ depending on a firm’s particular situation. Furthermore, investigation should include potential as well as actual channels since important improvements may be secured by removing existing barriers to knowledge transmission. Network overlap may also be identified. Carlsson (2006) lists four major sets of institutional structures within which innovation occurs: ‘national innovation systems’, ‘technological systems’, ‘regional innovation systems’, and ‘sectoral innovation systems’. Of these, ‘national innovation systems’ have been extensively explored for nearly 20 years (Lundvall, 1992; Nelson, 1993) but have proved hard to operationalize at the level of the firm. ‘Technological systems’ (Carlsson and Stankiewicz, 1995) centre on the role of techno-economic relationships in the innovation process. Many of the underlying concepts of ‘regional systems of innovation’ can be traced back to Marshall (1920 and earlier editions). More recently, the study of regional systems has gained popularity

105

Distributed knowledge bases

X

D

C

A

L

D

Sectoral system network S

F

X

P P R

Regional system network

N Technological system network

Figure 6.1

T

X

Innovation networks

through surveys of Silicon Valley (Saxenian, 1994) and broader conceptual statements (Storper, 1997). Finally, studies of ‘sectoral systems of innovation’ have gained momentum in recent years (for example Malerba, 2004, 2005). All of these have been attempts to codify activities that have been evident for decades if not centuries. Moreover, all four of these sets of studies lead to results that are messy in the sense that individual experiences vary considerably no matter what dimension is under scrutiny.11 Analysts must therefore carefully compare maps in order to achieve as much reconciliation as possible. One segment of the process is illustrated in Figures 6.1 and 6.2. The first figure shows stylized versions of a number of innovation networks in which firms are grouped according to the common organizing principles of location, sector and technology. A few firms, such as ‘D’ and ‘X’, belong to more than one network, permitting them to be conduits for knowledge between innovation networks (Figure 6.2). As a result, innovative knowledge developed in one network can spread to others, although quite likely with time lags as three stages are required for dissemination within the originating network, across networks and, finally, within the receiving network. To appreciate fully the sources of information available to firms, it is therefore necessary to trace indirect connections between networks as well as to look at the immediate knowledge environment of each firm. The mapping process also makes it possible to identify certain firms that are especially important in the dissemination process. From a policy point of view, mapping firms within and across networks can both expose structural holes

106

Technological diffusion and interrelationships

Technological system network

Regional system network

Sectoral system network Connections between nodes in three networks Connections between nodes in two networks

Figure 6.2

Layers of innovation networks

(Burt, 1992) and provide suggestions for ways of building shortcuts for improving the speed and strength of flows.

EMPIRICAL EVIDENCE Practical difficulties arise in the empirical analysis of content. How can we describe the content of various knowledges/bodies of knowledge across particular industries, and how are they integrated? Although there is insufficient space for full-scale studies, we turn now to an illustration of this question by looking at the innovative experiences of a major sector whose knowledge base we seek to map. The main issue is the forms of knowledge involved in a sector or industry, the articulation of these knowledges and their flow across industries.

SHIFTING BASES OF INTERNAL R&D The range of technologies used by firms in established sectors has increased substantially in recent decades (Granstrand et al., 1997). Table 6.1 is based on the US patenting activities of more than 500 of the technologically most active firms in the world.12 It shows clearly that patenting by these firms, which was already diverse, tended to shift even further away from core

107

Source:

Robertson and Patel (2007).

Aerospace and Defence Chemicals Electrical/Electronics Food, Drink and Tobacco Instruments IT-Related Machinery Materials Metals Mining and Petroleum Motor Vehicles and Parts Paper Pharmaceuticals Photography and Photocopy Rubber and Plastics Telecommunications

10.7 47.0 6.6 8.1 2.2 1.9 5.0 50.5 21.7 42.9 3.3 19.1 33.7 11.0 50.0 5.2

81–90 9.8 45.6 5.5 8.6 3.0 1.4 4.6 48.9 22.1 45.7 3.3 25.1 23.2 8.5 54.0 1.8

91–00

Chemicals

0.3 14.3 0.1 10.7 0.6 0.0 0.3 2.2 1.5 3.2 0.0 1.9 46.0 1.6 3.2 0.1

81–90 0.5 16.2 0.2 25.9 2.9 0.0 0.5 2.9 3.1 3.0 0.1 2.0 60.1 0.9 2.1 0.1

91–00

Drugs and Biotechnology

32.0 8.0 61.7 2.6 47.4 74.2 21.1 9.0 11.9 5.5 21.2 12.3 2.7 63.3 6.1 72.2

81–90 33.2 8.2 67.4 2.1 42.4 83.2 22.7 11.9 16.8 5.3 26.3 7.5 1.7 67.6 5.0 82.9

91–00

Electrical and Electronics

47.6 26.7 28.5 30.2 47.9 20.8 54.5 31.3 56.9 45.8 45.3 38.6 15.1 22.8 32.8 21.2

81–90 46.0 25.7 24.3 24.7 49.7 14.4 52.8 31.1 49.5 44.4 43.2 37.2 13.0 21.5 31.5 14.4

91–00

Machinery and Process

Changing technological competencies of 500 large firms: 1981 to 2000 (percentage shares)

Product Groups

Table 6.1

7.2 0.2 1.1 0.1 0.7 1.2 5.4 0.3 2.3 0.9 25.2 0.3 0.0 0.0 2.0 0.4

81–90

8.3 0.7 1.2 0.1 0.7 0.5 5.5 0.6 3.1 0.5 22.9 0.2 0.0 0.0 2.0 0.3

91–00

Transport

108

Technological diffusion and interrelationships

technologies between 1981 and 2000.13 Not surprisingly the fields that gained the most were in the fast-developing areas of drugs and biotechnology and electronics, even among firms whose core businesses were in neither of these areas. To cope with this increasing reliance on distributed knowledge, firms in many industries have had to broaden their technological activities to deal extensively with areas that were previously of comparatively little importance.

FOOD-PROCESSING While it is not possible to present a full analysis in the limited space available, the food-processing sector offers excellent case studies of the growing importance of distributed knowledge in an LMT sector. There has been innovation along the links in the food-processing value chain for centuries. Using patent statistics, the changing technological activities of large food manufacturing firms based in the US, Japan and Europe are illustrated in Table 6.2. The number of US patents increased by over 80 per cent from 1981–90 to 1991–2000, and (in common with the other sectors in Table 6.1) the fields of patenting activity changed significantly. Although in all cases the absolute number of patents grew, the proportion of patents related to food-processing and products decreased from 38 per cent of the total to 29.2 per cent, and the share of patents related to chemicals and chemical processes also decreased. By contrast, patents in the ‘drugs and bioengineering’ class almost quadrupled from one decade to the next, their share growing from 13.6 to 29.3 per cent. This is not unexpected in view of the recent increase in relevance of bioengineering for agriculture and food-processing, and clearly demonstrates a major acquisition of new scientific and technical skills in a sector that is evolving rapidly despite its long history. In fact, it is likely that these figures understate the absorption of new technologies in food and related industries, especially in food-processing. Many technologies are also imported in the form of technology embodied in equipment or other inputs such as packaging, as food-processing is only part of a long chain of production, all of whose links are subject to improvements in quality and customer satisfaction (Peri, 2005). As an editorial in the first issue of the journal Innovative Food Science and Emerging Technologies noted: Food science and technology by nature are multidisciplinary. Many publications cover two or more of a range of disciplines, such as nutrition, microbiology, structure, physics (high pressure, ultrasound), electrical engineering (pulsing electric fields, radiofrequency heating), protein and lipid chemistry and membrane technology. (Lelieveld, 2000)

109

Distributed knowledge bases

Table 6.2 Changing technological competencies of large firms in food: 1981 to 2000 Technical field

1981–1990

1991–2000

Number of Per cent Number of Per cent US Patents US Patents Drugs and bioengineering 356 Food and Tobacco (processes and 997 products) Chemical processes 391 Organic chemicals 261 Non-electrical specialized 151 industrial equipment Miscellaneous metal products 62 Dentistry and surgery 32 Apparatus for chemicals, food, 80 glass etc. Other 47 Assembling and material-handling 33 apparatus Bleaching dyeing and disinfecting 25 General non-electrical industrial 24 equipment General electrical industrial apparatus 32 Instruments and controls 40 Metallurgical and metal-working 19 equipment Materials (incl. glass and ceramics) 20 Image and sound equipment 4 Plastic and rubber products 13 Textile, clothing, leather, wood 9 products Inorganic chemicals 7 Agricultural chemicals 5 Total 2623

13.6 38.0

1399 1392

29.3 29.2

14.9 10.0 5.8

586 331 293

12.3 6.9 6.1

2.4 1.2 3.0

121 119 111

2.5 2.5 2.3

1.8 1.3

65 42

1.4 0.9

1.0 0.9

40 40

0.8 0.8

1.2 1.5 0.7

39 37 33

0.8 0.8 0.7

0.8 0.2 0.5 0.3

23 19 17 14

0.5 0.4 0.4 0.3

0.3 0.2 100

13 12 4769

0.3 0.3 100.0

Source: Robertson and Patel (2007).

CASE STUDY 1: FISHING, FISH-FARMING AND MEAT-PROCESSING On the basis of extensive research undertaken by the STEP14 (Science Technology Economic Policy) group in Oslo, more detailed, if still preliminary, maps may be extracted from segments of the food-processing sector.

110

Technological diffusion and interrelationships

Consider fishing and fish-farming in Norway, both of which are apparently low-technology sectors in terms of internal R&D. This is a large industry worldwide, with aquaculture growing particularly strongly; this is moreover an important growth sector for developing countries. Examples of embodied flows in fishing include use of new materials and design concepts in ships, satellite communications, global positioning systems, safety systems, sonar technologies (linked to winch, trawl and ship management systems), optical technologies for sorting fish, computer systems for real-time monitoring and weighing of catches, and so on. Within fish-farming, these high-technology inputs include pond technologies (based on advanced materials and incorporating complex design knowledges/designs), computer imaging and pattern recognition technologies for monitoring (including 3D measurement systems), nutrition technologies (often based on biotechnology and genetic research), sonars, robotics (in feeding systems), and so on. These examples are not untypical of ‘low-technology’ sectors – on the contrary, most such sectors cannot only be characterized by such advanced inputs, but are also arguably drivers of change in the sectors that produce such inputs. The disembodied flows and spillovers are also significant. Underlying the technologies for fishing and fish-farming mentioned above are advanced research-based knowledges. Ship development and management rely on fluid mechanics, hydrodynamics, cybernetic systems, and so on. Sonar systems rely on complex acoustic research. Computer systems and the wide range of IT applications in fisheries rest/are built on computer architectures, programming research and development, and ultimately on research in solid-state physics. Even fishponds rest on/may utilize wave analysis, CAD/CAM design systems and the like. Within fish-farming the fish themselves can potentially be transgenic (resting ultimately on research in genetics and molecular biology), and feeding and health systems have complex biotechnology and pharmaceutical inputs. In other words a wide range of background knowledges/knowledge, often developed in the university sector, flows into fishing: Mathematical algorithms for optimal control, molecular biology, and a wide range of sub-disciplines in physics, for example. A similar breadth of scientific and technological fields underpins innovation in the very old industry of meat-processing. The abstracts of papers presented at the 52nd International Congress of Meat Science and Technology testify to the wide range of fields that now contribute to innovation (Troy et al., 2006). Extended sessions were devoted to ‘Meat quality – genomics and biotechnology’ (17 papers) and ‘Meat quality – muscle biology and biochemistry’ (28 papers). ‘Hot topics’ included ‘Polarimetric ohmic probes for the assessment of meat aging’; ‘Investigating the behavioural properties of adipose tissues using confocal laser scanning microscopy’, and ‘Influence of

Distributed knowledge bases

111

pelvic suspension and RN-genotype on shear force and sensory quality in pork loin’.

CASE STUDY 2: IMPROVEMENTS IN PACKAGING The extent to which the meat, fish and vegetable sectors have been affected by changes in the field of packaging gives a taste of the breadth of current developments in the industry. Food packaging presents major challenges and opportunities for food processors, as reflected in the many papers on packaging and preservation reported by Troy et al. (2006) and Marsh and Bugusu (2007). Non-biodegradable polymers such as polyvinyl chloride not only contribute to major disposal problems but are vulnerable to increases in petroleum prices (Bucci et al., 2005). The use of alternative forms of packaging may also increase the shelf-life of products (Avella et al., 2006; Del Nobile et al., 2008), which reduces the importance of speed in transportation. Moreover, longer shelf-life can enable a broadening of markets, continuing the trend to global supply that was begun in the nineteenth century, and enhance economies of scale. For example, improved packaging that allows refrigerated, rather than frozen, Norwegian or New Zealand fish to be sold in distant markets could generate greater demand and higher prices for producers in Norway and New Zealand (as well as to overlapping markets and new types of competition). Several different scientific and technological bases for improving the effectiveness of food packaging are under consideration. A recent study by Cannarsi et al. (2005) for example compared the use of two biodegradable films for wrapping freshly cut beef steaks with the results obtained from polyvinyl chloride, the plastic that is currently used. The outcomes from the three films were compared following extensive tests designed to simulate normal storage conditions. As the authors found that there was no substantial difference in the performance of the three products, they concluded that a switch to biodegradable films was desirable on environmental grounds. Similarly, Del-Valle et al. (2005) have reported on longer shelf-life with reduced use of non-biodegradable packaging. The scientific base/ basis, however, is different in this case. They found that scientists created a mucilage-based coating derived from prickly pear cacti as an edible coating for strawberries that also offered the possibility of reducing losses during handling and transport. Nanotechnology has also become important in the packaging industry. Nanocomposite films are now used to increase the shelf-life of minimally processed fruits such as apples (Avella et al., 2006). As nanomaterials

112

Technological diffusion and interrelationships

present a high risk of toxicity however, scientists have established that special care is needed to minimize the migration of inappropriate substances into foods and their subsequent ingestion (Powell and Kanarek, 2006a, b). As technologies enter into the food-processing supply chain at different points as well as from different sources, the possibilities for change are manifold. In the early stages of the chain for example, new processes such as fish-farming and new products (or modified versions of existing products) can lead to cost reductions that then force changes in subsequent stages such as distribution, and the same applies to changes in other links that then reverberate throughout the chain. Taken together, these pose considerable challenges for firms that need to coordinate responses to change. Mapping increases the awareness of food processors to developments in food production, packaging and transportation as well as in the technologies that their own firms develop internally and use directly.

CONCLUSION Firms in established or LMT industries must often operate in unstable and uncertain environments that require them to manage a diverse and changing array of knowledge bases. The message that emerges from the varied experiences that we have discussed is not that the problem is too complex to be analysed, but that the place to begin is with detailed empirical mappings of the management of distributed knowledge bases in order to determine which transmission routes are the most important and under which circumstances. The pervasiveness of distributed knowledge bases accounts for much of the diversity that the maps reveal since different firms belong to different, if overlapping, networks as a result of many factors including different social connections, perhaps derived from using different suppliers and catering for different customers. The outlooks and training of owners and managers also vary across firms in the same sector or region and among firms using similar technologies and drawing on similar scientific bases. It is naïve to believe that the study of any particular dimension or network structure can adequately capture how knowledge bases are managed in respect to innovation.15 Any firm operates in a region or regions, belongs to a sector or sectors, and employs one or more technologies. And conditions will often vary across firms because of their own internal characteristics. A global firm will probably be differently placed in many of these types of networks than a highly localized firm would be (which is not to deny that even one-plant SMEs can also be embedded in

Distributed knowledge bases

113

international networks in important ways). Similarly, multi-product and single-product firms may have access to different knowledge bases, although this can be altered to some extent by investments in absorptive capacity. Moreover the possibility of diminishing returns must be recognized since search costs would surely put the detailed mapping and analysis of knowledge flows for all firms beyond the capacity not only of firms but of governments. Sampling strategies need to be developed in order to maximize the value of mapping exercises. In the end it is likely that only a very limited number of firms and sectors could be explored in depth. The information gathered however could be used together with the results of the broader but necessarily more shallow innovation surveys that many governments have conducted in recent years. This would also enable analysts to generate a far richer picture of how distributed knowledge bases are currently managed, as well as offer better insights into how management could be improved in certain cases.

NOTES 1.

2. 3.

4.

5. 6.

7. 8. 9.

This is not invariably true. Some industries that have been around for long periods, such as automobiles and aerospace, are very active in R&D, although they tend to concentrate more on engineering and applied science than on pure science. As it ages, the electronics sector increasingly falls into the same category. For an extended treatment of the continuing importance, see Edgerton (2007). It may also be complicated by the fact that some sections of ICT and other high-tech sectors have become less exotic in a technological sense as they have aged but continue to be considered cutting-edge in statistical surveys. At the same time, other, newer, activities have been added to the high-tech category. Thus without an occasional but rigorous pruning, the proportion of activity classed as high-tech would continue to grow even if the proportion of technologically new activities in reality remained the same or even decreased. Similarly, Grinstein and Goldman (2006: 121) have recently written that ‘Technology firms occupy a central position in modern economies. They drive economic growth [and] productivity gains and have created new industries and innovative products and processes.’ Bewilderingly, they justify their claim by noting that the importance of technology firms ‘is reflected in the wide coverage they receive in the mass media and in the business literature.’ Firms with fewer than 250 employees. As Hirsch-Kreinsen et al. (2006) have shown, because of problems of definition, many of the innovative activities of firms in established industries are not classified as being R&D even though they lead to significant changes from the viewpoint of an individual firm. The most recent innovation survey conducted in Australia (ABS, 2006) has shown that similar definitional issues prevail there. This kind of differentiation goes back quite a long way in economics, but has been significantly developed in recent years (for an early account, see W. Salter (1966: 13–16)). Richard Nelson (1987: 75) calls this the ‘generic’ level of a technology. The use of the word ‘distributed’ has become popular in recent years to describe something that has multiple sources of inputs. The types of activity that are distributed are

114

10.

11. 12.

13. 14. 15.

Technological diffusion and interrelationships diverse and include purchasing and marketing as well as production, but on further analysis most sooner or later involve information and knowledge. See Coombs et al. (2003) and Coombs and Metcalfe (2000). For a recent treatment that advocates wider distribution of R&D and other technological activities, see Chesbrough (2003). This is also true of many small- and medium-sized enterprises (SMEs) that are too small to be able to afford R&D activity as conventionally classed. The main exception of course is the small but important number of SMEs set up explicitly to exploit new knowledge. Some firms of this type are highly specialized in knowledge creation but lack other capabilities needed to produce and market the fruits of their research (Dahmén, 1989; Robertson et al., 2003). For example, Malerba (2004, 2005) notes the high degree of variations among the sectors that he and his colleagues have mapped. A patent is granted when a patent examiner believes that the applicant has the competence to improve technology in a given field, despite the fact that it may be difficult to foresee its degree of usefulness at the time. Patent data therefore reflect corporate capacity to generate change and improvement in a given area of technology. In this respect, their main drawback is that – until recently – they did not cover software inventions, and that firms sometimes use other methods than patenting to protect their technological lead. As a result, our findings should be taken as an understatement of the importance of new technologies to LMT sectors because they may fail to catch many other aspect of technological upgrading. There is a residual category containing all the patents that are not in these five categories. Consequently, the percentages within each product group reported in Table 6.1 do not add up to 100. Now NIFU-STEP. Groenewegen and van der Steen (2006) discuss layering within each type of innovation system. This suggests that another of the challenges of mapping is to relate layers within a network to similar layers in parallel and overlapping networks.

REFERENCES ABS (2006), Patterns of Innovation in Australian Businesses 2003, Canberra: Australian Bureau of Statistics. Avella, M., G. Bruno, M. E. Errico, G. Gentile, N. Piciocchi, A. Sorrentino and M. G. Volpe (2006), ‘Innovative packaging for minimally processed fruits’, Packaging Technology and Science, 20, 325–35. Baldwin, Carliss Y. and Kim B. Clark (2000), Design Rules: The Power of Modularity, vol. 1, Cambridge, MA: MIT Press. Bucci, D.Z., L.B.B. Tavares and I. Sells (2005), ‘PHB packaging for the storage of food products’, Polymer Testing, 20, 1–8. Burt, Ronald S. (1992), Structural Holes: The Social Structure of Competition, Cambridge, MA: Harvard University Press. Cannarsi, M., A. Baiiano, R. Marino, M. Singaglia and M.A. Del Nobile (2005), ‘Use of biodegradable films for fresh cut beef steaks packaging’, Meat Science, 70, 259–65. Carlsson, Bo (2006), ‘Internationalization of innovation systems: a survey of the literature’, Research Policy, 35 (1), 56–67. Carlsson, Bo and Richard Stankiewicz (1995), ‘On the nature, function and composition of technological systems’, in Bo Carlsson (ed.), Technological Systems and Economic Performance: The Case of Factory Automation, Boston, MA and Dordrecht: Kluwer, pp. 21–56.

Distributed knowledge bases

115

Chesbrough, Henry (2003), Open Innovation: The New Imperative for Creating and Profiting from Technology, Boston, MA: Harvard University Press. Christensen, Jens Frøslev (2006), ‘Wither core competency for the large corporation in an open innovation world?’, in Henry Chesbrough, Wim Vanhaverbeke, and Joel West (eds), Open Innovation: Researching a New Paradigm, Oxford: Oxford University Press, pp. 35–61. Coombs, Rod and J. Stan Metcalfe (2000), ‘Organizing for innovation: coordinating distributed innovation capabilities’, in Nicolai Foss and Volker Mahnke (eds), Competence, Governance, and Entrepreneurship, Oxford: Oxford University Press, pp. 209–31. Coombs, Rod, Mark Harvey and Bruce S. Tether (2003), ‘Analysing distributed processing of provision and distribution’, Industrial and Corporate Change, 12 (6), 1125–55. Dahmén, E. (1989), ‘ “Development blocks” in industrial economics’, in Bo Carlsson (ed.), Industrial Dynamics: Technological, Organizational, and Structural Change in Industries and Firms, Boston and Dordrecht: Kluwer. Del Nobile, M.A., A. Conte, M. Cannarsi and M. Sinigaglia (2008), ‘Use of biodegradable films for prolonging the shelf life of minimally processed lettuce’, Journal of Food Engineering, 85, 317–25. Del-Valle, V., P. Hernández-Muñoz, A. Guarda, and M.J. Galotto (2005), ‘Development of a cactus-mucilage edible coating (Opuntia ficus indica) and its application to extending strawberry (Fragaria Ananassa) shelf life’, Food Chemistry, 9, 751–6. Edgerton, David (2007), The Shock of the Old: Technology and Global History Since 1900, Oxford: Oxford University Press. Granstrand, Ove, Pari Patel and Keith Pavitt, (1997), ‘Multi-technology corporations: why they have “Distributed” rather than “Distinctive core” competencies’, California Management Review, 39, (4), 8–25. Griliches, Z. (1990), ‘Patent statistics as economic indicators: a survey’, Journal of Economic Literature, 28 (4), 1661–707. Grinstein, Amir and Arieh Goldman (2006), ‘Characterizing the technology firm: An exploratory study’, Research Policy, 35 (1), 121–43. Groenewegen, John and Marianne van der Steen (2006), ‘The evolution of national innovation systems’, Journal of Economic Issues, 40, 277–85. Hatzichronoglou, T. (1997), ‘Revisions of the high technology sector and product classification’, OECD, STI working papers 1997/2. Hirsch-Kreinsen, Hartmut, David Jacobson and Paul L. Robertson (2006), ‘ “Lowtech” industries: innovativeness and development perspectives – a summary of a European research project’, Prometheus, 24 (1), 3–21. Kline, Stephen J. and Nathan Rosenberg (1986), ‘An overview of innovation’, in Ralph Landau and Nathan Rosenberg (eds), The Positive Sum Strategy, Washington, DC: National Academy Press. Kodama, Fumio (1992), ‘Technology fusion and the new R&D’, Harvard Business Review, 70 (4), 70–78. Lelieveld, H. (2000), ‘Foreword’, Innovative Food Science and Emerging Technologies, 1, 3. Lundvall, B.Å. and B. Johnson, (1994), ‘The learning economy’, Journal of Industry Studies, 1 (2), 23–42. Lundvall, Bengt-Åke (1992), National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, London: Francis Pinter.

116

Technological diffusion and interrelationships

Malerba, Franco (2004), ‘Sector systems of innovation: concepts and issues’, in Franco Malerba (ed.), Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe, Cambridge: Cambridge University Press, pp. 9–41. Malerba, Franco (2005), ‘Sectoral systems: how and why innovation differs across sectors’, in Jan Fagerberg, David C. Mowery and Richard R. Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 380–406. Marsh, Kenneth and Betty Bugusu (2007), ‘Food packaging – roles, materials, and environmental issues’, Journal of Food Science, 72, R39–R55. Marshall, Alfred (1920), Principles of Economics, 8th edn, London: Macmillan. Marx, Karl (1976 [1883]), Capital, vol. 1, Harmondsworth: Penguin. Nelson, Richard R. (1987), Understanding Technological Change as an Evolutionary Process, Amsterdam: North Holland. Nelson, Richard R. (1993), National Innovation Systems: A Comparative Analysis, Oxford: Oxford University Press. Organisation for Economic Co-operation and Development (OECD) (1984), OECD Science and Technology Indicators, No. 2: R&D, Innovation and Competitiveness, Paris: OECD. OECD (1993), Proposed Standard Practice for Surveys of Research and Experimental Development (‘Frascati Manual’), Paris: OECD. OECD (1995). The Knowledge Based Economy, OEC/GD (96) 102, Paris: OECD. OECD (2005), OECD Science, Technology and Industry Scoreboard, Paris: OECD. Patel, Parimal R. and Keith Pavitt (1995), ‘Patterns of technological activity: their measurement and interpretation’, in: Paul Stoneman (ed.), Handbook of the Economics of Innovation and Technical Change, Oxford: Blackwell. Patel, Pari and Keith Pavitt (1998), ‘The wide (and increasing) spread of technological competencies in the world’s largest firms: a challenge to conventional wisdom’, in Alfred D. Chandler, Jr., Peter Hagström and Örjan Söjvell (eds), The Dynamic Firm: The Role of Technology, Strategy, Organization, and Regions, Oxford: Oxford University Press, pp. 192–213. Pavitt, Keith (1984), ‘Sectoral patterns of technical change: towards a taxonomy and a theory’, Research Policy, 13 (6), 343–73. Peri, C. (2005), ‘The universe of food quality’, Food Quality and Preference, 17, 3–8. Porter, Michael E. (1990), The Competitive Advantage of Nations, New York: Free Press. Powell, Maria C. and Marty S. Kanarek (2006a), ‘Nanomaterial health effects – part 1: background and current knowledge’, Wisconsin Medical Journal, 105 (2), 16–19. Powell, Maria C. and Marty S. Kanarek (2006b), ‘Nanomaterial health effects – part 2: uncertainties and recommendations for the future’, Wisconsin Medical Journal, 105 (3), 18–23. Powell, Walter W. and Kaisa Snellman (2004), ‘The Knowledge Economy’, Annual Review of Sociology, 30, 199–220. Robertson, Paul L. (1998), ‘Information, similar and complementary assets, and innovation policy’, in Nicolai J. Foss and Brian J. Loasby (eds), Economic Organization, Capabilities and Co-ordination: Essays in Honour of G. B. Richardson, London, Routledge, pp. 261–88. Robertson, Paul L. and Parimal R. Patel (2007), ‘New wine in old bottles. Technological diffusion in developed economies’, Research Policy, 36 (5), 708–21.

Distributed knowledge bases

117

Robertson, Paul L., David Jacobson and Richard N. Langlois (forthcoming), ‘Innovation processes and industrial districts’, in Giacomo Becattini, Marco Bellandi and Lisa De Propis (eds), Handbook of Industrial Districts, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Robertson, Paul L., Eduardo Pol and Peter Carroll (2003), ‘Receptive capacity of established industries as a limiting factor in the economy’s rate of innovation’, Industry and Innovation, 10, 457–74. Salter, W.E.G. (1966), Productivity and Technical Change, 2nd edn, Cambridge: Cambridge University Press. Sanchez, Ron and Joseph T. Mahoney (1996), ‘Modularity, flexibility, and knowledge management in product and organizational design’, Strategic Management Journal, 17, 63–76. Saxenian, AnnaLee (1994), Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Cambridge, MA: Harvard University Press. Schulz, Martin (2001), ‘The uncertain relevance of newness: organizational learning and knowledge flows’, Academy of Management Journal, 44 (4), 661–81. Storper, Michael (1997), The Regional World: Territorial Development in a Global Economy, New York: The Guilford Press. Troy, Declan, Rachel Pearce, Briege Byrne, and Joseph Kerry (2006), 52nd International Congress of Meat Science and Technology: Harnessing and Exploiting Global Opportunities, Wageningen, Netherlands: Wageningen Academic Publishers.

7.

LMT innovations in a high-tech environment: human-factor ‘tools’ for the airline industry David Jacobson and Bernard Musyck

INTRODUCTION Until recently the focus of innovation research has been on high-tech industries, their products and their processes. Together with this focus was an association between innovation on one hand, and research and development (R&D) on the other. The PILOT (Policy and Innovation in Low-Tech Industries) project has been at least in part responsible for correcting this1. It is now clear that there is substantial innovation, often unrelated to R&D, in and among firms in low- and medium-tech (LMT) sectors (HirschKreinsen et al., 2006). It is also clear that in most major sectors of the economy high-tech and low-tech methods are closely related, and that, indeed, ‘the continued health of LMT sectors is crucial to the prosperity and growth of high tech industries’ (Robertson and Patel, 2005). The present chapter provides a case study of this interrelationship between LMT and high tech. The focus is on elements of the airline industry. The discussion adds new evidence from a case that has not been studied before in the context of the literature on innovation and LMT. It also adds a focus on primarily service-based activities, rather than the manufacturing focus of the PILOT and other research on LMT. The aviation sector consists of a variety of manufacturers of products and providers of services, from companies designing and/or assembling aircraft and their components to those providing airline services/flights for consumers, and/or maintenance services for the airlines.2 There is a complicated and often global chain of buyers and suppliers in this sector. The sector in general is complicated for a number of reasons. First, despite the fact that among the major innovations in the sector there are those relating to computerization of the flight deck, and these have huge implications for the work of pilots, such innovations were created and introduced by manufacturers that do not operate airlines; there is 118

LMT innovations in a high-tech environment

119

therefore no direct link between the innovation and the people who end up using the innovation. Second, in addition to innovations on the flight deck (avionics) there is also a wide variety of other innovations in the aviation industry. New technologies are frequently implemented, driven by R&D (like the emerging research activity on alternative fuels) or by the final customer (the airline operator) who may be motivated by the demands of passengers (e.g. flight entertainment and communication systems). Other, process innovations, have included most notably the new business model among airline operators. Third, there are in many of the sub-sectors both independent companies and subsidiaries of companies operating in other sub-sectors. And these two types of firms compete with one another. For example in maintenance there are both specialist MROs (maintenance and repair organizations) and those owned by airlines. Moreover, in relation to large civil aircraft the actual maintenance procedures are determined by the aircraft manufacturer, mostly either Boeing or Airbus.3 All are subject to regulatory approval. Recently large aircraft manufacturers have also started offering their own all-inclusive maintenance contracts, as is the case for the ‘Gold Care’ service offered by Boeing for its next generation aircraft, the 787 (Sobie, 2007). While this extraordinary variety of types of competitors should arguably drive innovation in maintenance, the system of strict guidelines of the aircraft manufacturer, approved by the regulator, at least to some extent impedes innovation.4 Fourth, the sector is also complicated by high technology, for example in the electronics inherent in the avionics systems and the research into and incorporation of composite materials in modern aircraft. These are also subject to the safety requirement and the intense and dense regulatory regime under which virtually all parts of the sector operate. Any changes in design or components of aircraft require regulatory approval. This strongly influences the implementation of innovations and the rate and nature of the evolution of the sector. The combination of the nature and cost of innovations and the rigorous regulation of aviation results in the possibility of ‘prisoners’ dilemma’type impediments to the implementation of innovation. Prisoners’ dilemma impediments arise, for example, where the costs of an innovation are borne by a single company. Once the innovation becomes certified, however, it becomes available to competitors. The end result for the innovator is that, because of extra costs, it loses competitive advantage. Only with high levels of trust that facilitate risk-sharing partnerships, or the intervention of an additional agent, can prisoners’ dilemma impediments be overcome. The importance of the prisoners’ dilemma issue thus varies among aviation sub-sectors. In the aircraft development and production

120

Technological diffusion and interrelationships

sectors, for example, high development costs and concomitant risks create incentives for risk-sharing. The situation is different for airlines which traditionally have collaborated much less, perhaps because of the predominance of legacy carriers operating within closed national contexts (Jacobson and Musyck, 2007). Projects like HILAS and the Lean Flight Initiative (Ward and de Brito, 2007) provide airlines with new opportunities for collaboration. Fifth, and particularly significant in the context of this book, while the aviation sector, and aerospace in particular, is characterized as high tech (Carrillo, 2005), in addition to the parts of the sector in which there are very high levels of sophisticated R&D and innovation, many of the products and services that make up the sector are essentially low- and medium-tech products and services. Even among these relatively low-tech products and services, in which there is little or no R&D, such as passenger airline services, there are significant – albeit unpatentable – innovations. An important illustration of this is the innovation in the airline business model that resulted in the low-cost airline with the 25-minute turnaround. In this chapter we provide a number of other examples. We also raise questions about the nature of the sector’s transformation and diffusion of knowledge associated with and reflected in these innovations. For example, it is instructive that even in a sector like this one, generally seen as high tech, important innovations can arise from on-the-job experience. Also, some of the apparently sophisticated technology consists of standard computers. The innovations are in the use to which they are being put and the applications that are being developed, particularly to monitor and reduce the risk of human error. Among our aims is to identify drivers of change at the human level, and their contribution, if any, to safety and productivity. This leads us to the concept of human factors which will be briefly introduced below.

WHAT ARE HUMAN FACTORS? The concept of human factors dates from World War II when psychologists were asked to investigate airplane accidents (Wogalter and Rogers, 1998). It was revealed that pilots had certain expectations of how things should work (for example how they could locate the landing gear instrument and activate it) and that these expectations were violated by aircraft designers (Ward and McDonald, 2006). It was in reference to the work of psychologists and physiologists at that time that the terms ‘applied psychology’, and ‘ergonomics’ were first used. According to the International Ergonomics Association (IEA) Council,

LMT innovations in a high-tech environment

121

Ergonomics (or human factors) is the scientific discipline concerned with the understanding of interactions among humans and other elements of a system, and the profession that applies theory, principles, data and methods to design in order to optimize human well-being and overall system performance. (International Ergonomics Association, 2008)

Human factors is thus a systems-oriented discipline, which promotes a holistic approach in which considerations of physical, cognitive, social, organizational, environmental and other relevant factors are taken into account (International Ergonomics Association 2008). Cognitive ergonomics includes the study of mental workload, decision-making, human-computer interaction, human reliability, work-stress and training as they may relate to human-system design, all crucial in aviation safety. Organizational ergonomics includes communication, crew resourcemanagement and design of working times (for example airline operation rostering). Human error is present in 100 per cent of aviation accidents (Braithwaite et al., 1998: 57). Errors can occur at the level of the maintenance of aircraft, their design or their operation. The major cause of all aviation accidents is pilot-error (McFadden and Towell, 1999), though accidents are often a result of a chain of events in which the pilot is the last link in the chain. Indeed, systematic accident investigations have traditionally led research to focus on finding factors that are related to pilot-error. These traditional approaches have been challenged in recent years; as Maurino (2000: 952) put it, ‘there is a need to attack causes rather than symptoms of safety deficiencies’. Following Reason (1997) human-factors studies need to focus on what makes organizations (and culture) relatively safe instead of on their dangerous moments. Thus there is a need for safety investigations to go beyond accident investigation and to look at how positive human factors analysis can help design much safer socio-technical systems. One of the conclusions of the work of Braithwaite et al. (1998) is the importance of culture at different levels: work groups, organizational level, industry level and national level. These cultural factors can influence the strength or weakness of airline operations, and harmonious working relations are desirable not only for the sake of safety but also to improve operational efficiency. This means that improved safety may lead to better efficiency. It is not only a matter of reducing incidents and avoiding accidents; safety culture is far more than a cost to the airline, in fact it can even become a ‘profit centre’.5 The LMT innovation presented in this chapter aims at what has been described above: improving human factors and operational safety. It is also hindered by various issues that make human-factor implementation in aviation problematic. As we shall see later, these issues include reluctance

122

Technological diffusion and interrelationships

and/or fear by crew members to report, complacency, crew fatigue, and conservatism and reluctance on the part of airline operators.

HUMAN-FACTORS IMPROVEMENTS: LMT INNOVATION This chapter focuses on a single type of innovation: human factors in the field of airline operations. As discussed earlier, human factors are just one form of the wider innovation landscape in the aviation industry. The chapter draws on fieldwork carried out in the context of a European FP6 project called HILAS (Human Integration into the Lifecycle of Aviation Systems, see www.hilas.info). The aim of the project is to develop a model of good practice for the integration of ‘human factors’ across the lifecycle of aviation systems. HILAS aims at the transformation of the aviation industry by improving flight safety through the integration of human-factor knowledge into all aviation-related activities. What this means is the reduction of risk of human error and the danger that this can pose for safety in the operation of aircraft. In substance, ‘the project is effectively a formal network intentionally focused on innovation – innovation in human factors (HFs) management and integration, innovation in flight operations and maintenance processes, innovation in technological design driven by operational and human factors data, and innovation in organizational and industrial change’ (Ward, 2006: 5). The project is based on four main strands of activity – Flight Operations (FO), Maintenance (MX) and Flight Deck (FD) technologies, and Knowledge Integration (KI). KI is the means of gathering and making available throughout the network the knowledge gained in all the strands; the aim is to drive innovations that will integrate operational and flight-safety approaches. Here we concentrate on FO. Focusing on HFs and innovation is important in the context of our own agenda regarding LMT innovations in the aviation sector. The aim of HFs research is to improve safety; safety and innovation can be improved concurrently, at least this is a strong hypothesis underlying the HILAS project. In MX and FO improving safety and efficiency cannot be done in big leaps because of the safety-critical human and machine interaction. Regulations are very important and this limits change and innovation (Jacobson and Musyck, 2006).6 Unlike other industries, airlines around the world are part of ‘a globally similar industry with little difference in the extensive regulations and safety management systems, aircraft technology and crew training. This standardization may be a contributory factor for limitations on innovation and abilities for change’ (Ulfvengren,

LMT innovations in a high-tech environment

123

2007; see also Steele and Pariès, 2007 on barriers to innovation in the aviation industry). If innovations occur, they do so in a rather restricted way and unlike that of other sectors. As Ulfvengren (2007) notes, ‘extensive regulations and safety concerns make it difficult for operators to do more than to suggest ideas and let the safety department do risk analysis on suggested improvements’. Moreover, unlike other sectors in manufacturing industries, there is little spatial proximity between groups of airline employees (like pilots, dispatchers and other support staff) since flight crews are always on the move and do not share physical space on a daily basis. There is a rich body of literature in innovation studies pointing to the importance of close collaboration, physical proximity, shared culture, collaborative problem-solving and close communication in support of innovation in general. All the above are HFs, and HILAS aims to provide the tools and methods to collect relevant information and to facilitate processes that will influence communication and culture and subsequently impact on safety and productivity, in the absence of spatial proximity. A key factor in such network-based innovation is trust. Heanue and Jacobson (2002), among others, have shown that trust can evolve within spatially diffuse networks. Such innovations are being developed and tested in HILAS; they are of an incremental nature and in general not a result of technological R&D. HILAS is a collaborative research project that uses the experience of various partners to map out ways to improve HFs and eventually lead to industry-wide implementation of innovation. Thus HILAS can be seen as an R&D consortium aiming to generate innovation. However, while this is in some sense a ‘normal’ process (R&D leading to innovation), it is also unusual, first, in that it is both highly collaborative, and second, in that a significant proportion of the research is being undertaken by social scientists. Both the collaborative and social science nature of the research have led to a need to understand the role of intellectual property rights (IPRs) and of trust in sharing knowledge, at least some of which is not susceptible of protection through IPRs. As noted by Ward (personal communication, 2008) all HILAS software and hardware that will underpin improved HF processes will be subjected to/protected by IPRs. However, the innovative, though nontechnical, organizational and informational aspects of the HILAS project are just as important. In any case the IPR issue raises the question whether sharing best practices and probable subsequent improvements of safety outcomes would not be more desirable than the protection of such innovations. Much of the research in HILAS is in industrial and organizational psychology, in culture, cultural differences and their impact on HFs, and in innovation and its management both within and between organizations. The focus is on finding ways, technically, organizationally and behaviourally, of

124

Technological diffusion and interrelationships

gathering information on operations in many companies, across all the strands. This is a great deal more and potentially more useful information than could be generated by any one company alone. If sufficient information within all the strands can be gathered to identify patterns and generate knowledge, and if ways of sharing that knowledge can be agreed, then all the members of the network (and possibly others) could use that knowledge to reduce the HF risks – accidents and incidents. However for commercial reasons it is important that the information be available generically, so that the data specific to an individual company are not identifiable. It is in this sense that the major innovation of HILAS is collaborative: it requires that the information be furnished by each company, but its value is only in the aggregate of information provided by all the companies together. The IPR issue arises not only in the question of what specific technologies owned by individual companies those companies are willing to contribute to the process of achieving the HILAS aims, but also in the determination of who will own and control the outcome of the HILAS project. Given the socioscientific – and therefore probably unpatentable – nature of the probable outcomes of the project, what IPRs are identifiable remains to be determined.

KEY INNOVATIONS IN HILAS Airline operators are faced with the double task of improving their competitiveness through lean and cost-effective flight operations while at the same time enhancing safety and reliability. To achieve this, any tool or methodology used must encompass HF-related information to continuously improve processes. ‘Research suggests that the design of such tools takes second place to continuous improvement behaviour itself. This involves a suite of behaviours, which evolve over time rather than a single activity’ (Bessant et al., 2001). These behaviours cluster around several core themes, for example, the systematic finding and solving of problems, monitoring and measuring processes, and strategic targeting (Cahill et al., 2007: 3). Thus in aviation too continuous improvement of behaviour is most important. Cahill and Losa (2007) show how research is being carried out to ‘produce’ new ‘task support tools’7 for crew within the HILAS framework; it is very much a participatory process involving several airlines, gathering information in a variety of ways (including in-depth interviews, jump-seat observations and workshops) from participants working in flight planning, active flight operation (dispatch, cabin crew and maintenance), safety and quality. For a long time the airlines’ approach to safety was concentrated on reactive measures (compliance with regulatory framework and measures

LMT innovations in a high-tech environment

125

to prevent recurrence of accidents). More recently, preventive safetymanagement approaches have been applied which include: scientifically based risk-management methods, a non-punitive culture of incident and hazard reporting, company commitment to the management of safety, the collection and analysis of safety-related data from normal operations and the sharing of safety lessons and best practices. To do this airlines have developed safety/risk management systems which are often coupled with performance monitoring and evaluation tools. These tools collect written or electronic information on human performance through self-reporting and observer-based methodologies, but also data on aircraft performance. One limitation of the above tools is that they do not provide a real-time picture of daily routine operations which in turn could support predictive risk management. Another problem is that when data are gathered, analysed and stored in different formats, this poses an information management challenge and makes it difficult to obtain an integrated safety/risk picture. Thus there is a need for new systems that will allow airlines to gather, integrate, analyse and communicate all airline information (commercial, operational and safety) in real time. Such systems would then support the airline’s safety strategy by providing adequate analysis of information on safety and risk, and also improve the airline’s safety culture through improvement of reporting and sharing of safety related information (Cahill et al., 2007). The result of the FO research, and related research in the other strands of HILAS, has been the definition of four HF ‘tools’ (A, B, C and D). These tools, the key innovations of HILAS, aim to help airline operators and maintenance companies understand why certain events take place (that is, what went right or wrong and why) and to draw conclusions and learn lessons that can lead to the design of new systems that take into account human needs and characteristics. The first tool (Tool A) concentrates on communication and information gathering.8 The tool offers: (a) task support; (b) performance feedback; and (c) reporting capability. It can be used by flight and cabin crews as well as maintenance and ground operators. Tool A includes three applications used by pilots both when off duty and on duty at different points in the operational process: a remote web application accessible on a PC, a crewroom application accessible on a PC, and an Electronic Flight Bag (EFB)9 application. Together these applications make it possible both to channel information to the flight crew and to obtain information from the flight crew. This can happen at different points in the crew roster and operational timeline and from different locations. The EFB’s contribution is important because it completes the information loop by linking up ground and air information (Cahill and Losa, 2007).

126

Technological diffusion and interrelationships

In substance, Tool A is the main tool for allowing operational crew (both in the air and on the ground) to communicate (including feedback and reporting) amongst themselves.10 To perform these functions, various hardware devices can be used including EFBs, workstations in the office, Personal Digital Assistants (PDAs) and mobile phones. Flight data will be recorded continuously during each flight and feedback will be provided to the crew if deviations from the norm are identified. In these cases – when reporting is mandatory – pilots will be able to report electronically instead of on paper as is the case today. The tool makes it possible for pilots to explain why their performance did not comply with the benchmark parameters. This means that in addition to providing raw data, the system aims to collect qualitative data which could reveal ‘important latent conditions and human factors issues’ (Ulfvengren, 2007). These crucial qualitative data may explain why certain flight parameters or (unsafe) events deviate from the norm. In turn, once this is done, it opens the way to establishing common causes for certain situations; procedural changes to tackle problems at the systemic level can then be developed. The tool also allows pilots to see, either immediately after the flight or later, what their performance was; this facilitates their own learning process (a kind of real-time benchmarking) (Ulfvengren, 2007). The performance feedback is especially relevant when pilots are required to complete a mandatory HF report possibly linked to a safety-critical event discovered in the Flight Data Monitoring (FDM) where flight technical data are recorded (Cahill, personal communication, 2007). The above process of data collection and benchmarking is also complemented with the ability to file optional reports electronically. In these reports, pilots can offer feedback, highlight weaknesses in operations, suggest known ‘workarounds’ or other, experience-based, constructive suggestions. Archived reports are also available (crews have access to prior reports) and can update and edit them and get information as to what the safety department is doing about the issues raised. This is an improvement in the information loop (reporting and getting feedback from the report). Tool B is actually a piece of equipment: it is the ground server and the data base that allows the integration of all data from Tool A including aircraft technical data and all relevant flight data (from all participant airlines and other external sources). Schematically, Tool B would typically work as follows: B receives data from A and integrates this data with other airline data sent to B (from other airline tools); B receives data from C. Tool C is a data analysis/mining and reporting tool; it offers risk analysis, process improvement and information flow analysis. It can be used by several different departments including safety, training, flight planning, flight operations control, maintenance, quality control, fuel and others.

LMT innovations in a high-tech environment

127

The hardware used is a simple workstation in the office. Schematically this database query/report tool would typically be used to query Tool B, to view, analyse and report based on data in Tool B, to add new information to Tool B (which is sent to A or other users of C), and to automatically send information through B to A. Specifically, in the case of Threat and Error Management (TEM) reports provided by pilots using Tool A, these are analysed in conjunction with other information from Tool B and then fed back to Tool A as TEM-specific information customized for a particular flight. This feedback information flow between A and C links to task support as well as reporting (Cahill, personal communication, 2007). Overall, Tool C is both an organizational and technology tool for safety and risk management. It facilitates both short- and long-term risk analysis (strategic safety analysis, improvement of decisions for action and change) and offers information updates to support tasks and feedback to aircrew. ‘The idea is to be able to track latent conditions from flight operations to organizational processes and to link actions upstream in the process to outcomes and consequences downstream in operations’ (Ulfvengren, 2007). Tool C is intended to provide some basic trends and statistical analysis. For instance it will be able to answer a query regarding the frequency of delays of certain flights to a particular destination and point out whether the cause is linked to Air Traffic Control (ATC), or delays in receiving the load sheet,11 or technical issues. The idea is that at different levels, operational issues can be linked to HF issues. A similar query regarding the frequency of unstabilized landing approaches12 could also be made to Tool C. The first step will be to find the percentage of such approaches at a given airport and then to query the reasons for this occurrence. One trend might find for example that Crew Resource Management (CRM) reports indicate that the majority of the cases can be linked to crew problems13 but that ATC issues are also relevant to a certain extent (specific ATC instructions due to traffic congestion for instance). Tool C is part of a new system of monitoring and responding to ‘events’. If an event happens, as soon as information from the flight deck is synchronized with the main system in the operations room (that is, only when the aircraft is on the ground), an alarm bell will ring to request that the crew fill in a HF report. The on-board system (such as the Flight Operation Quality Assurance, FOQA14) does not communicate with operations in real time (EFB Class 1 is used and not Class 2 as mentioned elsewhere) and that is why the synchronization is only possible after the aircraft lands. It is only when the pilot synchronizes with the company’s main system that Tool B ‘realizes’ that an event has taken place and that an HF report is necessary. Tool D allows flight planning, active flight operation and changes and improvements to processes. It is not a technology tool but rather an overall

128

Technological diffusion and interrelationships

organizational system that defines safety management, process improvements, procedures, roles, safety culture and responsibilities. Tool D defines the information flow logic for the other three tools and its users will include everybody in the airline. Ulfvengren (2007) notes that Tool C is only defined at a very high level of functionality and she suggests that it should facilitate the management of change, for example, identifying potential improvements in processes, which could be based on information derived from Tool C. Overall, Tools A, B and C provide technology support for Tool D which is the main contribution of the research project. The other tools and technology will just act as facilitators (Cahill, personal communication, 2007). It is clear from the above that flight-crew task-support tools as described will have to be linked to broader organizational safety and risk management tools and processes. While safety will need to be managed both at an organizational and flight level, cockpit tools of the future might need to integrate the three functions of task/performance support, performance feedback and performance reporting (Cahill and Losa, 2007).

HILAS INNOVATION AND EXISTING TECHNOLOGY The FO strand of HILAS is aiming to improve airlines’ operations through the analysis of each airline’s risk and operational performance profile. To do this, the project uses a combination of software (Tool A) and hardware (the EFB). The HILAS hardware is a Class 1 EFB,15 in essence a conventional laptop or tablet computer that can be used inside the flight deck of an aircraft. The EFB, it should be emphasized, is not a HILAS innovation but had already been developed prior to HILAS and is in the process of being introduced in a number of airlines. Since the onset of commercial aviation, all the onboard work has been performed on paper and very little evolution has been witnessed in this field. The introduction of EFBs will allow the development of an almost ‘paperless’ cockpit and eliminate all manual/paper processes between aircraft systems and the airline backing systems. Thus the device will allow operators to minimize the time needed to produce and manage reports on board the aircraft and reduce the risk of error as well. In addition to the savings produced by electronic documents it will allow airlines to align the airline back office operation with the flight deck. EFBs offer a large variety of applications: journey log, flight plans, route briefing data, performance calculation, weight balance calculations, check lists, technical log, load sheets, flight manuals, aircraft documents and others. Navigation charts (global en-route, airport and terminal charts) are amongst the most

LMT innovations in a high-tech environment

129

important applications; airlines spend a considerable amount of money and effort to maintain updated paper sets of airport and navigation charts in the flight decks of each of their aircraft (for large fleets of aircraft, this can represent substantial amounts). A Class 2 EFB can be used to store these maps and charts electronically. More sophisticated EFBs which are directly connected to the avionics of the aircraft will allow the use of dynamic maps: moving maps of an airport surface with a pilot’s ‘own-ship’ position marked for Class 3 devices, and display of the position of all other aircraft equipped with Automatic Dependent Surveillance Broadcast when they transmit their GPS position for Class 3 devices.16 Other functions will include real-time weather information using satellite weather, video surveillance from the passenger cabin or enhanced vision imagery of the runway ahead using an infrared camera (Hughes, 2007). Croft (2007) notes that a project is underway in the US to restructure the Federal Aviation Authority Notices to Airmen (Notam) system so it will provide pilots with real-time electronic updates to airport facilities and diagrams, including changes to runway configurations. Currently most airlines still use air safety reports filled in manually on paper forms. If an incident occurs, the captain is supposed to fill in a form and send the document to the safety department. There, somebody enters the data manually into an IT application that helps the company aggregate and analyse the incidents. If the airline had EFB systems installed in its aircraft, pilots could just submit the reports electronically and this would eliminate the manual data entry. One problem currently faced by airlines is that pilots do not really report as much as they should. One of the reasons for this is deficiencies in the information loop; pilots feel that they are not always listened to, their reports are not evaluated, and they do not expect prompt feedback (Serradas, personal communication, 2007). If pilots were given the opportunity to file reports electronically, they could actually trace their statements on an EFB and check whether the report has been read and by whom, its evaluation status and possible actions taken. By promoting the use of the EFB, HILAS may entice pilots to report more often. In a sense this underlines the importance of motivation in achieving change. A culture of non-reporting is anathema to HF progress. Observing that reporting has an effect is much more likely to engender a culture of reporting than when it gets no obvious response. In addition this could be beneficial to the airlines by improving their overall awareness of processes, their performance monitoring and feedback. EFBs can be used to monitor operational processes as shown above. However HILAS’s unique contribution is to use EFBs to collect HF related data and to structure these data for further processing and analysis. Large

130

Technological diffusion and interrelationships

airlines can collect thousands of reports every day and such large amounts of data cannot be analysed manually without some kind of IT support (such airlines face huge problems in structuring the data and often information remains unused for months or years). The use of EFBs in conjunction with the various tools developed by HILAS will make the exploitation of these data possible. Once data have been collected and sorted, an HF expert is needed to evaluate the information and make sense of it. The data are then applied to a ‘risk model’ that will help the airline’s HF expert provide helpful feedback to the airline operator regarding performance, risk and safety. HILAS is proposing something new. The whole idea of the project is to link the ‘what’ and the ‘why’. Technology can provide information on what is going on but cannot answer why. The ‘what’ is to some extent already known from FOQA and FDM on the ground; the ‘why’ is necessary to answer such questions as why a landing approach is unstabilized, or why a certain crew exceeded recommended speed. HILAS aims to provide the tools to equip people to analyse HF situations. This is not simple because there are various departments in a company and they all have different views about what to do and how. People and departments have different interests and priorities and negotiate their interests as they go along. Thus once a problem has been identified, people might have different views on what should change and how to achieve the change. The complex and sophisticated individual and organizational processes involved in identifying the causes and solutions to such problems cannot be done by computation alone; human mediation is essential. HILAS will address concerns such as: Does training need to be improved and, if so, in what specific respects? Do flight plans need to be prepared differently? If so, what specific additional information is required? Are changes needed in procedures relating to interactions between ATC and flight crews, and, if so, in what specific respects? More directly commercial issues will also be covered. For example, Tool C will facilitate identification of causes of delays and this, in turn, will make possible the prevention of those delays in future. The data and information generated by, and analysis made possible through, the HILAS tools are new because they are comprehensive, or system-wide. Thus changes and improvements can be assessed not only at the level of the crew’s behaviour but more importantly at the level of the whole system. This is a shift in perspective (Cahill, personal communication, 2007). To sum up, what are the key innovations offered by the HILAS tools? First, the bundle of tools proposes to create a new type of report – a mandatory company HF report which will be linked to the flight technical data. When an event occurs, the pilot will be automatically requested to

LMT innovations in a high-tech environment

131

complete this company HF report.17 It is worth emphasizing that, for this to work properly, a pilot’s report about an event should not be used against him or her but only for safety/risk management purposes because it is sensitive information and must be treated as protected data. Ultimately the report would be integrated with other data and reports in Tool B to make possible a system evaluation – a complete picture of the system’s performance at any one time. Second, the project aims to provide access to flight technical data so that pilots can learn from their own experience without the need for a middle person. In practice, a version of the FDM may show highlights, perhaps a visualization of the flight and any potential or actual problems that may be associated with it. If crews can access information about previous flights, they can learn from past experience. For this to happen, new industrial relations agreements need to be put in place to allow the use of FDM data and make flight information available to promote learning and improvements. Third, the task support and reporting tools will have flight-crew training concepts embedded in them. For instance the tools will provide access to an intelligent flight plan providing a summary of information about flight threats (categorized into aircraft, crew and environment). Crews will be able to review information about threats and discuss them in their planning and briefing before proceeding to the aircraft. Crews will also be able to obtain customized CRM-oriented briefing information (the data will link to the crew composition). At the end of their flight (using the EFB) or at a later stage (using the Internet or a mobile communication device) crews will be able to complete an optional ‘Threat and Error Management’ (TEM) report on threats encountered and how these were handled. This then can be fed into Tool C, used for TEM for future flights or fed into training sessions. Fourth, the task-support tools will allow crews to update and edit reports through the use of the archive facility of the system and also allow them to track the status of their own reports. This should provide an incentive and motivation to report. Overall, the task-support tools aim to improve the flow of information at both the level of task information and that of process improvements and organizational learning. Regarding task information, there will be a significant improvement of information flow (when crews are on or off duty, on the ground or in the air) and handover information (such that, if for example a bird strikes an engine, the next crew will get feedback information). Improved information flow will benefit operations from a tactical perspective, and at a more strategic level it will promote organizational learning. The value added of the project is that it adopts a network-centric perspective rather than a department- or process-centric perspective (Cahill, personal communication, 2007).

132

Technological diffusion and interrelationships

Will HILAS, or even EFBs, be introduced into airlines in the near future? Among the reasons why airlines have been slow to adopt an EFB type solution is because they find it difficult to define a business case for implementing EFBs as part of a process of improving HFs in the long run. What airlines want, after all, is to improve productivity and it is difficult to prove that safety and productivity are positively related.18 For larger airlines, an EFB solution is a big investment because the whole fleet has to be equipped, not just one aircraft. Moreover, apart from capital investment, training is also expensive. In general airlines are probably more interested in flying than implementing such projects. As yet there has been no airline claiming to have made substantial savings through the implementation of EFBs, so most are waiting for evidence of success from the ‘first-movers’ (Serradas, 2007). Despite the fact that the Federal Aviation Authority in the US has decided to certify Class 2 electronic flight bags, ‘the market for these products on existing airline aircraft hasn’t taken off in a big way yet according to avionics executives’ (Hughes, 2007: 158). It should be emphasized that HILAS goes beyond EFBs; it is possible, indeed, that HILAS’s HF module, as part of the EFB package, makes EFBs more attractive (Ward, personal communication, 2008). Another major factor is the airlines’ resistance to change. As documented in earlier work, airlines like any other companies are resistant to change, therefore the implementation of new technology onboard in the cockpit has been slow (Jacobson and Musyck, 2007). This inertia may arise from regulatory impediments to change in aviation and/or from organizational culture, which – aside from that of the low-cost airlines (and maybe even in those) – is not particularly flexible. Among the ultimate aims of HILAS is to change this, particularly in relation to HFs.

A NEW MODEL FOR AVIATION IN EUROPE? As highlighted by McDonald the conceiver and director of the HILAS project, its general objective can be expressed as a knowledge-transformation process that leads to changes in the way things are done in the aviation sector: The HILAS project will develop a ‘system life-cycle model’ in which knowledge generated about the human aspects of the system at the operational end is transformed into an active resource for the design of more effective operational systems and better, more innovative use of technologies. (2007: 1)

The innovation of HILAS is the development of new ways of obtaining data to create and use human-factor knowledge. The tools described above

LMT innovations in a high-tech environment

133

are consistent with this, enabling the collection of data that in aggregate contribute to new HF knowledge. Through participation of all relevant stakeholders, and supported by appropriate software and web interfaces, the process driven by HILAS aims to provide a dynamic model of ‘a full and valid agenda for change’. How can this agenda for change, with all the concomitant changes in the way things are done in aviation, be accepted by the diverse companies and other organizations, given that as we have shown they are often reluctant to share data and information about their operational system?19 There are two answers to this. As we have argued, spatial proximity can contribute to trust and rapid diffusion of innovation but is frequently absent in aviation. However, ‘organizational proximity’ can be an alternative to spatial proximity. Organizational proximity arises from ‘interactions in intra-industry relations, co-operation and collective learning processes’; it ‘creates a capacity to assemble fragmented information, tacit knowledge and other non-material and non-standardized resources’; it may be ‘a prerequisite for collective learning processes, and for co-operation among different organizations in the creation of new resources and innovation’ (Heanue and Jacobson, 2002). Heanue and Jacobson (2002) go on to show how organizational proximity contributes to the evolution of trust. This description of the origins and consequences of organizational proximity closely accords with HILAS. All the firms in HILAS are in or associated closely with aviation; HILAS itself is a consortium or network collaborating in a collective learning process; its aim is indeed to create a capacity to assemble fragmented information, tacit knowledge and the like, in order to contribute to HF innovation. Thus the first factor contributing to a willingness to collaborate in the agenda for change among the diverse companies even in the sub-set of aviation that is HILAS, is the organizational proximity in HILAS and its contribution to trust. The second factor is that by jointly developing the innovation, and/or providing for the sharing of the costs of the innovation, HILAS in a sense creates an additional agent that obviates the ‘prisoners’ dilemma’ problem. If the HILAS innovation is shared with a trusted agent in return for benchmarking services and access to valuable knowledge on how to improve their operations, airline operators may be convinced and participate. It is likely that airlines will be interested in transforming their own operational data into knowledge that they can use to promote innovation in their operation. Companies that recognize the value of the knowledge they stand to acquire will be more inclined to foster trust and share knowledge. Thus knowledge could be transformed into a product with an inherent commercial value, especially if and when it can become a product that exists independently of the author. Such a ‘knowledge business’ could be of interest to companies

134

Technological diffusion and interrelationships

wanting to benchmark their operational organization; it could also foster improved models of human-factor implementation or develop common operational models between sub-systems of the aviation sector (such as MX, dispatch or FO), promote the development of new technologies and also help regulating authorities implement new rules. In other words, the ongoing improvement of the ‘Operational Process Model’20 which has to be done cooperatively by all the stakeholders involved in the process of feeding in information, becomes knowledge that can be sold. The success of such a model would hinge on two factors: the fact that there is no free sharing of information between firms of the sector and the fact that the process of knowledge transformation has to add real value to the data that was fed into the system. The idea is that, if it is possible to sell codified knowledge in this way, this could form the basis of a post-HILAS activity that would be self-sustainable. The challenge at this stage would be to clarify how all the knowledge from HILAS organizations’ and individuals’ experiences, training and education, can be transformed into a saleable product. If indeed the above model can be achieved it will indicate the ability of the firms to collaborate to foster innovation and learning despite the existence of barriers to the sharing of information and data. There are serious impediments to the above changes (indeed, to any change in aviation). As Steele and Pariès (2007) point out – relating a consultancy company’s experiences implementing a new safety management technique during two large-scale projects in the aviation industry – there are significant impediments to change and cooperation in the sector. Change in relation to HFs is impeded because the regulatory regime and industrial structure discourage the adoption of anything beyond minimum safety requirements. Differences in national, professional and organizational cultures make mutual understanding and cooperation difficult. This is exacerbated by the fact that intense competition can put limits on the cooperation between firms necessary for safety. The key is in the ‘ambiguous commercial value of “safety” ’ (Steele and Pariès, 2007: 19). However – returning to the general objective of HILAS – if knowledge about human aspects of the system can become an ‘active resource’ facilitating subsequent change, then this knowledge will have clear commercial value and will help the diffusion of innovation in the sector.

CONCLUSION In this chapter we have addressed LMT innovation in the context of an industry not normally associated with LMT. Aviation, broadly understood,

LMT innovations in a high-tech environment

135

includes aerospace – including aircraft design and manufacture – as well as airlines and all their work in providing the service of air transport, and the maintenance of aircraft. The aim of HILAS is to transform aviation by improving flight safety through the integration of human-factor knowledge into all aviation-related activities. To achieve this a great deal of tacit and disparate knowledge must be formalized and aggregated. For this, and to achieve the integration and accessibility of HF knowledge across the sector, new tools have been designed. We have described the tools and the relationship between the so-called ‘Electronic Flight Bag’ and the first tool, Tool A. Neither the tools nor the implementation of the EFB solution in the cockpit are high-tech. Even if we accept that laptop computers are high tech (which is highly questionable), their application in EFB solutions is not. Neither the HILAS tools nor the means of implementing them are any more high tech than the use of computers and software for word-processing in offices. As Robertson and Patel have argued: all levels of technology, from the newest to the most traditional, are inextricably linked in a modern developed economy and . . . they feed off each other by being mixed in various combinations in different sectors. (2005: 274)

These innovations are basically LMT innovations. Their aim is to generate HF knowledge and to facilitate its diffusion throughout the aviation sector. Through the vast amounts of data that it will be possible to collect, collate and analyse, knowledge of patterns of relationships between human factors, airline (and maintenance) performance, incidents and accidents will be created. The new tools will contribute to changing the ‘non-reporting culture’ by providing the means for a more rapid response to appropriate reporting. They will also offset the ‘prisoners’ dilemma’ impediment to some innovation by providing an immediate means for diffusing new HF knowledge.

ACKNOWLEDGEMENT The authors wish to thank Joan Cahill and Yvonne Ward from the Department of Psychology of Trinity College Dublin and Diogo Serradas from Aircraft Management Technologies, Dublin for their constructive comments on earlier drafts of the chapter. The authors gratefully acknowledge European Commission funding for HILAS: Human Integration in the Life-cycle of Aviation Systems under the 6th Framework Programme. Priority No. 1.4 – Aeronautics and Space Research Area 3, Priority Title – Improving Aircraft Safety and Security (IP8), Contract N0. 516181.

136

Technological diffusion and interrelationships

NOTES 1.

2.

3. 4.

5. 6. 7. 8.

9.

10.

The PILOT project was a FP5 project, the details of which can be found at www.pilotproject.org. That PILOT and its participants were not the only ones to focus on lowtech innovation is clear from Maskell (1996), Lund Vinding (2001), Palmberg (2001), Von Tunzelmann and Acha (2003) and Quatraro (2005), to mention just a few. Aircraft manufacturing and assembly are very different from airline operations; they are fundamentally different activities relying on significantly different knowledge bases. For the former, the national, regional and sectoral systems of innovation play a crucial role. National governments for example promote aerospace industries, and regional and local clusters of related industries form the backbone of the dense network of aerospace companies that act as first- and second-tier suppliers to the aircraft assembler. For operations socio-economic and institutional (regulatory) factors are important, and key suppliers (and their organization) are very different. There are also several other manufacturers serving the civil aviation market (e.g. regional jets, turboprops, general aviation). There may be different perspectives on this, because different companies in all subsectors of aviation adopt new innovations at different rates. Thus very innovative companies may consider themselves held back by regulations, while others find it difficult to achieve the technological and other standards required by regulation. The job of regulation in setting minimum standards for safety, though requiring some cooperation with aviation companies, clearly also requires independence from them. That improved safety can improve profitability, however, remains to be proved. This is one of the objectives of the HILAS project, on which much of this chapter is based. However, see again Footnote 5. The introduction of ‘tools’ involves changes to task practices and their overall process; it is not just a matter of installing a new piece of technology into the cockpit. Here we mean information ‘in’ and ‘out’. When on duty, information ‘in’ includes information about flight threats, latest flight-plan information, latest operational updates and communications, briefing guidance, and flight relevant documents. When off duty, information ‘in’ includes access to future flight-plan information, performance feedback for previous flights, access to safety and training information and access to reports archives. Information ‘out’ relates to reporting and communications. Reports can be completed either on or off duty. This includes mandatory HF reports (e.g. Air Safety Reports – ASRS, incident reports), mandatory operational reports (for example flight log/voyage report) and optional reports. Optional reports can be anonymous or confidential. As far as communication is concerned, flight crews can talk to operations control and maintenance (Cahill, personal communication, 2007). In recent years flight information that was traditionally presented in paper format and carried in the pilot’s bag is now able to be supported by digital means. Such EFBs can include electronic maps and documents, be used to calculate flight performances of aircraft and to draft crew reports, and are also becoming the new means of communication between ground-support staff and the crew (Cahill, personal communication, 2007). Class 1 EFBs are not connected to the aircraft systems; here crews are required to synchronize EFB information with ground systems while on the ground. Class 2 EFB devices offer connectivity with the aircraft systems and more advanced EFB devices (Class 3) have enhanced navigation and guidance functionalities such as intelligent map displays providing terrain, traffic and weather overlays (Cahill and Losa, 2007; Learmount, 2007; Moores, 2007; Croft, 2007; Hughes, 2007). Communication facility will allow crew to communicate with other operational and management personnel in various processes, including some not directly related to flight. Indeed Tool A has reporting functionality for other roles, for example cabin crew and dispatch, and thus the tool should be considered in a broader perspective instead of a ‘flight-crew-centric perspective’. The tool also includes an observer-reporting tool (Cahill, personal communication, 2007).

LMT innovations in a high-tech environment 11.

12.

13. 14.

15.

16.

17.

18.

19.

137

Load control is part of each airline’s operational responsibility. The primary function of load control within aircraft handling is to provide the load sheet which shows the number of ‘souls’ onboard and baggage/cargo/mail figures. This document also shows that the centre of gravity of the aircraft is within balance limits and that the maximum aircraft weights are also within their limits. The captain has to sign off the load sheet before the aircraft can take off. Unstabilized approaches are frequent factors in approach-and-landing accidents. An approach is considered stabilized only if all the criteria in company standard operating procedures are met before or when reaching a given minimum stabilization height. The criteria include the aircraft’s correct flight path, speed, sink rate, power setting between given reference values, correct landing configuration, and all conducted briefings and checklists. Any time an approach is not stabilized, a go-around should be conducted (approach and landing are cancelled). Most unstabilized approaches are attributed to fatigue, pressure of flight schedule and crew-induced or ATC-induced circumstances resulting in insufficient time to plan, prepare and conduct a stabilized approach (Flight Safety Foundation, 2007). Crews may not be getting on well, be doing proper briefings or rushing the briefings, may not operate ‘like a team’ because for instance the captain is a very authoritarian person, or the first officer is finding it difficult to agree decisions with that particular captain. FOQA is a data-intensive program that collects, stores and analyses recorded flight data. The goal is to improve the organization’s overall safety, maintenance effectiveness and overall operational efficiency. ‘FOQA programs are an attempt by operators to identify data from flights where deviations from the norms take place. The object is to pick out potential problems and correct them before they lead to accidents. A FOQA program typically requires the installation of a Quick Access Recorder (QAR) on board an aircraft to record flight parameters. Data are collected to determine whether the aircraft is deviating from standard procedures or operating limitations. A computer analysis is performed to identify deviation trends and take action to prevent accidents’ (Kolczynski, 2006: 1). One of the fundamentals of FOQA is that when ‘events’ or ‘severe events’ are queried this should always be done in a non-punitive way. Fear that it may not be can prevent reporting (Cahill, personal communication, 2007). A Class 1 EFB was chosen for the purpose of the HILAS project because 95 per cent of the requirements of Tool A are covered by the device. In other words there were no benefits associated with the use of Class 2 and 3 EFBs within the framework of HILAS. Moreover certification for Class 2 and 3 equipment would have required at least 6–12 months, so in essence the HILAS project would be finished before anything could be implemented and tested. There are important cost differences between the three classes of devices: Class 1 EFB devices cost between 2500 and 7000 USD, Class 2 EFB devices cost between 15 000 and 80 000 USD and Class 3 EFB devices can cost between 200 000 USD and 400 000 USD (Serradas, personal communication, 2007; Hughes, 2007). Unlike a Class 3 EFB, a Class 2 device cannot exchange data with a flight management system unit. While Class 2 can display navigation charts and approach plates, the pilot has to enter data into the flight management system manually using a control display unit (Hughes, 2007). In this report the crew could state whether the cause of an event is a single event or a chain of events. The crew’s feedback is essential; it could also provide evaluative information (was it possible to follow standard procedures in this situation; if not, why not?), and suggestions for improvements and opinions on whether procedures are appropriate or not. As pointed out by an industry observer (Serradas, personal communication, 2007), ‘airlines are looking for a killer application that will increase productivity and safety. The problem is that there is no one such single application that will generate significant cost savings. It is actually a set of integrated applications across different departments of an airline that will achieve this. As a result airlines have struggled to understand and develop a rationale to implement EFBs’. It should be said however, that the fact that a number of airlines are working together within the context of HILAS shows that at least some companies are ready to collaborate for the greater good of the industry.

138 20.

Technological diffusion and interrelationships ‘The Operational Process Model (OPM) represents, as comprehensively as possible, current knowledge about how the system functions.’ The OPM is being developed as a fully functional model which collects and manages data, information and knowledge about the system, represents the system graphically, reports on the current state of the model and enables simulations of current, future or alternative versions of the operational process (McDonald, 2007: 2).

REFERENCES Bessant, J., S. Caffyn and M. Gallagher (2001), ‘An evolutionary model of continuous improvement behaviour’, Technovation, 21, 67–77. Braithwaite, G.R., R.E. Caves and J.P.E. Faulkner (1998), ‘Australian aviation safety – observations from the “lucky country”’, Journal of Air Transport Management, 4 (55), 55–62. Cahill, J. and G. Losa (2007), ‘Flight crew task performance and the design of cockpit task support tools’, Proceedings of the European Conference on Cognitive Ergonomics, 83–7. Cahill, J., N. McDonald, P. Ulfvengren, F. Young, Y. Ramos and G. Losa (2007), HILAS Flight Operations Research. Proceedings of HCI International 2007, Bejing, China, Berlin: Springer. Carrillo, J.E. (2005), ‘Industry clockspeed and the pace of new product development’, Production and Operations Management, 14 (2), 125–41. Croft, J. (2007), ‘Warning signals’, Flight International, 10–16 July, 30–3. Ergonomics (2007), 28–31 August, London, UK, pp. 83–8. Flight Safety Foundation (2007), ‘Approach-and-landing accident reduction briefing note 7.1 – stabilized approach’, accessed 15 November at www.flightsafety.org/alar/alar_bn7-1stablizedappr.pdf. Heanue, K. and D. Jacobson (2002), ‘Organizational proximity and institutional learning’, International Studies of Management and Organization, 31 (4), 56–72. Hirsch-Kreinsen, H., D. Jacobson and P.L. Robertson (2006), ‘ “Low-tech” industries: innovation and development perspectives – a summary of a European research project’, Prometheus, 24 (1), 3–21. Hughes, D. (2007), ‘Finding yourself: lower cost electronic flight bags, with ownship position, may proliferate on airlines’, Aviation Week & Space Technology, 18, 158–62. International Ergonomics Association (2008), ‘The discipline of ergonomics’, accessed 15 January at www.iea.cc. Jacobson, D. and B. Musyck (2006), ‘HILAS: A sectoral system of innovation in aviation’, project deliverable, October, accessed at www.hilas.info/mambo/. Jacobson, D. and B. Musyck (2007), ‘HILAS: Implementation requirements – Innovation’, project deliverable, July, accessed at www.hilas.info/mambo/. Kolczynski, P. (2006), ‘FOQA 2006 – Aviation safety versus legal exposure’, mimeo, accessed 15 November at www.aviationlawcorp.com/content/foqasafety.htm. Learmount, D. (2007), ‘Lufthansa systems tests class 2 EFB on A340-600’, Flight International, 12–18 June, p. 16. Lund Vinding, A. (2001), ‘Firms and knowledge institutions: the innovation potential in low-tech sectors and small firms’, paper presented to the Nelson and Winter Conference, Aalborg, 12–15 June.

LMT innovations in a high-tech environment

139

Maskell, P. (1996), ‘Localised learning in the furniture industry’, DRUID working paper 96. Maurino, D. (2000), ‘Human factors and aviation safety: what the industry has, what the industry needs’, Ergonomics, 43 (7), 952–9. McDonald, Nick (2007), ‘Knowledge transformation process’, Trinity College Dublin working document, 29 July. McFadden, K. and R. Towell (1999), ‘Aviation human factors: a framework for the new millennium’, Journal of Air Transport Management, 5, 177–84 Moores, V. (2007), ‘Approval nears for ATR 42/72 Class 2 EFB’, Flight International, 17–23 July, p. 13. Palmberg, C. (2001), Sectoral Patterns of Innovation and Competence Requirements – A Closer Look at Low-tech Industries, Sitra Report Series 8, Helsinki: Sitra. Quatraro, F. (2005), ‘A Schumpeterian approach to innovation clustering in a lowtech technology in a peripheral region: the case of garments in Mezzogiorno’, Innovation: Management, Policy & Practice, 7 (4), 435–50. Reason, J. (1997), Managing the risks of organizational accidents, Aldershot: Ashgate. Robertson, P.L. and P.R. Patel (2005), ‘New wine in old bottles: technological diffusion in developed countries’, Perspectives on Economic, Political and Social Integration, 11 (1–2), 271–304. Sobie, B. (2007), ‘Outer limits. Maintenance outsourcing has become a crucial part of the business model for low-cost and smaller carriers. But there is no one size fits all solution’, Airline Business, October, 46–53. Steele, K. and J. Pariès (2007), ‘Experiences of Dédale implementing a safety model and risk-based decision aid approach in past projects’, HILAS work package WP1.3.6. Ulfvengren, P. (2007), ‘HILAS tools for continuous improvement in aviation?’, mimeo, KTH, The Royal Institute of Technology, Stockholm, Sweden. Von Tunzelmann, N. and V. Acha (2003), ‘Innovation in “low-tech” industries’, TEARI working paper no. 15, SPRU, University of Sussex. Ward, M. and N. McDonald (2006), ‘Human factors in aircraft maintenance – problems and possibilities’, HILAS theoretical discussion paper, Trinity College Dublin. Ward, Y. (2006), ‘Integrating operational and safety improvement in aviation through a European innovation network’, European Operations Management Association (EurOMA) Conference Proceedings, Glasgow, 19–21 June. Ward, Y. and M. de Brito (2007), ‘Lean-safe operations for the aviation industry’, European Operations Management Association (EurOMA) Conference Proceedings, Turkey, June. Wogalter, M.S. and W. Rogers (1998), ‘Human factors/Ergonomics: using psychology to make a better and safer world’, Eye on Psi Chi, 3 (1), 23–6.

8.

Technology fusion and organizational structures in lowand medium-tech companies Daniela Freddi

INTRODUCTION In this chapter we develop the argument recently advanced by a few authors (Hirsch-Kreinsen et al., 2003; Sandven et al., 2005; Von Tunzelmann and Acha, 2005) that the role of medium- and low-tech sectors in economic growth has been undervalued or even neglected1. As Sandven et al. (2005: 57) point out, ‘medium and low-tech industries have persisted over the past decades despite the claims that we are undergoing a kind of structural revolution’. According to them the cause of this persistency is that growth is based much more on the ‘internal transformation’ of existing sectors than on their substitution by new sectors (ibid.: 58). Therefore in line with these arguments we maintain that the attention of scholars and policy-makers should move away from the concern with how to increase the production of high technologies, to other issues, for example, that of how to increase the application of high technologies in existing sectors in order to promote their internal transformation.2 For this reason we illustrate here the possible evolution of LMT sectors that might occur thanks to the fusion of traditional and radical new technologies3, in particular the shift from traditional mechanics towards mechatronics, and try to asses the consequences of this on companies’ organizational structures. As we have shown elsewhere (Freddi, 2007, forthcoming), ‘mechatronics’ refers to a blend of mechanics and electronics originating in Japan at the end of the 1970s with the first applications of electrical and electronic devices to mechanical components (Berardinis, 1990; Kodama, 1992). Later, progress in electronics and computer science led to the need for a real synergy between mechanics, electronics and informatics because, in mechatronic products, ‘the whole is greater than the sum of the parts . . ., the design process has led to a product which not only performs at a better 140

Technology fusion and organizational structures

141

level than previously, but also would have been completely unobtainable hitherto’ (Hewitt, 1993: 3). The very high level of interdependence between mechanics, electronics and informatics prompted Kodama (1992) to present mechatronics as one of the key examples of technology fusion, similar to technological complementarities (Carlsson and Stankiewicz, 1991; Carlsson, 1997), where combining different technologies can enhance the performance of existing products as well as widen the possibilities for creating completely new solutions. In technology fusion, however, the interdependency between different technologies is so great that actually a (new) body of knowledge emerges as a fusion of the previous knowledge bases. One of the main consequences of this evolution, testifying to the deepness of the transformation, has been the change manifested in the design phase. Tomkinson (1991) and Hewitt (1993) maintain that all three technologies embedded in mechatronics, as well as all the existing interdependencies between them, have to be considered simultaneously during the design phase in order to obtain efficiently functioning machines. The integration of different technologies is the key distinctive element of mechatronic products, however it is necessary to underline that there exist degrees in the level of integration. For example a mechanical machine with a fully integrated electronic device whose software controls its operations can be called mechatronic. Nevertheless in some cases, electronic or informatic applications have only a minor task such as measuring the machine’s working performance or warning the operator of failure, while in other cases they might adapt current performance to the external environment. Therefore our analysis is of firms that have integrated the different technologies (not of firms that have not) and the extent or degree to which their products are mechatronic. As stated above, in this chapter we are interested in assessing the impact of this evolution on companies’ organizational structures. Therefore we analyse the possible consequences that complexity – defined in terms of extension and strength of interdependencies – might have on companies’ organizational structures. For, as already noted, the main feature of technology fusion and mechatronics is the high level of interdependence between the embedded technologies. In particular, we analyse the relationship between complexity and the boundaries of the firm, exploring to what extent complex problems and products are more likely to be realized within or between companies, and to what degree the design and production of complex systems can be decomposed into separate tasks or components to be realized independently in order to increase the level of specialization of components as well as overall efficiency.

142

Technological diffusion and interrelationships

The chapter is organized into three main parts. The first part reviews the traditional explanations of the boundaries of the firm, transaction costs and modularity, and illustrates their limits in explaining organizational structures. Therefore we introduce the systems-integration branch of literature and present the factors that, in line with this view, can enhance or limit vertical integration. We discuss separately the organization of R&D activities and production because, as we will illustrate, they can be organized according to similar as well as opposing models. In the third part we give details of our methodology and present the empirical results deriving from ten case studies on mechatronic companies.

EFFICIENCY, COMPLEXITY AND ORGANIZATIONAL STRUCTURES Dealing with Efficiency: Transaction Costs and Modularity The different possible modes of organizing production, in terms of degree of verticalization and coordination, have been extensively analysed in the literature. The seminal contribution of Coase (1937) introduced the dichotomic distinction between the transactions conducted through markets or within firms. This view was extended by the transaction-costs approach (Williamson, 1975, 1985), which considers that markets are characterized by imperfect information with bounded rational agents tending to take advantage of this, behaving in an opportunistic way. These external conditions contribute to transaction costs, thus causing firms to decide to internalize a certain number of activities and processes if and when this is cheaper than buying them on the market (Williamson, 1975, 1985; Klein et al., 1978; Grossman and Hart, 1986; Hart and Moore, 1990). Some scholars have pointed out that, although it is very useful, this view does not account for all the diverse modes of production organisation (Richardson, 1972; Robertson and Langlois, 1995): markets and hierarchies are not the only two possible ways of coordinating production, but they can be seen as the extreme cases on a wide spectrum. Between these extremes, there are ‘complex networks of co-operation and association’ (Richardson, 1972: 892) among firms. In view of these arguments, vertical integration should be considered in terms of degree, with some factors enhancing it and others limiting it. For example, according to the transaction-costs view, asset specificity is one of the most enhancing factors in vertical integration (Klein et al., 1978). The first contributions to transaction-cost theory were later revised and integrated in order to take into consideration not only the physical

Technology fusion and organizational structures

143

production of goods, but also the dynamics related to the organisation of knowledge-intensive activities (Pisano, 1990; Monteverde and Teece, 1982; Mowery, 1983). The common element in these is that vertical integration is driven by efficiency incentives, in line with the bulk of transaction-cost theory. In fact, knowledge is treated as a production input with some peculiar characteristics that can lead to market failures (Brusoni et al., 2001). In the last 20 years coordination problems have been extremely amplified by the increasing complexity of products. In fact, products have enlarged considerably the number of components embodied as well as the range of technologies embedded. The definition of complexity for products and systems depends strongly on the framework in which the concept of complexity is used. Some authors underline that the high level of ‘customised components, the breadth of the knowledge and skills required and the degree of new knowledge involved in production’ (Hobday, 1998: p. 690) are the principal features of complex systems, which could make it impossibile for a single person to develop the whole product (Baldwin and Clark, 2000). Other scholars have pointed out that the distinctive characteristic of complex products is the high level of mutual interdependence between the different components and sub-systems (Schilling, 2000). In this view, the level of separability of the system components is linked to the level of its complexity and can be seen as a continuum: the higher the level of separability of the system, the smaller its complexity. In the present work, in referring to complex products, we adopt this approach, thus giving a particular emphasis to the presence of several strong interdependencies between components. To help firms deal with the very high level of complexity of certain products, scholars and engineers have developed in the last decade the concept of modularity (Baldwin and Clark, 2000; Ulrich, 1995). These authors maintain that it is possible to split up the design phase of a whole complex system into smaller design modules, so that, instead of one design team developing the entire project, several design teams can do the work in parallel. Of course since the product is complex, with several interdependencies between the modules, problems of coordination can easily arise, which can be solved by use of so-called ‘design rules’ (Baldwin and Clark, 2004: 8). These rules are interfaces on which the different design teams agree before starting the design phase. In this way each team is free to experiment and develop any solution as long as it respects the design rules. It is important also to clarify that product modularity and organisational modularity are two different concepts, but they are closely interrelated (Langlois and Robertson, 1992; Sanchez and Mahoney, 1996). The reason is that the advantages generated by modularity can be exploited

144

Technological diffusion and interrelationships

through a modular organization: ‘although organisations ostensibly design products, it can also be argued that products design organisations, because the coordination tasks implicit in the specific product designs largely determine the feasible organisation designs for developing and producing those products’ (Sanchez and Mahoney, 1996: 64). Therefore a firm’s choice of modularity-in-design implies to a certain extent its choice of modularizing the organization. The traditional explanations for the boundaries of firms mentioned thus far – transaction costs and modularity – have greatly contributed to the comprehension of organizational structures, in particular in terms of degree of verticalization. Nevertheless, they exhibit limits in explaining certain organizational choices generated by two main factors. On one hand, new organizational forms have recently emerged that cannot be fully understood only on the basis of these contributions (Prencipe, 1997; Brusoni et al., 2001). On the other hand, in both approaches, knowledge has been treated similarly to other production factors, implying for example that cognitive labour can be divided or integrated for efficiency reasons as are other productive activities (Brusoni and Prencipe, 2001). In the following section we illustrate some contributions intended to overcome the limits of these traditional approaches by taking into consideration the nature of knowledge-intensive activities and their consequences for industrial organizations. Dealing with Complexity: Systems Integration, Problem Decomposability and Organizational Coupling Recently a group of scholars (Prencipe 1997, 2000; Brusoni et al., 2001; Dibiaggio, 2007), drawing on the resource-based and competence-based perspective of the firm (Penrose, 1959; Prahalad and Hamel, 1990; Dosi et al., 1999), have argued that both in transaction-cost and modularity approaches, the boundaries of the firm are defined on the basis of what companies do and not on what companies know. These authors – as we do in this chapter – keep products and technologies distinguished, whereby the latter are ‘understood as the bodies of knowledge, or understanding and practice, that underpin product design and manufacturing’ (Brusoni et al., 2001: 597). This distinction allows them to illustrate, through empirical analysis, how companies producing complex products or systems have progressively increased their level of in-house technological capabilities, while augmenting the outsourcing of components. For example Prencipe (1997, 2004) shows that companies in the aeroengine industry insource certain component technologies and R&D activities in order to conserve the possibility of generating new knowledge, and not for

Technology fusion and organizational structures

145

reasons of efficiency. Thus in his view the explanation of vertical integration decisions based only on economic factors are not appropriate. In his words, technological knowledge is, in fact, ‘localised’, path-dependent in its development and context-dependent. This means that ‘contexts’ provide the sine qua non for generating new knowledge. As a result, hiving off components technologies (and even worse, R&D activities) means losing contexts which generate new knowledge or, in other words, damaging the firm’s change-generating capacity. (Prencipe, 1997: 1274)

At the same time Prencipe finds that there is an increasing outsourcing of components through broad use of agreements with external partners (Prencipe, 2004). Similar evidence has been found by other scholars in different sectors (Brusoni, 2001; Dibiaggio, 2007; Miller et al., 1995; Von Tunzelmann, 1998). All these studies have contributed to making clear that the division of labour and the division of knowledge can follow different paths. In particular the level of decomposability of products and knowledge can be different and thus lead to diverse organizations. As a consequence the boundaries of the same firm can differ sharply when considered from the perspective of physical production, or in terms of knowledge production. For these reasons, and on the basis of the evidence illustrated above, a new model of industrial organization has been theorized (Miller et al., 1995; Prencipe, 2003; Prencipe et al., 2004). The systems integration branch of literature points out that in several industries producing complex products there are leading firms that act as integrators. These firms manage a broad network of suppliers, often hierarchically organized in tiers, while keeping in-house their products’ design and development as well as the production of key components. We have pointed out that verticalization should be seen in terms of degree and mentioned factors in the traditional explanation of firm boundaries that enhance or limit vertical integration. Now we would like to highlight elements that can enhance verticalization and de-verticalization in design, R&D activities and production, from other perspectives that, differently from traditional views, consider the nature of knowledge-intensive activities. For the reasons stated in the systems-integration approach, we shall discuss separately the activities related to production and R&D activities. To begin, we stress that the focus is on firms that deal with problems and products that can be defined as complex. While the definition of complex products has been already provided, complex problems will be delineated in the following part. However in both products and problems the distinguishing element of complexity is the presence of numerous, strong interdependencies between components and sub-problems (Schilling, 2000;

146

Technological diffusion and interrelationships

Simon, 1981) which can have important organizational consequences, to be illustrated in the following. Problem Decomposability and the Organization of R&D Activities We present here some contributions that have attempted to provide an explanation of the relationship existing between problem-solving activity performed during R&D activities and organizational structures. In particular we consider those contributions trying to determine to what extent cognitive labour accomplished during design activities can be decomposed into separate tasks, in order to reduce its complexity and develop it through collaborative agreements. The main reference among these contributions is to the seminal work of Herbert Simon (1981), related to the problem-solving strategy for the resolution of complex problems. According to Simon (1981) the extremely high level of interdependence between sub-problems or tasks is the main characteristic of complex problems. This means that those who deal with a complex problem have to consider contemporaneously several interdependent elements because any decision related to one component may affect the whole system or several other components. In fact, while single components might be well known and understood, the interdependencies between them are instead in large part unknown. For this reason Simon (ibid.) points out that complex problems are characterized by the difficulty in being decomposed into separate tasks or sub-problems to be solved independently of other components. According to Simon, complex problems can be solved by tackling the more peripheral issues and then moving towards the top or the centre of the system, advancing through trial and error. However, the level of problem complexity can vary with the level of decomposability into separate tasks. Simon (ibid.) in fact discusses the characteristics of decomposable and neardecomposable problems. In the first case the independence of the subsystems is total and the separation in different components is extremely easy, while in the second case there exists a certain level of functional interrelatedness among sub-problems and a complete separability is not possible. The issue of to what extent problems can be decomposed, the positive as well as negative aspects deriving from decomposition, and the organizational structures linked to the level of problem decomposability is under discussion. For example Marengo et al. (2000, 2005) consider the degree of problem decomposability as a continuum, ranging from problems fully decomposed into a set of the smallest problems, to totally undecomposed problems. The authors point out that there is a trade-off in enhancing problems’ decomposition and that this can affect the organization of

Technology fusion and organizational structures

147

problem-solving activities. While the decomposition can reduce the search space, thus leading to tackling simpler problems, at the same time it reduces the range of possible solutions, and even to the point of eliminating the possibility of optimality. We wish here to underline particularly the explicit attempt in these contributions to link the level of problem decomposability to organizational structures, as these authors point out: ‘in our perspective diverse organisational forms map into (1) problem representations, (2) problem decomposition, (3) task assignment, (4) heuristics for and boundaries to exploration and learning, and (5) mechanisms for conflict resolution over interests but also over alternative cognitive frames and problem interpretations’ (ibid., p. 783). These elements contribute to developing diverse organizational structures, that can vary from ‘an archetype involving complete, hierarchical representations, precise task assignment; tight boundaries to exploration . . .’ (ibid.: 783) to ‘somewhat similar to a university department, with a number of representation at least as high as the number of individuals, fuzzy decompositions and conflict resolution rules, little task assignment and loose boundaries to exploration’ (ibid.: 783). The impact of the degree of problem decomposability on the organisational structures has been also empirically explored. For example, Valentin and Jensen (2003), in studying the case of biotechnology in food processing, have shown that Simon’s theory can have important consequences for the organization of R&D activities: in problems composed of highly interdependent tasks, R&D activities will tend to be developed in-house, rather than with external partners. According to these empirical results therefore, R&D development through the collaboration of different organizations is possible only in cases of highly decomposable problems. The main point we retain from the contributions illustrated in this section is that companies, in dealing with complex problems, may have to consider the possibilities for decomposing them into separate subproblems. Because the level of problem decomposability can vary, companies can face cases of full separability, total undecomposability as well as intermediate levels. Therefore we can expect, as Valentin and Jensen (2003), that companies tend to develop in-house design and R&D activities when the technologies embedded in the developing products are highly interdependent, but where the level of mutual interrelatedness is lower, a higher degree of collaboration among partners. Product Decomposability and the Organization of Production In order to provide a conceptual framework for exploring to what extent a system can be decomposed and thus produced through collaboration

148

Technological diffusion and interrelationships

between firms, we rely on the contribution of Brusoni et al. (2001) based on the concept of loosely coupled systems (Weick, 1976; Orton and Weick, 1990). As Orton and Weick claim (1990: 204). This concept was developed to explain the fact that organizations appear to be both closed, rational, and determined systems dealing with known environments, as well as open systems able to tackle uncertainty. According to these authors (ibid.), apparently incompatible behaviours are actually complementary and able to coexist within the same organizations, which in such cases can be described as loosely coupled systems. Here they provide the definition of this kind of organisational structure: loosely coupling suggests that any location in an organisation . . . [c]ontains interdependent elements that vary in the number and strength of their interdependencies. The fact that these elements are linked and preserve some degree of determinacy is captured by the word coupled in the phrase loosely coupled. The fact that these elements are also subject to spontaneous change and preserve some degree of independence and indeterminacy is captured by the modifying word loosely. The resulting image is a system that is simultaneously open and closed, indeterminate and rational, spontaneous and deliberate. (ibid.: 205).

Loosely coupled systems can be seen as in an intermediate position between tightly coupled and decoupled systems, and their distinctiveness and responsiveness are two co-varying factors that contribute to shaping an organization after one of these models. The former indicates the degree of specialization of the single independent components of the system, while the latter, the degree of integration between the interdependent components. Brusoni et al. (2001) applied this approach to the study of multitechnology and multi-component companies in order to explore the determinants of their organizational structures. In particular they posit that the degree of predictability of product interdependencies and the rate of change of component technologies are the two key elements determining which organizational structure, among those listed by Orton and Weick (1990), is going to prevail. As Brusoni et al. (2001: 611) illustrate in a matrix, when system interdependencies are known and predictable, and the rate of change of component technologies is even, the system is decoupled. In fact, these circumstances are similar to those generated by modularity: with stable interfaces there is no specific need for integrators, since the integration is obtained automatically once the interfaces are established and maintained. The opposite is the case when the product interdependencies are unpredictable and the rate of change of component technologies is uneven. Then product decomposability is almost impossible, thus we have tightly coupled

Technology fusion and organizational structures

149

systems, where production is realized via vertical integration. These instances are particularly common where some of the technologies embedded in products are new, implying that the interdependencies have to be understood and there is a wide range of possible solutions that have to be considered (Chesbrough, 2003). In the remaining two cases, either when the interdependencies are unpredictable, but the rate of change of component technologies is even, or when this is uneven but with predictable interdependencies, systems integration is the more likely organizational structure to emerge. On the basis of the approach proposed by Brusoni and colleagues (ibid.), companies can of course change their organizational choices over time, according to the dynamics of technological change, for example moving from tightly coupled to decoupled or loosely coupled systems. We retain two main points from these contributions. The first one is clearly stated by Pavitt (2003, p. 79): ‘the appropriate organisational process for generating and exploiting advances in technological knowledge – in particular, markets or integration – have been heavily influenced by the nature of technical change itself’. In other words the nature and characteristics of technological change can directly influence companies’ organizational structures. This phenomenon has not been captured by traditional explanations of the boundaries of firms. The different features of technological change in different historical periods have therefore pushed towards diverse and changing coordination mechanisms over time. The second point is related to the case of mechatronics. As explained in the introduction to this chapter, in technology fusion the level of interdependencies between technologies is particularly high, and mechatronics was generated by the blend of the quite stable technology of mechanics and two, newer and fast-changing technologies – electronics and informatics. Therefore at least in the initial phases of mechatronic production we can expect that product interdependencies still have to be fully understood and that coordination within companies thus occurs in tightly coupled systems which later might evolve into more open organizational forms.

METHODOLOGY AND EMPIRICAL RESULTS Data and Methodology The present work draws on data collected during the second phase of a larger research activity. The whole research project, begun in 2003, intended to capture the extent and characteristics of the evolution of a traditionally mechanical industrial district4 towards the production of

150

Technological diffusion and interrelationships

mechatronics. The district under analysis is the Province of Reggio Emilia (Emilia Romagna region of northern Italy). This region is characterized by strong specialization in manufacturing. In its last official report, Unioncamere Emilia-Romagna (2005) identifies, on the basis of production specialisation, the 18 most industrialized regions of Europe with large shares of manufacturing in total value-added. Within this group, Emilia Romagna is third in terms of GDP per capita and second in terms of employment rates. Emilia-Romagna reports a very low percentage of hightech industries,5 and the shares of low-, medium-low and medium-high tech are equally distributed around 30 per cent. In this region there are 24 officially recognized industrial districts, and six of them are specialized in the production of mechanics. Two of these six mechanical districts have been located in Reggio Emilia Province since the end of the Second World War. The entire project proceeded through the following steps: 1. 2.

3.

realization of a census of all mechatronic companies located in the traditionally mechanical district under analysis; effectuation of a survey among mechatronic companies to assess the level of integration among different technologies, their organizational structures and economic performance, and so on; and qualitative, micro-level analysis of the determinants and possible evolution of the organizational structures of mechatronic companies.

The procedures followed for the census and survey and the empirical results obtained, which constitutes the first phase of the overall project, have been already described and reported elsewhere (Freddi, 2007, 2008). In the next part we just briefly recall the main results obtained in this phase, in order to provide a background to the qualitative analysis that represents the bulk of the present work. The analysis reported here can stand alone, however the use of both quantitative and qualitative evidence is useful to obtain a solid construct validity (Eisenhardt, 1989; Yin, 2003). To do this, three different sources of data have been used. First, a review of the technical literature related to mechatronics was made to help identify the characteristics of and problems caused by the process of technology fusion. Also, meetings with engineers and professors from the University of Reggio Emilia and Modena, specializing in mechatronics, were extremely useful in clarifying complex technical issues, in outlining the approach to the research project and providing definitions to the main concepts. Second, a census and a survey of mechatronic companies were conducted. The census counted 139 producers of mechatronic machines or components, called ‘core mechatronic companies’, and

Technology fusion and organizational structures

151

81 firms that were their direct suppliers (Industriali Reggio Emilia, 2004). Once the mechatronic companies were identified, they were interviewed to assess to what degree they integrate the different technologies, their economic and innovative performance, and so on. The third kind of data was gathered in the interviews. A comparative, holistic, multiple-case study (Yin, 2003) was performed to assess and comprehend changes in the organizational structures possibly resulting from the shift towards mechatronics. As we illustrate in the next part, the earlier phase of the research project revealed a high level of heterogeneity among firms in terms of degree of integration between different technologies and verticalization of the organizational structures. Therefore a total of ten companies, different from each other in these two aspects, was interviewed following an in-depth, semi-structured questionnaire. Companies were selected following a ‘replication’ rather than ‘sampling’ logic, in order to predict ‘contrasting results but for predictable reasons’ (Yin, 2003: 47). The interviews were conducted with technical design managers or, in the case of micro-companies, with the owners, and lasted approximately two hours. Interviewees were asked to report possible changes that occurred in their organization after the beginning of mechatronic production, on their original competencies and on their evolution, and on the main technical and organizational problems encountered in the process. The data obtained from the interviews were elaborated into case histories that could allow a better comprehension of the issues under analysis. Empirical Results As we reported elsewhere (Freddi, 2007, 2008) the quantitative analysis of the primary data collected through the survey of the firms identified via the census revealed two key results. First, there is a very high level of heterogeneity among companies in terms of the degree of integration of the different technologies. In fact it was possible to identify, through the weight that companies assigned to the importance of mechanics, electronics and informatics, the existence of three groups, which we name advanced integrators, weak integrators and the in-between group. Focusing our attention on the advanced integrators, we concluded that new companies appeared to enter the traditional mechanical cluster when mechatronics started to diffuse, and that the integration process seemed to take place mainly within companies rather than between companies. This was evident with respect to the design activity of mechatronic products (Freddi, 2008), while the conclusions related to the organisation of the production phase were slightly fuzzier (Freddi, 2007).

152

Technological diffusion and interrelationships

The evident results as well as the ambiguous conclusions gave rise to new issues that we tried to tackle thoroughly the in-depth case studies reported here. In particular we wanted to understand the determinants of the heterogeneous results, as well as to identify possible different paths and modes in the organizing of mechatronic design and production. For this reason out of the ten selected companies, two were weak integrators and eight were advanced integrators. Since in the overall research project we mainly focused on technologies, mechatronic producers could be very different in terms of size, products realized, origin and history. The weak integrators interviewed (Company A and B) share various characteristics: both were founded at the beginning of the 1970s; they produce mechanical components such as hydraulic control systems, gearboxes and pumps mainly for the agricultural sector; they are large (about 800 employees) and mechatronic production accounts only for 7 per cent of their turnover. Both maintain that at the moment they are not experiencing real integration, but just combining different technologies using applications developed for other sectors. They admit that they ‘take a mechanical product and then just add electronic equipment’. The main factor that pushed them towards the production of mechatronics was their willingness to meet some client requests and to sell complex systems. However these companies experience significant limits in the process of integration; in fact real mechatronic solutions are not yet reliable enough and too costly to be produced. As Company B stated: ‘real integration can be imagined but has not yet occurred because it is too expensive. In fact, real integration would imply a completely new design of the product. The costs associated with this choice are too high, considering the R&D required; our tests of the prototypes and products would not be, at least at the beginning, as reliable as they are in the industry at the moment. Some of our competitors have tried this, but then stopped and went back to traditional, well-known solutions’. In both cases the original competencies were and at the moment still are mainly in the mechanical field, even if these began slowly to evolve at the end of the 1990s. This is evident if we consider the specialisations of the engineers and technicians responsible for the design of mechatronic products: approximately 80 per cent are specialized in mechanics, and the remaining in electronics, informatics or mechatronics. However in both companies mechanics and informatics and, only in part, electronics are designed in-house because companies want to be able to control all the different technological aspects of the products. Also in terms of production, mechanics and informatics are fully realized internally while electronics is mainly externalized. Differently from the weak integrators, who appeared to have very similar histories and organizational structures, it is possible to separate the

Technology fusion and organizational structures

153

advanced integrators into two clearly distinguishable groups. Five advanced integrators were born as electronic companies while the remaining three originally had mechanical knowledge bases. The advanced integrators that in origin had mainly electronic competencies are very small (approximately 15 employees) and entered the market at the end of the 1980s for the realization of electronic applications. All of them originated as spin-offs of other electronic companies and the decision to begin a new activity was due to problems encountered in proposing innovative ideas in the previous companies. All these firms have progressively deepened their knowledge in electronics but in the meantime have also acquired knowledge related to mechanics and informatics. In some cases this evolution occurred almost immediately after the beginning of the activity, while for others the process started later. Mechatronic production accounts for approximately half of turnover in all cases. Because of the small size of this group of advanced integrators, their competencies in informatics and mechanics were acquired directly by their employees, without significant change in the companies’ structure. Apart from one firm, all these companies design mechatronic solutions with the cooperation of their clients, which are mainly mechanical companies. In other words, these companies can be seen as specialized suppliers of large mechanical companies that ask for mechatronic components to be inserted into larger systems. It is possible to say that there is a division of cognitive labour in the design phase between the group of advanced integrators and their clients. The companies do not consider this choice efficient, and in fact it is extremely time consuming and a full integration is limited by this sort of organization. Thus one of the companies even admitted aiming to be acquired by a mechanical company wanting to enhance its mechatronic production. The physical production of the mechanical components of mechatronic products is entirely realized by the clients or by suppliers. In fact, it is almost impossible for these companies to realize internally the production of mechanics. In fact this would imply the acquisition of extremely expensive physical capital. As mentioned, the three remaining advanced integrators were originally mechanical companies that have evolved towards mechatronics. It is noticeable that these companies are at three different stages: one of them (Company C) is at the very beginning of the process; another one (Company E) is at a very advanced level, while the third (Company D) is somewhere in the middle. In fact, Company C was founded in 1957 to produce grinders and woodworking machines, and started to integrate electronics and informatics ten years ago. Company C maintains that its machines are still mainly mechanical, however, it wants to enhance the integration process. To do this the company has employed electronic and

154

Technological diffusion and interrelationships

mechatronic engineers and have acquired an external technical office that until then had been working for them as an independent company. The acquisition was made in order to shape the competencies of the technical group according to their specific needs, that is, they specialized their general competencies in mechanics, electronics and informatics on the basis of its own needs. Currently the design of mechatronic products is fully internal; the company produces only the mechanical elements and buys on the market both electronic equipment and software. However it aims to produce informatics internally in order to enhance the integration of the different technologies. Similarly Company D was founded at the beginning of the 1960s to produce machines for the ceramic sector. Fifteen years ago mechatronic production was started. This implied employing new computer and electronics engineers as well as upgrading the competencies of existing personnel. The major problems encountered in the integration process were related to the ability to make all the different technologies work together, ‘as [if] they were musicians that have to play in the same orchestra’. For this reason mechatronic solutions are designed only internally, while production is focused on mechanics and informatics. Finally, Company E produces end-of-line automation solutions and handling systems with palletizing robots, laser-guided vehicles and high-speed wrappers, and reaches a very high level of integration of mechanics, electronics and informatics. In fact due to the complexity of the systems, full integration is required. The company was founded in the early 1980s with mainly mechanical competencies and a few years later completely transformed itself into a mechatronics company. To do that it acquired a local company producing robots and employed a large number of technicians and engineers (which presently is 50 per cent of total employees, with 70 per cent of them specialized in informatics). It is interesting to note that this company is similar to all the other companies interviewed in that it develops completely internally the design of mechatronic products, but production is almost fully externalized. In short, it acts as an integrator of different technologies and components or sub-systems. In conclusion, the qualitative analysis revealed the existence of very different paths that lead to mechatronic production. Considering the origin of the companies, a sort of convergence of different knowledge bases (mainly mechanical or electronic) towards mechatronics is noticeable. In fact the traditionally mechanical companies acquired the competencies related to electronic and informatic fields, while electronic firms increased their knowledge in the mechanical area. Almost all companies, independently of their previous history, design mechatronic products fully inhouse. This result is consistent with our expectations on the basis of

Technology fusion and organizational structures

155

Simon’s (1981) contribution. In other words, it is evident that in order to manage and blend together different technologies, which implies resolving a large number of interdependencies, it is necessary to directly manage the entire design phase. If we consider instead the production phase, the picture is more complex. In general we can conclude that mechatronic production can be organized via decoupled or loosely coupled systems. The companies that have reached an advanced level can focus their activity on integrating technologies and systems rather than directly manufacturing them. In line with the findings of Brusoni et al. (2001), during the first phases of mechatronic production companies’ organizational structures could be described instead as tightly coupled systems. Only at the beginning in fact, as the companies stated, were product interdependencies hard to predict and the rate of change of component technologies was uneven. Thus they internalized the production of the unknown applications mainly to acquire experience and knowledge. However we stress that in our analysis, this path was not viable for all firms. In fact, while this kind of evolution was possible in companies with a mechanical engineering origin, electronic firms were limited by the excessive costs of physical capital.

CONCLUSIONS In this chapter we discuss the relationship between problem and product complexity and companies’ organizational structures. In doing so we point out that the traditional and widely applied explanations of the boundaries of the firm, transaction costs and modularity appear to be inadequate to explain productive organizations. Therefore we present the main building blocks of the systems-integration branch of literature because in our opinion, it offers a better framework for the analysis of organizational structures in the case of high-level problem and product complexity. In particular we illustrate that, according to some authors, the level of complexity of problems and products can affect respectively the organisation of R&D activities and production activities. In particular we emphasize that in the case of highly complex problems and products – implying a very high level of interdependency of tasks and components – organizational structures tend to be more integrated. Our evidence reveals that this is true of R&D and design activities of mechatronic products, while it applies to production only in the initial phases. Therefore we conclude that in our analysis technology fusion appears to occur within companies with regard to what companies know, while it takes place mainly between companies in terms of what companies do.

156

Technological diffusion and interrelationships

NOTES 1. According to the definition provided by the OECD (1994), high-tech sectors are those in which R&D intensity is above 5 per cent, while in medium high-tech sectors it ranges from 3 to 5 per cent, in medium low-tech sectors between 0.9 and 3 per cent, and in low-tech sectors it is less than 0.9 per cent. 2. The possibilities offered by the application of high-tech products, such as IT, in existing industries have been extensively studied, focusing in particular on possible improvements in companies’ administrative functions (Sriram et al., 1997; Scott and Westbrook, 1991; Sher and Lee, 2004) The focus of our analysis differs sharply from these previous contributions: we analyse the integration of different technologies, seen as separate bodies of knowledge, that allow the transformation of already existing products or the creation of completely new ones. 3. In this chapter we define technologies following the approach proposed by Dosi (1982: 152), where technology is defined ‘as a set of pieces of knowledge, both directly “practical” (related to concrete problems and devices) and “theoretical” (but practically applicable although not necessarily already applied), know-how, methods, procedures, experience of success and failures’. This definition implies that products do not overlap with technologies, but can be seen instead as expressions and embodiments of technologies. 4. Industrial districts are identified through a precise methodology set forth in the Italian National Law number 317 of 1991. On the basis of this methodology which considers several aspects such as the size of firms and industrial specialisation the National Institute of Statistics calculates the number of industrial districts located on Italian territory. 5. According to Eurostat classification (based on the OECD taxonomy), industries with NACE codes from 15 to 22 and from 36 to 37 are defined as low-tech; with codes 23 and 25 to 28 as medium-low; with codes 24, 29, 31, 34 and 35 as medium-high tech; and with codes 30, 32 and 33, as high-tech.

REFERENCES Baldwin, Carliss Y. and Kim B. Clark (2000), Design Rules. The Power of Modularity, Cambridge, MA: MIT Press. Baldwin, Carliss Y. and Kim B. Clark (2004), ‘Modularity in the design of complex engineering systems’, Harvard Business School working paper, January. Berardinis, L.A. (1990), ‘Mechatronics: a new design strategy’, Machine Design, 62 (8), 50–8. Brusoni, Stefano (2001), ‘The division of labour and the division of knowledge: the organisation of engineering design in the chemical industry’, unpublished PhD thesis, SPRU, University of Sussex at Brighton. Brusoni, S. and A. Prencipe (2001), ‘Unpacking the black box of modularity: technology, products and organisation’, Industrial and Corporate Change, 10 (1), 179–205. Brusoni, S., A. Prencipe and K. Pavitt (2001), ‘Knowledge specialization, organisational coupling, and the boundaries of the firm: why do firms know more than they make?’, Administrative Science Quarterly, 46, 597–621. Carlsson, Bo (ed.) (1997), Technological Systems and Industrial Dynamics, Boston, MA: Kluwer Academic. Carlsson, B. and R. Stankiewicz (1991), ‘On the nature, functions and composition of technological systems’, Journal of Evolutionary Economics, 12, 93–118.

Technology fusion and organizational structures

157

Chesbrough, Henry W. (2003), ‘Towards a dynamics of modularity. A cyclical model of technical advance’ in Andrea Prencipe, Andrew Davies and Mike Hobday (eds), The Business of Systems Integration, Oxford: Oxford University Press, pp. 174–98. Coase, R.H. (1937), ‘The nature of the firm’, Economica, 4 (16), 386–405. Dibiaggio, L. (2007), ‘Design complexity, vertical disintegration and knowledge organisation in the semiconductor industry’, Industrial and Corporate Change, 16 (2), 239–68. Dosi Giovanni, Richard R. Nelson and Sidney Winter (1999), The Nature and Dynamics of Organisational Capabilities, Oxford: Oxford University Press. Dosi, G. (1982), ‘Technological paradigms and technological trajectories’, Research Policy, 11, 147–162. Dosi, Giovanni, Mike Hobday, Luigi Marengo and Andrea Prencipe (2003), ‘The economics of systems integration. Towards an evolutionary interpretation’, in Andrea Prencipe, Andrew Davies and Mike Hobday (eds), The Business of Systems Integration, Oxford: Oxford University Press, pp. 95–113. Eisenhardt, K.M. (1989), ‘Building theories from case study research’, Academy of Management Review, 14, 532–50. Freddi, Daniela (2007), ‘In which ways and through which channels do firms in traditional mechanical clusters learn, absorb and apply radical new technologies? Evidence from an Italian mechatronic cluster’, paper presented at the DRUID Summer Conference, (Copenhagen, June). Freddi D. (forthcoming), ‘The integration of old and new technological paradigms in LMT sectors: the case of mechatronics’, Research Policy, special issue on ‘Technological Change in Low- and Medium-Technology Industries’. Grossman S.J. and O.D. Hart (1986), ‘The cost and benefit of ownership: a theory of lateral and vertical integration’, Journal of Political Economy, 94 (4), 691–719. Hart, O.D. and J. Moore (1990), ‘Property rights and the nature of the firm’, Journal of Political Economy, 98 (6), 1119–58. Hewitt, J. (1993), ‘Mechatronics: more than just a name’, The Industrial Robot, 20 (6), 3–5. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2003), ‘Low-tech industries and the knowledge economy: state of the art and research challenges’, EU 5th Framework project, ‘Pilot: Policy and Innovation in Low-tech’, online, accessed October 2007 at www.pilotproject.org/publications/publications.html. Hobday, M. (1998), ‘Product complexity, innovation and industrial organisation’, Research Policy, 26, 689–710. Industriali Reggio Emilia (2004), La meccatronica a Reggio Emilia, Collana ‘Meccanica e Meccatronica’ dell’Associazione Industriali di Reggio Emilia. Klein, B., R. Crawford and A. Alchian (1978), ‘Vertical integration, appropriable rents and the competitive contracting process’, Journal of Law and Economics, 21 (2), 297–326. Kodama, Fumio (1992), ‘Japan’s unique capability to innovate: technology fusion and its international implications’, in Thomas S. Arrison, C. Fred Bergsten, Edward M. Graham and Martha Harris (eds), Japan’s Growing Technological Capability. Implications for the US Economy, Washington, DC: National Academy Press, pp. 147–64. Langlois, R.N. and P.L. Robertson (1992), ‘Networks and innovation in a modular system: lesson from the microcomputer and stereo component industries’, Research Policy, 21, 297–313.

158

Technological diffusion and interrelationships

Marengo, Luigi, Corrado Pasquali and Marco Valente (2005), ‘Decomposability and modularity of economic institutions’, in Werner Callebaut and Diego RasskinGutman (eds), Modularity: Understanding the Development and Evolution of Natural Complex Systems, Cambridge, MA: MIT Press, pp. 383–408. Marengo, L., G. Dosi, P. Legrenzi and C. Pasquali (2000), ‘The structure of problem solving knowledge and the structure of organisations’, Industrial and Corporate Change, 9 (4), 757–88. Miller, R., M. Hobday, T. Leroux-Demers and X. Olleros (1995), ‘Innovation in complex systems industry: the case of flight simulation’, Industrial and Corporate Change, 4, 363–400. Monteverde, K. and D.J. Teece (1982), ‘Supplier switching costs and vertical integration in the automobile industry’, Bell Journal of Economics, 13 (1), 206–13. Mowery, D. (1983), ‘The relationship between intrafirm and contractual forms of industrial research in American manufacturing 1900-1940’, Exploration in Economic History, 20, 351–74. Organisation for Economic Co-operation and Development (OECD) (1994), Science and Technology Policy-Review and Outlook 1994, Paris: OECD. Orton J.D. and K.E. Weick (1990), ‘Loosely coupled systems: a reconceptualization’, Academy of Management Review, 15, 203–23. Pavitt, Keith (2003), ‘Specialization and systems integration. Where manufacture and services still meet’, in Andrea Prencipe, Andrew Davies and Mike Hobday (eds.), The Business of Systems Integration, Oxford: Oxford University Press, pp. 78–91. Penrose, Edith (1959), The Theory of the Growth of the Firm, London: Blackwell. Pisano, G. R. (1990), ‘The R&D boundaries of the firm: an empirical analysis’, Administrative Science Quarterly, 35, 153–76. Prahalad, C. and G. Hamel (1990), ‘The core competence of the corporation’, Harvard Business Review, 68 (3), 79–92. Prencipe, A. (1997), ‘Technological competencies and product’s evolutionary dynamics: a case study from the aero-engine industry’, Research Policy, 25 (8), 1261–76. Prencipe, A. (2000), ‘Breadth and depth of technological capabilities in CoPS: the case of the aircraft engine control system’, Research Policy, 29 (7–8), 895–911. Prencipe, Andrea (2003), ‘Corporate strategy and systems integration capabilities. Managing networks in complex systems industries’ in Andrea Prencipe, Andrew Davies and Mike Hobday (eds), The Business of Systems Integration, Oxford: Oxford University Press, pp. 114–32. Prencipe, Andrea (2004), Strategy, Systems and Scope: Managing Systems Integration in Complex Products, London: Sage. Prencipe, Andrea, Andrew Davies and Mike Hobday (eds) (2003), The Business of Systems Integration, Oxford: Oxford University Press. Richardson, G.B. (1972), ‘The organisation of industry’, The Economic Journal, 82 (327), 883–96. Robertson, P.L. and R.L. Langlois (1995), ‘Innovation, networks, and vertical integration’, Research Policy, 24, 543–62. Sanchez, R. and J.T. Mahoney (1996), ‘Modularity, flexibility, and knowledge management in product and organisation design’, Strategic Management Journal, 17 (Winter special issue), 63–79. Sandven, Tore, Keith Smith and Aris Kaloudis (2005), ‘Structural change, growth and innovation: the roles of medium and low-tech industries, 1980-2000’, in

Technology fusion and organizational structures

159

Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), Lowtech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang. Sher P.J. and V.C. Lee (2004), ‘Information technology as a facilitator for enhancing dynamic capabilities through knowledge management’, Information & Management, 41(8), 933–45. Schilling, M.A. (2000), ‘Towards a general modular system theory and its application to inter-firm product modularity’, Academy of Management Review, 25, 312–24. Scott, C. and R. Westbrook (1991), ‘New strategic tools for supply chain management’, International Journal of Physical Distribution & Logistics Management, 21, 23–33. Simon, Herbert (1981), The Sciences of the Artificial, 2nd edn, Cambridge, MA: MIT Press. Sriram, V., R.L. Stump and S. Banerjee (1997), ‘Information technology investments in purchasing: an empirical study of dimensions and antecedents’, Information & Management, 33, 59–72. Tomkinson, D. (1991), ‘Mechatronics: a new flavor in integration’, Automation, 38 (12), 24–8. Ulrich, K.T. (1995), ‘The role of product architecture in the manufacturing firm’, Research Policy, 24, 419–40. Unioncamere, Emilia Romagna (2005), Rapporto sull’economia regionale 2005 e previsioni per il 2006, Ufficio Studi Unione Camere di Commercio, ondine, accessed October 2007 at www.rer.camcom.it/datiestudi.html. Valentin, F. and R.L. Jensen (2003), ‘Discontinuities and distributed innovation: the case of biotechnology in food processing’, Industry and Innovation, 103, 275–310. Von Tunzelmann, G.N. (1998), ‘Localised technological search and multitechnology companies’, Economics of Innovation and New Technology, 6, 231–55. Von Tunzelmann, Nick and Virginia Acha (2005), ‘Innovation in “low-tech” industries’ in Jan Fagerberg, David C. Mowery and Richard R. Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 407–32. Weick, K.E. (1976), ‘Educational organisations as loosely coupled systems’, Administrative Science Quarterly, 21, 1–19. Williamson, Oliver E. (1975), Markets and Hierarchies, New York: Free Press. Williamson, Oliver E. (1985), The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting, New York: Free Press. Yin, Robert K. (2003), Case Study Research: Design and Methods, 3rd edn, London: Sage.

9.

Industrial innovations in relation to service sectors Marja Toivonen

INTRODUCTION Services play a dominant role in the present-day economic landscape. They are also pivotal sources of growth for production, and particularly for employment, within ‘post-industrial economies’. Besides the growth taking place in service sectors, services play a critical role in the advancing development of the manufacturing sector. The underlying reason is that the share of intangible assets – relationships, information and knowledge – as sources of value for firms is increasing in relation to tangible assets – physical materials and goods (Tomlinson, 1997; Hipp and Grupp, 2005; OECD, 2006). Linking services to manufacturing is not a new phenomenon if we mean services that support products and production. Many industrial firms have for decades produced or purchased marketing and sales services as well as services linked to research and development or human resources management (Martinelli, 1991; Mathieu, 2001a). What is new is the idea that services can be an important source of competitive advantage in manufacturing. Industrial firms are today increasingly adopting strategies in which services do not fulfil a supporting function but are seen as a key growth area for future businesses. Services are sold either separately or, more commonly today, combined with physical products into integrated solutions (Davies, 2003a; Howells, 2000). In addition to the increasing importance of intangible assets, manufacturers have several other reasons to favour service-oriented strategies. First, the markets for many manufactured products are maturing. Second, the production of services provides steadier revenue, as many service sectors are less sensitive to economic fluctuations than manufacturing, and cash flows from different types of sectors can balance the effects of economic cycles. Third, customers increasingly desire a fuller coverage of their needs, which can be provided by goodsservices combinations. Fourth, services provide opportunities to deepen 160

Industrial innovations in relation to service sectors

161

and prolong the relationships of manufacturing firms with customers (Davies, 2003b). There is a growing interest in the process whereby manufacturers become service providers. Researchers have warned against too simplistic thinking in this respect. Services should not be linked to the existing business as ‘add-ons’, but the business model of the company should be carefully reconsidered and its organization refocused (Brax, 2005). The need for this kind of approach derives from the characteristics of services, which deviate from those of goods: the inseparability of the process and the outcome, and the participation of the client in the production process (De Brentani, 1991; Sundbo and Gallouj, 2000). Moving to a service business requires innovative behaviour from the manufacturing company, whereby actual product, organizational or market innovations may also emerge. This chapter discusses in more detail the possibilities and challenges linked to the combination of services and manufacturing. Our focus is on the production of services in manufacturing, but we also briefly touch upon new features in the purchase of services by manufacturers. We start by introducing a service model which makes clear the special nature of services and how they differ from goods. Thereafter we present different ways in which goods and services can be combined into value-producing offerings. In the next section we move from the product level to the business level: we analyse the issues emerging in the transition from pure manufacturing to the provision of services. The business-level analysis also includes the opposite scenario: the purchase of services from outside. Throughout these analyses, our perspective is that of innovation: what can be renewed and how the renewals can be made. (In this chapter, we use the terms ‘renew’ and ‘renewal’ to refer to [seeking] new solutions, not to bringing back the original state of things.)

THE NATURE AND COMPONENTS OF A SERVICE Services can be defined as activities or performance provided to satisfy customer needs (Grönroos, 1990). In services the production process directly contributes to the benefit provided to the client, whereas in manufacturing the production and delivery systems are separate from the benefitproviding outcome. Services as entities thus differ considerably from goods. Several researchers have aimed to describe the basic components of services, that is, to model a service. Here we present the model of Edvardsson and Olsson (1996; see also Edvardsson, 1997) which suits particularly well those service studies including the perspective of innovation and new service development.

162

Technological diffusion and interrelationships

The starting point of the Edvardssonian model is that each individual service consists of both a customer-perceived outcome and a customerunique process. Besides the perspective of the customer, the service can be considered also from the perspective of the provider. Here the question is not of the actual provision of the service – the service company cannot do it without the customer. What it can do, however, is to develop the best and right prerequisites for well-functioning customer processes and attractive customer outcomes. The prerequisites include three basic components: the service concept, the service process and the service system. The term ‘service concept’ refers to the description of the customer’s needs and how they are to be satisfied. The content and structure of the service are specified here, including, for instance, the specification of the core service and supporting services. The service process is the prototype for every customer process and describes the chain of activities that must function properly when the service is actually produced. The service system constitutes the resources that are required by, or are made available to, the service process in order to realize the service concept. It includes components like the service company’s staff, the physical/technical environment and the organizational structure; also customers function as a resource. The model is summarized in Figure 9.1, which we have redrawn on the basis of the texts and figures from the above-mentioned articles (Toivonen et al., 2007). The Edvardssonian model shows that there are abundant possibilities to renew a service. Any one of the different components included may be the target of innovation. The model can be supplemented with an analysis of the nature of the change. Gallouj and Weinstein (1997; see also Gallouj, 2002) have identified the following ways to renew a service: ● ● ● ● ●

improvement of some individual component; addition/subtraction of some component; substitution of some component; recombination of components from different services; and clarification of the way in which different service components contribute to the benefit provided to the client.

The combination of Edvardsson’s and Gallouj’s views provides us with a reasonable tool for examining innovations in ‘pure’ services. In addition to the renewals that can be made in services as such, a manufacturer moving into the service business can find innovation opportunities in different kinds of combinations of goods and services. In the next section we consider these opportunities.

Industrial innovations in relation to service sectors

163

The perspective of the customer and actual service provision

UNIQUE OUTCOME PERCEIVED BY THE CUSTOMER

UNIQUE PROCESS WHERE THE CUSTOMER PARTICIPATES

PREREQUISITES SERVICE CONCEPT

SERVICE SYSTEM

SERVICE PROCESS

Analysis of the customer’s needs and the ways in which they are met; the content and structure of the service

Resources: staff, physical/ technical environment, the organizational structure; customers as a resource

Prototype of customer processes, describes the chain of activities

The perspective of the service provider and service development

Source:

Toivonen et al. (2007).

Figure 9.1

The service model of Edvardsson et al., redrawn

DIFFERENT WAYS TO COMBINE GOODS AND SERVICES Kotler (2003) has distinguished five basic possibilities for goods-services combinations: (a) pure tangible goods; (b) tangible goods with accompanying services; (c) hybrids where goods and services are equal in balance; (d) major service with accompanying goods and services; and (e) pure services. Today, manufacturing firms typically offer tangible goods with accompanying services, but the trend is towards the hybrid form; for some manufacturers, it is towards the product with a major service focus (Brax, 2005). Industrial services have been typified in several ways. On the basis of the relation of these services to physical products, Homburg and Garbe (1999) make a division into pre-purchase, at-purchase and after-sales services. Prepurchase services include, for example, different kinds of analysis services. At-purchase services cover, among others, different kinds of logistics and storing services. Repair and maintenance are typical after-sales services. Adopting new services in the purchasing context has been referred to as a spatial expansion or reconfiguration of the offering, whereas the adoption

164

Technological diffusion and interrelationships

of new after-sales services means that the offering is expanded or reconfigured temporally (Sawhney et al., 2004). Mathieu (2001b) focuses on the dimension of service specificity and makes a categorization into customer service, product services and service products. Customer service is aimed at facilitating a company’s sales at the general level; here the costs emerging are mostly treated as overhead costs. Product services are aimed at facilitating the sale of a product provided by the company and supporting its operation. The tangible good can be the price carrier for the bundled benefits and cover the costs of service, or the added value may be charged separately. Service products are independent from the company’s tangible offerings and can be purchased separately from other transactions. Product services are services which supplement a good, but which cannot be purchased separately. Here, it is primarily a good which satisfies the client’s need. The services included are of two types: enabling services and facilitating services. The former are essential for the benefit to be provided and gained, whereas the latter are not necessary – they can, however, improve the quality of the total offering in an important way. In a service product, the core as well as the enabling and facilitating parts are services. Yet, any one of these parts can include goods. Goods are commonly used as resources in service production, but they can also be an element in the final outcome (see Grönroos, 1990; Vaattovaara, 1999). The above-described categorizations show that there are abundant options for combining goods and services (material and immaterial products) in an innovative way. Even more opportunities can be found when the target of industrial services is included in the analysis. Paloheimo et al. (2004: 23) have divided these targets into three groups and present some examples of services belonging to each of them: ● ● ●

equipment – delivery, installation, calibration, maintenance, upgrades; equipment use in a process – user training, maintenance management, equipment availability, process optimization; business – equipment financing, asset management, consultancy.

The more the service concerns not only the equipment, but also its use, and business in a broader sense, the more attention deserves how the manufacturer as service provider can support its clients. This development direction of industrial services goes hand-in-hand with the general idea of value offerings. While analysis of the competitive conditions was earlier the usual starting point in the development of business, the main question set today is more and more often: What new value can a firm offer its clients and how

Industrial innovations in relation to service sectors

165

can it achieve this? Outperforming the competition and profitable growth result from successful value offerings. If a firm only concentrates on how to match or beat competitors, it easily restricts itself to the conventional context, which also the competitors know and therefore all seek to obtain a competitive advantage by means of minor improvements (Hoover et al., 2001; Kim and Mauborgne, 1999). In a strategy based on value offerings, an innovative orientation is inherent. This kind of strategy considerably extends the creative scope of individual firms and gives them a wide range of options, even irrespective of the general situation in their industry. A firm need not compete for a share of a given demand, but it can redefine clients’ problems and discover hidden demand through systematic pursuit of innovations. In these, the arrangement or rearrangement of existing items is often the core of renewal. The perspective should, however, be wide enough to include the crossing of conventional borders. For instance the application of operational models used in other industries can be beneficial if done skilfully (Kim and Mauborgne, 1999; Normann and Ramirez, 1998). The value-offering approach can lead to the production of so-called CoPS (complex products and systems), which are project-type offerings consisting of tangible goods and services. Turnkey-contracts of capital goods are a typical example here (Davies and Brady, 2000). The value offering approach can also lead to the provision of so-called solution offerings. A thorough analysis of the client’s needs may show that even complex combinations of goods and services are not enough, but that customers are interested in resolving their problems without taking the responsibility for organizing the means, which thus becomes the task of the provider. The provision of solutions involves a long-term orientation: the provider becomes more or less a part of the customer’s operations or business system (Brax and Jonsson, 2007). The provider of industrial services can also adopt different strategies as regards its positioning in the markets: it can change the value-offering point, that is, the point at which the offering is linked into the value chain of the client (Hoover et al., 2001). When the value-offering point is moved upstream in the demand process, it is often possible to increase both the client’s benefit and the provider’s profitability.

CONDUCTING SERVICE BUSINESS IN A MANUFACTURING FIRM Several researchers have described the typical steps that a manufacturing company takes when moving from goods-focused production to

166

Technological diffusion and interrelationships

service-oriented business. One of the most detailed analysis of the nature of this transformation is presented by Oliva and Kallenberg (2003: 161–5). According to them, the following recurring pattern of actions can be found in successful transitions along the goods-services continuum: 1. 2.

3. 4.

The companies consolidate their product-related services and often relocate services to a newly-created service unit. The companies enter the installed base (that is, their products possessed by customers) service market through defining and analysing this market, creating an infrastructure for marketing and delivering services, and responding to local service demand. The companies expand to relationship-based services, or they focus on process-centred services. In the final stage, the companies take over parts of the end-users’ operations.

Even when the move towards an increasing service focus takes place in steps or stages, it means an essential change in the business model of the company. The bigger the service element in the products of the company, the deeper this change is. Figure 9.2 presents a model combined by Brax (2007) from the analyses of Cova et al. (2000) and of Mathieu (2001a, b). Mathieu (2001a) categorizes industrial services into those supporting a good and those supporting customer activities in relation to the good. Cova et al. (2000) extend this division by adding a new category for those services that support not only the client, but also the stakeholder network surrounding the client. (They further divide services supporting the client into those that support action in relation to the provider’s goods and those without a direct link to the provided goods.) In Figure 9.2, the three-part (simpler) categorization of Cova et al. (2000) has been positioned in Mathieu’s (2001b) two-dimensional framework where service specificity is one dimension, and the organizational intensity of the service manoeuvre, the other. In the preceding section we have already discussed the dimension of service specificity as a characteristic of the service output. The dimension of organizational intensity describes the strength and scope of the service from the organization’s point of view. Like service specificity, the organizational intensity is divided into three levels: tactic, strategic and cultural. Considering the categorization of Cova et al., (2000) we can say that starting the provision of services that support a good usually means the least intense (tactic) transfer from the organizational perspective. At the other end are services that support not only the client, but also the network of the client – entering this kind of service business has the most intense cultural impact within the organization. Between the two

167

Industrial innovations in relation to service sectors

services supporting a client ’s network

Service as a product

Service specificity

services supporting a client

Product service

Customer service

services supporting a good Tactic

Strategic

Cultural

Organizational intensity

Source: Brax (2007), based on Cova et al. (2000) and Mathieu (2001a, b).

Figure 9.2

A model of service strategies in manufacturing

extremes are services that support the client; their adoption means a strategically new attitude in the manufacturing company. Mathieu (2001b) argues that the more specific and the more culturally intensive the service manoeuvre is, the greater the financial and other benefits. On the other hand, moving from goods-related services to customers’ business-related services is challenging. It implies, among other things, that the provider sells availability, that is, assumes some of the equipment-related risk that the customer normally carries. Part of the risk can be avoided or at least reduced by offering additional services to customers. Thus it is common that manufacturers provide their customers with instructions and training, offer tools for maintenance management, or monitor the condition of the equipment remotely (Paloheimo et al., 2004). At the same time as the trend towards services that support the customer’s business strengthens, the more traditional, goods-related services have also gained new features. Applying the life-cycle concept in the context of after-sales services has become more and more common today. Instead of one-off contacts via a product sale, manufacturers aim to extend their customer relationships to cover the whole life-cycle of products, which means that consumption is becoming a continuing process. This practice enables the service provider to offer customers some additional value.

168

Technological diffusion and interrelationships

Fulfilling what has been promised is however essential and requires careful consideration when the service contract is long-lasting. In a long-term perspective the optimization of total costs is one of the key issues (Cohen and Whang, 1997). Maintenance and other after-sales support services are an important part of life-cycle services, and they often positively influence the customer’s purchase decision (Cohen and Whang, 1997). Maintenance can be conducted on the basis of different approaches: run-to-failure, preventive maintenance, condition-based maintenance, and design improvement (Tsang, 2002). Despite the longer history of maintenance services, their production is challenging too. Brax (2005) has identified six types of challenges which are partially overlapping: the challenges of production, delivery, product-design, marketing, communication and relationship. The last three challenges have to do with the fact that, as distinct from goods, services require a common understanding of the product concept as well as customer motivation in coproducing the service. The vast amount of information needed about the installed base is one central challenge in the production and timing of delivery. In the product design of services, cultural differences have to be taken into account more seriously than in the production of goods. Brax concludes that even the first steps towards a service business require a change in the background philosophy. The transaction-oriented attitude, which can persist even when manufacturers adopt services into their product repertoire, is not supportive of successful service offerings. Regarding skills requirements, researchers have pointed out that good technical skills, emphasized in traditional manufacturing, are not enough in the service business; in addition, partnering competence and market knowledge become fundamental (Shepherd and Ahmed, 2000). When services are provided together with goods as bundles or solutions, integrative skills are also essential. Further, sector- and client-specific knowledge is needed particularly when the purpose is to support the client’s action. In the business-to-business context, the provision of services requires a comprehensive understanding of the value-creation process of clients and the role of services in it (see Hoover et al., 2001). A business model in which the realities of customers are taken as a starting point and are responded to with value offerings, is tightly linked with the development of networking practices. The many components included in value offerings, as well as the long-lasting nature of the client relationship, require that the provider has a network of partners who supply complementary assets, capabilities, products and services. (Kim and Mauborgne, 1999) The need for cooperation has to be analysed regarding both the supply and the demand chain. The supply chain involves subcontractors and horizontal networking partners; through collaboration

Industrial innovations in relation to service sectors

169

with these, efficiency in production and distribution of equipment and services can be achieved, maintained and improved.

THE PURCHASING OF SERVICES THAT SUPPORT MANUFACTURING As stated in the introduction, this chapter focuses mainly on service business conducted by manufacturers. However, the discussion of industrial innovations in relation to service sectors is incomplete if we neglect the ‘other side of the coin’ – the purchasing of services by manufacturers. More services that were earlier provided in-house in manufacturing companies are today acquired from external providers. The services may target the manufacturer’s core production process, other essential value-creating functions (research and development, sales and marketing, strategic planning or logistics), or the support functions (information technology, legal functions, human resources management or accounting) (Paloheimo et al., 2004). Since the mid-1980s there has been a great number of studies examining the reasons behind the expanding practice of outsourcing. Four main groups of factors influencing the ‘make or buy’ decisions in the service context have been identified (for a more comprehensive analysis and detailed references, see Toivonen, 2004): ● ● ● ●

cost-efficiency factors; factors connected with the quality of service; characteristics of the purchasing firm; characteristics of the service purchased.

Regarding cost-efficiency, two causes especially have been thought to speak in favour of outsourcing: scale economies in specialized free-standing service companies, and the avoidance of risks and fixed costs in purchasing companies. The latter means that outsourcing is preferred when the demand for a service is insufficient, sporadic or unpredictable, resulting in inefficient use of resources if organized on an in-house basis. As examples, legal counsel and tax accounting in small firms have been mentioned. Without denying the significance of costs, several researchers have suggested that access to knowledge and expertise is an even more important motive for outsourcing. Especially the increasing technical complexity and specialization of many service functions result in firms often lacking internal capability to meet the demand at a sufficient level of efficiency or quality. Of the characteristics of the purchasing firm, size has often been mentioned as a factor causing differences in the level of outsourcing. The use

170

Technological diffusion and interrelationships

of external services has been found to be most common in medium-sized establishments. Large companies can quite easily provide services in-house in a profitable way, which restricts their willingness to outsource. On the other hand, in small firms the use of external services is restricted by the undeveloped differentiation of functions. Outsourcing also varies according to the type of service. Some functions such as legal services have generally had a much higher level of externalization than others, for example accounting. The strategic significance of the service is, however, the most important characteristic influencing the outsourcing decision. Services vital to business or related to a firm’s core operations tend to be carried out in-house; this holds true also for services containing commercially confidential or firm-specific information. On the other hand, even though the use of external services is a ‘strategic decision’ in the sense described above, firms having an outsourcing strategy have long been a small minority. Most manufacturing firms have adopted a pragmatic purchasing policy, externalizing and internalizing services on an individual basis. In addition, cultural factors and established practices have played a role here: a great part of firms have continued on the basis they are used to, and undertake no change in the organization of their services. In recent years the situation has been changing in this respect. Outsourcing decisions have become more conscious and companies specializing in outsourcing consultancy have emerged (Bragg, 1998). It has been emphasized that the level of externalization is not the only significant issue. What matters is how the external markets are used, that is, the actual mode of organization of outsourcing is important. The outcome can be quite different depending on, for example, whether each element of a service is purchased separately, or whether a service is acquired as a whole or even bundled together with other services. In recent discussions, the different tasks included in skilful purchase and use of external services have also been analysed. Successful outsourcing includes, among others, the following tasks (see OECD, 2006): ● ● ● ● ● ●

recognizing the need for a service; defining the content of the assignment; finding an appropriate service provider; managing the cooperative relationship during the service provision; putting the outcomes of the service into use; evaluating the service process and outcomes.

There are abundant possibilities for innovative solutions when in-house and external services are combined. In addition, one group of external

Industrial innovations in relation to service sectors

171

services is in a special position from the viewpoint of innovation: the socalled knowledge-intensive business services (KIBS). These services are provided by expert firms in the areas of software and hardware consultancy; research and development; engineering and other technical consultancy; legal, financial and management consultancy; and marketing communications. Several studies have shown that the use of KIBS can facilitate innovation activities of the user-company in an important way. Besides the direct provision of expert knowledge, KIBS diagnose and clarify problems in user-companies, and act as agents for change by offering a neutral outside perspective. They also carry experiences and ideas from one context into another, identify good practices, and fulfil a brokering function (putting different knowledge sources and users in contact with each other) (Bessant and Rush, 1995; Miles, 1999).

CONCLUDING REMARKS In this chapter we have shown that moving to the service business offers manufacturing companies abundant innovative opportunities. Firstly, innovations can concern a service as such – its concept, process or resource properties. Second, there are numerous ways in which goods and services can be combined. Third, in addition to the product level, a successful adoption of services into the product repertoire almost necessarily leads to some renewal at the organizational level. Finally, there are innovative possibilities at the business level, as in the positioning of a firm in the markets. Until now we have not made explicit what we mean by innovation in services. A long discussion of this topic is not possible in the present context. However, we want to warn against too loose a ‘definition’ of innovation. Since Schumpeter, the concept of innovation has referred to particular economic phenomena having specific criteria. Even though Schumpeter concentrated on material goods, his thoughts are also useful when we discuss innovations in services. There are three criteria in particular, on the basis of which innovations can be separated from other kinds of renewal in organizations. First, innovation is something which is carried into practice (Schumpeter, 1934: 88). An idea without an application – and in the case of private firms, without the acceptance of the markets – is not yet an innovation. Second, innovation is something which provides benefit to its developer (Schumpeter, 1912/2002: 111). Distinctiveness as regards competitors and the accompanying entrepreneurial profit are central factors which motivate the pursuit of innovations. In more recent writings, the added value that innovations can provide to the customers is often emphasized. Taking the needs of customers as a starting point has been perceived to

172

Technological diffusion and interrelationships

promote the achievement of profits in a sustainable manner (Kim and Mauborgne, 1999). Third, innovation is something which is reproducible; in other words, it has more than one specific application (Schumpeter, 1934: 228). Whereas the first two criteria are important in the case of both goods and services, this third criterion is particularly relevant when we try to define a service innovation. (As regards goods, it is usually self-evident.) Even though each service act includes some unique features, not all of them are innovations. An essential feature in the Schumpeterian theory is that innovation is analysed as an economic concept. A true innovation is not only beneficial to the company that has created it, but as a way of doing things more viably it leads other companies to follow. Thus innovation gets diffused through imitation, and in this way promotes the development of the branch, even the development of the whole economy (Schumpeter, 1934; see also Drejer, 2004). In some innovation analyses an interpretation which clearly deviates from Schumpeterian thinking has gained ground: also solutions which are new to the firm adopting them are counted among innovations, even though already commonly used elsewhere (see Tidd et al., 2005: 12). Without denying the fact that ‘newness’ is always a relative concept, we argue that ‘new to a firm’ should be excluded from the definition of innovation because it leads to the strange conclusion that backward companies are innovating when they adopt well-known practices. However, we do not want to underrate the value of firm-specific renewals. They just belong to the sphere of a different phenomenon: continuous adaptation and the diffusion of innovations linked with it. Innovations represent a break in business-as-usual – a discontinuous change, however small it may be (Schumpeter, 1942: 83).

REFERENCES Bessant, J. and H. Rush (1995), ‘Building bridges for innovation: the role of consultants in technology transfer’, Research Policy, 24, 97–114. Bragg, S.M. (1998), Outsourcing, New York: John Wiley & Sons. Brax, S. (2005), ‘A manufacturer becoming service provider – challenges and a paradox’, Managing Service Quality, 15 (2), 142–55. Brax, S. (forthcoming), ‘Towards an understanding of the service offering’, Helsinki University of Technology, IMI working paper. Brax, S. and K. Jonsson (2007), ‘From products to integrated solution offerings in capital goods maintenance’, Helsinki University of Technology, IMI working paper no. 3. Cohen, M.A. and S. Whang (1997), ‘Competing in product and service: a product life-cycle model’, Management Science, 43 (4), 535–45.

Industrial innovations in relation to service sectors

173

Cova, B., E. Dontenwill and R. Salle (2000), ‘A network approach to the broadening of the offering’, paper presented at Industrial Marketing and Purchasing 2000 Conference, accessed at: www.bath.ac.uk/imp/conference/ htm#papers. Davies, A. (2003a), ‘Integrated solution: the changing business of systems integration’, in A. Prencipe, A. Davies and M. Hobday (eds), The Business of Systems Integration, New York: Oxford University Press. Davies, A. (2003b), ‘Are firms moving “downstream” into high-value services?’ in J. Tidd and F. M. Hull (eds), Service innovation – Organizational Responses to Technological Opportunities and Market Imperatives, London: Imperial College Press. Davies, A. and T. Brady (2000), ‘Organizational capabilities and learning in complex product systems: towards repeatable solutions’, Research Policy, 29, 931–53. De Brentani, U. (1991), ‘Success factors in developing new services’, European Journal of Marketing, 25 (2), 33–59. Drejer, I. (2004), ‘Identifying innovation in surveys of services: a Schumpeterian perspective’, Research Policy, 33, 551–62. Edvardsson, B. (1997), ‘Quality in new service development: key concepts and a frame of reference’, International Journal of Production Economics, 52, 31–46. Edvardsson, B. and J. Olsson (1996), ‘Key concepts for new service development’, The Service Industries Journal, 16 (2), 140–64. Gallouj, F. (2002), Innovation in the Service Economy: The New Wealth of Nations, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Gallouj, F. and O. Weinstein (1997), ‘Innovation in services’, Research Policy, 26 (4/5), 537–56. Grönroos, C. (1990), Service Management and Marketing, Lexington, MA and Toronto: Lexington Books. Hipp, C. and H. Grupp (2005), ‘Innovation in the service sector: the demand for service-specific innovation measurement concepts and typologies’, Research Policy, 34, 517–35. Homburg, C. and B. Garbe (1999), ‘Towards an improved understanding of industrial services: quality dimensions and their impact on buyer-seller relationships’, Journal of Business-to-Business Marketing, 6 (2), 39–71. Hoover, W.E., E. Eloranta, J. Holmström and K. Huttunen (2001), Managing the Demand-Supply Chain: Value Innovations for Customer Satisfaction, New York: John Wiley & Sons. Howells, J. (2000), ‘The nature of innovation in services’, report presented to the OECD Innovation and Productivity in Services Workshop, 31 October– 3 November, Sydney, Australia. Kim, W.C. and R. Mauborgne (1999), ‘Strategy, value innovation, and the knowledge economy’, Sloan Management Review, (Spring), 41–54. Kotler, P. (2003), Marketing Management, New York: Pearson Education. Martinelli, F. (1991), ‘A demand-orientated approach to understanding producer services’, in P.W. Daniels and F. Moulaert (eds), The Changing Geography of Advanced Producer Services, London and New York: Belhaven Press. Mathieu, V. (2001a), ‘Product services: from a service supporting the product to a service supporting the client’, Journal of Business & Industrial Marketing, 16 (1), 39–58.

174

Technological diffusion and interrelationships

Mathieu, V. (2001b), ‘Service strategies within the manufacturing sector: benefits, costs and partnership’, International Journal of Service Industry Management, 12 (5), 451–75. Miles, I. (1999), ‘Services in national innovation systems: from traditional services to knowledge intensive business services’, in G. Schienstock and O. Kuusi (eds), Transformation Towards a Learning Economy, The Finnish National Fund for Research and Development – Sitra, Report No. 213, Helsinki: Sitra. Normann, R. and R. Ramirez (1998), Designing Interactive Strategy – from Value Chain to Value Constellation, Chichester: John Wiley & Sons. Organization for Economic Co-operation and Development (OECD) (2006), Innovation and Knowledge-Intensive Service Activities, Paris: OECD. Oliva R. and R. Kallenberg (2003), ‘Managing transition from products to services’, International Journal of Service Industry Management, 14 (2), 160–72. Paloheimo, K-S., I. Miettinen and S. Brax (2004), Customer-Oriented Industrial Services, BIT Research Centre, Helsinki University of Technology. Sawhney, M., S. Balasubramanian and V.V. Krishnan (2004), ‘Creating growth with services’, Sloan Management Review, (Winter), 34–43. Schumpeter, J.A. (1912/2002), ‘The economy as a whole – seventh chapter of the theory of economic development’, Industry and Innovation, 9, 93–145. Schumpeter, J.A. (1934), The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle, Cambridge, MA: Harvard University Press. Schumpeter, J.A. (1942), Capitalism, Socialism and Democracy, New York and London: Harper & Brothers Publishers. Shepherd, C. and P.K. Ahmed (2000), ‘From product innovation to solutions innovation: a new paradigm for competitive advantage’, European Journal of Innovation Management, 3 (2), 100–6. Sundbo, J. and F. Gallouj (2000), ‘Innovation as a loosely coupled system in services’, in J.S. Metcalfe and I. Miles (eds), Innovation Systems in the Service Economy – Measurement and Case Study Analysis, Boston, MA, Dordrecht and London: Kluwer Academic Publishers. Tidd, J., J. Bessant and K. Pavitt (2005), Managing Innovation: Integrating Technological, Market and Organizational Change, Chichester: John Wiley & Sons. Toivonen, M. (2004), Expertise as Business: Long-term Development and Future Prospects of Knowledge-intensive Business Services (KIBS), Department of Industrial Engineering and Management, Helsinki University of Technology. Toivonen M., T. Tuominen and S. Brax (2007), ‘Innovation process interlinked with the process of service delivery – a management challenge in KIBS’, Economies et Sociétés, Economics and Management of Services series, 3/2007, 355–84. Tomlinson, M. (1997), The Contribution of Services to Manufacturing Industry: Beyond the Deindustrialisation Debate, Manchester University Centre for Research on Innovation and Competition (CRIC), working paper no. 5. Tsang, A.H.C. (2002), ‘Strategic dimensions of maintenance management’, Journal of Quality in Maintenance Engineering, 8 (1), 7–39. Vaattovaara M. (1999), Transforming Services into Products in a Systems Engineering Company, Helsinki University of Technology Department of Industrial Management and Work and Organisational Psychology report no. 9.

10.

The relevance of services for high-, medium- and low-tech firms – an empirical analysis in German industry Eva Kirner, Gunter Lay and Steffen Kinkel

INTRODUCTION High-, Medium- and Low-tech Classification of Industries In the past two decades the notions of ‘high-tech’ and ‘low-tech’ have become frequently-used categories which are generally applied to classify different industries. The distinction between high- and low-technology sectors has been given increased consideration, especially in relation to innovation and competitiveness. It is often assumed that high-tech sectors are superior to low-tech sectors as regards their innovativeness and competitive strength. The classification of high-, medium- and low-technology sectors was initially proposed by the OECD during the 1980s and has been widely adopted in different contexts today (Hirsch-Kreinsen et al., 2005). This classification is based on the share of sales spent on research and development (R&D) in different industry sectors. Considering the average expenditure on R&D in a certain sector, the sector is classified either as a high-, medium- or low-tech sector. The usefulness of this classification has been criticized by various scholars (Hirsch-Kreinsen et al., 2005; Von Tunzelmann and Acha, 2005). The arguments of these critical remarks can be summarized as follows: Traditionally, the concept low-tech and high-tech refers to industry sectors in general, not to single firms. However, depending on the degree of intrasectoral heterogeneity, a sectoral approach might be misleading because it reflects a sectoral average and ignores differences within the sector. Furthermore, it remains unexplained how low-tech industries especially have kept their competitiveness for decades without being innovative in terms of products. 175

176

Technological diffusion and interrelationships

Even in highly developed industrialized countries, low- and medium-tech sectors still count for the major part of the economic growth and employ a significant share of the workforce (Hirsch-Kreinsen et al., 2003), and this finding clearly shows the competitiveness of these sectors. ‘Servicizing’ Manufacturing Industries Parallel to the ‘high-tech/low-tech debate’, the process of ‘servicizing’ manufacturing industries has become a subject of research. By ‘servicizing’ (Rothenberg, 2007) or ‘tertiarization’ (Leo and Philippe, 2001; Lay, 2002) manufacturers gradually start to change their business focus. In order to respond to new challenges of markets, manufacturing firms integrate additional services in their B2B offerings. Thus manufacturing firms are expected to achieve and increase competitiveness (Sontow et al., 1997). Servicizing implies that manufacturing industries can offer their customers additional value by providing complete solutions and bundles of products and product-related services (Davies, 2003: Lay and Nippa, 2005; Lay and Jung Erceg, 2002; Markeset and Kumar, 2005). In the literature especially the rationales behind this trend (see for example Wise and Baumgartner, 1999; Goffin, 1999 or Oliva and Kallenberg, 2003), the challenges of managing the processes of providing industrial services (see for example Johansson and Olhager, 2003; Johansson and Olhager, 2006; Kumar and Kumar, 2004), the problem of measuring the quality of product-related services in B2B-markets (Parasuraman, 1998), and the key success factors in strategic marketing and management of industrial services (Neu and Brown, 2005; Matthyssens and Vandenbembt, 1998) have been analysed so far. In addition several classification schemes have been developed to distinguish between different types of product-related services (Lalonde and Zinszer, 1976; Frambach et al., 1997; Boyt and Harvey, 1997; Mathieu, 2001). These classification approaches describe product-related services emerging between low-end or traditional services and high-end or advanced services. Some authors characterize especially this second kind of service as ‘high-value integrated solutions’ (for example Davies, 2003) and attribute increasing importance to this way of combining products and services. By developing advanced services manufacturers change the focus of their B2B concepts from selling products to providing solutions. Service becomes a crucial factor in the process of creating superior value for customers (Matthyssens and Vandenbembt, 1998; Viardot, 1999). Due to the scarce statistical coverage of services in manufacturing industries however, there is little information available on the economic relevance of services for manufacturing firms. Furthermore, there are no research

The relevance of services for high-, medium- and low-tech firms

177

results dealing with the question of the competitive impact of servicizing based on large-scale surveys. Research Questions and Methodology Against the above background, this chapter examines the relevance of product-related services in the specific light of low-, medium- and high-tech industries. However, the level of analysis is not the industrial sector but the single firm. For an analysis of different innovation and economic behaviours related to different levels of R&D intensity, micro-level analysis seems preferable over a sectoral analysis because it can be assumed that industry sectors are composed of a variety of low-, medium- and high-tech firms which might each follow different innovation and competitive paths. To capture this diversity, firm-level data from Germany will be analysed. Our research focus tackles the question whether – besides an R&D-based innovation strategy – an alternative innovation approach might exist based on product-related services and whether such an approach might be an explanation for the survival of low- and medium-tech firms under the competitive conditions in high-wage countries. Given that the initial distinction between high-tech and low-tech is based on R&D expenditures, it is to be expected that product innovation activity is greater in high-tech than in low-tech sectors. However, this reflects only one possible type of innovation besides a number of other innovation types and competitive strategies (Kinkel et al., 2004). It seems therefore important to analyse also other competitive strategies and innovation paths which are pursued in low- and medium-tech sectors and more specifically, in low- and medium-tech firms. Our analyses are based on the German Manufacturing Survey 2006 conducted by the Fraunhofer Institute for Systems and Innovation Research (ISI) and part of the European Manufacturing Survey (EMS) comprising surveys in 12 countries. The objective of this regular, questionnaire-based, mailed survey, conducted in Germany, is to systematically monitor manufacturing industries. The survey addresses firms with 20 or more employees from all manufacturing sectors (NACE 15-37). The questionnaire includes questions on the implementation of innovative manufacturing technologies, organizational innovations, cooperation, relocation, performance indicators, products and services, as well as general company data. The German Manufacturing survey was first launched in 1993 and is conducted every two to three years. In 2006, 13 426 firms in manufacturing industries in Germany were contacted, whereupon 1663 firms returned a useable questionnaire, giving a response rate of 12.4 per cent (Lay et al., 2007). The dataset represents a cross-section of the manufacturing sectors. Using this database we analyse in detail whether:

178 ●

● ●

●

Technological diffusion and interrelationships

the classification of industry into low-, medium- and high-tech sectors creates homogeneous firm groups with regard to their R&D expenditures; there are differences in productivity between low-, medium- and high-tech firms (not sectors); low- and medium-tech firms compensate for their innovative weakness in terms of product innovations by increased efforts in the field of product-related services; and low- and medium-tech firms derive their competitiveness from service sales which are above the average.

LOW-, MEDIUM- AND HIGH-TECH FIRMS IN LOW-, MEDIUM- AND HIGH-TECH SECTORS: DISTRIBUTION AND PERFORMANCE On the basis of 1663 datasets of firms in the German manufacturing sector, in this section the distribution of low-, medium- and high-tech firms within low-, medium- and high-tech sectors is analysed. Furthermore, the overall performance of low-, medium- and high-tech firms is compared by measuring their respective capital and labour productivity. The results highlight possible differences in the competitive strength of low-, medium- and hightech firms. The following analysis draws on the latest low-, medium- and high-tech classification proposed by Legler and Frietsch (2007) based on sectoral R&D expenditure. Unlike the older OECD definition (1994) which distinguishes between four different sectoral R&D intensities, this classification consists of three classification categories. The thresholds for separating low-, medium- and high-tech sectors identified by Legler and Frietsch are: >7 per cent average sectoral expenditure for R&D in case of high-tech sectors; between 2.5 per cent and 7 per cent for medium-tech sectors; and less than 2.5 per cent for low-tech sectors. The same thresholds are applied in the following analysis also to the firm-level data when distinguishing between low-, medium- and high-tech firms. Distribution of Low-, Medium- and High-tech Firms in German Industry Regarding the difference between a sectoral- and firm-level distinction between the categories low-, medium- and high-tech, the level of intrasectoral heterogeneity has been analysed. As Figure 10.1 shows, each industry sector is composed of a mix of low-, medium- and high-tech firms. Although sectors are classified as either low-, medium- or high-tech sectors,

179

The relevance of services for high-, medium- and low-tech firms Food products, beverages and tobacco Textiles, textile products or leather Paper products, publishing or printing Chemical industries Rubber and plastic products Basic metals and fabricated metal products Machinery and equipment Electrical and optical equipment (excl. Nace 33) Medical, precision and optical instruments, etc. Transport equipment Other sectors of manufacturing

Up to 49 employees 50 to 249 employees 250 and more employees 0%

50%

100%

Share of firms Low tech firms

Figure 10.1

Medium tech firms

High tech firms

Low-, medium- and high-tech firms by sector and firm size

they are to a significant extent composed of firms which do not match the respective classification, being either more or less R&D intensive than the sector on average. Given that around half the firms of any sector do not match the respective sectoral classification as low-, medium- or high-tech (Kirner et al., 2007) with regard to R&D intensity, a high degree of intrasectoral heterogeneity can be determined. Thus it seems crucial to focus on the micro-level of the single firm rather than on sectors when analysing different competitive strategies or innovation paths and their relationship to R&D intensity. Furthermore, the data shows that low-tech firms are mostly small- and medium-sized firms (SMEs) while medium-tech firms are most often middle-sized enterprises. High-tech firms, however, can be found in equal proportions both among small and large firms, which indicates that a significant share of small firms is also characterized by high R&D intensity. Many of them might be spinoffs or high-technology start-ups. Given the diversity shown of high-, medium- and low-tech firms as regards sector and firm size, the following analysis will be conducted on the firm level.

180

Technological diffusion and interrelationships

Performance of Low-, Medium- and High-tech Firms: Capital and Labour Productivity The effect which is most prominently related to the benefits of high R&D intensity is increased competitiveness. It is assumed that high-technology sectors or firms are more successful than their low- or medium-technology counterparts because they are expected to be more innovative. This argument obviously holds if we look at the innovation output measured by new products. Given that R&D expenditures are important resources and inputs for the development of new products, it is not surprising that R&Dintensive firms indeed perform better than low-tech firms in product innovation output. This assumption has also been confirmed by recent empirical analysis (Kirner et al., 2007). However, product innovation success is only one of the possible ways to achieve competitiveness: different innovation paths exist which might equally lead to market differentiation and economic success, and other performance indicators such as productivity can measure economic success and innovativeness, not only the amount of new-product sales. Using our data, we analysed the productivity of low-, medium- and hightech firms. Productivity can be measured in different ways: either as labour productivity (total sales minus total inputs divided by total labour costs) or capital productivity (total sales minus total inputs divided by depreciation). These two different productivity measures capture different dimensions of efficiency. Labour productivity measures the value added generated per euro of labour costs and can therefore be considered a suitable performance indicator for the efficiency of the use of the human resources of a company. Capital productivity measures the value added generated per euro capital costs for machines and equipment and can be regarded as an indicator of the efficiency of the use of the technical equipment of the firm. Together they allow an assessment of the total factor productivity to compare the overall competitiveness of low-, medium- and high-tech firms. Given that productivity is known to be highly dependent on firm size (large firms tend to be more productive than small firms, mostly due to economies of scale and scope), the analysis was performed differentiating between three firm size classes. As Figure 10.2 shows, our analysis revealed that low-tech firms do not display a lower labour productivity than medium-tech and high-tech firms, which is equally true for all three firm size categories. In the case of small companies of fewer than 50 employees, low-tech firms even show a statistically significantly higher labour productivity than their medium- and high-tech counterparts. One explanatory factor for these results might be that most low-tech firms are facing stronger global competition than medium- or high-tech firms within their customer

181

The relevance of services for high-, medium- and low-tech firms 3.00

Firm size (number of employees) Low-tech firms

Note:

More than 250 employees

1.00

0.00

More than 250 employees

0

50 to 249 employees

10

50 to 249 employees

20

2.00

Up to 49 employees **

Labour productivity

30

Up to 49 employees

Captial productivity

40

Firm size (number of employees)

Medium-tech firms

High-tech firms

** Significant group difference, p  0.05

Figure 10.2 Capital and labour productivity of low-, medium- and hightech firms by firm size or technology niches. This leads to higher price competition for low-tech firms, particularly smaller ones, with companies from low-wage countries, resulting in a greater need to improve labour productivity in their highwage home base. As our results show, low-tech firms seem to be able to organize their value-adding processes at least as efficiently and profitably as medium- and high-tech firms. Another explanation might be that low-tech firms are able to use less qualified and thus less expensive personnel to run their less complex and possibly more standardized production processes than medium- and high-tech firms with more complex products and production processes are able to do. Furthermore, the high-price pressure on the global markets of many lowtech firms such as textile companies might lead to reduced personnel intensity and thus to strictly capital-intensive production structures. However, the capital productivity in low-tech firms is obviously not lower than in medium- or high-tech companies. Overall, low-tech companies seem to put a stronger emphasis on a very efficient use of modern production equipment and their production workforce. They seem to be in most cases able to compensate for their lower returns – which consequently lead to lower levels of sales and productivity measures – in their globally competitive markets.

182

Technological diffusion and interrelationships

In the end, however, it is not most decisive for low-tech firms in highwage countries like Germany whether they are as productive as medium- or high-tech firms from their country, but rather whether they manage to be at least as productive as their global competitors in other (particularly lowwage) countries. Analyses in this field might be an interesting topic for further research, particularly if focusing on still unexploited productivity potentials in technical and organizational process innovations of low-tech firms in high-wage countries.

THE RELEVANCE OF SERVICES FOR THE COMPETITIVENESS OF LOW-, MEDIUM- AND HIGH-TECH FIRMS The previous analysis of productivity revealed that no significant difference could be found between low-, medium- and high-tech firms as regards the overall efficiency of their business processes. In the following, the role of services and service innovations in low-, medium- and high-tech firms will be more closely analysed. The key question is on the one hand whether low-, medium- and high-tech firms pursue different strategies as regards product-related services or, on the other hand, whether there is a difference regarding their innovation activity related to services. The Offer of Product-related Services in Low-, Medium- and High-tech Firms As mentioned, a number of different product-related services, which manufacturing firms offer customers, can be distinguished. Different services might be suitable for the respective products and target markets. Besides services which have been traditionally offered by manufacturing firms for a long time, such as consulting, technical documentation, training or maintenance, other types of services are also evolving and starting to gain relevance. Manufacturing firms can provide customized software, financing models or can even operate the equipment for the customer. Figure 10.3 shows how many of the low-, medium- and high-tech firms are offering each of these various forms of product-related services. As is to be expected, traditional services like consulting or training are much more frequently offered by all firms than are, for example, software or financing models. High-tech firms prove to be generally more active in offering services than low-tech firms. Medium-tech firms range between, with the exception of consulting, while medium-tech firms seem to be the

183

The relevance of services for high-, medium- and low-tech firms

Design/consulting/project planning Technical documentation Training Assembly/initial start-up Maintenance/repair Software development Build-operate-own Leasing/renting/finance

0%

20%

40%

60%

80%

100%

Share of firms Low-tech firms

Medium-tech firms

High-tech firms

Figure 10.3 Type of product-related services offered by low-, medium- and high-tech firms most active group. Between one-half and four-fifths of manufacturing firms offer traditional services, while initial set-up and maintenance are offered by one-third to two-thirds of the surveyed firms, with the extent depending on their different R&D intensities. Software, built-operate-own models or financing are comparatively rare, offered only by less than onetenth of low-tech and up to almost one-fifth of high-tech firms. There are substantial differences between different types of services. However, previous analysis has revealed that – though medium-tech and high-tech firms are clearly significantly more active than low-tech firms at providing services – still as much as 81 per cent of low-tech firms offer at least one product-related service, and 64 per cent of them, two different types of product-related services to their customers (Kirner et al., 2007). The above analysis shows that low-tech firms are less active compared to their medium- and high-tech counterparts in offering product-related services. However, this difference seems to result mainly from a stronger focus,

184

Technological diffusion and interrelationships

since four-fifths of low-tech firms do offer at least one product-related service. The substantial difference between low-tech and medium- or hightech firms is apparent in the simultaneous offer of different types of services. Here, medium- and high-tech firms prove to be more active than low-tech firms, offering a broader range of different services to customers. Service Innovators Among Low-, Medium- and High-tech Firms One specific aspect which seems particularly interesting when looking at alternative competitive strategies, such as a service strategy in manufacturing firms, is the aspect of innovative services. It can be assumed that service offers need to be new and innovative in order to provide a competitive advantage against competitors. If a manufacturing firm is following a service strategy by offering its customers product-related services, those services need to be ‘up to date’. Thus it would be interesting to know whether low-, medium- and high-tech firms differ regarding the share of service innovators among them. Service innovators are understood to be firms which obtain sales from services which have been introduced within the last three years of the reference period. This definition includes the criterion of market implementation because only those firms are considered to be innovators which actually generate turnover with new services. The analysis shows that around half of both high- and medium-tech firms are service innovators, compared to just one-third of low-tech firms. These results indicate that although low-tech firms still participate in service innovations to a considerable degree, medium- and high-tech firms are clearly more frequently engaged in service innovations. Furthermore, we have analysed the share of product and/or service innovators among low-, medium- and high-tech firms. As Figure 10.4 shows, two-thirds of low-tech firms but almost 90 per cent of medium- and high-tech firms are product and/or service innovators, offering either product or service innovations or both. While the share of firms offering only new products (but no new services) to their customers is quite similar among low-, medium- and high-tech firms, the share of firms providing only service innovations (but no product innovations) is the greatest among low-tech firms. One-tenth of the surveyed low-tech and 9 per cent of the medium-tech firms are exclusively service innovators, compared to only 6 per cent of high-tech firms. However the most striking difference is the share of firms which are simultaneously product and service innovators, offering both new products and new services to their customers. Here, high-tech firms are clearly at the forefront: 43 per cent of them offer new products and also new services. This share is somewhat smaller among medium-tech firms (37 per cent) but only half as large among low-tech firms (22 per cent).

The relevance of services for high-, medium- and low-tech firms

185

High-tech firms

Medium-tech firms

Low-tech firms

Per cent 0

20

40

60

80

100

Share of firms No product or service innovator Only product innovator

Only service innovator Product and service innovator

Figure 10.4 Share of product and service innovations among low-, medium- and high-tech firms These results reveal that high-tech firms are the most active group in product and service innovations, closely followed by medium-tech firms, especially as regards the simultaneous offering of these two types of innovation. Nevertheless, still one-fifth of low-tech firms are product and service innovators and another 10 per cent of them, service innovators only. Thus, in spite of the clearly greater service innovation activity among high-tech and medium-tech firms, service innovations also occur among one-third of the surveyed low-tech firms, particularly among those which also follow a product innovation strategy. Share of Turnover with Services and Service Innovations in Low-, Mediumand High-tech Firms The share of service innovators provides an assessment of the extent of innovation activity among low-, medium- and high-tech firms. However, to assess the actual economic relevance of services for low-, medium- and high-tech firms and the specific role of service innovations therein, further indicators are needed. The share of firms engaging in an innovation activity is one aspect of measuring innovativeness. The other aspect is measuring the performance or market success achieved with services and service

186

Technological diffusion and interrelationships

innovations. This can be assessed through performance indicators similar to those used to measure the success of product innovations. Similarly to the share of turnover with new products, the share of turnover with services and service innovations can be regarded as a useful direct indicator of the market success of services. As Figure 10.5 shows, a considerable share of turnover in manufacturing firms – ranging from one-tenth to one-fifth of total sales – originates from services (on average 17 per cent). This share tends to be particularly large in small- and medium-sized companies (SMEs) and in high-tech firms. Thus, services play a relevant role for manufacturing firms and contribute in a substantial way to their financial success, with one-third to two-thirds of the turnover with services resulting from service innovations – services which have been introduced within the past three years of the surveyed reference period. This indicates that service innovations contribute strongly to the total turnover with services. Firms are continually renewing their service offers and thus providing improved and innovative services to their customers. It can be assumed that new products are often accompanied by new services specifically adapted to the characteristics of the new products and to customer demands. The fact that manufacturing SMEs seem to be particularly successful in generating turnover with product-related services might be related to their greater proximity to customers, which allows them to complement their products with suitable, customized services. Furthermore, in the empirical analysis high-tech firms (especially hightech SMEs) achieve the largest shares of turnover with services. This might be related to the great technological complexity of products offered by high-tech firms. Users of complex products are more likely to require various forms of services such as training, special maintenance or software – which are probably less important for products of low-tech firms. To identify the factors which influence the share of turnover with services in firms, we have also computed a multivariate regression model. The dependent variable here was the share of turnover with services, whereas various structural factors such as firm size, sector, product complexity, batch size, age of firm (start-up or not), qualification level of the workforce, export quota and R&D intensity have been considered as control variables. The regression model revealed that the share of turnover with services in manufacturing firms is statistically significantly positively influenced by single-unit or smallbatch-size production (p  0.001), highly skilled workforce (p  0.001) and low firm age (start-up) (p  0.05), whereas negatively influenced by high export quotas (p  0.05) and large-batch-size production (p  0.05). Regarding the effect of R&D intensity, a negative relationship was found (p  0.1) between low R&D intensity and the share of turnover with

187

Firm size (number of employees)

The relevance of services for high-, medium- and low-tech firms

up to 49

L M H

50 to 249

L M H

250 and more

L M H 0

5

10

15

20

25

Share of sales with product related services (%) Product related services – offered for more than 3 years Product related services – introduced within last 3 years

R&D intensity: L – low-tech firms M – medium-tech firms H – high-tech firms

Figure 10.5 Turnover with sales of product-related services (introduced within the last three years versus offered for longer than three years) by firm size and R&D intensity services, even if only on a lower level of significance when compared to the other results. This multivariate model revealed that a large share of turnover with services is dependent on close relations to customers, both in terms of customized products (single-unit or small-batch-size production) and in terms of geographic proximity (low export quotas). Furthermore, large shares of turnover with services seem to be linked to a comparatively highly skilled workforce and are more likely to be achieved by start-ups rather than by established firms. Interestingly, product complexity did not prove to be an influence on the share of turnover with services in a statistically significant way. However, the multivariate analysis confirmed that lowtech firms are indeed weaker performers compared to their medium- and high-tech counterparts as regards the share of turnover with services. This result is independent of other influencing structural differences between low-, medium- and high-tech firms, since the model was controlled for the structural variables mentioned above.

188

Technological diffusion and interrelationships

Another interesting aspect of the relevance of services for manufacturing firms is the pricing of product-related services. Manufacturing firms are faced with the problem that customers tend to perceive only the material product as the main offer which they are willing to pay for. Services are regarded as additional features around the core product. However, services might be just as important for the quality of functioning and use of the product as the product itself (such as software [support] or training). Thus, product-related services are often sold in a bundle with the material product, whereby the material and the immaterial part of the offer are not clearly differentiated. If product-related services are mingled this way with the core product, they are not separately calculated. The customer pays for the entire product-service bundle, the total price of which already contains the price of product-related services. The other alternative is to calculate services as a separate offer, differentiating between the price of the material product and the price of the product-related service. Our analysis shows (Figure 10.6) that the majority of turnover with services is achieved by ‘hiding’ the service value within the total price of the product-service bundle. This is especially the case in low-tech firms. It can be assumed that low-tech firms are even less perceived to be service providers than high-tech firms might be. These detailed results on the way services are calculated and accounted for in manufacturing firms indicate that customers of manufacturing firms seem to be more prepared to pay for services offered in the form of a full package or product-service bundle than for services which are offered separately, in addition to the core product. Thus, it seems crucial to consider the ‘hidden’ part of turnover with services, especially in the manufacturing sector. Given that a large part of turnover with services is not as obvious and recognizable as turnover generated with products alone might be, the economic relevance of services for the manufacturing sector can only be adequately assessed by considering the specific ways of pricing.

CONCLUSIONS This chapter aimed to combine two strands of the literature dealing with the competitiveness of companies, industries and mature economies: the debate on high-, medium- and low-tech and that on ‘servicizing’ in manufacturing industries. First the distinction between low-, medium- and high-tech sectors versus low-, medium- and high-tech firms was discussed. On the basis of firm-level survey data from the German manufacturing sector it was shown that each sector is composed of a variety of low-, medium- and high-tech firms. Thus

F rm s ze (number of emp oyees)

The relevance of services for high-, medium- and low-tech firms

189

L up to 49 M H L 50 to 249 M H L 250 and more M H 0

5

10

15

20

25

Share of sales with product related services (%) Share of sales (directly accounted) Share of sales (indirectly accounted)

R&D intensity: L – low-tech firms M – medium-tech firms H – high-tech firms

Figure 10.6 Share of turnover with product-related services (directly and indirectly accounted) by firm size and R&D intensity it was argued that any detailed empirical analysis of R&D intensity needs to be carried out also at the firm level in order to adequately capture the influence of R&D intensity on innovation and competitiveness. Sectorlevel analysis necessarily aggregates low-, medium- and high-tech firms and thus cannot distinguish possible different competitive and innovative patterns of low-, medium- and high-tech firms. Second the performance of low-, medium- and high-tech firms was analysed. Here no significant productivity gap could be identified between high-, medium- and low-tech firms. These empirical results indicate that no substantial general efficiency gap between low-, medium- and high-tech firms can be postulated. They seem to be similarly efficient in their business processes; however, they might emphasize different competitive strategies and innovation paths. One of these possible other innovation paths could be associated with services. Thus in the third section the service activities of low-, medium- and hightech firms were depicted. Services are commonly associated with the service

190

Technological diffusion and interrelationships

sector; however, they are also becoming increasingly important for manufacturing firms. Manufacturing firms can achieve competitive advantage through offering complete solutions to their customers. Such productservice combinations or bundles can contain in addition to the core product various product-related services which help customers implement and use the product. The type and quality of service offer can differentiate manufacturing firms from their competitors and serve as a distinct competitive strategy. The empirical analysis presented in this chapter shows that the vast majority of manufacturing firms already offers product-related services to their customers. High-tech firms seem to be more active in this field compared with low-tech firms, especially as regards the number of different services offered simultaneously. Innovations also play an important role in the area of services. Around half of the surveyed medium- and high-tech firms and one-third of low-tech firms are service innovators, achieving more turnover with new services. This indicates that services, like products, are updated and further and/or newly developed. Given that the different R&D intensity of low-, medium- and high-tech firms is expected to be closely related to product-innovation capacity, the share of product and service innovators was analysed for each. The data shows that high-tech firms are most often either only product innovators or simultaneously product and service innovators, whereas low-tech firms engage less frequently in product innovation or both innovation forms simultaneously. Interestingly, around one-tenth of low-tech firms are only service innovators, firms which innovate in their services but not their products. Furthermore, the share of turnover with services proves to be greatest in high-tech firms and also in SMEs in general. On average, almost one-fifth of total sales in manufacturing firms results from services. As much as around half of the turnover with services results from service innovations, which shows the innovative dynamics in this area. Thus, both services in general and service innovations in particular are significantly contributing to the economic success of manufacturing firms and especially so in the case of high-tech SMEs. Remarkably, more than half of the turnover with services results from incorporating the price of the service into the total price of the productservice bundle. Customers of manufacturing firms are not always prepared to pay directly for services. This seems to be particularly the case regarding customers of low- and medium-tech firms, whereas high-tech firms are more often able to calculate their services as a separate offer with separate pricing. In conclusion, services prove to be an important source of competitive advantage and economic success for manufacturing firms. High- and medium-tech firms are particularly active in the area of product-related services, both in terms of offering various services and also as regards the

The relevance of services for high-, medium- and low-tech firms

191

share of turnover with services and service innovations. The superior performance of high-tech firms can be attributed to the assumption that service innovations are closely related to product innovations. High-tech firms are more often product innovators, which necessarily inspires new forms of services adapted to the new products. The hypothesis that low-tech firms compensate for their weakness in R&D and product innovation by strengthening their service innovation activities in order to remain competitive could not be supported by empirical findings. All in all, low-tech firms seem to be inferior to medium- and high-tech firms both in product and service innovations. However, at the same time, no productivity gap can be found between low-, medium- and high-tech firms. They all seem to be similarly efficient in their business processes. Hence, future research should investigate whether low-tech firms base their competitiveness on yet other innovation strategies, perhaps on innovative processes instead of innovative products and services. Technical process innovations as well as innovative ways of organizing production and work processes might be a further source of competitive advantage. Superior quality, flexibility and productivity could be important effects of innovative processes.

REFERENCES Boyt, T. and M. Harvey (1997), ‘Classification of industrial services – a model with strategic implications’, Industrial Marketing Management, 26, 291–300. Davies, Andrew C. (2003), ‘Are firms moving “downstream” into high-value services?’, in Joe Tidd and Frank M. Hull (eds), Service Innovation – Organizational Responses to Technological Opportunities & Market Imperatives, London: Imperial College Press, pp. 321–42. Frambach, R.T., I. Wels-Lips and A. Gündlach (1997), ‘Proactive product service strategies – an application in the European health market’, Industrial Market Management, 26, 341–52. Goffin, K. (1999), ‘Customer support – a cross-industry study of distribution channels and strategies’, International Journal of Physical Distribution & Logistics Management, 29 (6), 374–94. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2003), ‘Low-tech industries and the knowledge economy. State of the art and research challenges.’ EU 5th Framework project ‘PILOT: Policy and Innovation in Low-tech’. accessed 31 October 2007 at http://129.217.232.4/is/dienst/de/textonly/ content/V4/V42/pdf/ap-soz01.pdf. Hirsch-Kreinsen, Hartmut, David Jacobson and Staffan Laestadius (eds) (2005), Low-tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang. Johansson, P. and J. Olhager (2003), ‘Industrial service profiling: matching service offerings and processes’, International Journal of Production Economics, 89, 309–20.

192

Technological diffusion and interrelationships

Johansson, P. and J. Olhager (2006), ‘Linking product-process matrices for manufacturing and industrial service operations’, International Journal of Production Economics, 104, 615–24. Kinkel, S., G. Lay and J. Wengel (2004), ‘Innovation: Mehr als Forschung und Entwicklung. Wachstumsstrategien auf anderen Innovationspfaden’, Mitteilungen aus der Produktionsinnovation serhebung, 33, Karlsruhe: Fraunhofer-Institut für System- und Innovationsforschung. Kirner, Eva, Steffen Kinkel und Angela Jäger (2007), ‘Innovationspfade von Low-, Medium- und High-Tech-Unternehmen in der deutschen Industrie’, in J. Abel and H. Hirsch-Kreinsen (eds), Lowtech, Unternehmen am Hightech-Standort, Berlin: Sigma, pp. 165–92. Kumar, R. and U. Kumar (2004), ‘A conceptual framework for the development of a service delivery strategy for industrial systems and products’, Journal of Business & Industrial Marketing, 19 (5), 310–19. Lalonde, Bernard and Paul H. Zinszer (1976), ‘Customer Service: Meanings and Measurement’, Chicago, IL: National Council of Physical Distribution Management. Lay, Gunter (2002), ‘Service provider industry. Industrial migration from manufacturing to selling products and services’, working paper, Fraunhofer ISI, Karlsruhe. Lay, Gunter and Petra Jung Erceg (eds) (2002), Produktbegleitende Dienstleistungen. Konzepte und Beispiele erfolgreicher Strategieentwicklung, Berlin: Springer. Lay, Gunter and Michael Nippa (2005), Management produktbegleitender Dienstleistungen. Konzepte und Praxisbeispiele für Technik, Organisation und Personal in serviceorientierten Industriebetrieben, Heidelberg: Physica. Lay, Gunter, Angela Jäger and Spomenka Maloca (2007), Dokumentation der Umfrage Modernisierung der Produktion 2006 des Fraunhofer-Instituts für System- und Innovationsforschung, Karlsruhe. Legler, H. and R. Frietsch (2007), Neuabgrenzung der Wissenswirtschaft. Forschungsintensive Industrien und wissensintensive Dienstleistungen (NIW/ISI Listen 2006). Studien zum deutschen Innovationssystem, 22, Bundesministerium für Bildung und Forschung (BMBF). Leo, P.Y. and J. Philippe (2001), ‘Offer of services by goods exporters: strategic and marketing dimensions’, The Service Industries Journal, 21, 91–116. Markeset, T. and U. Kumar (2005), ‘Product support strategy: conventional versus functional products’, Journal of Quality in Maintenance Engineering, 11, 53–67. Mathieu, V. (2001), ‘Product services: from a service supporting the product to a service supporting the client’, Journal of Business & Industrial Marketing, 16 (1), 39–58. Matthyssens, P. and K. Vandenbembt (1998), ‘Creating competitive advantage in industrial services’, Journal of Business & Industrial Marketing, 13 (4/5), 339–55. Miles, I. (2005), ‘Innovation in services’, in Jan Fagerberg, David Mowery and Richard R. Nelson, The Oxford Handbook of Innovation, Oxford: Oxford University Press pp. 433–58. Neu, W.A. and S.W. Brown (2005), ‘Forming successful business-to-business services in goods-dominant firms’, Journal of Service Research, 8 (1), 3–17. Organisation for Economic Co-operation and Development (OECD) (1994), Science and Technology Policy: Review and Outlook. Paris: OECD. Oliva, R. and R. Kallenberg (2003), ‘Managing the transition from products to services’, International Journal of Service Industry Management, 14, 160–72.

The relevance of services for high-, medium- and low-tech firms

193

Parasuraman, A. (1998), ‘Customer service in business-to-business markets: an agenda for research’, Journal of Business & Industrial Marketing, 13 (4/5), 309–21. Rothenberg, S. (2007), ‘Sustainability through servicizing’, MIT Sloan Management Review, 48 (2), 82–91. Sontow, Karsten, C. Jaschinski and Robert Kurpiun (1997), Industrielle Dienstleistungen – Neue Ertragschancen für den Maschinen- und Anlagenbau, Aachen: Forschungsinstitut für Rationalisierung an der RWTH. Viardot, Eric (1999), Introduction to Information-based High-tech Services, Boston, MA: Artech House. Von Tunzelman, Nick and Virginia Acha (2005), ‘Innovation in “low-tech” industries’, in Jan Fagerberg, David C. Mowery, Richard R. Nelson, The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 407–32. Wise, R. and P. Baumgartner (1999), ‘Go downstream – the new profit imperative in manufacturing’, Harvard Business Review, September-October, 77 (5), 133–41.

PART III

Local versus global perspectives in innovation

11.

Innovation activities versus competitiveness in low- and medium-technology-based economies: the case of Poland Anna Wzia˛tek-Kubiak

INTRODUCTION The collapse of the central planning system resolved the geographicoeconomic division of Europe into a model with the former socialist Central and Eastern Europe (CEE) at the periphery and the market economies of the European Union (EU) at the centre. This division was one result of the long isolation of CEE from the world economy. The planned economic system left CEE countries with distorted economies, a low level of technology – resulting in low-quality goods – poor competitiveness of most manufacturing industries and an anti-innovation bias. The structure of incentives that was passed on from the old system tended to discourage technological change and innovation, and any improvements in efficiency tended to occur on the basis of system-specific institutional characteristics. With the new member-states’ progress in institutional reforms and their adjustment to market-economy institutions, the major issue that has emerged is how to sustain further growth. These countries’ preaccession progress and their increasing capacity to withstand the competitive struggle within the single market have shifted their focus towards the problem of creating knowledge-based economies, and in particular, of increasing the role of innovation. Changes in the structure of production in terms of the level of technology is discussed in this chapter. Low- and medium-technology industries are commonly regarded as based on a low level of knowledge with limited, if any, impact of innovation on their competitive performance, and thus with no potential for future development. The high-tech focus of discussions prevalent in the literature, which tend to examine differences in competitiveness across firms, sectors and countries, as well as the high-tech obsession – also apparent in Poland – have 197

198

Local versus global perspectives in innovation

meant that the role of innovation in growth and the competitive performance of low- and medium-tech sectors has been somewhat neglected. This has resulted in the focus of Polish government policies on high-tech industries and its neglect of the issue of innovativeness in medium- and lowtechnology industries. However, there are three problems: first, the new member-states do not abound in innovation resources and their quality is rather low. Second, their economies are much more based on low and medium levels of technology than those of the ‘old’ EU countries. Improvements in the innovativeness of low- and medium-tech sectors have a stronger effect on the economy than improvements in the innovativeness of high-technology sectors. And third, low- and medium-tech industries face increasing competitive pressure from their developing country counterparts. The question that arises out of these three problems is related to the role of innovation compared to the role of costs in raising the competitiveness of the new member states in the pre-accession period. An answer to this question is attempted in the second issue raised in this chapter, namely that of the differences in the innovativeness of low- and medium-tech industries as compared with that of high-tech industries of catching-up countries. This chapter discusses the role of traditional versus knowledge-based determinants of change in the competitiveness of Polish manufacturing in low- and medium-tech industries. The shift in factors that determine changes in the competitiveness of ‘catch-up’ economies is at the heart of the discussion raised in this chapter, on the innovativeness of low- and medium-tech as compared with high-technology industries, the specifics of innovation sources, forms and outputs and their role in economic growth. The chapter starts with a presentation of changes in the competitiveness of Polish manufacturing from 1998–2004 and the extent of progress made. Discussion of the role of ‘cost versus non-cost’ competitiveness in changes in the competitive pressure of Polish manufacturing on the EU market is included. Innovation performance, its inputs and the forms of innovation that are dominant, are presented in the second part of the chapter. The third part compares the innovation activities of large firms that originated in the socialist system with those of small firms that are newcomers to the Polish economy, as well as of foreign-owned companies. The differences in innovativeness between high-technology and medium-high-technology industries as compared with low- and medium-low-technology industries and linkages between them that impact on changes in their share in manufacturing production are presented in the fourth section. The final section concludes the chapter.

Innovation activities versus competitiveness

199

THE COMPETITIVE PERFORMANCE OF POLISH MANUFACTURING AND ITS FACTORS Since the beginning of the 1990s, Poland has remained on a path of fast economic growth. From 1995–2004 its annual GDP growth exceeded the EU15 average in every year except 2001, allowing Poland to continuously converge towards European levels of economic development. In terms of GDP growth Poland is a ‘catch-up’ country. The catching-up of the Polish economy has been accompanied by a shift from an inward to outward orientation. From 1995–2003, the share of exports in GDP increased from 23.7 to 34.7 per cent, while that of imports rose from 21.5 to 36.9 per cent. Because growth of Polish manufacturing exports to the EU151 have been far greater than those of the EU25,2 intraEU15 exports and extra-EU15 import growth,3 Polish manufacturing has taken a considerable chunk of all the EU’s trade.4 Poland also contributes more than a quarter of the value of the accession countries (ACs) exports to the EU15 and since 1999 this share has continued to increase. These improvements have accompanied liberalisation and transition to market conditions of the Polish economy. Much greater liberalization of the Polish than the EU economy has resulted in an equalization of market access for all firms and has revealed the low competitiveness of many domesticoriented firms, resulting in many being pushed out of the domestic market. Their position is being taken over by foreign companies, competitive domestic firms and new domestic entrants. The relatively steep fall in Polish manufacturing’s position on the domestic market reflects the growing extent of foreign competitive pressure on domestic firms and their adjustment to market conditions. However liberalization of the EU market has also created conditions conducive to Polish competitive exporters and new entrants and helped their expansion on the single market. As Table 11.1 demonstrates, Polish manufacturing industry is mainly lowand medium-tech. However, even though the share of high-technology (HT) industries in Polish manufacturing production and exports accounts for only 5 per cent and has remained largely unchanged, some changes in the structure of production in terms of level of technology have taken place. First, the very large share of low- technology (LT) industries in total manufacturing production and exports has diminished, with a corresponding increase of medium-technology industries. Second, the improvement in the latter sector in production and exports was to a greater degree the effect of an increase in the share of medium-high technology (MHT) rather than of medium-low-technology (MLT) industries’ production. Third, the structure of Polish manufacturing production by technology level and its changes differ from that of exports. In the 2000s the share of

200

Local versus global perspectives in innovation

Table 11.1 Changes in the structure of Polish manufacturing production and exports by technology level (based on OECD classification 1997) from 1998–2004 (in percentage) 1998

2002

2003

2004

Manufacturing production High technology 4.8 Medium technology including: 53.4 ● Medium-high technology 22.6 ● Medium-low technology 30.8 Low technology 41.9

5.4 51.0 21.2 29.8 43.6

5.1 53.5 23.4 30.1 41.5

4.5 56.9 25.6 31.3 38.6

Manufacturing exports High technology 5.2 5.4 Medium technology including: 55.5 62.0 ● Medium-high technology 27.1 32.9 ● Medium-low technology 28.4 29.1 Low technology 38.7 32.6

5.9 62.8 33.7 29.1 31.3

5.6 63.6 35.2 28.4 30.8

5.2 66.8 36.2 30.6 28

Source:

2001

Nauka i technika (2005) and Marczewski (2006).

LT industries in Polish manufacturing production has been much greater, while that of MHT industries has been far smaller than their share of exports. In the 2000s Polish manufacturers of LT industries have produced much more, and MHT industries far less, than was proportionate to their share of exports. The decrease in the share of LT industries in Polish exports was greater than that of their manufacturing production. The pace of change in the structure of manufacturing production in terms of technology level lags behind changes in the structure of manufacturing exports. As most changes in the structure of exports have taken place in trade with EU countries, a deepening of Poland’s trade relationships in the EU also supports the structural change in Polish manufacturing production. In the literature, two types of competitiveness are distinguished: shorterterm price competitiveness and longer-term, non-price-, technology- or knowledge-based competitiveness. Changes in the former are based on ‘relative unit labour costs’ (RULC), the traditional determinants of growth, rather than technological change, although the latter also affects labour productivity indirectly by impacting on changes in unit labour costs (ULC). However, decrease in RULC allows a drop in prices. In this chapter RULC are defined as the relationship between ULC in Poland and the EU15, while ULC are calculated as labour compensation (wages and salaries plus socialsecurity contributions) of manufacturing industry relative to its turnover. This shows whether changes in productivity counterbalance increases in

201

Innovation activities versus competitiveness

Table 11.2 Changes in the relative unit labour costs (RULC) and relative unit export values (RUEV) of Polish manufacturing from 1996–2003

RUEV weighted RULC

1996

1997

1998

1999

2000

2001

2002

2003

0.43 0.77

0.45 0.79

0.45 0.81

0.50 0.79

0.53 0.75

0.60 0.77

0.64 0.71

0.7 0.62

Source: New Cronos database. Eurostat.

wages and salaries. Relatively high productivity growth may have been accompanied by even higher wage increases. In such a situation, despite productivity growth, ULC would have worsened.5 When RULC is above one (indicating that ULC in Poland is higher than in the EU) the efficiency of the use of labour in Poland is lower than in the EU. Since 1998 Polish manufacturing’s RULC have dropped considerably (Table 11.2) and been accompanied by improvement in its EU market share. In contrast to developed market economies, where RULC contribute little to overall growth in market share, in Poland any improvement in EU market share tends to follow a drop in RULC (see the results of the logit model, Wzia¸tek-Kubiak and Magda, 2007). RULC have also been an important determinant of FDI inflows (Bevan and Estrin, 2004). However over time investment has played an increasingly important role in terms of changes in market share at the cost of RULC (Wzia¸tek-Kubiak and Magda 2005), suggesting a drop in the role of traditional factors of growth and price competitiveness. Since a drop in RULC can be the result of the operation of nontraditional factors of growth impacting changes in productivity, we have also disaggregated changes in RULC components: wages per employee and labour productivity. Table 11.3 presents the drop in the RULC of Polish manufacturing resulting from the combination of a closing of the productivity gap with the EU15 and a rate of growth for wages slower than for productivity accompanying the fall in employment. The labour force in Poland, in contrast to that of the EU15, did not absorb all the productivity gains in the form of higher wages. More growth of wages than of productivity in the EU15 also supported an improvement in RULC in Polish manufacturing and served to increase the competitive pressure of its goods on EU goods. However the narrowing of the productivity gap of Polish manufacturing vis-à-vis the EU was due to the very strong decline in labour input, rising output and a higher investment rate than that of the EU15.

202

Local versus global perspectives in innovation

Table 11.3 Average changes in wages, productivity, turnover and employment of Polish and EU manufacturing in 1998–2003 (at current prices) Country Wages per Labour Turnover Employment Difference between employee productivity changes in wages and productivity (in percentage points) Poland EU15

38 30

82 20

48 16

19 6

44 10

Source: Author’s estimates based on Central Statistical Office of Poland and Eurostat data.

The fact that increases in productivity have been accompanied by a considerable decrease in labour inputs (Table 11.3) could mean that firms in the Polish manufacturing sector are still more engaged in defensive, passive restructuring based on a shedding of labour force (Konings, 2003) and low wages, that is to say, on traditional factors of growth, rather than on active restructuring. However this could also be explained by the fact that, first, aggregate data for manufacturing as a whole hide variations in the kind of restructuring introduced and factors responsible for changes in market share across industries. In Poland the industries that most increased their market share also started defensive restructuring earlier and then shifted to strategic, active restructuring based on higher investment intensity (Wzia¸tek-Kubiak and Magda, 2005). From 1998–2003 they undertook less defensive and more active restructuring. The expansion of their production and exports was accompanied by job creation. The opposite was the case for industries that started defensive restructuring later, mainly after 1999. This means that during the downturn in the Polish economy (1999–2001) the unproductive labour force was being replaced by more productive workers, inefficient firms were shrinking and efficient ones expanding. The simultaneous creation and destruction of jobs in Polish manufacturing suggests that a process of Schumpeterian competition has been at work. Variation in the types of restructuring undertaken by firms in different industries implies variation in the role of price versus non-price competitiveness across firms and industries. Second, the steep drop in employment that accompanied the considerable increase in turnover in Poland suggests that Polish firms have also introduced new ways of organizing work, and that ‘organizational process innovation’ has taken place. Third, a high investment rate, which is the main carrier of innovation, has had a strong positive effect on labour productivity growth (Sztaudynger, 2003).

Innovation activities versus competitiveness

203

The impact of investment and innovation (Kolasa, 2005) on changes in the competitiveness of Polish manufacturing firms has increased over time. Improvement in the productivity of Polish manufacturing has also been the result of an improvement in the quality of goods. As a proxy for product quality we use ‘unit export value’ (UEV). Although this is not free from some deficiencies6 it is recognised (Aiginger, 1998) as a good ‘overall’ quality indicator. UEV is defined as the export’s euro value divided by its physical weight (per kg) (OECD 1998: 94), while the ‘relative unit export value’ (RUEV) is the relationship between UEV in Poland and the EU15. We assume that an increase in RUEV is a sign of an improvement in Polish export-product quality in comparison to intra-EU15 exports. The low average RUEV level (Table 11.2) of Polish exports to the EU, especially when compared with that of Hungary, suggests their lower technology level. The considerable upward shift in the quality of Polish goods, reflecting improvements in RUEV, has been accompanied by a high level of variation across manufacturing industries. As RUEV in a relatively large number of industries is low and is very high in only a few,7 there are considerable differences in technology levels across Polish manufacturing industries. However, this tends to be typical of the new member states and in Poland such differences are even smaller than in the case of Hungary. Specialization at low product-quality levels implies a limited potential for exports of Polish manufacturing industries to the EU15 and may also hamper their growth opportunities in the long run (Murphy and Schleifer, 1991). This is all the more so given that Poland is also experiencing a rapid change towards demand for goods of higher quality. If Poland’s process of catching-up is not accompanied by improvement in the quality of products, that will restrain growth in production which enhances competitiveness. In effect, Poland may find itself in a ‘low-quality trap’ and could generate a ‘path dependence in which industrialised countries have the advantage of already having the capacity to produce high quality’ (Dulleck, 2002: 15). Therefore in Poland production growth dynamics will tend to be strengthened if accompanied by improvements in quality and technological level of goods.

INNOVATION INPUTS AND OUTPUTS IN POLISH MANUFACTURING8 Following the Oslo Manual, innovation-input indicators9 cover expenditures on R&D, acquisition of embodied and disembodied technology and training, as well as marketing.

204

Local versus global perspectives in innovation

A key feature of the Polish economy inherited from the period of central planning is the crucial role of investment in technology upgrading. From 1990–2004 investment growth in Poland was very high, exceeding 7 per cent per year. Investment expenditure in fixed assets (at constant prices) increased by 79.3 per cent. This was accompanied by considerable inflows of foreign direct investment (FDI) into Poland. Because the level of firms’ expenditure on innovation depends on expected profit, any changes in GDP affect the dynamics of innovation expenditure over time. The slowdown of the Polish economy resulted in a heavy drop in innovation (Table 11.4) and investment expenditure as a ratio of GDP, while the upturn of the Polish economy was accompanied by an increase in expenditure on innovation and investment, in the case of the latter with some time lag. We should bear this in mind while analysing data covering the period of slowdown (1999–2001) and recovery of the Polish economy (since 2002). The introduction of new technology carried by investment requires improvements in skills. With the relative shift towards demand for a highly qualified labour force, it has become critically important for it to improve its skills. Although the growth of education expenditure and GDP have been similar, the qualifications of the Polish labour force have improved considerably. The increase in the number of private universities has resulted in a large increase in the number of students in higher educational institutions. The gross enrolment rate in tertiary education10 is 47.8 per cent, meaning that every second pupil in the 19–24 age range is studying at an institution of higher education. Although the mature learning rate11 in Poland is almost twofold lower than in the EU25, youth educational attainment12 levels are higher than in EU25 (89.5 and 76.7 per cent respectively). These improvements serve innovation activity, reduce the anti-innovation bias of some Polish employees and create a good basis for future development based on innovation. However, the catch-up of the Polish economy has not been accompanied by an increase in expenditure on R&D. Although the number of research scientists and engineers in Poland is relatively large (given their income level), the portion of Gross Expenditure on R&D (GERD) in Poland’s GDP is low and diminished slightly from 0.65 in 1996 to 0.58 per cent in 2004. In 2004 only 0.5 per cent of the total number of Polish manufacturing firms launched R&D on a permanent basis. A large amount of government money in total R&D spending is typical for a peripheral type of economy. The key question is whether or not there are spillovers from research and higher education institutions into industry. From 2002 to 2004 only 11.6 per cent of the total number of innovating firms cooperated with domestic scientific and higher education institutions, and this was far lower

205

Innovation activities versus competitiveness

Table 11.4 Dynamics of innovation expenditure in Polish manufacturing and their structure in 2000–4 (in percentage)

Changes in innovation expenditure

2000

2001

2002

2003

2004

20.0

6.0

20.4

12.0

1.0

9.3 3.0 20.1 64.8

11.1 4.8 15.6 63.3

7.5 2.8 23.2 59.8

32.5

27.2

24.9

0.2 1.5

0.2 1.4

0.3 2.6

Structure of innovation expenditure In-house R&D 12.8 10.2 Acquisition of disembodied technology 2.4 1.9 Building 22.4 26.1 Acquisition of domestic and imported 54.0 54.8 machinery and equipment Acquisition of imported machinery 25.6 26.2 and equipment Training No data 0.7 Marketing 3.2 2.8 Source: Author’s estimates based on GUS (2005 and 2006) data.

than from 1998 to 2000. This low collaboration rate (that is, the number of firms cooperating with research and education institutions among all innovative firms) between research and higher education institutions and manufacturing firms reflects the low level of diffusion of governmentfunded R&D. Bearing in mind the crucial role of networks and spillovers in the innovation activities of low- and medium-technology sectors, this raises the issue of the sources of innovation in a peripheral economy. Following the Oslo Manual, innovation activities are broken down into intramural R&D, acquisition of embodied (buildings as well as domestic and imported machinery) and disembodied (licensed) technology, training and marketing of upgraded and new products. We concentrate on the period 2000 to 2004, that is, the period of slowdown and then an upturn in the Polish economy. From 2000 to 2004 expenditure on innovation in Polish manufacturing increased by 25 per cent, far more than the investment in fixed assets and GDP. The share of intramural R&D in the total innovation outlays of manufacturing industries was small, and diminished in favour of expenditure on disembodied, and especially embodied, technology (Table 11.4). The small portion of acquisitions of disembodied technology in total innovation expenditures hides the considerable increase in the number of innovations introduced. In 2004, in comparison with 2000, the number of foreign licences introduced in manufacturing production increased threefold, while that of domestic licences increased even more. This resulted in an increase in the

206

Local versus global perspectives in innovation

sales of products based on licences, within total sales of Polish manufacturing, from 9.5 in 2000 to 11 per cent in 2004, and an increase in exports from 7.2 to 11.8 per cent. This suggests that the Polish economy’s trade integration into the ‘Single Market’ has served to upgrade its manufacturing products. Moreover, the relative quality of Polish goods exported to the EU15 was not only much higher than that of those exported to other ACs, but has also kept increasing, from 0.51 in 1999 to 0.66 in 2003. The relative quality of Polish exports to ACs was constant, although higher in terms of intra-AC export quality overall. The quality of Polish goods sold on the domestic market was also lower than the quality of goods exported to the EU15. Given the process of deepening integration into the EU and shifts in domestic demand towards higher-quality goods, this raises the question of the basis of the future development of many domestic-oriented Polish manufacturing firms. The main carrier of technology is the acquisition of machinery and equipment for innovation, the share of which in total expenditures on innovation has been increasing at the expense of R&D and expenditures on training and marketing. Acquisition of domestic disembodied and embodied technology and its increase has been greater than that from foreign sources. This suggests that the innovation activities of Polish manufacturing firms are mostly based on domestic sources, although this is the peripheral type of economy and home to the highest inflows of FDI in Central and Eastern Europe. A considerable drop in the amount of expenditures on training in total innovation outlays may suggest that either Polish firms highly value the skills of the labour force, and that internal training predominates, and/or that the level of the technology introduced has not been very high. As consumers tend to be conservative and may initially reject new products from new producers, the role of marketing in the innovation activity of the LMT sector is very important (Hirsch-Kreinsen et al., 2003; HirschKreinsen, 2004). The small and diminishing part of marketing costs in the total innovation budgets of Polish manufacturing firms has hampered their expansion of production. This raises the question of whether they have the financial capacity for the marketing of product innovations, or perhaps product differentiation does not play such an important role. In other words, do innovative Polish firms (those that have carried out product or process innovation) focus on process innovation driven by cost reduction, or product innovation which seeks product differentiation? Of course, all firms strive to find ways of whittling down the cost of production. Although process innovation does enhance firms’ competitiveness and allows product innovation, a focus on the latter enables the penetration of new markets by a displacement of other goods and also enables the capture of new markets ahead of competitors or the acquisition of greater market share.

Innovation activities versus competitiveness

207

The considerable drop in RULC of Polish manufacturing industries helped drive an improvement in their competitive position from 1998 to 2003 on the EU market. However the increasing impact of investment on improvements in the competitive position of Polish manufacturing over time, especially since 2001, reflects the increasing role of innovation. Thus, during the upturn of the Polish economy from 2002 to 2004, what was more significant for Polish manufacturing firms: cost-reducing static efficiency or competition for markets with product differentiation? From 2002 to 2004 about 21 per cent of the total number of Polish manufacturing firms carried out process innovation, marginally ahead of those that carried out product innovation exclusively (18 per cent). Most of the firms carried out combined product-process innovation. The question is, what is the effect of combined product-process innovation and process innovation in respect to product differentiation? Availability of data allows us only to consider those innovating firms that said that the impact of innovations on their activity was ‘considerable’. As many as 37.5 per cent of these firms managed to increase the number of products, while 42 per cent improved product quality and almost 30 per cent entered new markets. On the other hand, as an effect of introducing process innovation only, 17.5 per cent of the above-mentioned firms diminished unit labour costs, and 16 per cent, intermediate costs, which is two times less than the number of firms that introduced new standards or increased production capacity. This suggests that process innovation introduced by Polish firms serves product differentiation to a greater extent than lowering costs. This has resulted in an increase in the share of new and upgraded products within total sales of manufacturing production, from 18.5 per cent in the period 1998 to 2000 to 22.3 per cent in 2004. This suggests that the impact of traditional cost factors on improvement in competitiveness has been diminishing and the role of innovation, which serves product differentiation, has strengthened. All in all the increase in the competitive pressure of many industries in Polish manufacturing on their counterparts in the ‘old’ EU member states has been based not only on improvements in costs but also on product innovation. However this process is highly differentiated across industries and firms.

VARIATIONS IN INNOVATION ACTIVITY ACROSS MANUFACTURING FIRMS From 2002 to 2004, the propensity to innovate (that is, the number of firms that have introduced innovation out of the total number of firms) in Polish

208

Local versus global perspectives in innovation

manufacturing accounted for only 25.6 per cent. Although this rate is similar to that of most new member states, it is much lower than the average in the ‘old’ EU countries (42 per cent). Improvements in the competitiveness of much less Polish than the EU incumbent countries manufacturing firms has been based on knowledge-based factors of growth. However, since the period 1998 to 2000 the propensity to innovate has increased considerably (by 8.5 percentage points). This increase and an increase in the share of innovation expenditure resulting in sales (innovation intensity rate), has been uneven across firms in terms of their size and ownership. In this chapter we compare the innovation of different types of firms – large ones with over 249 employees, small ones with 10–49 workers, and private domestic firms – with the innovation of foreign-owned companies (FOCs). The former tend to date from the socialist period and some of them are still state-owned. Most of them have been restructured, while some are still in the process of restructuring, especially in terms of costs.13 Expenditure on innovation in Polish manufacturing is highly concentrated in large firms. Large Polish firms’ share of spending, out of total Polish manufacturing expenditures on all types of innovation inputs (except for training and buildings) is greater than that in ‘old’ EU countries – over 72 per cent. In the case of R&D and marketing expenditure, where economies of scale play a role, their share is even higher. This implies very low expenditure on innovation per small firm, different innovation strategy and structure of expenditure than in large firms. Differences in the share of various firms, in terms of their size, in total innovation expenditure of Polish manufacturing tend to be commensurate with differences in their propensity to innovate and their innovation intensity rates. Propensity to innovate in Polish manufacturing, in terms of product and process innovation as well as innovation rate, tends to increase with the size of the firm. The bigger the firm, the greater the propensity to innovate and the higher the innovation rate. In 2002 to 2004, the propensity to innovate of small firms was only 17.7 per cent. This was threefold lower than for medium-size firms (50–249 employees) and fourfold lower than for large firms, although in all groups of firms this rate increased considerably. The bigger the firm, the greater the share of its spending on R&D, licences, imported machinery and marketing in total expenditure on innovation (Table 11.5), and the higher the collaboration rate. The smaller the firm, the greater the proportion of acquisitions of embodied, especially domestic-origin technology and the share of combined R&D and acquisition of licences out of total innovation expenditure, and the lower the collaboration rate. Basing their production on domestic machinery, small firms operate at lower levels of technology and innovation networks play a limited role. This, as well as the lower skills-level of employees in small

209

Source:

6.1 5.1 6.5 8.8

R&D

0.4 1.6 2.6 3.4

Acquisition of licenses 21.6 23.3 36.6 18.6

Expenditure on buildings

Author’s estimates based on GUS (2005 and 2006) data.

10–49 50–249 250–499 Above 499

Firm size (by number of employees) 65.2 64.1 49.0 61.8

Acquisition of machinery 16.3 25.5 19.8 26.4

Acquisition of imported machinery

Table 11.5 Structure of expenditure on innovation by firm size in 2002–4 (in percentage)

0.5 0.4 0.2 0.2

Training

1.7 2.3 1.3 3.2

Marketing

210

Local versus global perspectives in innovation

firms, results in a very large share spent on training within innovation expenditures, which suggests management’s awareness of the skill barrier when introducing innovation, and their constructive rather than adverse approach to innovation. Summing up, in the innovations of large firms, the crucial role is played by disembodied technology, while in small firms it is embodied technology stemming from the purchase of domestic machinery. Although the level of technology of small firms is lower than that in large firms, in the innovation strategy of the former a drop in costs is less important than it is for large firms, and product upgrading and differentiation play a more crucial role in improving their competitiveness. In new member states, FOCs play an increasingly important role. Their share of total Polish manufacturing firms’ income increased from 26.8 per cent in 1998 to 38.8 per cent in 2003, while their part in total investment expenditures in fixed assets rose from 3.6 per cent in 1994 to 28 per cent in 2003. For FOCs, the export intensity rate (the amount of exports in total sales) is much higher than for domestic firms, but this gap has also increased considerably over time. From 1998 to 2003 the share of export sales in total FOC sales rose from 14.1 to 20.6 per cent, while for domestic firms it fell from 7.1 to 6.7 per cent. This raises the question of whether FOCs differ from domestic firms in terms of innovation. This is all the more important given that FOCs are the main agents in the shift in Polish manufacturing production from the LT to MT sector. Not only has the share of HT and MHT industries’ production been greater in total FOC manufacturing production, but it has also increased much more (from 42.7 per cent in 2001 to 49 per cent in 2004) than has HT and MHT among domestic private firms (rising from 27.7 to 30.7 per cent) and the share of LT in the manufacturing production of FOCs has dropped (from 40.7 to 31.8 per cent) more than that of domestic private firms (from 44.8 to 41.2 per cent). The above shift in the types of production of FOCs has accompanied the increase in FOCs’ share of total spending on innovation in Polish manufacturing. In 2002 to 2004, although the domestic firms in manufacturing production were more numerous than FOCs, expenditures on innovations in the former were lower than in the latter. Average expenditure on innovation per FOC was fourfold higher than for domestic private manufacturing firms and reflects the expansion of FOCs in Poland.14 This demonstrates the weakness of Polish private domestic companies in creating knowledgebased competitiveness. Surprisingly, the greater proportion of HT and MHT industries in FOC production was accompanied by a low amount of R&D in their total innovation expenditure.15 This was counterbalanced by a very frequent acquisition of licences. This, as well as the small difference between the

211

Source:

2.8 0.8 2.9

6.0 4.2

Acquisition of licenses

6.5

R&D

21.5 27.1

23.9

Expenditure on buildings

Author’s estimates based on GUS (2005 and 2006) data.

Private sector including: Domestic firms FOCs

Sectors by ownership type

65.0 60.9

60.6

Expenditure on machinery and equipment

28.0 28.3

27.4

Expenditure on imported machinery

0.2 0.3

0.2

Training

Table 11.6 Structure of expenditure on innovation in firms by ownership type in 2002–4 (in percentage)

3.5 2.2

2.8

Marketing

212

Local versus global perspectives in innovation

amount of training in FOCs and domestic firms would seem to suggest that they both still operate at a lower level of technology than in the ‘old’ EU countries. FOCs have brought to Poland a mature product stage of medium-technology-industry production. However the opening in Poland of a lot of R&D units by FOCs in recent years may serve to increase the share of HT industries in Polish manufacturing. In general however, though FOCs are an important source of technology, domestic technological potential embodied in technology accumulated in Poland is even more important than foreign sources of technology upgrading (Weresa, 2004).

VARIATIONS IN THE INNOVATION ACTIVITY OF FIRMS ACROSS INDUSTRIAL SECTORS IN TERMS OF TECHNOLOGY LEVEL Because Polish manufacturing is based on LMT industries, questions arise related to the differences in the innovativeness of HT and LMT industries and the sources of those differences. One should keep in mind that innovation in LMT sectors differs from that of HT and MHT industries, and that some types of innovation activities in the former are difficult to measure (Hirsch-Kreinsen et al., 2006; Bender et al., 2005; Palmberg, 2001; Hansen and Göran, 1997). As LT and MLT industries are significant users of output from the HT and MHT sectors, a process of diffusion of technology from the latter to the former takes place. As competitiveness reflects the ability to use knowledge, ‘technological competition leads rather directly to inter-industry diffusion of technologies and therefore to the inter-industry use of the knowledge which is “embodied” in these technologies’ (Smith, 2002: 20). The greater the inter-industry diffusion of knowledge, the greater the increase in innovativeness and ability to compete of firms into which the knowledge spreads. Any difference in the ability to use knowledge among firms, industries and countries affects differences in their competitiveness. In the innovation activities of LT industries, spillovers from HT and MHT industries play a crucial role. This is accompanied by a different type of knowledge that is needed, used and developed in LT and MLT industries as compared to HT and MHT industries (Bender et al., 2005). The impact of HT and MHT industries on LT industries has a supply-side technological push character. The impact of MLT and LT industries on the innovation activity of HT and MHT industries has a demand-side pull character and changes in demand in the former impact on innovation, production and effectiveness of the latter (Carroll et al., 2000). LM and MLT industries also play an important role in the innovation activities of HT and MHT industries

Innovation activities versus competitiveness

213

(Brusoni and Sgalari, 2006). However as some innovation activities of LT industries are difficult to measure, not all their innovations are revealed in an analysis based on the Oslo Manual approach. As in this chapter we explore data based on a two-digit level, that data for the HT sector is underestimated (it does not cover data on pharmaceuticals, aircraft and spacecraft), while the data on the MHT sector is overestimated. Polish HT and MHT industries differ from MLT and LT industries in terms of the propensity to innovate. The higher the technology level of a given Polish industry, the greater the propensity to innovate in the firms. The propensity to innovate of HT-industry firms ranges from 27 to 48 per cent, while that of LT-industry firms ranges from 14 to 17 per cent; below the average level of Polish manufacturing. The propensity to innovate of MLTindustry firms is lower than MHT/HT-industry firms as well. This implies that the overall poor propensity to innovate of Polish manufacturers results from the scant propensity to innovate of LT- and MLT-industry firms, which nevertheless do a large very part of all Polish manufacturing (see Table 11.1). In light of this, the Polish innovation policy of concentration (at least as declared) on HT industries is surprising. LT and MLT industries are not even mentioned in Polish policy and, because the share of LT and MLT industries in Polish manufacturing is relatively high, improvements in the propensity to innovate of these industry firms could strongly impact on the propensity to innovate of Polish manufacturing generally, create new spillovers, and generate expansion of HT and MHT industries. Polish HT industries differ from those of the ‘old’ EU countries mainly in technology level. However HT industries located in Poland, as those in the ‘old’ EU countries, are much more R&D-intensive than LT and MLT industries. The higher the technology level of Polish industry, the greater the share of R&D in expenditure on innovation, and the greater the propensity to launch R&D. The portion of R&D in total innovation expenditure in Polish HT and MHT industry firms is almost threefold greater than the average for Polish manufacturing, and twelvefold more than in LMT and LT industry firms (Table 11.7). The share of HT-industry firms that launched R&D on a permanent basis, from the total number of HTindustry firms, is also much higher than that of LT-industry firms. From 11 to 20 per cent of innovating firms in HT industries launched R&D on a permanent basis, while this was the case for 4 to 20 per cent in MHT-industry firms and from 0.1 to 1.3 per cent in LT-industry firms. The higher the technology level of Polish industry, the lower the amount of embodied technology, especially in domestic machinery. The expenditure on domestic machinery of HT industries was not only less than expenditure on R&D, but also twofold lower than expenditure on imported machinery. The higher expenditure of MHT industries on licences

214

Source:

HT MHT MLT LT

24.9 14.6 4.9 2.0

R&D

2 4 3 2

Acquisition of licenses 9 27 23 22

Expenditure on buildings

Author’s estimates based on GUS (2005 and 2006) data.

Sectors by technology level 56 47 66 68

Expenditure on machinery and equipment 38 17 29 30

Expenditure on imported machinery

0.5 0.2 0.2 0.2

Training

Table 11.7 Structure of expenditure on innovation in terms of technology level in 2002–4 (in percentage)

2.0 4.6 0.4 3.3

Marketing

Innovation activities versus competitiveness

215

counterbalances their lower-level R&D and suggests that they are operating in the maturing product stage. In contrast to HT industries, innovation activities in LT and MLT industries in Poland are based mainly on the diffusion of embodied technology, especially domestic machinery. Because firms in LT industries are users of components and machinery mainly from domestic MHT- and HT-industry firms, their greater expenditure on domestic machinery means greater demand for Polish MHT- and HT-industry production. Not only the improvement in the competitiveness of HT and MHT industries, but also their expansion, depends on their ability to serve the needs and innovation activities of LT industries (Robertson and Patel, 2007). The high and increasing expenditures of LT and MLT industries on domestic machinery and equipment serves as an impulse for the development of Polish HT and MHT industries and the increase in their share of Polish manufacturing production (see Table 11.1). As the proportion of LT industries in Polish manufacturing production has been diminishing, the future expansion of HT and MHT industries will depend on export expansion as well as on their ability to meet the demand for embodied technology from other industries, including LT and MLT industries that innovate. The higher level of technology in HT industries is accompanied by a greater amount of both acquisition of imported machinery and R&D in total innovation expenditure, and demands a higher-skilled labour force. This results in these industries having the highest level of spending on training within innovation expenditure (see Table 11.7). In light of the large expenditures on R&D and licences of MHT industries, their expenditure on training is surprisingly low (compared to other types of industries). Considering that they are operating at a mature product stage, low expenditure on training may reflect the high and continuously improving skills of the Polish labour force. The predominance of ‘internal training in LT and MLT industries’ firms/LT- and MLT-industry firms, which is mostly supplied unsystematically during daily work and at the workplace’ (HirschKreinsen et al., 2006: 12) may result in lower spending on training in Polish LT and MLT firms than in HT industries. The higher the technology level of Polish industries, the more firms collaborate with other institutions in developing innovations. Not only have more HT firms launched R&D, but also more HT firms collaborate with other sources to help develop innovation. As much as 17 to 22 per cent of innovative firms in HT industries, that is, twofold more than innovative firms in LT industries, have different types of partnerships. The system of innovation of Polish LT- and MLT-industry firms is much more based on internal than external firm activity. This hampers diffusion of innovation to Polish LT and MLT industries and impacts their innovativeness negatively.

216

Local versus global perspectives in innovation

Although process innovation is important in the innovation activities of LT and MLT industries, the demand (consumer) focus – that is, marketinduced product innovations opening up new sales opportunities – plays a crucial role. Changes in demand, interactions with customers and suppliers are seen as main sources of LT-industry innovation, and the perception of innovation opportunities affects competitiveness there. The more consumer- and demand-oriented focus of LT industries, and concentration on other forms of innovation such as marketing and design (Sterlachini, 1999) means that the proportion of expenditures on marketing (see Table 11.7) and the rate of protection of some types of rights (trademarks, industrial and consumer design) have been quite high. The picture of innovativeness across Polish industries in terms of technology appears unclear when one analyses the effects of innovation. On the one hand, far more HT firms (from 50 to 80 per cent of the total number of HT-industry innovating firms) than LT-industry firms (from 30 to 37 per cent) increased their range of goods or improved their quality of goods (from 52 to 60 per cent and from 22 to 47 per cent respectively), or increased their market shares or entered new markets (from 32 to 48 per cent and from 14 to 35 per cent respectively). On the other hand, however, if we consider other types of innovation output measures such as the number of patent applications, registration of different types of (such as industrial) designs, trademarks and intellectual property rights – that is, when we consider measures that reflect the protection of some types of rights that are the effects of the innovation activities of firms – we get a different picture of the innovativeness of industrial sectors in terms of technology. In terms of the above-mentioned innovation outputs, HTindustry firms lag behind low-technology firms. The number of HTindustry firms that registered designs, trademarks, intellectual property rights as well as patent applications among the total number of innovating manufacturing firms was very small. This accounted for 4 to 6 per cent of the total number of HT-industry innovating firms, while that of lowtechnology-industry firms was between 35 (in the case of registration of designs) to 57 per cent (in the case of trademarks) (author’s estimates based on GUS, 2006: 98). This shows that LT-industry firms are not innovation-adverse and that they focus on different types of innovation than HT industries do. Summing up, the introduction of various measurements of innovativeness gives different pictures of the innovativeness of industries in terms of technology level. This raises the issue of the suitability, for all types of industries, of some of the measurements of innovativeness – such as those that do not take into account the differences in types of innovation or innovation inputs and outputs across industries.

Innovation activities versus competitiveness

217

CONCLUSIONS Poland is an example of an economy that, being based on LMT and with almost no improvement in its HT-sector share of manufacturing production, has nevertheless considerably increased its competitive pressure on the EU market. This improvement has been accompanied by a shift in the structure of production from LT to MT and especially to MHT, supported by a considerable drop in unit costs and an improvement in the quality of goods. The role of traditional determinants of improved competitiveness has diminished in favour of an increased role for innovation. Though innovation expenditure has been even more dynamic than that of investment, and the skills of the labour force improved, accompanying expenditure on R&D has been small. Weak linkages between manufacturing firms and publicly-funded R&D implies little diffusion of these R&D results. This makes the role of embodied technology and licensing in the innovation strategies of Polish manufacturing firms crucial, and does not support a shift from mature to new product stage, nor from medium- to hightechnology industry. The restructuring of Polish manufacturing industries has been accompanied by the increasingly important role of product innovation. Although process innovation is important, it impacts more on the improvement in the quality of goods and their differentiation than on a drop in costs. This indicates an awareness in Polish management of the role of product differentiation in competition. In terms of some innovation indicators used in the Oslo Manual, Polish HT and MHT industries seem to be more innovative than LT and MLT industries. They have a greater propensity to innovate and their innovation rate is higher than the latter. However the share of number of patent applications, registrations of different types of designs (such as industrial), trademarks and intellectual property rights of LT industries represented in their total number of Polish manufacturing has been very high, while that of HT industries has been very low. Since some of the effects of innovation in LT industries have been impressive, they have to have been innovative. The focus of their innovation activities has been different from those of HT industries: the innovation activities of HT industries are based on R&D and imported machinery; the innovations of MHL industries are based more on acquisition of licences than on R&D, with embodied technology playing a more important role than in the case of HT industries; the innovations of LT industries are based on embodied technology mainly of domestic origin. This shows the important role of the diffusion of technology from HT and MHT industries into LTindustry innovations, indicating that much of the improved efficiency and

218

Local versus global perspectives in innovation

product differentiation in MLT and LT industries is also derived from MHT- and HT-industry improvements, which confirms Robertson and Patel’s (2007) conclusion that the health of HT and especially MHT industries depends on their ability to serve the needs of LMT and LT sectors. Linkages between Polish HT and MHT on the one hand, and LT and MLT industries on the other, make the health and innovativeness of the latter (almost 70 per cent of Polish manufacturing) crucial to the future development of Polish manufacturing, and this fact warrants changes to Polish innovation policy that would improve the innovativeness of the LT and LMT sectors.

NOTES 1. 2. 3. 4. 5. 6.

7. 8.

9. 10. 11. 12. 13. 14. 15.

The data on EU trade comes from the Comext database. It is based on trade data expressed in EUR. At the time, the EU15 and ten future new member-states. The share of Polish manufacturing exports to the EU25 in the total EU25 internal exports increased from 1 per cent in 1996 to 1.8 per cent in 2003 and in the EU25 external imports from 2.4 per cent to 3.8 per cent respectively. From 1995–9 it was behind the other nine new member states. This took place in some Polish manufacturing industries from 1996–1998. Changes in unit export values for a given product category may reflect both changes in product quality and in the product bundle. The problem becomes more serious the more aggregated the product is. It may be different from unit prices since it represents a unit of weight rather than the price of any unit (Rosati, 1998). Where ‘high’ means RUEV exceeding an average of  3x std. deviations. This section is based on data extracted from Nauka i Technika w 2004 (Science and technology in 2004), GUS (Central Statistical Office), Warsaw 2006; Dzialalnos´ c´ innowacyjna przedsiébiorstw przemyslowych w latach 2002–2004, (Innovation activities of industrial firms in 2002–4), Central Statistical Office, Warsaw 2006. Innovation that is linked to introducing new or significantly improved products and/or production processes. The ratio of the number of people studying at university level to the population aged 19–24. The population aged 20–64 participating in education and training in the same age group. The population aged 20–24 having completed at least vocational school in the population of the same age group. Many more large than small firms have diminished unit costs as an effect of innovation. This is confirmed by the high level of their spending on building. However since 2004 an increasing number of FOCs have opened R&D units in Poland.

REFERENCES Aiginger, K. (1998), ‘Unit values to signal the quality position of CEECs, in the use of unit values to discriminate between price and quality competition’, in The Competitiveness of Transition Countries, OECD Proceedings, Paris, pp. 93–121.

Innovation activities versus competitiveness

219

Bender, Gerd, David Jacobson and Paul L. Robertson (eds) (2005), ‘Non-researchintensive industries in the knowledge economy’, Perspectives on Economic Political and Social Integration, XI (1–2), special issue. Bevan, A. A. and S. Estrin (2004), ‘The determinants of foreign direct investment into European transition economies’, Journal of Comparative Economics, 32, 775–87. Brusoni, S. and G. Sgalari (2006), ‘New combinations in old industries: the introduction of radical innovations in tire manufacturing’, Journal of Evolution Economics, 16, 25–43. Carroll, P., E. Pol and P. L. Robertson (2000), ‘Classification of industries by level of technology: appraisal and some implications’, Prometheus, 18 (4), 417–36. Dulleck, U. (2002), Trade and Transition – is there a low quality trap? WIFO. Hansen, P.A. and S. Göran (1997), ‘Will low technology products disappear? The hidden innovation processes in low technology industries’, Technological Forecasting and Social Change, 55, 179–91. Hirsch-Kreinsen, H. (2004), ‘Low technology’: a forgotten sector in innovation polity’, paper presented at the International ProACT Conference, 15–17 March, Tampere, Finland. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2003), ‘Low–tech industries and the knowledge economy: state of the art and research challenges’, paper written within the context of the research project ‘PILOT: Policy and Innovation in Low-Tech’. Hirsch-Kreinsen H., D. Jacobson and P.L. Robertson (2006), ‘ “Low-tech” industries: innovativeness and development perspectives. A summary of a European research project’, Prometheus, 24 (1), 1–21. GUS (2005), Nauka i technika w 2004 [Science and Technology in 2004], CD, Warsaw: Central Statistical Office. GUS (2006), Dzialalnos´c´ innowacyjna przedsie¸biorstw przemyslowych w latach 2002–4, [Innovation activities of industrial firms in 2002–4], Warsaw: Central Statistical Office. Kolasa, M. (2005), ‘What drives productivity growth in the new EU member states? The case of Poland’, ECB working paper series no. 486, Frankfurt am Main. Konings, Josef (2003), ‘Restructuring of firms in central and Eastern Europe’ in Gertrude Tumpel-Gugerell and Peter Mooslechner (eds), Economic Convergence and Divergence in Europe. Growth and Regional Development in an Enlarged European Union, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 168–82. Marczewski, Krzysztof (2005), ‘Zmiany struktury ekonomicznej polskiego eksportu towarów przetworzonych w latach 1998–2004’ in Jan Koty´nski (ed.), Polityka gospodarcza Polski w integruja¸cej sie¸ Europie 2005–2006. Raport roczny, Warsaw: IKCHZ, pp. 110–15. Murphy, K.M. and A. Shleifer (1991), ‘Quality and trade’, Journal of Development Economics, 53, 1–15. Organisation for Economic Co-operation and Development (OECD) (1998), The Competitiveness of Transition Economies, OECD proceedings, Paris: OECD. Palmberg, C. (2001), ‘Sectoral pattern of innovation and competence requirements. A closer look at low-tech industries’, Sitra Report Series 8, Helsinki: Sitra. Rosati, D. (1998), ‘Emerging trade patterns of transition countries: some observations from the analysis of “unit values” ’, MOCT-MOST, 8, 51–67. Robertson, P.L. and P.R. Patel (2007), ‘New wine in old bottles: technological diffusion in developed economies’, Research Policy, 36, 708–21.

220

Local versus global perspectives in innovation

Smith, K. (2002), ‘What is the “knowledge economy”? Knowledge intensity and distributed knowledge bases’, INTECH discussion paper series, The United Nations University, June. Sterlachini, A. (1999), ‘Do innovative activities matter to small firms in non-R&Dintensive industries? An application to export performance’, Research Policy, 28, 819–32. Sztaudynger, Jan J. (2003), ‘Nieliniowos´c´ wplywu inwestycji na wzrost gospodarczy’ [‘The nonlinear impact of investment on economic growth’], in Eugeniusz Kwiatkowski and Tomasz Tokarski (eds), Wzrost gospodarczy, restrukturyzacja i rynek pracy w Polsce. Uj z´ cie teoretyczne i empiryczne, [Economic growth, restructuring and labour market in Poland. Theoretical and empirical approach], Lódz: University of Lódz´. von Tunzelmann, Nich and Virginia Acha (2006), ‘Innovation in “low-tech” industries’ in Jan Fagerberg, David C. Mowery and Richard R. Nelson (eds), The Oxford Handbook of Innovation, Oxford, Oxford University Press: pp. 407–32. Weresa, T. (2004), ‘Can foreign direct investment help Poland catch up with the EU?’ Communist and Post Communist Studies, 37, 413–27. Wzia¸tek-Kubiak, Anna and Iga Magda (2007), ‘How do new member states cope with competition in the EU market’ in Hoshi Iraj, Paul J.J. Welfens and Anna Wzia¸tek-Kubiak (eds), Industrial Competitiveness and Restructuring in Enlarged Europe, Basingstoke, Hampshire: Palgrave Macmillan. Wzia¸tek-Kubiak, A. and I. Magda (2005), ‘Differentiation of changes in competitiveness among Polish manufacturing industries’, CASE Foundation Studies and Analysis no. 314, accessed at www.case.com.pl/upload/publikacja_plik/ 9939339_sa314.pdf.

12.

Low-tech industries between traded and untraded interdependencies: a dynamic concept of industrial complementarities Martin Heidenreich

INTRODUCTION The current debate on low-tech innovation patterns develops from criticisms of ‘high-tech myopia,’ the idea that economic growth and employment is mostly the result of research-intensive industries (Hirsch-Kreinsen et al., 2005a; Von Tunzelmann and Acha, 2005). This kind of critique is particularly present in the discussion of public research and technology policies. Rightly, one is reminded of the importance of medium- and low-technology industries for growth, employment and innovations. It is emphasized that ‘Learning and innovation can take place without research and development (R&D), for example through acquisition of tacit and practical knowledge, and through formal and informal diffusion between firms’ (Jacobson and Heanue, 2005: 315). Technological upgrades, better designs or customerspecific applications or, on a more general level, learning-by-interacting and practical, experience-based, often implicit knowledge is considered to be an essential source of innovation especially for low- and medium-technology industries (LMT).1 LMT industries therefore are considered to be an integral part of advanced industrial regions (Robertson and Patel, 2007). Their specific innovative behaviour (focus on process innovation, design and marketing, weak internal R&D and engineering capabilities, external acquisition of knowledge; see Heidenreich, 2008) is complementary to other knowledgeintensive industries and services. These complementarities between companies with a different technological basis are at the core of Pavitt’s taxonomy of different patterns of technological change. LMT companies are analysed as supplier-dominated firms2 which rely on the capabilities of 221

222

Local versus global perspectives in innovation

external technology suppliers in order to produce a product as cheaply as possible, or a design-intensive product. Recently, this complementarity of high- and medium-high-tech industries (HMHT) and LMT industries has been analysed by Robertson et al. (2003: 471) as complementarity between enabling and recipient sectors: ‘[T]he ability of innovative firms to survive and prosper depends on the extent to which they are able to fit into the context of established industries that dominate mature economies . . . the recipients provide the markets that innovative firms need to take advantage of economies of scale and amortize their R&D expenses.’ This concept of industrial complementarities is based on two assumptions: on the one hand, that they are limited to a national arena. The possibility of shifting technologically less demanding production processes to low-wage countries is not taken into account because it undermines the complementarities between enabling and recipient industries. On the other hand, the complementarities between high- and low-technology industries are assumed to be mostly based on market relations:3 low-tech sectors are dependent on efficient suppliers of advanced technologies and solutions, while the success of high-tech companies depends also on the size of the market: [T]he viability of high-tech sectors and the levels of resources devoted to research and development are often directly related to the rate of diffusion because the main customers for many high-tech products are in the established sectors, and therefore the rates of return to R&D in high-tech areas are also a direct function of rates of diffusion. (Robertson and Patel, 2007: 711)

These two assumptions are contradictory: if industrial complementarities are based mainly on market relations, it cannot be explained why labourintensive production processes in a common market are not completely concentrated in the countries with the lowest wages. Why are there still clothing or shoe companies in Italy, France or Germany? Why are these complementarities not transformed into complementarities between advanced regions which develop the required machinery, services and software, and low-wage regions which use these inputs? And if the complementarities between enabling and recipient industries are limited to only one country, a relocation of production should not be expected at all. On the basis of the concept of industrial complementarities proposed by Robertson et al. (2003), Pol et al. (2002) and Robertson and Pavitt (2007), we would therefore expect either no relocation of production at all (first assumption) or a complete outsourcing of LMT companies to East European countries (second assumption). Thus, we need a more dynamic and ‘socially embedded’ concept of complementarities which takes into account also other forms of

Low-tech industries between traded and untraded interdependencies

223

interdependencies and the dynamic balance between the advantages of spatial proximity of low- and high-tech industries and outsourcing (next section). On the basis of such a concept of industrial complementarities we analyse, on the basis of the ‘Fourth Community Innovation Survey’ of the European Union (EU), the traded and untraded interdependencies between low-, medium- and high-tech companies in Europe (third section). In the fourth section the territorial dynamics of LMT industries are described on the basis of the regional data of the EU. In the fifth section the attempts to face the risks and limitations of mostly traded interdependencies between LMT and HMHT companies are analysed taking the example of the newly designed Polish regional policies. Finally, our results are briefly summarized.

INDUSTRIAL COMPLEMENTARITIES BETWEEN LOW- AND HIGH-TECH SECTORS Complementarity refers to ‘a constellation in which two (or more) elements must be combined to produce a particular outcome’ (Höpner, 2005: 332). Complementary goods are goods whose sales are strongly tied to each other (for example Microsoft Windows and Microsoft Office). The sales of iron ore and coal are another example: when steel companies reduce their steel production, this will affect both the demand for iron ore and coking coal. Entrepreneurial strategies may also be complementary to each other: marketing strategies based on flexible, customized products, for instance, can be complementary to the use of flexible computer-based production technologies and logistics (see Milgrom and Roberts, 1995). In political economy, institutional complementarities are characterized by the fact that ‘the performance of a configuration increases when its elements assume specific properties’ (Höpner, 2005: 333; see also Crouch et al., 2005). A complementarity between regulated labour markets and regulated product markets implies that for example the highly regulated German labour market (which is shaped by a restrictive employment protection legislation, a vocational training system and the system of co-determination) increases the marginal ‘efficiency’ of regulated product markets facilitating the specialization of German industry in a ‘diversified quality production’ (Streeck, 1991). Also industries may be complementary to each other. The success of the German textile machine or mining equipment companies is also a result of the former strength and proximity to textile and mining companies. This sectoral dimension of complementarity has been analysed by Pavitt (1984), who distinguishes four sectoral patterns of innovation which are complementary to each other. First, routinized technological regimes are

224

Local versus global perspectives in innovation

dominated mostly by scale-intensive firms which have their competence in the coordination and organization of complex production processes. Second, specialized equipment suppliers transfer their knowledge to other businesses in the form of machinery and installations. Third, science-based firms are the core of high-tech sectors based on radical innovations, and supplier-dominated firms which are recipients of advanced technological equipment produced by other companies. Pavitt (1984: 364) describes the main technological flows between these sectors as follows: Supplier dominated firms get most of their technology from production intensive and science-based firms (e.g. power tools and transport equipment from the former; consumer electronics and plastics from the latter). Science-based firms also transfer technology to production intensive ones (e.g. the use of plastics, and of electronics, in the automobile industry). And . . . science-based and production intensive firms both receive and give technology to specialized suppliers of production equipment.

For Pavitt, the competitive advantage of these companies and their place in the industrial division of labour depends on specific capabilities; these are: research and development (science-based firms), organizational competence (scale-intensive firms), customer-specifically developed highquality machinery (equipment suppliers), and customer contacts and innovative design (supplier dominated firms). On this basis, industrial complementarities can be defined by the fact that the performance of an industrial sector (e.g. its innovativeness or profitability) increases due to the exchange of ideas, products, services and employees with another industrial sector. These complementarities may be either facilitated by territorial proximity or a common institutional environment supporting the exchange of knowledge between these industries or by market relations within or between countries. In general, such a complementarity is based on distributed capabilities (Robertson and Smith, 2007). Robertson and Patel (2007: 711) explain industrial complementarities mostly by economies of scale: Perhaps the most important backward linkage from LMT to high-tech industries comes simply from the revenue that sales provide, which helps to cover the substantial fixed costs that arise out of the innovation process and engenders economies of scale . . . Diffusion can be crucial at this stage because the larger the number of LMT industries that adopt an innovation, the quicker the rate of amortisation of development costs will be. These economies of scale can then be translated into lower prices of innovative products for the LMT industries (greater pecuniary externalities), further economies of scale for the hightechnology industries.

Low-tech industries between traded and untraded interdependencies

225

They assume that the interdependencies between low- and high-technology industries are mostly based on traded, market-based relations. This assumption can be justified by the fact that LMT industries are characterized mainly by mature products limiting the scope for product innovations. Therefore, the competitive emphasis favours ‘those firms, large or small, which are able to achieve greater skills in process innovation and process integration, and with more highly developed internal technical and engineering skills’ (Utterback and Suarez, 1993: 2). After fundamental process innovations, the process of standardization reduces even further the possibilities of product or process innovations. Cost minimization (for example by the relocation of production to low-cost countries) thus becomes a major strategy for LMT companies (Utterback and Abernathy, 1975). But traded interdependencies between low- and high-tech industries are only one dimension of industrial complementarities. Another dimension is ‘untraded interdependencies’ which are defined by Storper (1997: 5) as ‘conventions, informal rules, and habits that coordinate economic actors under conditions of uncertainty; these relations constitute region-specific assets in production’. These interdependencies facilitate mutual learning processes within a regional or national innovation system, that is, a ‘set of distinct institutions which jointly and individually contribute to the development and diffusion of new technologies . . . it is a system of interconnected institutions to create, store and transfer the knowledge, skills and artefacts which define new technologies’ (Metcalfe, 1995: 462–3). The relational networks between high- and low-tech industries thus may be based on traded and untraded interdependencies. The relationships between traded and untraded interdependencies are dynamic, and have also a territorial dimension. While untraded interdependencies are linked and limited to a specific community or society (for example a region or a country), traded interdependencies are not. The equilibrium between the concentration of LMT and HMHT industries within the same national or regional territory and the relocation of production, especially of LMT industries (which are especially sensitive to high wage differentials and lower transportation and communication costs), is therefore is a dynamic one, as shown in Figure 12.1. It can be expected that a general reduction of transportation and communication costs or a reduction of coordination costs shifts the equilibrium from E to E´ (a new equilibrium characterized by relocation of LMT industries to low-wage countries). An example of such a reduction is the enlargement process of the EU, which requires the implementation of the socalled community acquis and its 20 000 legal acts. The corresponding supranational regulation of the common market reduces trans-border

226

Local versus global perspectives in innovation

Transaction costs (due to untraded or traded interdependencies) Mostly untraded interdependencies (conventions, informal rules and habits that coordinate economic transactions faced with uncertainty)

E

Enlargement of the EU

Technical maturation of a sector

E‘

Relocation advantages (a function of high wage differentials between high- and low-wage countries and lower coordination, transportation and communication costs)

Traded, marketbased interdependencies (suppliercustomers for relations)

Territorial concentration of LMT and HMHT industries (due to the importance of spatial proximity)

Territorial decoupling of LMT and HMHT industries (outsourcing to low-wage countries)

Figure 12.1 Dynamic complementarities between low- and high-tech industries (as a function of traded and untraded interdependencies)

transaction costs and thus may cause a shift especially of LMT industries towards the East. Such a shift may also be caused by the technical maturation of a sector.

TRADED AND UNTRADED INTERDEPENDENCIES BETWEEN LOW- AND HIGH-TECH INDUSTRIES In this section we analyse the industrial complementarities between highand low-tech industries and their two dimensions (traded and untraded interdependencies) on the basis of the Fourth Community Innovation Survey (CIS 4).4 As an indicator of traded interdependencies we focus on the acquisition of machinery, equipment and software; as an indicator for untraded interdependencies we have chosen the information sources and the patterns of cooperation between innovative companies. The most important innovation activities of European companies are the acquisition of machinery, equipment and software (82.3 per cent) and

Low-tech industries between traded and untraded interdependencies

Table 12.1

227

Type of innovation activities in the EU (2002–4) Enterprises engaged in intramural R&D (%)

Enterprises Enterprises engaged in engaged in acquisition of acquisition of machinery, other external equipment and knowledge software (%) (%) (2004; in % of all innovative enterprises)

High-technology industries (HT) Medium-high-technology industries (MHT) HMHT-industries Medium-low-technology industries (MLT) Low-technology industries (LT) LMT-industries Manufacturing Services (excluding public administration) Total

83.2 77.5

75.8 76.8

28.7 23.4

78.6 55.2

76.6 82.9

24.5 17.6

48.3 51.3 61.2 45.4

85.4 84.3 83.0 78.2

19.2 18.5 20.9 26.2

57.3

82.3

23.7

Source: CIS4. Data for EU-27 countries without Austria, Ireland, Latvia, Slovenia, United Kingdom and Finland.

intramural R&D (57.3 per cent; see Table 12.1). The acquisition of machinery, equipment and software is even slightly more important in LMT industries in comparison with HMHT industries. At least quantitatively the external acquisition of machinery and software is the most important source of knowledge for all types of industrial sectors. The HMHT industries which have been designated ‘enabling industries’ are to an overwhelming extent also recipient industries characterized by the purchase of novel efficiency-enhancing products as predicted already by Robertson et al. (2003: 465). The expenditures for the acquisition of machinery, equipment and software constituted over half of the innovation expenditures of the LMT industries (in 2004: 51 per cent) and only one-fifth of the expenditures of the HMHT industries (in 2004: 19 per cent). In LMT industries most innovations originate from suppliers of equipment and materials (see Pavitt, 1984: 356). Internal sources of information are the most important source for innovative companies (45.6 per cent; see Table 12.2 and Faulkner, 1994). However, other important sources of knowledge are external, for example customers, suppliers, competitors and other companies. Companies are

228 18.2 23.3 26.2 24.9 22.8 25.2 23.1

39.9

40.6 44.8 44.1

45.6

18.0

53.1

54.3 41.5

18.9

59.0

Suppliers (%)

26.5

24.1 27.3 19.3

22.6

34.2 26.1

32.5

41.4

Clients or customers (%)

12.2

11.4 11.8 10.5

12.1

12.4 10.5

11.8

15.0

Competitors or other enterprises of the same sector (%)

5.8

6.3 6.3 6.4

6.6

5.7 5.9

6.1

4.3

Consultants, commercial labs or private R&D institutes (%)

3.7

3.2 4.1 2.9

2.9

6.2 3.6

6.1

6.3

2.7

2.7 2.9 2.5

2.6

3.2 3.0

3.3

3.1

UniverGovernsities ment or or other public higher research education institutes institu(%) tions (%)

11.5

13.5 12.7 9.2

14.6

11.6 12.0

10.8

14.9

Confer-, ences trade fairs, exhibitions (%)

8.3

8.1 8.1 8.0

9.4

8.1 6.3

7.5

10.8

Scientific journals; trade/ technical publications (%)

5.5

5.5 5.1 6.6

6.3

4.4 4.4

4.5

3.8

Professional and industry associations (%)

Sources: CIS4. Cf. Table 12.1. EU-27 member states without Austria, Ireland, Portugal, Latvia, Slovenia, Sweden and the United Kingdom. Suppliers: Suppliers of equipment, materials, components or software.

High-technology industries (HT) Medium-hightechnology industries (MHT) HMHT-industries Medium-lowtechnology industries (MLT) Low-technology industries (LT) LMT-industries Manufacturing Services (excluding public administration) Total

Within the enterprise group (%)

Table 12.2 Highly important sources of information for innovation, as a percentage of innovative enterprises, 20 EU member states

Low-tech industries between traded and untraded interdependencies

229

linked to these external sources also by untraded interdependencies. The most widely used external sources are customers (26.5 per cent), suppliers (23.1 per cent) and competitors (12.2 per cent). The role of clients or customers as sources of information is lower in LMT in comparison to HMHT industries (24.1 in comparison to 34.2 per cent). The role of suppliers however is much higher in LMT industries (26.2 per cent) in comparison with HMHT industries (18.2 per cent). This indicates that LMT companies rely strongly on the R&D and innovative activities of their suppliers (especially in the printing, food production, refinery, paper and clothing industries). Supplier-buyer relations are the most important basis for traded and untraded interdependencies of LMT companies. Another dimension of industrial complementarities are cooperation and innovation partnerships with suppliers, customers, competitors, advisors, universities and research institutes (see Table 12.3). LMT companies are involved in innovation partnerships less frequently than HMHT companies (21.6 instead of 32.6 per cent). The proportions of innovating and cooperating enterprises in 22 industrial sectors are highly correlated (r2: 0.48). This indicates that cooperation is a major dimension of untraded interdependencies. Suppliers are the most important external innovation partners in LMT industries. Their role is considerably higher than the role of suppliers in HMHT industries (24.9 in comparison to 18.2 per cent), where clients or customers are more important. The most important arena for cooperation is still the national level (19.5 per cent) – and this includes the regional level, which has not been analysed separately. Innovation cooperation with other European (9.1 per cent of all innovative enterprises) and extra-European companies (3.9 per cent) are much less important – an indicator for the territorial confinement of untraded industrial complementarities (see Table 12.4). In conclusion, the most important sources of knowledge for LMT industries are the supply of high-quality machinery and installations, sophisticated software, and intensive and frequent cooperation with suppliers, customers, research institutes and competitors. The innovativeness of LMT industries depends mostly on the suppliers of machinery, equipment and software, for which half of their innovation expenditures are spent. The role of suppliers as source of information, as providers of technology and innovation (and to some extent as cooperation partner) is so prominent that supplier relations can be considered the most important basis of traded and untraded interdependencies. Empirically, these two dimensions of industrial complementarities can hardly be distinguished. Only in the exceptional case of the EU enlargement which implies the integration of a country into a common market and the regulatory harmonization of two formerly separated territories, may the resulting decrease of coordination

230

Food products and beverages Tobacco products Textiles Wearing apparel; dressing Tanning, dressing of leather; luggage Wood and products of wood and cork, except furniture Pulp, paper and paper products Publishing, printing, reproduction of recorded media Coke, refined petroleum products and nuclear fuel Rubber and plastic products

21.6 28.6 21.7 14.9 15.2 19.2

23.0 14.4

44.0

24.3

38.0

31.2 38.7 18.0

25.8

32.1

45.1

50.2

22.5

51.6

Enterprises All types of with cooperation; innovation in % of all activity. innovative % of all enterprises enterprises (%) (%)

8.1

23.1

4.9

11.2

3.7

2.0

20.4 5.7 2.9

6.4

15.7

29.9

9.7

19.3

14.8

7.4

20.4 13.9 9.8

14.1

13.6

17.2

7.1

11.6

8.9

8.3

4.1 10.9 10.0

9.8

6.2

15.7

3.6

8.0

5.7

3.4

2.0 6.4 4.4

6.4

8.2

17.9

4.3

8.0

4.5

4.1

8.2 7.6 4.3

7.9

9.7

21.6

2.2

6.3

5.4

2.4

6.1 7.0 1.9

6.7

5.1

18.7

1.3

3.1

2.8

4.0

2.0 4.7 2.2

4.3

Other Suppliers Clients Competitors Consultants, Universities Government enterprises of or or other commercial or other or public within the equipment, customers enterprises labs, or higher research same materials, (%) of the private education institutes enterprise components same R&D institutions (%) group (%) or software sector (%) institutes (%) (%) (%) (Cooperation partners; in % of all innovative enterprises)

Table 12.3 Innovation activity and cooperation during 2002–4 (in percentage of all innovation enterprises)

231

18.3 34.1 20.7

20.5 19.1 37.4 31.3

32.6 23.3

20.3 21.6 25.2 27.4 25.5

42.6

48.1 42.9

41.4

37.4 63.0

54.1

55.7 41.6

34.0

37.0 41.7 26.8

39.5

9.5

6.3 8.5 10.7

5.8

12.7 6.8

12.6

2.4 13.3

4.2

11.8 4.8

6.3

16.5

14.3 16.1 18.9

14.2

19.8 14.5

19.1

14.1 22.6

15.1

19.6 11.3

11.9

13.9

11.3 13.7 12.2

10.1

19.0 12.9

18.0

6.8 22.8

9.9

19.7 11.3

8.1

8.3

6.2 7.3 9.3

5.9

9.7 6.6

9.4

4.9 11.2

5.4

7.8 6.0

5.7

8.9

7.4 8.9 9.0

6.9

11.7 8.1

11.3

8.3 13.3

6.0

9.2 6.2

7.1

8.8

6.6 9.6 6.6

5.2

15.8 8.5

15.0

8.9 18.9

5.7

18.1 7.2

7.0

5.7

4.1 5.8 5.3

3.4

8.9 5.0

8.8

3.2 9.3

2.8

10.1 4.2

4.8

Source:

CIS4. See Table 12.1. EU-27 Member States without Ireland.

Question: ‘During the three years 2002 to 2004, did your enterprise cooperate on any of your innovation activities with other enterprises or institutions? Innovation cooperation is active participation with other enterprises or non-commercial institutions on innovation activities. Both partners do not need to commercially benefit. Exclude pure contracting out of work with no active cooperation.’ Sources and notes: see Table 12.1.

Other non-metallic mineral products Basic metals Fabricated metal products, except machinery Furniture; manufacturing n.e.c. Recycling High-technology industries (HT) Medium-hightechnology industries (MHT) HMHT-industries Medium-lowtechnology industries (MLT) Low-technology industries (LT) LMT-industries Manufacturing Services (excluding public administration) Total

232

Local versus global perspectives in innovation

Table 12.4 Territorial dimension of innovation cooperation (2002–4; in percentage of all innovation enterprises) All types of Enterprise cooperation; engaged in in % of all any type of innovative innovation enterprises cooperation, national (%)

Enterprise Enterprise engaged in engaged in any type of any type of innovation innovation cooperation cooperation within other within United European States or other countries (%) countries (%)

(Cooperation partners; in % of all innovative enterprises) LMT-industries High- and mediumhigh technologies Manufacturing Services (excluding public administration) Total

21.6 32.6

17.3 25.0

7.9 13.9

2.1 7.4

25.2 27.4

19.8 27.4

9.9 25.3

3.9 9.7

25.5

19.5

9.1

3.9

Source: Author’s calculations on the basis of the Fourth Community Innovation Survey; database accessed on 02 March 2007. Unweighted averages for the EU.

costs for international trade relations drastically influence the balance between the untraded and the traded dimensions of industrial complementarities. A sharp increase in LMT industries in low-wage countries after an enlargement therefore should indicate a strong impact of traded interdependencies (as assumed by Robertson and Patel, 2007); a moderate or small increase indicates a great importance of untraded interdependencies also between high- and low-technology industries (as assumed by Storper, 1997).

ENLARGEMENT AND DYNAMICS OF LMT INDUSTRIES In the following we analyse to what extent the spatial distribution of HMHT and LMT industries has been changed by the enlargement of the EU – a process started in 1993 with the summit of Copenhagen and which resulted in formal EU accession in 2004 and 2007 of 12 mostly Central and Eastern European countries. If the assumption of Robertson and Patel

Low-tech industries between traded and untraded interdependencies

233

(2007) is correct, that industrial complementarities are mainly based on traded, market-based interdependencies, a significant shift of LMT industries to Eastern European countries could be expected during this period. Otherwise we should expect a limited relocation of LMT industries to lowwage countries. This question is discussed below on the basis of the employment shares of LMT and HMHT industries in Europe. Table 12.5 shows that in Europe, 11.7 per cent of employees worked in LMT industries in 2006 – almost twothirds of all the employees in manufacturing industry. The proportion is particularly large in the less prosperous Central and Eastern European countries, but also in Portugal and Italy. The size of the gross domestic product (purchasing power adjusted) correlates quite negatively with the proportion of LMT industries in 2004 (r  0.80). The proportion of LMT industries is lowest in Luxembourg, Ireland and the United Kingdom, that is, in countries with a very high GDP that have concentrated on knowledge-based services and industries. In the last 11 years the proportion of LMT industries in Europe has decreased from 13.8 in 1995 to 11.7 per cent in 2006, that is, by approximately one-sixth. The number employed in LMT industries in the more prosperous EU-15 countries (especially Denmark, United Kingdom, Luxembourg and Ireland) has decreased by more than 15 per cent, while its share has slightly increased in the 12 new member states (especially Romania, Bulgaria, the Czech Republic, Lithuania, Latvia and Poland). The share of high- and mediumhigh technologies remains stable in both Western and Eastern Europe. Therefore the contraction of the manufacturing sector is restricted to LMT industries. The reduction of the employment share of LMT industries in the more prosperous Western European countries suggests that the eastern enlargement of the EU contributed to the territorial decoupling of highand low-tech industries. The increasing territorial differentiation of the European industrial structure by the outsourcing of LMT industries to the European periphery is an indicator for the importance of traded, marketbased and cost-driven interdependencies also in comparison to untraded interdependencies. On the other hand 10.6 per cent (2006) of the employees in the 15 former member states are still employed in LMT industries, in comparison with 13.3 per cent in 1995. It can be assumed that the LMT industries which are still active in Western Europe are also linked by untraded interdependencies with HMHT companies – and that they play a crucial role in the competitiveness of industrial regions in Western Europe. Another indicator of the relative weight of the untraded and traded dimensions of industrial complementarities is the positive but low correlation between the regional employment shares of LMT industries and that of high- and medium-high-technology industries (2004: r  0.32 for 249

234

Luxembourg United Kingdom Ireland Sweden Denmark Cyprus Netherlands France Greece Belgium Malta Finland Spain Germany Austria Italy Latvia Hungary Lithuania

13.8 18.9 18.7 18.7 20.0 20.0 22.0 19.3 24.7 20.1 16.0 15.1 20.5 18.8 20.1 22.9 13.8 23.8 18.6

: 13.0 13.2 15.0 15.0 10.5 12.8 16.4 12.7 17.0 17.2 18.0 15.8 22.1 18.6 20.7 15.0 21.9 17.6

2.1 7.7 6.6 7.6 7.4 7.4 6.7 5.2 7.8 7.8 5.0 2.2 6.7 7.2 7.8 7.5 2.1 7.6 3.8 5.5 5.7 6.3 5.8 1.0 3.3 6.3 2.2 6.6 6.1 6.8 4.5 10.8 6.8 7.4 1.6 8.4 2.4

2006

1995

1995

2006

High- and medium-hightechnology manufacturing sector

Manufacturing sector

11.7 11.3 12.1 11.1 12.6 12.6 15.4 14.1 16.9 12.4 10.9 12.9 13.8 11.7 12.4 15.4 11.7 16.2 14.8

1995 7.5 7.6 8.7 9.2 9.5 9.5 10.0 10.4 10.4 11.1 11.2 11.2 11.4 11.8 13.4 13.4 13.5 15.2

2006

Low- and medium-lowtechnology manufacturing sector

3.1 7.1 8.7 6.9 8.7 8.7 9.5 9.5 11.0 7.5 7.9 10.0 9.4 7.3 7.5 10.0 3.1 11.9 12.7

1995 4.6 5.3 5.1 6.2 7.1 6.8 5.8 7.6 6.2 8.4 7.1 7.0 6.5 6.4 7.8 11.0 8.9 12.3

2006

Low-technology manufacturing sector

Employment in high- and low-technology sectors (2006; in percentage of total employment)

Sorted by share of LMT industries

Table 12.5

30.5 36.8 29.2 44.2 39.0 39.0 25.6 22.2 24.1 32.9 36.7 20.1 37.3 33.5 32.9 24.0 30.5 25.3 23.4

1995

43.0 34.9 47.5 43.8 28.3 42.3 36.4 24.9 38.6 31.2 41.1 27.0 34.3 30.4 30.1 24.5 28.5 25.0

2006

Knowledgeintensive service (KIS)

235

20.3 22.8 21.8 25.7 18.7 25.0 23.4 34.0 20.5 20.1 22.0

20.7 18.9 21.7 26.3 21.3 28.3 23.7 27.4 18.3 17.4 22.2

4.9 4.1 6.7 6.6 6.6 11.0 5.6 9.2 6.7 6.8 6.4

5.1 3.2 5.7 9.7 3.5 10.3 4.8 8.5 6.6 6.8 6.5

15.4 18.7 15.1 19.0 12.1 14.0 17.8 24.8 13.8 13.3 15.6

15.5 15.7 16.1 16.6 17.8 17.9 18.9 18.9 11.7 10.6 15.7

10.9 14.3 9.7 11.7 8.7 7.8 13.4 16.5 9.0 8.7 10.5

10.5 11.5 12.2 9.7 13.8 9.3 14.8 11.2 7.4 6.5 10.2

24.3 21.7 11.7 24.2 29.2 26.9 21.2 20.8 28.9 30.7 22.0

24.6 22.7 14.5 24.9 26.8 25.0 21.7 26.3 32.6 35.1 25.4

Source: Eurostat, regional database (accessed 25 August 2007).

Data for 1995 partially estimated on basis of later years. Total knowledge-intensive services: NACE Rev. 1.1 codes 61, 62, 64 to 67, 70 to 74, 80, 85 and 92.

Poland Portugal Romania Slovakia Estonia Czech Republic Bulgaria Slovenia EU-27 EU-15 New member states

236

Local versus global perspectives in innovation

European NUTS2-regions). As assumed by Robertson and Patel (2007) the performance and innovativeness of high-tech industries is also dependent on the knowledge accumulated in low-tech industries – but this relationship is rather weak and it does not prevent outsourcing and relocation decisions because market-based interdependencies can also be extended beyond national boundaries. In conclusion: regions with a high proportion of LMT industries are also characterized by a greater share of high-technology sectors, thus confirming the observation of Von Tunzelmann and Acha (2005: 429): ‘[W]hat we observe is a varying degree of permeation of high technologies into low-tech and medium-tech as well as into high-tech sectors’. However the complementarities between low-, medium- and high-tech sectors are mostly market-based, that is, they are not limited to interdependencies within the same national configuration. This explains the fact that the disparities in the distribution of LMT industries between Eastern and Western European countries have clearly increased over the last 11 years. This indicates an increasing specialization of the Central and Eastern European countries in this sector: ‘A number of industries that were initially spatially dispersed have become more concentrated. These are mainly slow growing and unskilled labour intensive industries whose relative contraction has been accompanied by spatial concentration, usually in peripheral low wage economies’ (Midelfart-Knarvik et al., 2000). Our results thus indicate that the complementarities of high- and low-tech industries which are based on territorial proximity have been weakened by the fifth EU enlargement. The impact of untraded interdependencies between HMHT and LMT companies does not prevent the relocation of production. The shift of LMT industries to Eastern Europe is the result of the comparative competitive disadvantage of LMT industries in high-wage countries; it can be interpreted as an indicator of the primarily traded complementarities between low- and high-tech industries because these complementarities can also be maintained by trade relations beyond national boundaries. On the one hand LMT industries are therefore linked by untraded complementarities to HMHT industries within the same region or country. On the other hand however, low- and medium-low technology branches are able to exploit lower labour costs in Eastern Europe, thus renouncing on the advantages of territorial proximity and untraded interdependencies. But this is not the end of the story: Central European countries try to transform the traded interdependencies between their newly created or relocated LMT companies and their regional and national environment once again into untraded interdependencies. This is discussed in the next section.

Low-tech industries between traded and untraded interdependencies

237

REGIONAL POLICIES IN POLAND: AN ATTEMPT TO TRANSFORM TRADED INTO UNTRADED INTERDEPENDENCIES? The Central and Eastern European countries are now facing the other side of the coin of traded interdependencies: Initially they were the (relative) winners of the relocation processes from Western to Eastern Europe, but now they are threatened by the same dynamics because their labour costs are rapidly increasing and other countries offer even lower wage costs. Faced with the threat of capital mobility and wage competition they may attempt to transform the traded interdependencies – which link the newly created LMT companies to their new national environment, but also to their home countries – once again into untraded interdependencies. The new regional policies in Poland are such an attempt. These policies try to support companies not only by the provision of public subsidies, a good transportation infrastructure, the availability of qualified employees, the support of small- and medium-sized companies by legal and financial services, and the provision of real estate or business incubators, but also by the stimulation of mutual learning processes between regional companies, agencies and political authorities. Below, we briefly describe the revival of sub-national Polish regions and discuss the question as to what extent these regional policies already succeed in transforming traded interdependencies into untraded ones. In the 1990s Poland created three sub-national, democratically legitimated administrative levels (voivodship, powiat, gmina) which have competency also in the field of economic and spatial planning (Gorzelak, 2000; Hausner et al., 1995). The creation of 16 large sub-national regions (voivodships) in 1999 was a decisive step towards the decentralization of Polish economic policy and could lead to the development of regional innovation systems (for this concept see Cooke et al., 2004). In 2000, the first ‘National Strategy for Regional Development’ was submitted. In 2001–3 the negotiations between the economic and labour ministries and the voivodships on the funding of regional programmes took place. The regionalization of economic policy is therefore still at the beginning, since state as well as regional actors must first get used to their new competencies (Pollock, 2004). The European Commission points out in particular three weaknesses: (a) each regional development plan was prepared independently without coordination between regional and national levels; (b) the fragmentation of the support system and the lack of regional flagship programmes weaken the regions; (c) the low level of experience of regional authorities and the limited interest of the private sector weakens the development of regional innovation policies (COM, 2005: 8). In 2004

238

Local versus global perspectives in innovation

however the decision was taken to further strengthen the regional level and its role in the development of regional innovation projects. The regional Marshal’s offices (and not the competing voivods) were assigned a crucial role in the conception and implementation of the regional development plans (COM, 2005: 7). This regionalization strategy will be continued in the current Regional Operational Programs (ROPs) for the period 2007–2013. Since 2000 numerous programmes for the economic revitalization of the Polish regions have been developed by the central government, especially by the Ministry of Economy and Labor and its different agencies. The most important are the ‘Polish Information and Foreign Investment Agency’, the ‘Polish Agency for Enterprise Development’ and the ‘Industrial Development Agency’ (ARP), which organize and monitor the regionalization process and provide the financial means for regional activities. The Industrial Development Agency for example has created and managed some of the ‘Special Economic Zones’ and ‘Industry and Technology Parks’ in Poland. The crucial advantages of the currently existing 14 special economic zones are state subsidies. Until 2017 investors can be refunded half their investment capital or half of their two-year labour costs. Three-quarters of the foreign capital (in Poland) is invested in such special economic zones. Industry and technology parks were created by ARP in 2002, but now they are mostly administered by regional development agencies. Another outcome of the centrally orchestrated decentralization of economic policies is the institutionalization of regional development agencies. The regionalization of economic policies has focused until now mostly on the provision of subsidies and other economic advantages especially to foreign investors, though recently it has shifted its efforts to the creation of regional networks (an essential dimension of untraded interdependencies). It is still uncertain to what extent this shift will be successful. Even the necessity of untraded interdependencies has not yet been unanimously accepted. Lorentzen (2007) for example shows that Polish companies use supranational and global sources of knowledge through the Internet as a functional alternative to local networks and localized interactive learning. From interviews conducted in 23 innovative, mostly private companies located in and around Cracow and Breslau, she states that ‘The national level did not play any outstanding role in knowledge sourcing . . . [l]ocal networks among firms . . . did not exist in the case studied here . . . The global scale represents the level towards which the innovative searches of the firms were directed.’ Also for Dornisch, (2002: 315) the ‘absence of embedded patterns of cooperative relations’ is an advantage, because this absence contributed to the ‘transitional capacity’ which has been crucial to the transformation of the Lodz region.

Low-tech industries between traded and untraded interdependencies

239

As a general argument against regional embeddedness, this argument is not convincing, as can be shown on the basis of the CIS4 (Table 12.6): The total share of innovative companies in Poland (24.8 per cent) is much smaller than the respective share in the EU (39.5 per cent). This is not only a result of a different pattern of industrial specialization. Also the Polish LMT companies are less innovative than their counterparts in the EU (23.5 in comparison to 37 per cent). There seem to be systematic barriers to innovation even in comparison with similar branches in other countries. This may be an indicator for the limitations of predominantly traded interdependencies between often foreign HMHT companies and Polish LMT companies. On the other hand if Polish companies are innovative, they are much more involved in cooperative relations, especially with their suppliers and clients, but also with competitors, consultants, universities and public research institutes (42.2 in comparison to 25.5 per cent). This is not only true for HMHT branches, but also for LMT branches (55 in comparison to 39 per cent). Also in Poland the most important arena for cooperation is the national level (36.1 per cent of all innovative Polish enterprises in comparison with 19.5 per cent of all European enterprises). The same pattern – a smaller share of innovative companies combined with a stronger involvement of innovative companies in (mostly national) cooperation networks – can also be observed in other post-socialist countries. This indicates that an involvement in cooperation networks in post-socialist countries is a crucial dimension of entrepreneurial innovativeness – especially in LMT industries. Due to the relatively weak impact of regional patterns of cooperation in Poland, Lorentzen (2007) advocates policies that support a global search for knowledge. In a similar vein, Dornisch (2002: 318) cautions mimicking the interactive learning of established western regions and recommends, to avoid lock-in, the continual generation of projects aimed at many complex tasks. Yet, the relative weakness and the strong role of national patterns of cooperation, also in innovative Polish companies, indicate that international forums of cooperation or the possibilities of non-embedded forms of radical innovation (traded interdependencies) are no sufficient alternative to nationally embedded forms of cooperation (untraded interdependencies). Therefore, the weak innovativeness of many Polish companies, even in comparison with other low- and medium-low-tech companies in Europe, may also be the result of the ‘absence of embedded trust-based patterns of cooperative relations’, that is, untraded interdependencies. This can only partially be overcome by foreign sources of knowledge, that is, traded interdependencies. The crucial question for the new Polish regional policies therefore is to what extent they will be able to integrate companies in LMT industries (but not only) into regional or national cooperation and innovation networks.

240 27.4 25.5 43.2 39.0 55.2 40.9 42.2

39.5

26.2 23.5 38.7

22.9

24.8

25.2 21.6 32.1

26.8

41.7 37.0 56.0

12.7

20.3

9.0 7.0 14.5

9.5

10.7

8.5 6.3 13.0

Other enterprises within the same enterprise group (%)

28.2

25.6

28.9 27.2 33.9

16.5

18.9

16.1 14.3 19.7

16.4

12.5

20.0 17.5 26.7

13.9

12.2

13.7 11.3 18.6

8.5

7.4

9.4 8.5 11.8

8.3

9.3

7.3 6.2 9.5

7.9

7.6

7.5 6.5 10.3

8.9

9.0

8.9 7.4 12.0

6.1

5.1

7.8 5.4 14.5

8.8

6.6

9.6 6.6 15.6

8.7

11.2

8.9 6.4 15.9

5.7

5.3

5.8 4.1 9.1

Suppliers of Clients Competitors Consultants, Universities Government equipment, or or other commercial or other or public materials, customers enterprises labs, or higher research components (%) in same private R&D education institutes or software sector institutes institutions (%) (%) (%) (%) (%) Cooperation partners; in % of all innovative enterprises

Source: Author’s calculations on the basis of the Fourth Community Innovation Survey; database accessed 2 March 2007. Unweighted averages for the EU.

23 EU member states Manufacturing LMT-industries High- and mediumhigh-technologies Services (excluding public administration) Total Poland Manufacturing LMT-industries High- and mediumhigh-technologies Services (excluding public administration) Total

Enterprises All types of with cooperation; innovation in % of all activity. innovative % of all enterprises enterprises

Table 12.6 Innovation activity and cooperation in Poland and the EU (2002–4; in percentage of all innovative enterprises)

Low-tech industries between traded and untraded interdependencies

241

CONCLUSION The aim of this chapter was to take a closer look at the concept of industrial complementarities which has recently been proposed by Robertson and Patel (2007) as a tool for the analysis of the relationships between high-, medium- and low-technology sectors. These complementarities are defined by the improvement of the performance of an industrial sector due to the exchange of ideas, products, services and employees with another industrial sector. Analytically, two types of these complementarities can be distinguished: first, complementarities based on traded interdependencies (economic transactions which are coordinated by the law of supply and demand); second, complementarities based on untraded interdependencies, or ‘conventions, informal rules, and habits that coordinate economic actors under conditions of uncertainty’ (Storper, 1997: 5) and facilitate the diffusion of tacit or implicit knowledge. While untraded interdependencies are based on common patterns of interpretation and communication which have been evolved at specific locations, traded interdependencies are market relations which can easily transcend national boundaries. The most important relationships for LMT industries are supplier relations characterized by both traded and untraded interdependencies: their innovativeness depends predominantly on the acquisition of machinery, equipment and software. Suppliers are also the most important cooperation partners and sources of information and knowledge. In contrast to high-tech industries, in low- and medium-low-technology industries the traded and untraded dimensions of industrial complementarities can be decoupled because they are characterized mainly by process innovation, limiting the scope for product innovations and increasing the need for cost minimization strategies (for example by the relocation of production to low-cost countries). This explains the contraction of LMT industries especially in Western Europe and their relocation to Eastern Europe. In this way the former complementarities between high- and low-technology sectors within the same region or country are increasingly replaced by other complementarities between advanced and peripheral industrial regions. The latter regions concentrate on low- and medium-low-technology industries, which are functionally and territorially separated from advanced industries and regions. LMT industries are increasingly concentrated in low-wage countries (see Midelfart-Knarvik et al., 2000) which are recipients of technology and new ideas. The new regional policies of the Polish government show that this danger has been recognized and that the decentralization and regionalization of economic policies is aiming for a stronger regional embeddedness of local companies and strengthening of untraded interdependencies also between high- and low-tech sectors, regional business

242

Local versus global perspectives in innovation

associations, universities, research and technology transfer centres and newly created or relocated regional companies.

NOTES 1. The distinction between LMT and medium-high-tech industries (HMHT) is based on the typology of industrial sectors proposed by the OECD (see Hatzichronoglou, 1997) which distinguishes four sectors: (a) high-technology; (b) medium-high technology; (c) mediumlow technology and (d) low-technology. This classification is based on the direct and indirect R&D intensity (including the R&D expenditure embodied in intermediates and capital goods purchased) of 22 manufacturing sectors. The first two categories are categorized as HMHT industries, the last two as LMT industries. 2. Pavitt (1984: 356) describes these companies as follows: ‘Supplier dominated firms can be found mainly in traditional sectors of manufacturing . . . They are generally small, and their in-house R&D and engineering capabilities are weak. They appropriate less on the basis of a technological advantage, than of professional skills, aesthetic design, trademarks and advertising. Technological trajectories are therefore defined in terms of cutting costs. Supplier dominated firms make only a minor contribution to their process or product technology. Most innovations come from suppliers of equipment and materials, although in some cases large customers and government-financed research and extension services also make a contribution . . ., in sectors made up of supplier dominated firms, we would expect a relatively high proportion of the process innovations used in the sectors to be produced by other sectors, even though a relatively high proportion of innovative activities in the sectors are directed to process innovations.’ 3. Only recently Robertson and Smith (2007) have shifted their focus to distributed knowledge bases which are another crucial dimension of the interaction between low- and hightech companies. 4. In the fourth wave of this survey in 2005, over 750 000 businesses with ten or more employees were asked about different innovation activities, innovation cooperation and sources of information on innovation, main obstacles to innovation activity, and so on. Only market activities in industry and in many service sectors (wholesale trade; transport, storage and communication; financial intermediation; computer and related activities, architectural and engineering activities; and technical testing and analysis) are included. The observation period covered three years, from the beginning of 2002 to the end of 2004. CIS4 was carried out in the current 27 EU member states, Iceland and Norway. A major reason for the selection of this European-wide survey was that in the fourth wave the industry classification of the enterprises questioned was declared, so that the data could be combined as low-, medium- and high-technology industries. For previous waves this was not possible. For some countries (often for Austria, Ireland, Latvia, Slovenia, the UK, Sweden and Finland) many data are missing.

REFERENCES COM/DG Enterprise (2005), ‘European trend chart on innovation’, annual innovation policy trends and appraisal report, Poland 2004–2005, accessed 14 March, 2006 at http://trendchart.cordis.lu. Cooke, Philip, Martin Heidenreich and Hans-Joachim Braczyk (eds) (2004), Regional Innovation Systems: The Role of Governance in a Globalized World, 2nd edn, London and New York: Routledge.

Low-tech industries between traded and untraded interdependencies

243

Crouch, Colin, Wolfgang Streeck, Robert Boyer, Bruno Amable, Peter A. Hall and Gregory Jackson (2005), ‘Dialogue on “Institutional complementarity and political economy” ’, Socio-Economic Review, 3 (2), 359–82. Dornisch, D. (2002), ‘The evolution of post-socialist projects: trajectory shift and transitional capacity in a Polish region’, Regional Studies, 36 (3), 307–21. Faulkner, W. (1994), ‘Conceptualizing knowledge used in innovation: a second look at the science-technology distinction and industrial innovation’, Science, Technology & Human Values, 19 (4), 425–58. Gorzelak, Grzegorz (2000), ‘The dilemmas of regional policy in the transition countries and the territorial organisation of the state’, in George Petrakos, Grzegorz Gorzelak and Gunther Maier (eds), Integration and Transition in Europe: The Economic Geography of Interaction, London: Routledge. Hatzichronoglou, T. (1997), ‘Revision of the high-technology sector and product classification’, STI working papers 2, Paris: OECD. Hausner, Jerzy, Tadeusz Kudlacz, and Jacek Szlachta (1995), Regional and Local Factors in the Restructuring of Poland’s Economy: The Case of South-Eastern Poland, Kracow: Kracow Academy of Economics. Heidenreich, Martin (2008), ‘Innovation in European low- and medium-technology industries’, paper submitted to the special issue of Research Policy on ‘Technological Change in Low- and Medium-Technology Industries’. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2003), ‘Low-tech industries and the knowledge economy: state of the art and research challenges’, University of Dortmund working paper 1. Hirsch-Kreinsen, Hartmut (2005), ‘Low-tech industries: knowledge base and organisational structures’, in Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), Low-Tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang, pp. 147–66. Hirsch-Kreinsen, Hartmut, David Jacobson and Staffan Laestadius (eds) (2005a), Low-Tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2005b), ‘Low and medium technology industries in the knowledge economy: the analytical issues’, in Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), LowTech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang, pp. 11–30. Höpner, M. (2005), ‘What connects industrial relations and corporate governance? Explaining institutional complementarity’, Socio-Economic Review, 3 (2), 331–58. Jacobson, David and Kevin Heanue (2005), ‘Implications of low-tech research for policy’, in Hartmut Hirsch-Kreinsen, David Jacobson and Staffan Laestadius (eds), Low-Tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang, pp. 315–31. Lorentzen, A. (2007), ‘The geography of knowledge sourcing. A case study of Polish manufacturing enterprises’, European Planning Studies, 15 (4), 467–86. Metcalfe, Stan (1995), ‘The economic foundations of technology policy: equilibrium and evolutionary perspectives’, in Paul Stoneman (ed.), Handbook of the Economics of Innovation and Technological Change, Oxford and Cambridge, MA: Blackwell Publishers, pp. 409–512. Midelfart-Knarvik K.H., H.G. Overman, S.J. Redding and A.J. Venables, (2000), ‘The location of European industry’, European Commission Economic Papers, 142.

244

Local versus global perspectives in innovation

Milgrom, P. and J. Roberts (1995), ‘Complementarities, industrial strategy, structure and change in manufacturing’, Journal of Accounting and Economics, 19, 179–208. Pavitt K. (1984), ‘Sectoral patterns of technical change: towards a taxonomy and a theory’, Research Policy, 13, 343–73. Pol, E., P. Carroll and P.L. Robertson (2002), ‘A new typology for economic sectors with a view to policy implications’, Economics of Innovation and New Technology, 11 (1), 61–76. Pollock, R. (2004), ‘The regional dimension’, CIMPAN, Rotterdam, accessed 14 November, 2006, at www.urenio.org/metaforesight/library/4.pdf. Robertson, Paul L. and Keith Smith (2007), ‘Technological upgrading and distributed knowledge bases’, draft version presented at the workshop ‘LowTechnology’: Innovativeness, Development and Perspectives in the Knowledge Economy, Dortmund, 15–16 March. Robertson, Paul L. and Parimal R. Patel (2007), ‘New wine in old bottles: technological diffusion in developed economies’, Research Policy, 36 (5), 708–21. Robertson, P.L., E. Pol and P. Carroll (2003), ‘Receptive capacity of established industries as a limiting factor in the economy’s rate of innovation’, Industry and Innovation, 10 (4), 457–74. Storper, M. (1995), ‘The resurgence of regional economies, ten years after: the region as a nexus of untraded interdependencies’, European Urban and Regional Studies, 2, 191–221. Storper, Michael (1997), The Regional World: Territorial Development in a Global Economy, New York and London: Guilford Press. Streeck, Wolfgang (1991), ‘On the social and political conditions of diversified quality production’, in Egon Matzner and Wolfgang Streeck (eds), Beyond Keynesianism: The Socio-Economics of Production and Full Employment, Aldershot, UK and Brookfield, VT, USA: Edward Elgar, pp. 21–61. Utterback, J.M. and W. J. Abernathy (1975), ‘A dynamic model of process and product innovation’, Omega, 3, 639–56. Utterback, J.M. and F. F. Suarez (1993), ‘Innovation, competition and industry structure’, Research Policy, 22, 1–21. Von Tunzelmann, Nick and Virginia Acha (2005), ‘Innovation in “low-tech” industries’, in Jan Fagerberg, David Mowery and Richard Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 407–32.

13.

High-tech innovation in catchingup countries: conditions and perspectives Staffan Laestadius, Linda Gustavsson and Vicky Long

INTRODUCTION In recent decades there has been an enormous increase in activities related to science and technology (S&T) and innovation outside OECD regions; in particular there is a dramatic expansion of S&T in Asian countries. China is probably the most well-known example of this process. After Japan, the first-tier ‘Asian Tigers’ (Korea, Taiwan and Singapore) are now followed by the second-tier ones (primarily China and India) – but these with significant home market potential and global-sized economies. Around the corner we can discern Asian countries like Vietnam – still poor but highly literate and dynamic, as well as Malaysia and Thailand which have been on the track for some years. Although this dynamic to a significant extent seems to be related to classical growth, it is obvious that it also contains non-incremental innovations as well as the creation of a variety of highly competitive capabilities. This chapter is about a high-tech profile in this dynamic which challenges our conventional wisdom on how the new industrializing countries should develop – and how ‘incumbent’ (since long developed) nations should act to stay competitive.1 The statistical picture of this new Asian dynamics is visible in many publications. International organizations like WBRD, IMF, UN, OECD and WTO nowadays publish analyses and statistical data which more or less all indicate the dramatic transformation taking place. Instead of presenting an enormous amount of data once again in a slightly revised version, the intention of this chapter is to go beyond the surface and analyse some of the data – and the processes behind them – in more depth. Following this introduction, the second section is illustrative rather than comprehensive and combines description of stylized facts (in a mode 245

246

Local versus global perspectives in innovation

inspired by Kaldor, 1985) with in-depth penetration for better understanding. The aim of this presentation is to provide a general common ground for the second section of the chapter. Using selected Chinese data we illustrate the difficulties of penetrating the aggregate statistics in fastgrowing giant economies, and we show that for countries like China (and India) the rise of their S&T and innovation efforts may even be severely underestimated. This underestimated dynamics may be analysed in still more dimensions. In the third section we discuss three foundations of our conventional wisdom and the reasons behind it. The fourth section focuses on knowledge-formation processes, on capabilities and on the characteristics of complexity. The final section contains a concluding discussion. The empirical parts of this chapter are primarily based on Chinese S&T statistical data and on in-depth interviews, discussions and observations (50) in Chinese ICT firms working in and outside China as well as foreign ICT multinationals working in China. As is the case in all analyses of rapid transformation processes, solid conclusions are difficult to draw; the empirically-based conclusions presented here are conjectures based on rapidly changing data rather than the final truth.

S&T AND INNOVATIVE EFFORTS AMONG THE NEW ASIAN TIGERS – DATA AND ANALYSIS As indicated above, this section is illustrative rather than comprehensive. For that purpose three sets of high-tech indicators have been chosen: ● ● ●

Foreign trade in high-tech products; Scientific publications and patent data; and R&D statistics with a certain focus on ICT and mobile communication data.

Foreign Trade in High-tech Products First of all, global trade2 has increased dramatically between 1995 and 2005; from $US5164 billion to $US10 431 billion – 202 per cent. On the regional level above-average growth of merchandise exports have been obtained, for instance by Latin America (238 per cent), the Middle East (356 per cent) and (the rest of) Asia (211 per cent). Among these is a set of success stories starting from very low levels. The most spectacular case is probably Vietnam which over the decade has increased its exports with 580 per cent, thus jumping to rank 50 among world exporters. Most spectacular in speed

High-tech innovation in catching-up countries

247

as well as size is of course the dramatic development of Chinese exports. The increase of more than 500 per cent from $US149 billion to $US762 billion – a speed which was visible already in the 1980s – has transformed China to become the third-largest exporting country in the world (after the US and Germany). Focusing on China we may, second, also note that the manufacturing content in Chinese exports, 92 per cent in 2005 and increasing, was not only significantly above world average but also above that of countries like the US, Germany and France. Chinese exports of office and telecommunication equipment (OTE) has during the period 2000 to 2005 increased 519 per cent from $US43.5 billion to $US226 billion making China an ICT exporter of the same magnitude as Japan and the US taken together ($US98 billion and $US126 billion respectively). This segment of Chinese exports grew during those years of rapid expansion of total exports from 17.5 to 29.7 per cent, making it twice that of the US and nearly the same as Japan’s (13.9 and 16.5 per cent respectively of total exports). As regards integrated circuits – normally assumed to be at the high end of the OTE segment – incumbent countries like the US still have a dominant position; but even here, their share of the world market is diminishing (from 20.4 to 14 per cent during 2000 to 2005). It is thus not so that this expansion by catch-up countries – taking place in many other although smaller countries as well – is made up of simple low-end products and assembly only. China is now a leading high-tech exporter and is increasing that lead. Chinese imports have also increased dramatically during the period discussed above (as is the case with all the Asian tiger economies). This is of course important from a general balance-of-trade perspective although that is not a focus of this chapter. Of importance from our perspective is that the dramatic increase in Chinese (and other ‘tiger’ economy) imports of ICT products – roughly of the same size and speed as the export figures – indicates that catch-up actors and countries are also rapidly introducing IC technology into the transformation of their industry and society and are not only repackaging them for export. Part of that transformation has taken the form of one of the fastest industrialization processes known in history: for a quarter of a century Chinese manufacturing output has increased at around 12 per cent annually. Scientific Publications and Patent Data Chinese scientific publications and citations in international journals have seen a dramatic increase recently. In 1995 China produced 11 100 internationally cited papers in English. Ten years later (2005) 63 700 papers were

248

Local versus global perspectives in innovation

recorded by SCI, an increase of almost 19 per cent per year (ISI database, 080109). In 2006, 77 700 papers were published – an increase of another 22 per cent. That was also the year in which China surpassed England in the number of citations in scientific journals. The rising productivity of Chinese scientists seems in line with the process in other Southeast Asian countries like South Korea and Singapore (see Choung et al., 2003; Leydesdorff and Zhou, 2005). In certain areas, for example nanotechnology, China some years ago surpassed the US in terms of numbers of publications (King, 2004). Chinese nanoscience publications in 2006 were almost twice as many as the US: 1500 versus 780 (ISI database 080109). A recent patent database study on Chinese patents in USPTO reveals a similar pattern for China as was the case for South Korea in innovative technology areas, namely that the technological sub-sectors of mechanical engineering and electronics are taking the lead in the catch-up process (Long and Palmberg, 2006). Interesting here is the ‘hockey stick’ development – slow progress followed by a rapid take off – during the first years of this century. Chinese patenting activity started late but seems to be coming on fast. R&D Performance with Focus on ICT and Mobile Communication Data Instead of just adding impressive figures this section analyses a small set of them in depth in order to reveal the complexity of understanding the processes taking place. As mentioned above we use China to illustrate our case. The intent is not to generalize from the most extreme case (which China probably is) but to use it as a platform for further discussions on how to improve our understanding of the processes beneath the surface. Starting some years ago Chinese authorities commenced publishing S&T data in English, thus making the development processes taking place more transparent also for non-Chinese-speaking analysts. It is recognized in many international publications that Chinese statistics still have severe problems with the quality of fundamental data (which statisticians are struggling with).3 But we have seen no recent indications of systematic intentional biases in published tables. In the following we use data from the Chinese MOST4 and MII.5 Chinese gross domestic expenditure on R&D (GERD) has during recent years (2003) grown to RMB 154 000 million or approximately 1.31 per cent of GDP. China is thus ahead of Ireland and Italy in R&D intensity. But is this the whole story? In this section we discuss four phenomena all of which indicate that this standardized way of normalizing R&D efforts – although the evident data are impressive – may hide significant parts of the dynamic segments of some giant economies. Two of them are illustrated

High-tech innovation in catching-up countries

249

with GERD statistics: the size phenomenon and the disequilibrium phenomenon. The third is based on R&D statistics for high-tech industry and indicates what we label a velocity phenomenon. The fourth phenomenon, finally, we relate to the transformation of the Chinese system. As regards the size phenomenon that is basically related to the heterogeneity of China: not all of China is developing fast – just a ‘smaller’ segment of it. China is not one innovation system: its modern and extremely dynamic sectors take in probably 300 million people – an economy comparable to the population of Western Europe – with incomes still lower though rising in PPP terms. The rest of the economy – some whole provinces, parts of other provinces and areas or segments of some dynamic regions, with approximately 1000 million people – is included in the aggregate national statistics, which makes averages like ‘per capita’ or ‘percentage of GDP’ of little value. The cellular mobile penetration in China may illustrate this point. China has a penetration rate of 29.9 subscriptions per 100 inhabitants, which is quite modest worldwide: Botswana has 46.6 (ITU, May 2006; these ratios are rapidly changing in both countries). In absolute numbers the Chinese mobile market and telephony system is however by far the most important in the world, and the Chinese hot spots for which we have data have cellular densities by far exceeding those of Scandinavia (Hong Kong: 122.7; Macao: 115.8) – data that give a totally different picture. Analysing R&D data at the provincial level we find that the three Chinese provinces having the highest R&D intensity are Beijing (7.0 per cent), Shaanxi (2.83 per cent)6 and Shanghai (2.06 per cent). Together these regions have a population of approximately 70 million inhabitants and would all rank very high compared to most European countries, of which many are of similar size. The differences across regions in terms of industry base, production-factor endowment and historical circumstances create many concentrations of R&D intensity. One important historical explanation for this phenomenon is the, originally Maoist, policy of locating important institutions in central spots in Western China where they were supposed to be sheltered in case of war. Xian (Shaanxi) is such a place, with an enormous software capacity in a science park surrounded by several advanced universities. Also Chengdu (Sichuan) and Chonqing have similar characteristics. As regards Xian (Shaanxi) we notice that the high R&D intensity on the one hand is a statistical phenomenon due to the very low GPP (gross provincial product) of the Shaanxi province. On the other hand, an in-depth look at Xian reveals enormous efforts in universities (35 universities and colleges with 380 000 students) and in the over

250

Local versus global perspectives in innovation

700 labs and R&D institutions (XSP, 2005). In addition Xian – as a consequence of classical Chinese centralist policy – is focusing on software. More than 30 000 software engineers graduate annually. This software incubator has more than 500 software firms with more than 40 000 employees (2004). Although the Xian software effort may be extreme even for China, it is far from unique. It is difficult to find clusters of similar magnitude in Europe. If the cluster concept has any meaning, the potential of the Xian area – still strongly supported by the government – must be assumed to be significant also by global standards. In addition costs are lower there than in coastal areas – though that aspect should not be overestimated. ‘The disequilibrium phenomenon’ in short may be illustrated by the fact that Chinese R&D has a totally different cost structure compared with what is the case in the OECD area. The portion of labour costs in R&D in China is 25 per cent rather than the 50 per cent common in OECD countries (MOST, 2004: 49). Digging deeper we may assume that the difference consists of two components: wage-cost differences and a residual due to fringe benefits normally paid to employees in China and probably not captured as labour costs in the accounting system. The wage-cost difference in turn may be explained by ‘productivity differences’ (using the standard economist’s argument that salaries – in equilibrium – reflect marginal productivity), and a different labour structure at research institutes (which in turn may be related to the wage structure). Without going into detail, we may assume that the labour-cost and salary structure is ‘too small’ for a given amount of R&D work performed. Although here discussed on the micro-level, this is, basically the same phenomenon as what in macro-level analyses is labeled PPP difference. If we focus on R&D and assume that R&D work is carried out roughly in the same way and with similar productivity in China as in Europe, and compensate for the lower labour-costs, it is easy to estimate that Beijing’s R&D expenditures of RMB 26 000 million would be equivalent to circa RMB 35 000–40 000 million if Chinese researchers were paid according to European standards. This is a very crude calculation but still illustrates the hidden magnitude of disequilibrium: the Beijing region alone probably has a ‘real’ R&D activity already half the size of that of Sweden. Our interviews with MNE-R&D labs (such as Ericsson) in China support the deductive arguments above. They reveal that R&D in China is not only cost-saving (about 50 per cent cheaper or even more), but also time-saving (up to 30 per cent shorter time-to-market for R&D projects according to our informants) because R&D staff work more hours per week (Long and Laestadius, 2006). The velocity phenomenon may be looked upon as analogue to the disequilibrium argument. In short we argue that the rapid growth of some

251

High-tech innovation in catching-up countries

4500 4000

3912

10 Million RMB

3500 3000 2656

2500 2172

2000 1500

1893 1477

1000 500 R&D investment

0 2001

2002

2003

2004

2005

Years Figure 13.1

Total R&D investment of the top 100 Chinese ICT firms

parts of the economy may overshadow other significant growth areas. This may be illustrated with data from MII on R&D expenditures in 100 leading Chinese ICT firms. Aggregate statistics show that R&D intensity among those firms is stagnant or even declining. This is however the result of a rapid increase of sales revenue (Figure 13.2), overshadowing the rapid increase in R&D investment (Figure 13.1) (see Long and Laestadius, 2006). Also other ‘ratios’ like the ‘technology dependency ratio’ and the ‘hightech balance of trade’ show similar characteristics. As regards the former, the enormous increase in Chinese acquisition of foreign technology during recent years (from about RMB 2000 to RMB 7000 million) still represents a clear decline of total spending on technology due to the still much faster rise of domestic R&D (MOST, 2004: 144f). And the steady and continuous deficit in the Chinese high-tech trade balance – a concept the importance of which in itself is far from obvious – definitely looses its importance in light of the fact that total Chinese foreign trade in high-tech products has multiplied during the period (MOST, 2004: 148). Although many of today’s successful countries developed rapidly – though not leapfrogging in the usual meaning – into fast-growing sectors, it has been argued that the ‘radical technological change in the last decades, with ICT-based solutions substituting earlier mechanical and

252

Local versus global perspectives in innovation

90

Sales revenue

80

10 million RMB

70 60 50 40 30 20 10

05 20

03 20

01 20

99 19

97 19

95 19

93 19

91 19

89 19

19

87

0

Years Figure 13.2 Total sales revenue of the top 100 Chinese ICT firms over 19 years electromechanical ones, and the derived change in the demand for skills and infrastructure’ has made it more difficult for latecomers to catch up, leaving that road open only for those countries which have invested massively in the formation of skills and R&D infrastructure (Fagerberg and Verspagen, 2002). Superficially, this is not the case for China. Although fast growing as shown in many recent reports, the country ranks at the bottom as regards enrolment in higher education, university degrees in natural sciences and engineering as well as in R&D intensity and patenting activities. These figures may all be a statistical illusion due to the aggregation of overall data from an enormous country divided between an advanced, basically coastal economy and an immense, backward ‘hinterland’, part of it still not fully integrated into the modernization process. As several analysts have argued, China in many aspects is too large a country to look upon as one unit (see Fairbank, 1987). To sum up thus far, not only is there a rapid development in formerly peripheral countries at the ‘high end’ of science, technology and industry as illustrated by Chinese data, it may also be argued that the dynamic processes taking place in such countries are more complex than is revealed by a superficial view of aggregate statistics. In the following section we analyse the resulting challenges to conventional wisdom which follow from this.

High-tech innovation in catching-up countries

253

THE FOUNDATIONS OF CONVENTIONAL WISDOM ON THE CHARACTERISTICS OF COMPLEXITY Three-phase or stage models dominate our mental map on how industrial and technical transformation takes place and how technologies are transferred between actors and locations. Together they contribute to the conventional wisdom on what constitutes competitiveness and thus also on how policy-makers in the ‘advanced’ industrialized countries should construct their knowledge-oriented policies to stay competitive in an era of intensified and fierce global competition. The Lisbon and Barcelona documents in the EU are good illustrations of that (see EC, 2002). In this section we briefly mention the three models – the stages of growth (Rostow) model, the product cycle (Vernon) model, and the linear (Bush) model – assuming that they are already familiar. This is followed by a discussion of the neglect, perhaps resulting from our implicit subordination to those models, of the characteristics of complexity. The Stages-of-growth Model Originally formulated by Walt Rostow (1960) this model formulates a general theory on stages followed by countries in their transformation from backwardness or underdevelopment to maturity. The five stages, assumed to be followed by all countries, are: (a) traditional society; (b) preconditions for take-off; (c) take-off; (d) drive to maturity; and (e) the age of mass consumption. Rostow does not discuss technologies in depth. It may however be argued that the Rostow model reflects a ‘Fordist’ view of technology and industrial processes as well as a lack of vision as regards consequences of the coming of globalization (which ‘took off’ in the 1970s and in the early twenty-first century is in the ‘maturing’ phase!). Related to the Rostow model, but inspired by the achievements of the rapid growth of certain Asian countries since the 1960s and 1970s, there have emerged less aggregated, but also more technology-focused analyses of catch-up processes, sometimes also focused on the possibilities of leapfrogging (Soete, 1985; Perez and Soete, 1988). As analysed by Hobday (1995), in the real world it is already far from easy to identify clear demarcation lines between leap-frogging and more goose-oriented processes taking place among Asian countries so far. The Product-cycle Model The traditional division of labour between industrialized and less industrialized countries, as well as the historical processes of industrialization, may

254

Local versus global perspectives in innovation

have contributed to a neglect of the study of what constitutes complexity in industrial and technical activity and knowledge formation. This neglect may well have been enforced by the many product-cycle, technology-cycle and industry-cycle theories in use by researchers following the original paper by Vernon (1966). As is well known, originally Vernon identifies three distinct phases with different characteristics and needs for factor inputs: new products, mature products and standardized products. In short the global localization of production is assumed to change from advanced countries (such as the US) via medium-advanced (such as European) countries towards low-cost (mostly Asian) countries as products become more mature or standardized. Vernon does not explicitly include innovation processes in his product cycle-model. That is done by Abernathy and Utterback (1978) in their model which not only identifies three phases but also a shift from product innovations to process innovations. Their model, developed in Utterback (1996 [1994]) also incorporates a theory on organizational change originally developed by Burns and Stalker (1961) (from organic structure to mechanistic structure) following the product-innovation cycle. The combination of a product-cycle model with cost differences between countries or regions made the Vernon model an attractive alternative to dominating theories on international trade and location of production. However its strong focus on factor costs may – implicitly at least – have contributed to the neglect of the fact that, in a world characterized by great disequilibrium, it is far from obvious that cost differences mirror differences in capabilities for handling complex processes or the availability of potentially advanced production factors. The Linear Model The third stage-model of relevance here is the linear model. Originally developed by Vannevar Bush (1945) and criticized by many academics (see Kline and Rosenberg, 1986; for an overview see Laestadius, 2006) it is still not only popular but dominant in S&T policy rhetoric (see EC, 2002; Sheehan and Wyckoff, 2003). The linear innovation sequences follow a path starting with basic research, succeeded by applied research, product and process development, and more market-related improvements and activities. Somewhat simplified, the message of the model is that the fundamental and most difficult tasks are performed in the basic research stage, the simpler ones at the end of the chain. Basically the factor costs (of researchers, product developers and workers) are not included in the model. The linear model provides the intellectual foundation for the high value of R&D in general

High-tech innovation in catching-up countries

255

and of basic science in particular. Those are activities which for a long time have been strongly concentrated in the old industrialized countries of Europe and America – and the US in particular – which may have contributed to a belief that they will stay there. Constructing the Conventional Wisdom The linearity and stages of these three models have a strong appeal. Although not formally linked to each other, they are attractive as building blocks in the conventional wisdom on growth and the diffusion of knowledge and industries. And even if these models relate to different realms of theory, they have a simple connection: it is easy to imagine a linear process starting with basic science and succeeded by Vernon’s stages of new products (characterized by product development), mature products and standardized products and his view on production processes. For the Rostow model the related interpretation is that the latecoming countries start at the assumed low end of industrial and technological complexity and, when maturing, successively upgrade their activities into more science-based ones. Not only the assumed skills and competencies needed, but also the cost structures, which fit into Vernon’s stages, thus favour a mental picture where the latecomers’ catchup processes may follow paths similar to those followed by those already ‘on top’. There is of course a tendency towards an unhistorical time-perspective in these models – at least in some of their standard interpretations. The conditions for industrial development differ today compared with what was the case when Rostow and Vernon wrote their texts. Globalization is one aspect of that conditional change, but is only part of the problem and has in addition been discussed in many other publications. The real problem emerges if it can be argued that the dominant view on complexity is wrong – that is, if the characteristics of knowledge, its complexity and the capabilities needed to acquire it differ from the basic (partly implicit) assumptions in the Rostow, Vernon and Bush models. If so, and if the costs of (almost) all kinds of human resources – due to a strong global disequilibrium – are significantly less in one part of the world than in the rest, we end up in an indeterminate situation. The restructuring following from that will – in addition to traditional resource endowments – be conditioned by the characteristics of learning processes and the knowledge base. This is discussed in the following section.

256

Local versus global perspectives in innovation

TOWARDS AN UNDERSTANDING OF COMPLEXITY AND THE CHARACTERISTICS OF THE KNOWLEDGE BASE Assuming that those firms and countries that can manage complex processes may develop competitive advantages in niches on all technology levels makes the complexity phenomenon a core issue. In short: if colonial heritage has been thrown off and the fetters of hierarchical structures loosened, why should we assume the catching-up countries necessarily face competitive disadvantages in science-based activities which to a large extent are labour- (human-resource) intensive, rather than in (traditional) engineering, crafts and manufacturing? Although this needs detailed investigation and detailed research of its own there are lot of recent stylized facts – some of which have been presented above – indicating that latecomers may catch up, and starting at the presumably most complex end of the economy, which challenges our conventional wisdom on innovative behaviour. In trying to grasp the complexity phenomenon we have to start the analysis by clearing our minds of the conventional wisdom that science and scientific knowledge are the most important drivers of the economy. Consequently we have to open our minds to other forms of knowledge creation as well and take a deeper look into cognition processes and the fundamentals of knowledge formation. One starting point for that is the twin concept of analytical–synthetic knowledge, originally formulated in Laestadius (1998) and Laestadius (2000) and further developed by Asheim and Gertler (2005) and Asheim et al. (2007). It resembles Simon’s (1962) distinction between the natural and the artificial, and Faulkner’s (1994) dichotomy between knowledge related to ‘experimental research’ and knowledge related to ‘design practice’. In addition there is a resemblance to the distinction made by Nightingale (1998) between science as ‘pattern seeking’ and ‘technology as artificial function’.7 Analytical knowledge is that kind of knowledge which dominates in most of science, such as in nanotechnology, chemistry and biotechnology. It relates to activities directed towards understanding and explaining natural systems by discovery and application of scientific laws. Analytical knowledge is based on deductive logic and is focused on deconstruction (‘analysis’) of complex structures to increase the understanding of them. Analytical knowledge is typically to a large extent codified. Synthetic knowledge is, in short, related to combinatorial skills, we may here include various entrepreneurial skills (creative combinations) as well as polytechnic, that is, engineering activities based on integrating different knowledge bases, technologies and systems into (more) complex systems or second-order technologies (designing artificial worlds).

High-tech innovation in catching-up countries

257

A synthesizing activity is essentially about achieving a ‘fit’ between entities, as diverse knowledge is integrated into new artefacts or processes. To a large extent this boils down to concrete technical problem solving (see Moodyson et al., 2008). As such it is related to handling, designing and creating new solutions through integrating diverse components into complex systems. The knowledge base is typically of a heterogeneous rather than singular character. Engineering practices are often highly synthetic in that they rely on trials, experimentation and testing – induction or parameter variation – in line with the Kline and Rosenberg (1986) chain-linked model and not the theory-led deductive process. This makes synthetic knowledge creation an inductive process primarily (Asheim et al., 2007). Both analytical and synthetic activities may include experiments. Successful design for example is often experimentally based (see Thomke, 2003).8 Synthetic activities are easily related to a systems approach, especially if we see complex knowledge-formation processes as ‘large number of parts that interact in a non simple way’ (Simon, 1962). Paoli (2003: 157) makes the point that ‘bringing together a partial solution to another partial solution (maybe in a different field) is therefore an act of construction of sense – which involves certain aspects that are completely invented – and not an act of mere recomposition’. This synthetic or systemic approach implies that technological knowledge is constructed by bits and pieces of knowledge building upon other, already acquired bits of knowledge, acquired both in the same specific context and in other adjacent contexts. Thus, no single actor carries the complete knowledge. Instead, technological knowledge is endogenous to the system in which each actor is rooted (Antonelli, 2005). In complex products, or multi-technological products, where ‘artefacts [are] made up of components and embody a number of technologies’ (Brusoni et al., 2001: 597), this becomes highly evident. We can probably say that the systemic character of a firm’s knowledge base is growing – a general trend across all industry sectors – as the inclination to streamline and specialize the activities seems to cut across all sectors resulting in a knowledge base exceeding the boundaries of the specific firm (see Brusoni et al., 2001). And with an increased knowledge specialization of firms, concurrent with an increasing amount of knowledge (often from dispersed knowledge fields and multiple knowledge sources) incorporated into new products, systems integration is becoming a strategically vital ability of firms today (see Hobday et al., 2003; Hobday et al., 2005). It should be noted that the dichotomy analytical–synthetic is epistemological in character and not an a priori description of differences between various industrial practices. The actual balance between analytical and synthetic knowledge in different industries (and academic disciplines) is consequently an empirical question open to study. We may thus assume that

258

Local versus global perspectives in innovation

academic work (and natural science in particular) typically has a significant dominance of analytical knowledge although we should not by definition exclude synthetic parts. Similarly we may assume that creating the artificial world – in manufacturing as well as in services – involves significant elements of synthetic knowledge although we can easily imagine that analytic elements may frequently be involved. Software design as well as innovations in biotechnology may serve as illustrations. In fact we may assume that activities – and thus industries – can, in principle at least, be ordered (classified) on the basis of their relative balance of analytical and synthetic knowledge respectively. Leaving aside the epistemological aspects of the distinction between tacit and articulated knowledge and its relation to the distinction between analytical and synthetic knowledge, we still have reasons to assume that analytical activities can be codified to a greater extent than non-analytical ones and in fact become so. In short we may thus assume that the balance between tacit and codified knowledge on average differs between industries and activities dominated by analytical and synthetic activities respectively, just as much as we may expect it to differ within industries, that is, between firms. This balance may change over time due to technical change, primarily within IC technologies. It has been argued that this is in fact the case: that there is a linear process of ICT-development causing and permitting more codification which in its turn will increase global diffusion of knowledge. According to Lundvall (1992), this ‘trend will be most important in science-based areas where the communication is easier to formalize and codify’ (p. 4). The ease of global diffusion and transfer of knowledge is a core issue here: what capabilities firms, plants and institutions in catching-up countries could most easily acquire. And how should international firms organize their global R&D networks (see Birkinshaw, 2002; Criscuolo and Narula, 2005; Gerybadze and Reger, 1999)? The implication of Lundvall’s statement is that analytical knowledge, due to the ease of codification, may be easier to transfer than synthetic. From the Analytical–Synthetic Dimension to Complexity, Capabilities and the Global Transfer of Knowledge The conventional wisdom, more or less explicit in the linear model literature (cf. Bush, 1945), product-life-cycle literature (see Vernon, 1966) and technology-policy literature (see EC, 2002; OECD, 1971; OECD, 1981) is that activities which in this chapter are assumed (and argued) to have a highly analytical knowledge content – that is ‘science’ – also are more complex than those being primarily synthetic. As to complexity, the definition given in Simon (1962) can serve as a starting point. There he classifies a piece of knowledge as complex if it

High-tech innovation in catching-up countries

259

comprises many elements that interact richly. Complexity can thus be defined in terms of the level of interdependence inherent in the subcomponents of a piece of knowledge (Simon, 1962; Sorensen et al., 2006). ‘A high degree of interdependence indicates that many ingredients influence the effectiveness of others so that a change in one may dramatically reduce the usefulness of the recipe . . . Low interdependence implies small crosscomponent effects and a corresponding opportunity to adapt and change ingredients independently’ (Sorensen et al., 2006: 998). What may be added is that complexity increases if the elements and interactions are widely different and also if the set of potentially useful combinations from a wide set is restricted (as in the search for drugs in the pharmaceutical industry). In order for an activity to create a competitive advantage (on national, regional or firm level), it must involve a certain measure of complexity; an activity that is easy to replicate or imitate does not provide a sustainable advantage against competitors. This complexity can be a feature of complex knowledge bases, where difficult-to-acquire knowledge or a systemic or integrative nature of the knowledge base makes it difficult to imitate. The complexity may be inherent in complex production processes where knowledge is built into advanced, highly integrated and capitalintensive production facilities (as in paper production). Another measure of complexity can be the market: some markets may be especially difficult to penetrate for new actors due to trade barriers, national regulations and the like. Products or processes protected by patents also provide an advantage; although these need not be complex they are difficult to imitate due to legislation and restricted access to knowledge and information. The complexity of synthetic knowledge formation can be found, for instance, in the translation of a problem into a solution, or in upscaling from prototype production to large-scale production (a significant problem in process industries). In any case, the complexity of competitive processes is difficult to grasp and therefore difficult to map. The general conclusion drawn from conventional wisdom is that advanced countries have comparative advantages in focusing their efforts on the analytical sector as these are assumed not to diffuse easily outside the advanced communities in control of them. That is the raison d´être, among others, of the ‘Lisbon/Barcelona process’ – creating a knowledgebased society with a strong focus on science and science-based activities. The position taken here, however, is primarily that the complexity of a process or an activity is an empirical question. Second, our discussion of the differences between analytical and synthetic knowledge gives us reason to assume that it is far from evident that analytical activities are always, or in general, more complex. Our approach does not a priori assume what mode is most advanced or difficult to learn, acquire or master. In fact,

260

Local versus global perspectives in innovation

instead of adopting a ‘linear model of complexity’ with highbrow analytical (scientific) knowledge at the high end, followed by the others in some order, we accept that these are different modes and that there is a variety of different mechanisms that influence cognitive and institutional processes. We can imagine complexity in both worlds of knowledge. To become industrially relevant, however, knowledge-formation processes have to be transformed into capabilities of firms – a concept which we use on the firm level. The creation of capabilities is in itself based on synthetic combinations of disciplines, of knowledge from different knowledge bases, of people with different knowledge profiles, of systems, of visions and the like, as well as the orchestration of these people and these processes so that the firm can maintain its long-term creativity and competitiveness. Adopting the capability concept obviously moves in the direction of the modern resource based theory of the firm and the ‘dynamic capability approach’ (see Dosi, 2000; Foss and Robertson, 2000; Kogut and Zander, 1992; Teece and Pisano, 1994; Zollo and Winter, 2002). Basically we have nothing to add to the dynamic capabilities discourse in this context: it falls generally in line with our discussion here. And the different capabilities among firms for developing and performing analytical as well as synthetic activities explain why firms differ (see Nelson, 1991). Returning to our original complexity concept, we conclude that capabilities are founded on several layers of knowledge-formation processes having different mixes of synthetic and analytical competence. High synthesizing ability is always needed in industrial activity however, as analytical (and academic) excellence is not enough to create capabilities and market success. Some of these capabilities developed by certain firms show a higher probability of staying profitable or avoiding challenging catchingup processes in the long term. Although we should not forget the importance of differences in factor costs (not the focus in this chapter), the capability of handling complexity (and the most complex processes) explains a lot. And, as we noticed above: it is far from obvious that those processes dominated by analytical knowledge formation are the most complex to handle. Even where that is the case, it is not self-evident that those processes normally are best handled by firms or institutions in the old industrialized countries.

CONCLUDING DISCUSSION Some people in incumbent (OECD or EU) countries may be worried about the recent and rapid catching-up processes of the Asian tiger economies.

High-tech innovation in catching-up countries

261

That is not the topic of this chapter. What is addressed here is basically that these catching-up processes seem to start at the ‘wrong’ end according to conventional wisdom. And although policy-makers and researchers rapidly learn about the stylized facts of these processes, and industrialists adapt to them in relocating trade and production, there is still a lack of fundamental understanding of the mechanisms behind these processes.9 Our chapter, based both on the stylized facts, primarily of Chinese development, and on an epistemological analysis of knowledge formation, intends to contribute to closing this gap in understanding. Instead of sticking to the conventional wisdom of the ‘three linearities’ (discussed in the third section above) which have been dominant in policy documents well into this millennium, we provide, in the previous conceptual discussion, a conjectural interpretation of the data as reported in the second section. Accordingly, the most important aspects of our interpretation are thus: 1.

2.

3.

There is a rapid growth of all traditional S&T-related indicators in catch-up countries, not least in China: R&D expenditure, publications, citations, patents and so on. Much of this is related to analytical knowledge formation. Either this kind of knowledge formation is not as complex as we are used to classifying it, or we have severely underestimated the cultural institutions responsible for its development, as in China. Under all circumstances our conclusion is that European policies on competitiveness should not realistically be based mainly on general R&D targets. The evidence that Europe has a sustainable competitive advantage based on its (actual or potential) R&D intensity is weak. The production-factor concept may reasonably allow us now to say that Chinese factor costs (although rapidly rising in some sectors and regions) are more or less favourable on all levels comparable to Europe and the US. It is thus to a large extent an open question as to in what areas China and some other recent ‘tigers’ (such as India) may successfully develop competitiveness in the present global economy, characterized by extremely large disequilibrium in factor costs and the distribution of knowledge. R&D activities – which in fact are more labour (human-resource) intensive than many manufacturing processes – are far from sheltered in Europe, even if this is an area where excellence may well outweigh factor costs. Recent industrial developments have not made very clear what technological areas will dominate in further international expansion. Although there is a ‘close-to-science’ profile in much of our collected data, there are also, as shown above, clear indications that the bulk of Chinese development is still not at the very high end of high-tech

262

4.

Local versus global perspectives in innovation

sectors. There is literature arguing that the hierarchical structure of Chinese firms reduces their creativity and thus their capability to develop high-end competitiveness (see for example Gilboy, 2004). As a general phenomenon in our Chinese interviews we could not confirm this; rather, we found variety, including some ‘Silicon-Valley-style’ firms – but that is a topic for further research. The recent, extremely competitive and profitable position of the Swedish medium-tech manufacturing industry may illustrate the present situation. Although these firms are definitely not known for having small R&D budgets, they do not belong to the world of R&Dintensive firms. They belong, we argue, to the classical high-quality engineering firms which integrate their R&D-efforts (which in itself may include a lot of ‘synthesizing’ product development) with strong synthesizing capabilities. They produce complex products and complex systems. That kind of synthesizing capability seems also resistant, within reasonable limits, to significant factor-cost disequilibrium. This seems also to be the case for many German firms who benefit from producing quality equipment for Chinese industrialization.

While we wait for more research on these topics, it is still possible to formulate a conjecture, a preliminary conclusion: sustainable competitiveness should be based on capabilities which go beyond the analytical knowledge base. Sustainable capabilities are founded on mixes of analytical and synthetic knowledge and we should not assume that they have to be primarily analytically based to be complex, or that analytically based knowledge processes favour old industrialized countries. The recent boom in Swedish engineering industries – and parts of the German as well – illustrates that the road to the competitive, knowledge-based economy may well take its starting point in the medium-tech segments of industry rather than in the so called high-tech sectors of the OECD classification. The mirror image of that is a continued catching-up of the Asian tigers at the high-end of the high-tech segment.

NOTES 1. This chapter, originally presented in Dortmund on 16 April 2007, is based on the authors’ joint and ongoing research and may contain passages/sections identical/similar to other texts produced by them. 2. Figures based on WTO statistics 2006 (www.wto.org). 3. There is a strong tendency to either underestimate or ignore the importance of statistics in China (mainly by Ming Ying-like firms who operate freely in a market economy), or to over-address and exaggerate them (by firms with strong state connections). 4. Ministry of Science and Technology

High-tech innovation in catching-up countries

263

5. Ministry of Information Industry 6. Xian, the capital of Shaanxi, ranks second in China in terms numbers of staff engaged in high-tech industries. 7. The analytical–synthetic dichotomy also fits well into ‘mode1/mode2’ concepts introduced more than a decade ago by Gibbons et al. (1994). Their ‘mode2’ resembles our ‘synthetic knowledge’, and their ‘mode1’ reminds us of the ‘analytical knowledge’ used here. An important difference between Gibbons et al. and the position taken here is that we do not argue that mode2 is a new form of knowledge. It has been there all the time. 8. Faulkner (1994) seems not to consider that aspect. She writes about ‘experimental research’ but not ‘experimental design’. 9. For recent academic analyses on the neglect of ‘low-tech’ industries, see Hirsch-Kreinsen et al., 2003 and Hirsch-Kreinsen et al., 2005.

REFERENCES Abernathy, W. and J. Utterback (1978), ‘Patterns of industrial innovation’, Technology Review, 80 (7), 40–7. Antonelli, C. (2005), ‘Models of knowledge and systems of governance’, Journal of Institutional Economics, 1, 51–73. Amin, Ash and Patrick Cohendet (2004), Architectures of Knowledge: Firms, Capabilities and Communities, Oxford: Oxford University Press. Asheim, B., L. Coenen, J. Moodysson and J. Vang (2007), ‘Constructing knowledge-based regional advantage – implications for regional innovation policy’, International Journal of Entrepreneurship and Innovation Management, 7 (2–5), 140–55. Asheim, Björn and Meric S. Gertler (2005), ‘The geography of innovation: regional innovation systems’, in Jan Fagerberg, David Mowery and Richard R. Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press. Birkinshaw, J. (2002), ‘Managing internal R&D networks in global firms – what sort of knowledge is involved?’ Long Range Planning, 35, 245–67. Brusoni, S., A. Prencipe and K. Pavitt (2001), ‘Knowledge specialization, organizational coupling, and the boundaries of the firm: why do firms know more than they make?’ Administrative Science Quarterly, 46, 597–621. Burns, Tom and G.M. Stalker (1961), The Management of Innovation, London: Tavistock. Bush, Vannevar (1945), Science, the Endless Frontier, Washington DC: Goverment Printing Office. Choung, J.-Y., H.G. Min and M.C. Park (2003), ‘Patterns of knowledge production: the case of information and telecommunication sector in Korea’, Scientometrics, 58 (1), 115–28. Criscuolo, P. and R. Narula (2005), ‘Using multi-hub structures for international R&D: Organizational inertia and the challenges of implementation’, DRUID working paper 05-13. Dosi, Giovanni (ed.) (2000), The Nature and Dynamics of Organisational Capabilities, Oxford: Oxford University Press. EC (2002), ‘More research to Europe – towards 3% of GDP’, Commission of the European Community, COM (202), 11 September 2002. Fagerberg, J. and B. Verspagen (2002), ‘Technology gaps, innovation-diffusion and transformation: an evolutionary interpretation’, Research Policy, 31, 1291–304.

264

Local versus global perspectives in innovation

Fairbank, John K. (1987), The Great Chinese Revolution, 1800–1986, London: Chatto & Windus. Faulkner, W. (1994), ‘Conceptualizing knowledge used in innovation: a second look at the science-technology distinction and industrial innovation’, Science, Technology and Human Values, 19, 425–58. Foss, Nicolai J. and Paul L. Robertson (eds) (2000), Resources, Technology and Strategy, London: Routledge. Gerybadze, A. and G. Reger (1999), ‘Globalisation of R&D: recent changes in management of innovation in transnational corporations’, Research Policy, 28, 251–74. Gibbons, Michael, Camille Limoges, Helga Nowotny, Simon Schwartzmann, Peter Scott and Martin Trow (1994), The New Production of Knowledge. The Dynamics of Science and Research in Contemporary Societies, London: Sage. Gilboy, G.J. (2004), ‘The myth behind China’s miracle’, Foreign Affairs, 83 (4), July/August, 33–48. Hirsch-Kreinsen, H., D. Jacobson, S. Laestadius and K. Smith (2003), ‘Low-tech industries and the knowledge economy: state of the art and research challenges’, Royal Institute of Technology, Stockholm, working paper. Hirsch-Kreinsen, Hartmut, David Jacobson ansd Staffan Laestadius (eds) (2005), Low-tech Innovation in the Knowledge Economy, Frankfurt am Main: Peter Lang. Hobday, Michael (1995), Innovation in East Asia – The Challenge to Japan, Aldershot: Edward Elgar. Hobday, Michael, Andrea Prencipe and Andrew Davies (2003), ‘Introduction’, in Andrea Prencipe, Andrew Davies and Michael Hobday (eds), The Business of Systems Integration, Oxford: Oxford University Press, pp. 1–12. Hobday, M., A. Davies and A. Prencipe (2005), ‘Systems integration: a core capability of the modern corporation’, Industrial and Corporate Change, 14, 109–43. ITU (2006), www.itu.org. Kaldor, Nicholas (1985), Economics Without Equilibrium, Armonk, NY: M.E. Sharpe. King, D.A. (2004), ‘The scientific impact of nations’, Nature, 430, 311–16. Kline, Stephen J. and Nathan Rosenberg (1986), ‘An overview of innovation’, in Ralph Landau and Nathan Rosenberg (eds), The Positive Sum Strategy – Harnessing Technology for Economic Growth, Washington DC: National Academy Press, pp. 275–305. Kogut, B. and U. Zander (1992), ‘Knowledge of the firm, combinative capabilities, and the replication of technology’, Organization Science, 3, 383–97. Laestadius, Staffan (1998), ‘Technology level, knowledge formation and industrial competence’, in Christopher Green and Gunnar Eliasson (eds), Microfoundations of Economic Growth – a Schumpeterian Perspective, Ann Arbor, MI: University of Michigan Press, pp. 212–26. Laestadius, Staffan (2000), ‘Biotechnology and the potential for a radical shift of technology in forest industry’, Technology Analysis and Strategic Management, 12, 193–212. Laestadius, Staffan (2006), ‘Beyond the high-tech/low-tech divide – towards a new taxonomy and new indicators to guide the transformation to a knowledge society’, final report from the PILOT project WP1: theoretical and conceptual perspectives, Royal Institute of Technology, Stockholm, research report. Leydesdorff, L. and P. Zhou (2005), ‘Are the contributions of China and Korea upsetting the world system of science?’, Scientometrics, 63 (3), 617–30.

High-tech innovation in catching-up countries

265

Long, Vicky and Staffan Laestadius (2006), ‘Globalization of Knowledge Formation – the ICT R&D Relocation’, Royal Institute of Technology, Stockholm, research report TRITA-IEO R 2006:15. Long, Vicky and Christopher Palmberg (2006), ‘Navigating IPR ticket from a latecomer’s perspective – the case of the emerging Chinese ICT industry’, in Jyrki Ali-Yrkkö and Christopher Palmberg (eds) Finland and the Globalization of Innovation, Helsinki: Taloustieto Oy. Lundvall Bengt-Åke (1992), National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, London: Pinter Publishers. Moodyson, J., L. Coenen and B. Asheim (2008), ‘Explaining spatial patterns of innovation: analytical and synthetic modes of knowledge creation in the medicon valley life science cluster’, Environment and Planning A, 5 (40), 1040–56. MOST (2004), China Science and Technology Indicators, 2004, Beijing: Ministry of Science and Technology of the People´s Republic of China. Nelson, Richard R. and Sidney G. Winter (1982) An Evolutionary Theory of Economic Change, Cambridge MA: Harvard University Press. Nelson, R.R. (1991), ‘Why do firms differ and how does it matter?’ Strategic Management Journal, 12, 61–74. Nightingale, P. (1998), ‘A cognitive model of innovation’, Research Policy, 27 (7), 689–709. Organisation for Economic Co-operation and Development (OECD) (1971), Science, Growth and Society, Paris: OECD. OECD (1981), Science and Technology Policy for the 1980s, Paris: OECD. Paoli, Massimo (2003), ‘The cognitive basis of systems integration: redundancy of context-generating knowledge’, in Andrea Prencipe, Andrew Davies and Michael Hobday (eds), The Business of Systems Integration, Oxford: Oxford University Press, pp. 152–73. Perez, Carlotta and Luc Soete (1988), ‘Catching up in technology: entry barriers and windows of opportunity’, in Giovanni Dosi, Chris Freeman, Richard R. Nelson, Gerald Silverberg and Luc Soete (2004), Technical Change and Economic Theory, London and New York: Pinter Publications. Prencipe, Andrea, Andrew Davies and Michael Hobday (eds) (2003), The Business of Systems Integration, Oxford: Oxford University Press. Robertson, Paul L. and Pari Patel (2007), ‘New wine in old bottles: Technological diffusion in developed countries’, Research Policy, 36, 708–21. Rostow, Walt (1960), The Stages of Economic Growth, Cambridge: Cambridge University Press. Sheehan, Jerry and Andrew Wyckoff (2003), Targeting R&D: Economic and Policy Implications of Increasing R&D Spending, OECD STI WP 2003/8, Paris: OECD. Simon, H. (1962) ‘The architecture of complexity’, Proceedings of the American Philosophical Society, 106, 467–82. Simon, Herbert (1969/1996), The Sciences of the Artificial, Cambridge MA: MIT Press. Schumpeter, Joseph A. (1969 [1934]), The Theory of Economic Development, Cambridge, MA: Harvard University Press. Soete, Luc (1985), ‘International diffusion of technology, industrial development and technological leapfrogging’, World Development, 13 (3), 409–22. Sorensen, O., J. Rivkin and L. Fleming (2006), ‘Complexity, networks and knowledge flow’, Research Policy, 35, 994–1017.

266

Local versus global perspectives in innovation

Teece, D., Pisano, G. (1994), ‘The dynamic capabilities of firms: an introduction’, Industrial and Corporate Change, 3, 537–56. Thomke, Stefan H. (2003), Experimentation Matters – Unlocking the Potential of New Technologies for Innovation, Boston, MA: Harvard Business School Press. Utterback, James (1996 [1994]), Mastering the Dynamics of Innovation, Boston, MA: Harvard Business School Press. Vernon, R. (1966), ‘International investment and international trade in the product cycle’, Quarterly Journal of Economics, 80, 190–207. WTO (2006), www.wto.org/english/res_e/statis_e/. XSP (2005), Xian Science Park official presentation document. Yusuf, Shahid, M. Anjum Altaf and Kaoru Nabeshima (eds) (2004), Global Production, Networking and Technological Change in East Asia, Washington, DC: The World Bank and Oxford University Press. Zollo, M. and S. Winter (2002), ‘Deliberate learning and the evolution of dynamic capabilities’, Organization Science, 13 (3), 339–51.

14.

Worshipping at the shrine of the knowledge-based society? James Wickham

INTRODUCTION For European policy-makers, it now seems to be axiomatic that we live in a ‘knowledge-based society’ (KBS), or if we do not, we are about to do so. The term functions as a description of contemporary reality, as a prediction of where we are about to go, and even as a prescription of where we should go. Furthermore, the term implies that, although the direction is apparently inevitable, failure to adapt now to this oncoming wave of the future would be reprehensible. As such the term shares many features with other over-arching accounts of social reality (information society, globalization, and the like). This chapter does not attempt a full-scale deconstruction of the concept of the KBS. Instead it begins by examining key texts by social scientists which each found a readership and influence beyond the academy, and each of which posits a particular relationship between ‘knowledge’ and social structure. A key political use of the term today is to argue for enhanced expenditure on research and development, and interwoven with this, for the ‘reform’ of European higher education along explicitly American lines. Accordingly the second part of the chapter points to some problems of ‘excellence’ in US education which European attempts to emulate that system ignore. The third part of the chapter explores another crucial part of the KBS thesis, the notion of the ‘learning organization’; it suggests that this is at variance with some developments that appear to be occurring in vocational education and which paradoxically amount to an undermining of the achievements of some European vocational educational systems. All these issues accept the basis of the KBS, namely the equation of ‘knowledge’ with knowledge that is used at work, but as the conclusion makes clear, this is an extremely narrow approach.

267

268

Local versus global perspectives in innovation

THE SOCIAL STRUCTURE OF THE KNOWLEDGEBASED SOCIETY A common theme of social scientific futurology from the 1970s on has been the theme of ‘knowledge’ as power. Those who occupy the socially prestigious and powerful occupations in the society do so because of what they know. In the literature that has developed since the 1970s we find an increasingly differentiated conception of the nature of this knowledge, which in part functions to make the link between knowledge and occupation more tenuous. At the same time however knowledge remains equated with formal education, and there remains the assumption that economic development depends upon the expansion of third-level education. There are three important authors here: Daniel Bell in the early 1970s, Robert Reich in the 1990s, and Richard Florida today. In his seminal The Coming of Post-Industrial Society Bell claimed that a key feature of the move towards the ‘post-industrial society’ was the growing importance of formal and abstract scientific knowledge. Indeed, he argued that ‘theoretical knowledge’ was one ‘axial principle’ of such societies (Bell, 1973). Economic growth now depends on the continuing and institutionalized development of such knowledge. Scientific knowledge is now decisive for economic growth, and the ‘post-industrial society’ institutionalizes such knowledge production to a hitherto unprecedented extent. Bell shares with later scholars an uncertainty about the relationship between ‘knowledge’ and occupations. Certainly, he is at pains to point out that some counts of ‘knowledge workers’ or ‘information workers’ often include many whose work bears little relationship to scientific knowledge; he acknowledges that the expansion of US scientific research in the midtwentieth century owed much to the specific demands of the military and the arms race. Yet despite such caveats, he moves from claiming that ‘knowledge’ is crucial to the society, to the claim that those who hold such knowledge fill the dominant occupations. As many commentators have pointed out (such as Webster, 2002), even if the first statement is true, the second hardly follows: just because something is essential, it does not follow that those who ‘own’ it comprise a distinct and dominant social group. Interestingly, Bell’s argument is actually subversive of nearly all contemporary theorizing about the knowledge society. Today, many writings on ‘globalization’ take it as axiomatic that the expansion of innovation in general and of information technology in particular is interwoven with the expansion of the market (see Friedman, 2005). For Bell by contrast, the growth of scientific knowledge involves the ‘subordination of the

Worshipping at the shrine of the knowledge-based society?

269

corporation’ and what he terms the decline of the ‘economizing mode’. In a phrase that would have seemed absurd in the decades immediately after he was writing, he remarked: It seems clear to me that, today, we in America are moving away from a society based on a private-enterprise market system toward one in which the most important economic decisions will be made at the political level, in terms of consciously defined ‘goals’ and ‘priorities’ (Bell, 1973: 297).

Like Bell’s work, Reich’s The Work of Nations relates social structure to a putative new role for knowledge. However, ‘knowledge’ in this account means something rather different to the relatively straightforward scientific knowledge that lies at the heart of Bell’s argument. Writing nearly 20 years after Bell, Reich (1993) is clear that the future belongs to the market: the subtitle of his book is Preparing Ourselves for 21st Century Capitalism. Yet in this capitalism the modern corporation has become a ‘facade’. Its work activity is carried out by webs of highly skilled individuals whose connections cross-cut the formal organization chart. These networks are global and are the organization’s main asset. Furthermore, the corporation is global rather than national, and this globalization is very different to the ‘trans-national’ or ‘multi-national’ corporation of 20 years ago. The multinational corporation was a national (usually US) company that operated in different countries: it was headquartered in one country, its shares were held by individuals and other institutions from that country, it carried out its key functions there. By contrast, argues Reich, the global corporation has no real national identity, for its ownership is dispersed across national boundaries. One of his most telling examples is how US government policy, attempting to renovate US industrial capacity in the defence industry, ran into the problem of trying to differentiate between ‘US’ and ‘foreign’ firms. For Reich this economic context produces a new social structure comprising three essential groups. The ‘routine production workers’ (about 25 per cent of all US jobs according to him) have standardized tasks such as assembly work or data entry. Since this can now be carried out anywhere in the globe, routine production workers are threatened both by automation which removes their jobs completely and by outsourcing which transports the jobs to cheaper labour markets. The ‘in-person servers’ have service jobs which are tied to the specific physical location – restaurant waiters or janitors have to do their work on site, since they have to interact directly with those whom they are serving. Globalization threatens these jobs not by relocation but by immigration which brings cheaper labour to the physical location. Finally, the ‘symbolic analysts’ (less than of 25 per cent all US jobs) are the beneficiaries of the new society. These new professionals can

270

Local versus global perspectives in innovation

work anywhere in the world because their networks are virtual rather than spatial: plugging into their network requires a modem, not an office. Location matters for them in terms of their physical environment and the quality of local services. Consequently in the new global society ‘as the rest of the nation grows more economically dependent than ever on the fortunate fifth, the fortunate fifth is becoming less and less dependent on them’ (Reich, 1993: 250). Unlike some techno-enthusiasts Reich does not fall into the trap of claiming that merely because people are working with information technology they are ‘knowledge workers’. As has become even clearer in the decade since the book was written, much service-sector work is as routine – and as mobile – as traditional manufacturing work. Studies of call centre workers for example have shown how formal educational standards are unimportant. When selecting recruits management looks for the ability to tolerate boring but stressful work and a personality that enables agents to ‘smile down the phone’ – what is not needed is knowledge. In the call centre information technology has simplified work, but it has also virtualized it. The combination of very low telephone and communication costs with data storage and data processing technologies enables the call centre to be located literally anywhere in the world, thus making employment all the more vulnerable to locational competition. Whereas during the 1990s call centres were a crucial growth area of employment in the UK, by 2000 jobs were already falling as centres were outsourced, in particular to India (Wickham and Collins, 2004). As Reich also makes clear, personal service work cannot be moved in this way, but this hardly means that it is high-skill or high-paid work. Even when jobs in personal services do involve ICT, they are not likely to be upskilled. Studies of customer-facing work in shops and banks for example show that such work involves less product knowledge. Thus workers in retail shops know less about what they are selling and have less responsibility for ordering and stock-keeping, partly because these tasks are now integrated into the computer-based stock system (Webster, 2004). In such ‘service sector Taylorism’ (Bosch and Lehndorff, 2004) the novelty, already visible in the call centre, is the commodification of personality: people have to do simplified and tightly controlled work, but look as if they are committed to the service. The work of ‘symbolic analysts’ is apparently very different. In fact Reich is actually rather unclear about both what ‘symbolic analysts’ do, and who they are. Certainly he is correct to point out that new features of contemporary managerial and professional work ensure that in the upper echelons work has become more individualistic and that the age of the ‘organization man’ is over. Today enterprises may try to create ‘corporate culture’, but

Worshipping at the shrine of the knowledge-based society?

271

their focus is too short to give corporate culture much traction for their own managerial employees. Just as firms stress corporate culture more, managers and professionals are becoming less integrated into their employment. Recent studies of managers’ self-understandings show that contemporary managers, at least in the Anglo-Saxon world, now understand themselves as making a career within a managerial labour market rather than climbing the hierarchy of a single enterprise (Martin, 2005; Martin and Wacjman, 2004); a similar market-based individualism has been documented as emerging in the professions (Hanlon, 2000). Equally, the actual work of much management has changed from administration to project work, with the consequence that collaborators are frequently changing (Grabher, 2004). These changes in career and work are interwoven with the partial move towards the ‘network firm’ in which market relationships with shifting groups of other enterprises replace some of the bureaucratic organization of the large corporation (see Marchington et al., 2005). Just like Bell’s white-coated technocrats, Reich’s freewheeling symbolic analysts do have some basis in reality. However just like Bell, Reich takes some key features of some occupations and essentially generalizes them to all professional and managerial occupations. The shift in the understanding of ‘knowledge’ away from scientific and technical knowledge continues further in Richard Florida’s very influential work The Rise of the Creative Class. Florida (2004) argues that ‘creativity’ is no longer just about artistic work, but is the common core of management, research and technological innovation. He notes a parallel between the expansion of spending on R&D and the expansion of the creative industries, while the growth in the numbers of ‘scientists and engineers’ is paralleled by a growth in the numbers of ‘bohemians’. Furthermore, he notes the rapprochement of business and bohemia in contemporary innovation: The great cultural legacy of the sixties, as it turned out, was not Woodstock after all, but something that had evolved at the other end of the continent. It was Silicon Valley. This place in the very heart of the San Francisco Bay area became the proving ground for the new ethos of creativity. (Florida, 2004: 202).

As we have seen, Reich’s symbolic analysts are marked by the social skills of networking, and in particular arbitraging between networks. Their skills are entrepreneurial rather than technological. Similarly, Florida’s ‘creatives’ are characterized by a fundamental social orientation (individualistic and meritocratic, but above all valuing diversity and novelty). For both authors therefore the work and indeed the effectiveness of their new dominant social groups are not defined by formal knowledge per se. Nonetheless for Florida, just as for Reich, these characteristics are more or

272

Local versus global perspectives in innovation

less directly linked to existing occupational categories. He claims that about 30 per cent of all American jobs can be defined as ‘creative’, with a ‘super creative core’ of about 12 per cent of those in ‘Science and engineering, computers and mathematics, education, and the arts, design and entertainment, [i.e.] people who work in directly creative activity’ (Florida, 2004: 74). Since access to these occupations involves educational qualifications, then the number of symbolic analysts or the extent of the creative class can be measured simply by examining changes in the occupational structure. It is on this basis that in his most recent book Florida argues that Ireland now has proportionally the largest creative-class employment, having achieved the highest annual growth rate in the group between 1995 and 2002 (Florida, 2005: 138). However, such arguments have to be treated with considerable scepticism. Reich himself was clear that his estimate of the number of ‘symbolic analysts’ was very approximate: it used aggregate occupational census categories and excluded the public sector. Recent and more detailed analysis of the same US census figures by Brown et al. (2004) overcomes both these problems; they estimate that ‘no more than 20 per cent’ of US jobs can be classified as ‘symbolic analysts’. Using similar definitions, they estimate the proportion of knowledge workers (their term for symbolic analysts) amongst jobs in the UK as roughly similar. Those people whose work really can be described by terms such as ‘symbolic analysts’ or ‘creative class’ probably comprise only a subset of all those employed in occupational categories such as ‘managers’ or ‘professionals’. Furthermore, the expansion of higher education has led to a situation where occupations which not even Florida would describe as ‘creative’ are now becoming filled with graduates. Given that the UK government aims to have 50 per cent of young people in higher education, this suggests that many graduates are over-educated for the jobs they fill. There is in fact some evidence that this is occurring. A study of graduates in the British service sector reports that many young graduates are employed in ‘non-graduate’ jobs. There is also evidence of recent qualification inflation paralleling the expansion of third-level education: in 1988 22 per cent of all graduates working in the UK retail sector were employed in ‘Other Occupations (mainly sales and clerical)’ but by 1998 fully 37 per cent of graduates in the sector were in these occupations (Mason, 2002). Other scholars have put forward a rather more optimistic interpretation of the expansion of higher education in the US and especially the UK. In the past it has been argued that the British economy was locked into a ‘low skill equilibrium’ (Finegold and Soskice, 1988) in which low standards of skill ensured that industry focused on low-quality products – and demanded only low levels of training – thus creating a self-perpetuating circle. By contrast

Worshipping at the shrine of the knowledge-based society?

273

today it is claimed that the UK, like the US, has developed an education system which is good at producing both specialist professional and managerial skills and an adaptable graduate workforce, while secondary and vocational education for the rest of the population remain of a very low standard (Brown et al., 2001). Combined with the shift of employment to the service sector – greater than in any other European economy – and a deregulated labour market, this has ensured high graduate employment. Importantly however, students seek out business and management courses while shunning subjects such as computer science which one might imagine would be more central to a ‘knowledge-based society’ as conventionally defined. In other European countries studies have produced estimates of 20 per cent and more young people considering themselves over-educated for their current jobs (Lindley, 2002: 125). In countries such as France, Germany and Italy – the European states that Anthony Giddens has recently termed ‘blocked societies’ (Giddens, 2007) – the labour-market situation of graduates is even worse. Here rigid labour-market regulation privileges insiders (largely middle-aged males) over outsiders (that is, young people). At the same time the qualifications produced by third-level institutions are not valued by employers. Consequently many graduates find themselves moving from one short-term temporary job to another, but with little sign that this is developing into anything resembling a traditional career. That this is not just a problem of weak demand is shown by the case of Spain, where this phenomenon is widespread despite an economic boom. The limiting case is Italy, where the extent to which young people stay on in education varies inversely with the level of employment. In affluent EmiliaRomagna with full employment and a flexible labour market based on small enterprises, young people are less likely to remain in education than in Calabria and Sicily. Confronted by graduate unemployment – or at least rigid entry barriers to ‘real’ jobs – many young Europeans now head to cities such as London and Dublin. Initially they staff the bars and restaurants, but also the foreign language call centres – the rapidly expanding employment in cities which are now importing young people from the entire European labour market (Gordon et al., 2007). Such young people are partly attracted by a more ‘creative’ life style, but also crucially by the belief that the more flexible labour markets of these ‘escalator cities’ will enable them to move upwards into the jobs for which they have been educated (Favell, 2006). Yet the paradox is that access to such jobs does not require the formal knowledge beloved of theorists of the KBS. Arguably it does require the social skills and flexibilities that, almost by mistake, are generated by some higher education systems. In order to understand what is involved, it is necessary to unpack higher education change in Europe and the US.

274

Local versus global perspectives in innovation

EUROPEAN EDUCATION REFORM AND THE AMERICAN FANTASY If knowledge is central to the KBS, and if knowledge is what is ‘produced’ in formal education, above all in universities, then it is obvious that we need more knowledge, and that we need more and, crucially, better universities. Today in Europe university reform has become part of the post-Lisbon ‘competitiveness’ agenda. One element of this has been the ‘Bologna process’ by which European universities are aligning their qualifications, creating a European third-level system with a standard degree structure (effectively undergraduate, masters and doctorate degrees) and much higher student and staff mobility. Rather like the ‘European Employment Strategy’, ‘Bologna’ is an example of ‘soft law’ – decision-making by mutual bench marking and governance rather than old-style ‘command and control’ government. Probably to the surprise of its original proposers, the Bologna process has been a dramatic success: countries outside the EU are joining up and access to Bologna has become one of the benchmarks used by candidate countries to gauge their own progress toward ‘European’ standards (Keeling, 2006). Meanwhile within the EU all universities are now moving rapidly towards a common degree structure; the consolidation of the ECTS (European Credit Transfer System) weighting of courses not only facilitates student mobility, it is becoming a lever with which students and politicians are beginning to force improvements in the arrogant incompetence of much of what passes for ‘teaching’ in so many universities. Two points should be noticed about the Bologna process. First, its origins, like the original Erasmus programme, lie in the attempt to use the European education system as a tool for the creation of European ‘identity’, and only more recently has it become linked to the rather different objective of ‘competitiveness’. Second, although the aim of the process is a greater integration of European universities’ national systems, this integration is of national and state-funded systems. While a common degree nomenclature can also facilitate market-based integration, this is not the driving force of the process and it does not depend on it. The Bologna process is therefore fundamentally an attempt to reform and integrate European public universities. However, what really excites university authorities across Europe is a rather different aspect of the knowledge-based society – the attempt to emulate the USA in its level of R&D spending and – very much linked to this – the attempt to emulate American ‘excellence’ in university research. Within the EU budget of 2005 for example, virtually the only increase proposed by the Commission was a doubling of the research budget. The new

Worshipping at the shrine of the knowledge-based society?

275

‘European Research Council’ was set up to fund a very small number of elite researchers (High Level Group on Frontier Research, 2005), while at national level governments are increasingly promoting the idea of a small number of elite universities within their mass third-level systems (Münch, 2007). In this version the KBS becomes a justification for ‘pure research’; the quality of the research is judged purely by other scholars within the discipline. The key – and indeed usually the only – criterion of quality is the citation count in peer-refereed journals. Such measures are of course hardly uncontroversial even within academe: scholars in arts and humanities tend to put more weight on scholarly books than journal articles, some areas of the social sciences see ‘applied’ work and other forms of publication as valuable in their own right, while in many areas of the social sciences and humanities the equation of excellence with ‘international’ (that is, Anglophone) publication means a denigration of national or regionally based knowledge and traditions (Wickham, 2004; Wickham and Collins, 2006). Intriguingly however, more is involved than a spat within academe. The economic justifications for pure research rely on what is known as the ‘linear model’ of innovation: pure knowledge is produced by ‘science’, turned then into ‘technology’ and finally becomes ‘innovation’ within the economy. While such a view is flattering to university scientists, it has long been widely challenged by scholars researching precisely these connections within innovation studies. It is also challenged by work within science studies which stress the multiple forms and different institutional contexts of scientific knowledge production. Thus Nowotny et al. (2003) differentiate between ‘Mode 1’ and ‘Mode 2’ science. In Mode 1, scientific knowledge is produced within universities which are effectively institutionally separated from the wider society. Both the course of the research process and the forms of scientific validation are endogenous to the academic community. The shift to Mode 2 involves opening the research actors, the research process and even the forms of validation to the wider society. Within science policy this shift can be seen in attempts to measure the social and economic impacts of research and to use these as criteria for research evaluation (see Spaapen et al., 2007) The rhetoric of excellence merges with that of the market as universities compete against each other for research funds. Where this process has gone furthest, namely in Britain and Ireland, university managers are beginning to behave like the managers of newly privatized state enterprises in the 1980s (see Florio, 2004: 202) – demanding ever higher salaries to justify their ‘leadership’ role that competition has allegedly made necessary. Frequently cited here is the need to emulate US universities.

276

Local versus global perspectives in innovation

The increasing national – and incipiently global – dominance of the elite US universities is a classic example of a winner-takes-all market. What seems to be completely ignored within this debate is the way in which the transformation of American universities has become interwoven with the increasing rigidity of the American social structure. The elite research universities are also elite undergraduate colleges and their graduates have been increasing their share of the top jobs on the graduate labour market. Accordingly, access to such colleges becomes more important and they can increase their fees. Rising fees in turn allow the ‘best’ colleges to increasingly monopolize the ‘best’ staff and thus enhance their research output, thus stretching out the hierarchy even further. On the one hand, elite American colleges are increasingly dominated by the children of the very rich; on the other hand, the cost of ‘college’ education has risen across the board, so that not surprisingly, the cost of education, like the cost of health-care, has become part of the nightmare of ‘Middle America’ (Frank and Cook, 1996). Indeed, the cost of a US university education has outpaced family income for 25 years. This has contributed to growing inequalities in access to higher education: In 1979 students from the richest 25% of American homes were four times as likely to attend college as those from the poorest 25%. Now, students from the richest 25% of American homes are more than 10 times as likely to attend colleges and universities as those from the poorest 25% (Reich, 2004).

Inequality is also growing within the university sector. Elite universities have always been to some extent both socially and intellectually elitist; only at rare moments does intellectual excellence go with an egalitarian recruitment of the student body. However, in America today social and academic elitism are becoming almost synonymous, for access to the leading research universities is becoming more and more restrictive. At the top 146 universities, fully 75 per cent of students come from the wealthiest quartile (Marcus, 2004). At Ivy League colleges only 3 per cent of all students come from the poorest 25 per cent of families, and African Americans have not increased their share of Harvard’s student population recently despite massive expenditures (Lewis, 2004). Two developments here are worth noting. First, universities’ reliance on private and philanthropic funding means that they cultivate their alumni and other individuals for gifts. Not surprisingly, they often fall into the temptation of giving the children of such donors privileged access. While such (covert) ‘legacy lists’ are often criticized, they are an almost inevitable consequence of the desperate search for private donations (Lewis, 2004). Second, the enthusiastic

Worshipping at the shrine of the knowledge-based society?

277

embrace of a market model means a public re-calibration of the nature of financial aid for students (Washburn, 2005). Originally intended as means to ensure that less privileged children could reach university, it increasingly is used as a means to simply attract the best and brightest students, whatever their background. A recent study from the US Century Fund reports: ‘Colleges are beginning to act less like non-profit educational institutions and more like market players, using financial aid as a way of attracting talented students away from competitors, rather than as a method of helping those who need it most’ (Marcus, 2004). The growing educational inequality within the USA is thus part and parcel of the growing inequality of that society. As the executive summary of a recent report states: ‘The growing inequality in family incomes is contributing heavily to the growing disparities in student achievement’ (NCEE, 2006). For most of the last 30 years, the real incomes of many Americans have stagnated while income disparities have grown dramatically (Ryscavage, 1999). Just as it is the very rich and those with the highest incomes who have benefited most from America’s economic growth since the 1970s, so it is these who have benefited most from the ‘success’ of American higher education. ‘The first adverse consequence from the marketization of higher education in the United States concerns social stratification. The most prestigious brands of higher education increasingly are available only to those who can pay for them’ (Reich, 2004). One bizarre feature of the current European enthusiasm for American universities is that it seems to judge the entire system by the achievements of a few elite institutions. This is as absurd as judging the UK secondary education system by the achievements of its elite private schools such as Eton. Indeed, claims that the USA is a ‘knowledge-based society’ sit uneasily with its mediocre levels of basic education and training. The top American universities and business schools may be rated the best in the world, yet in international comparisons a relatively large proportion of Americans fail to acquire a basic education (NCEE, 2006). US employers complain of the low literacy levels of young Americans with for example in one report, three quarters of incoming high-school graduates viewed as deficient in basic English writing skills’ (Knight, 2006). Indeed in terms of overall literacy levels the USA scores only about average, with as much as 18 per cent of 15-year-olds reaching only the basic Level 1 on the PISA reading literacy scale (OECD, 2003: 69). Worse still, the US has one of the highest proportions of 15-year-olds who achieve only the basic Level 1 in mathematical competence (OECD, 2006: 82). Finally, it is worth reflecting on the implications of defining the US as the most ‘knowledgeable’ society in the world. This after all is a country of mass religiosity and popular superstition, a country in which 68 per cent of

278

Local versus global perspectives in innovation

the population believe in the devil (Rifkin, 2004: 20) and in which creationism is actually taught in schools. The intellectual vacuity of the entire ‘knowledge-based society’ argument is highlighted by the fact that most Americans are notoriously ignorant of countries outside the USA, and that the American mass media has no real tradition of public-service broadcasting – just Jerry Springer. An American ‘knowledge-based society’ is really an oxymoron.

LEARNING ORGANIZATIONS AND EUROPE’S VOCATIONAL EDUCATION ACHIEVEMENTS Central to the rhetoric of the ‘knowledge-based society’ is the claim that knowledge is used and indeed developed within the workplace itself. However as we shall now see, the extent to which this is occurring is debatable; assessing the argument involves examining changes in the organization of work and in vocational education and training. One important strand of argument emerges from innovation studies and focuses on developments in the relationship between tacit and codified knowledge. Especially in Scandinavia the new economy is held to require a move towards the ‘learning organization’ (Lundvall and Borras, 1998). Such workplaces allegedly facilitate continuous learning by employees and so can continually adjust to new processes, products and services. The application of codified formal knowledge in the workplace is seen to require a parallel growth in tacit contextual knowledge. This in turn requires workplaces which are participatory and have high levels of mutual trust. This optimistic and at least implicitly democratic rhetoric parallels the rather different tradition within human resource management (HRM). A central concern of HRM has been to appropriate employee knowledge and creativity through the management of workplace culture. While researchers have debated the extent to which this actually occurs, there seems little dispute that this is the direction of management strategy. These two rather different approaches to the development and utilization of knowledge at work do appear to have some empirical validity. Using data from the third European Foundation working-conditions survey, Lorenz and Valleyre (2004) construct a typology of four different types of work organization. In both the ‘learning’ and the ‘lean’ types employees have autonomy, engage in learning in work and contribute to problemsolving, but in the second form employees are also subject to tight control and quantitative production norms. While 49 per cent of the sample works in ‘learning’ organizations, these are disproportionately concentrated in Sweden, Denmark (and to some extent Finland), but ‘lean’ organizations

Worshipping at the shrine of the knowledge-based society?

279

are disproportionately concentrated in the UK and Ireland. Lorenz and Valleyre suggest that the continued coordination of the labour market in the Nordic countries contributes to higher levels of trust within the workplace which in turn facilitate learning organizations; conversely the deregulated labour markets of the UK and Ireland push workplace innovation towards the ‘lean’ organizational form. Here learning occurs, but within parameters clearly defined by management – much indeed as required by the proponents of HRM. The importance of labour-market co-ordination is also a crucial issue for vocational education. While the German ‘dual system’ of apprenticeship was widely seen by the international literature of the 1980s as a decisive component of the (then) successful German economy (see Lane, 1989), vocational education has been largely ignored in the contemporary KBS discussion. This is at first sight surprising, since a developed system of vocational education would appear to be an obvious part of a ‘knowledgebased society’. This is particularly so given that both the French and German traditions (‘formation professionnelle’, ‘berufliche Ausbildung’) have broader connotations than the English ‘vocational education’. Especially the German system could be seen as contributing to a ‘professional’ work identity and perhaps even to the ideal of the ‘worker citizen’. It equips the apprentice for subsequent skilled employment and also contains a continuing general and theoretical education. Such a system is only possible because of tripartite structures involving government, employers’ organizations and trade unions, even though the latter play a subordinate role. It requires institutional structures which facilitate the aggregation of the needs of individual employers into longterm collective needs. Such an apprenticeship system therefore is both consequence and effect of strong economic institutions which in turn contribute to a stronger civil society. Conversely, one reason the UK has never been able to develop effective vocational training is the very weakness of its employers’ associations. Since the 1980s the German vocational education system has appeared to be under threat. Commentators highlighted the difficulties of developing new and more flexible curricula (Casey, 1991) and the growing competition from the expansion of conventional third-level education (Greinert, 1994). Perhaps surprisingly, it now seems that the curriculum has in fact been modernized, but nonetheless, fewer firms are offering apprenticeship places (Bosch et al., 2007). This in turn appears to be interwoven with the weakening power of German employers’ associations and the fundamental de-institutionalization of the economy. Since the 1970s vocational education and training has been an important component of EU social policy. While education and conventional social

280

Local versus global perspectives in innovation

policy such as health or social insurance remained a jealously guarded prerogative of member states, ‘training’ could be defined as an area of European policy since it involved employment and the labour market. In this policy space it was possible for CEDEFOP (the European Centre for the Development of Vocational Training) to be created in 1975 as a response to pressure, in particular from European trade unions, for a European social policy (Guasconi, 2004). Across Europe the national variety of training systems remains and there has certainly been no generalization of the German apprenticeship system. However, especially outside the UK, we can identify some features of a ‘European’ training system: the involvement of both unions and employers’ organizations (thus strengthening institutions of ‘social partnership’) and a curriculum that goes beyond the immediate needs of individual firms. Certainly there is no shortage of official European rhetoric. Since 2002 the ‘Lisbon Process’ is paralleled by the ‘Copenhagen Process’ to enhance collaboration in vocational education and training (European Commission, 2002). Like the Lisbon Process itself, achievement depends essentially upon the techniques of benchmarking, mutual persuasion and soft law first developed in the European Employment Strategy (Goetschky, 1999). Furthermore what little progress has been achieved appears to have been largely in terms of increasing the completion rates within secondary education. Despite the demographic imperatives of an ageing population, there has been little expansion of life-long learning or training for those already in work. The term ‘knowledge society’ suggests a society in which everyone is knowledgeable, but this turns out to be a peculiar distortion of contemporary reality. Indeed, this section of the chapter has suggested that current trends are actually undermining some crucial sites in which learning has occurred. While in a learning society it is plausible that work organizations are themselves learning organizations, this is not the norm in most European societies and worse, current trends actually undermine this. A learning organization, as opposed to an organization that is merely ‘lean’, requires great social trust and supportive state institutions. In France and (especially) Germany these are almost certainly being undermined by ‘Anglo-Saxon’ labour market reforms. At the same time vocational education is being undermined. This is especially clear in the EU’s new memberstates and accession-states where international organizations such as the World Bank have argued that training should be left to the individual and to private providers. Here the role of public institutions is limited to formal school system and there is little interest in creating strong representative organizations (employers associations, trade unions) which in turn could support a vocational training system.

Worshipping at the shrine of the knowledge-based society?

281

CONCLUSION As a concept the ‘knowledge-based society’ or even the ‘knowledge-based economy’ is peculiarly vague. While it is accepted as a self-evident truth by politicians and policy-makers, it is only loosely connected to any clearly formulated sociological arguments. As we have seen, those accounts of social change which do stress the importance of ‘knowledge’ for contemporary society are problematic. The expansion of higher education does not of itself ensure that graduates work as ‘symbolic analysts’ or become members of the ‘creative’ class. Instead there is some evidence that educational qualifications are simply being used to queue applicants for jobs rather than being in any sense necessary to actually do these jobs. Furthermore even when higher education is ‘relevant’ for the graduates’ work, it would seem that what matters in most cases is generic skills (in particular those relevant to social interaction) rather than the formal codified knowledge to which the KBS appeals. Discussion of ‘the’ KBS detracts from the differences between societies. In this it resonates with other sociological theories of convergence, from the industrial society theorists of the 1960s (see Kerr et al. [1960] 1973) to the contemporary globalization gurus (see Friedman, 2005). Parallel with these accounts of convergence is a long tradition in sociology that criticizes them for ignoring the evidence that differences between societies cannot always be explained in terms of a historical lag. Thus just as rich countries can differ in their income distribution, so such countries also differ in terms of their educational institutions and their educational achievements. In these terms the US stands out both for the high performance of its elite universities and the rather mediocre standard of the education achieved by most of the population. Given the concern with the ‘learning organization’, it is clear that not all KBS discussion concentrates exclusively on formal education. This movement away from formal knowledge brings the KBS discussion closer to the concerns of the PILOT project. However, some contemporary developments actually undermine learning within the workplace. If the workforce as a whole is going to ‘learn’, then this requires institutional structures that restrain employers from short-term hiring and firing. However, most European labour markets are being deregulated in order to facilitate such short-term employment. In the past ordinary employees have acquired skills and experience through national vocational educational and training systems, but these are now being eroded. The KBS is a fundamentally optimistic vision of contemporary society. Its assumption that there is inevitable dynamic towards ever-increasing knowledge ignores the possibility that some social groups may be less knowledgeable than in the past. Finally this discussion has only touched on

282

Local versus global perspectives in innovation

the most fundamental issue of all: within the KBS all that matters is knowledge as utilized in employment. At least until the middle of the twentieth century, it would have been self-evident that any discussion of ‘knowledge’ should have included or indeed primarily focused on what we would now term ‘high’ culture. Equally, the programmes of educational reform which were part of the consolidation of the European welfare states in the second half of the last century understood ‘knowledge’ as education for citizenship and social participation. Today the former may be considered unacceptably elitist, the second boringly worthy. Yet compared to such ideals, the knowledge of the KBS is restricted and utilitarian – the sort of knowledge that would have appealed only to Dickens’ Mr Gradgrind.

REFERENCES Bell, Daniel (1973), The Coming of Post-Industrial Society, Harmondsworth: Penguin. Bosch, Gerhard and Steffen Lehndorff (2004), ‘Service economies: high road or low road?’ in Gerhard Bosch and Steffen Lehndorff (eds), Working in the Service Sector – a Tale from Different Worlds, London: Routledge. Bosch, G., T. Haipeter, E. Latniak and S. Lehndorff (2007), ‘Demontage oder Revitalisierung? Das deutsche Beschäftigungsmodell im Umbruch.’ Kölner Zeitschrift fur Soziologie und Sozialpsychologie, 59 (2), 318–39. Brown, Philip, Andy Green and Hugh Lauder (2001), High Skills: Globalization, Competitiveness and Skill Formation, Oxford: Oxford University Press. Brown, Philip, Anthony Hesketh and Sara Williams (2004), The Mismanagement of Talent: Employability and Jobs in the Knowledge Economy, Oxford: Oxford University Press. Casey, B. (1991), ‘Recent developments in the German apprentice system’, British Journal of Industrial Relations, 29 (2), 205–22. European Commission (2002), ‘Copenhagen Declaration’, accessed 4 November, 2007 at http://ec.europa.eu/education/copenhagen/copenahagen_declaration_ en.pdf. Favell, Adrian (2006), ‘London as Eurocity: French free movers in the economic capital of Europe’, in Michael Peter Smith and Adrian Favell (eds), The Human Face of Global Capital, New Brunswick, NJ: Transaction Publishers, pp. 247–74. Finegold, David and David Soskice (1988), ‘The failure of training in Britain: analysis and prescription’, Oxford Review of Economic Policy, 4 (3), 21–53. Florida, Richard (2004), The Rise of the Creative Class: And How It’s Transforming Work, Leisure, Community and Everyday Life, New York: Basic Books. Florida, Richard (2005), The Flight of the Creative Class: The New Global Competition for Talent, New York: HarperCollins. Florio, Massimo (2004), The Great Divestiture: Evaluating the Welfare Impact of the British Privatisations 1979–1997, Cambridge, MA: MIT. Frank, Robert H. and Philip J. Cook (1996), The Winner-Take-All Society: Why the Few at the Top Get So Much More Than the Rest of Us, New York: Penguin Books.

Worshipping at the shrine of the knowledge-based society?

283

Friedman, Thomas L. (2005), The World Is Flat: A Brief History of the Twentieth Century, New York: Farrar, Strauss and Giroux. Giddens, Anthony (2007), Europe in the Global Age, Cambridge: Polity Press. Goetschy, J. (1999), ‘The European employment strategy: genesis and development’, European Journal of Industrial Relations, 5 (2), 117–37. Gordon, Ian, Tony Travers and Christine Whitehead (2007), The Impact of Recent Immigration on the London Economy, London: London School of Economics for City of London. Grabher, G. (2004), ‘Learning in projects, remembering in networks? Communality, sociality and connectivity in project ecologies’, European Urban and Regional Studies, 11 (2), 103–23. Greinert, W.D. (1994), ‘Vocational education and socio-economic change: causes of the crisis of the dual system of vocational education’, Zeitschrift für Pädagogik, 40 (3), 257–372. Guasconi, M.E. (2004), ‘The unions and the relaunching of European social policy’, European Journal of Vocational Training, 32, 55–62. Hanlon, G. (2000), ‘Seeking the New Jerusalem? – The New Right, social democracy and professional identities’, Sociological Research Online, 5 (1) accessed at www.socresonline.org.uk/5/1/hanlon.html. High Level Group on Frontier Research (2005), Frontier Research: The European Challenge, Luxembourg: Office for Official Publications of the European Communities. Keeling, R. (2006), ‘The Bologna process and the Lisbon research agenda: the European Commission’s expanding role in higher education discourse’, European Journal of Education, 41 (2), 203–23. Kerr, Clark, J.T. Dunlop, E. Harbison and C.A. Myers (1973), Industrialism and Industrial Man, first published 1960, Harmondsworth: Penguin. Knight, Rebecca (2006), ‘Entrants to workforce in US are deficient in basic skills’, Financial Times, 2 October. Lane, Christel (1989), Management and Labour in Europe, Aldershot: Gower. Lewis, P. (2004), ‘Dad’s old school tie may be just the ticket’, Times Higher Educational Supplement, 24–31 December. Lindley, Robert (2002), ‘Knowledge based economies: the European employment debate in a new context’, in Maria Rodriques (ed.) The New Knowledge Economy in Europe, Cheltenham UK and Northampton MA, USA: Edward Elgar, pp. 95–145. Lorenz, Edward and Antoine Valeyre (2004), ‘Organisational change in Europe: national models or the diffusion of a new “one best way”?’ DRUID working paper no. 04-04, accessed 4 November, 2007 at www.druid.dk/wp/pdf_files/0404.pdf. Lundvall, Bengt-Åke and Susana Borras (1998), The Globalising Learning Economy: Implications for Innovation Policy, Brussels: European Commission, accessed 4 November, 2007 at ftp://ftp.cordis.lu/pub/tser/docs/globeco.doc. Marchington, Mick, Damian Grimshaw, Jill Rubery and Hugh Willmott (eds) (2005), Fragmenting Work: Blurring Organizational Boundaries and Disordering Hierarchies, Oxford: Oxford University Press. Marcus, J. (2004), ‘US fees hike pose threat to access’, Times Higher Educational Supplement, 9 September. Martin, B. (2005), ‘Managers after the age of organizational restructuring: towards a second managerial revolution?’ Work Employment & Society, 19 (4), 747–60.

284

Local versus global perspectives in innovation

Martin, B. and J. Wajcman (2004), ‘Markets, contingency and preferences: contemporary managers’ narrative identities’, Sociological Review, 52 (2), 240–64. Mason, G. (2002), ‘High skills utilisation under mass higher education: graduate employment in services industries in Britain’, Journal of Education and Work, 15 (4), 427–56. Münch, Richard (2007), Die akademische Elite. Zur sozialen Konstruktion wissenschaftlicher Exzellenz, Frankfurt: Suhrkamp. National Center on Education and the Economy (NCEE) (2006), Tough Choices or Tough Times: The Report of the New Commission on the Skills of the American Workforce, New York: John Wiley. Nowotny, H., P. Scott and M. Gibbons (2003), ‘ “Mode 2” revisited: the new production of knowledge – introduction’, Minerva, 41 (3), 179–94. Reich, Robert (1993), The Work of Nations: Preparing Ourselves For 21st Century Capitalism, first published New York, 1991, London: Simon & Schuster. Reich, Robert (2004), ‘Higher education in the United States is coming to resemble any other kind of personal service industry’, Higher Education Policy Institute Lecture 2004, accessed 4 November, 2007 at http://news.bbc.co.uk/go/pr/fr//2/hi/uk_news/education/3564531.stm. Rifkin, Jeremy (2004), The European Dream, London: Polity. Ryscavage, Paul (1999), Income Inequality in America: An Analysis of Trends, Armonk, NY: M.E. Sharpe. Spaapen, Jack, Huub Dijstelbloem and Frank Wamelink (2007), Evaluating Research in Context: A Method for Comprehensive Assessment, The Hague: Consultative Committee of Sector Councils for Research and Development, accessed 4 November, 2007 at www.nwo.nl/nwohome.nsf/pages/NWOA_ 6V4BXK_Eng. Washburn, Jennifer (2005), University Inc.: The Corporate Corruption of Higher Education, New York: Basic Books. Webster, Frank (2002), Theories of the Information Society, London: Routledge. Webster, J. (2004), ‘Digitising inequality: the cul-de-sac of women’s work in European services’, New Technology Work and Employment, 19 (3), 160–76. Wickham, J. (2004), ‘Something new in old Europe? Innovations in EU-funded research’, Innovation: European Journal of Social Science Research, 17 (3), 187–204. Wickham, J. and G. Collins (2004), ‘The call centre: a nursery for new forms of work organisation?’ Service Industries Journal, 24 (1), 1–18. Wickham, J. and G. Collins (2006), ‘Involving users in social science research: a new European paradigm?’, European Journal of Education, 41 (2), 269–80.

Index accident investigations 121 actors external 9–10 innovative 26, 35 interactions between 25–6 pooling knowledge 37–8 reconfiguring 36–9 aerospace industry 120, 135 Air Traffic Control (ATC) 127, 130 American Association for the Advancement of Science 64 Anthony, R.N. 65 applied psychology 120 Aristotle 72 Arnold, Malcolm 73 Asia, innovation in 45, 46, 245 ‘Asian Tigers’ 19, 245, 246, 247, 260, 261, 262 ASIC manufacturers 52, 53 Austria 35 Automatic Dependent Surveillance Broadcast 129 automation 51 automation solutions 53 automobile industry 47–8, 50, 59 aviation industry 118 accidents and safety 121, 124–5, 134 data collection and interpretation 129–30 and high technology 119 innovations in 118–9, 123 models 132–4 regulation of 119, 122 see also EFB, HILAS, human factors Beijing, China 249, 250 Bell, Daniel 268–9 Bernal, J.D. 64, 65 biotechnology 5, 108, 110, 147 ‘blocked societies’ 273

‘Bologna process’ of university reform (EU) 274 Boretsky, M.T. 68 Botswana, cellular mobiles in 249 British Association for the Advancement of Science 71 Bulgaria 233 Bureau of Census (US) 69 Bureau of Labor Statistics (US) 66 business system standards, development of 49 business models 168 ‘capabilities’ (term) absorptive capacity/capabilities 28, 34, 39, 47 capability building 28 configurational capabilities 29–30, 34–5 creation of capabilities 260 defined 27–8 dynamic capabilities 28 reconfiguring 37 transformational 28, 32–4, 39 Capital (Marx, Das Kapital) 95 capital investment 88, 99 capital productivity 180–82 catching-up countries 256, 261 Cattell, J.M. 65 CEDEFOP (European Centre for the Development of Vocational Training) 280 Central and Eastern Europe (CEE) 195 and EU accession 232–3 foreign direct investment in 206 and interdependence 237 transition from planned to market economies 197, 237 Chengdu (Sichuan) China 249 285

286

Index

China 16, 245, 246 complexity of innovation and economic systems in 249, 250–51, 252 disequilibrium phenomenon 250–51 effects of physical size 252 ICT in 246, 247 increase in high-tech trade 246–7, 251 R&D 248–52, 261–2 S&T 246, 248 scientific publishing 247–8 velocity phenomenon 250–51 Chongquing, China 249 classification 72–3 Cleaning in Place (technique) 34 clusters of firms 15 ‘Community Innovation Survey’ (CIS, EU) 99 competences core 28, 29 high quality zero defect competence 28 technological 107 complexity characteristics of 253–60 complexity and interdependence 259 dealing with complexity 144–6 linear model of complexity 260 models 253–5 nature of complexity 258–60 understanding complexity 256–60 CoPS (complex products and systems) 165 ‘corporate culture’ 270–71 cost efficiency 169 Crew Resource Management (CRM) 127, 131 culture 73 significance of 121 Czech Republic 233 de-industrialization, debate on 5 Denmark 44, 233, 278 Department of Commerce (US) 68, 69 Department of Labor (US) 69 ‘design rules’ 143 developing countries 5 diffusion and innovation 87–8 and technology 102, 222

dyads and triads 72–3 ‘dynamic capability’ concept of 9, 16–17 economics of science 67 of innovation 95 economy economic development 251–2 giant economies 246 knowledge economy 94–5, 262 learning economy 4, 281 LMT sectors and economic growth 140 mature 188 moral economy 71–5, 76 shift from manufacturing to services 5 see also China ECTS (European Credit Transfer System) 274 education and jobs 272–8 EFB (electronic flight bag) 125, 127, 128–30, 132, 135 see also aviation industry, human factors efficiency 142–4 Emilia Romagna region, Italy 13, 273 research project 150–55 employment and innovation 88–9 Erasmus Programme 274 ergonomics 120–21 ‘escalator cities’ 273 Ethernet 52 EU Industrial R&D Investment Scoreboard 44–5 Europe 108 aviation in 132–4 balance of payments deficit 67 distribution of LMT firms in 233, 236 Eastern 6, 233 East-West relation 236 European economic landscape 15 industries in 43 industries innovation practices 226–7 markets fields in 44 Middle 6 policy makers in 267 R&D and innovation 45–6 relocation of companies within 237

Index social policy 280 traditional industrial strengths 50–51 Western 6, 12, 68, 233 see also European Union ‘European Employment Strategy’ 274 European Foundation 278 European Manufacturing Survey (EMS) 177 European Research Council 275 European Union 3, 4, 7, 16, 40, 69, 99, 197, 260 Accession Countries 199 Bologna process 274 collaboration rates 205 and education 278–80 enlargement of 225–6, 229, 232–6, 280 EU15 199, 200, 201, 203, 206 EU25 199, 204 European Commission 237 ‘Fourth Community Innovation Survey’ (CIS 4) 223, 226 high tech industries 199 Industrial R&D Scoreboard 44–5 innovation policies 40 as knowledge-based economy 5 as knowledge society 3 liberalisation of 199 Lisbon and Barcelona documents 253, 259, 274 ‘new’ states 18, 210 ‘old’ EU states 198, 207, 208, 212, 213, 233 Poland and 198 protectionism 3 and research budget 274–5 and R&D 5 Federal Aviation Authority, US 132 Federal Ministry of Research and Technology, Germany 52 Federation Aviation Authority Notices to Airmen (Notam) 129 Finland 33, 44, 278 FIP 51 firms absorbing new technologies 44 becoming service providers 161 165–6

287 boundaries of firms 141, 142, 144 capital and labour productivity 180–82 challenge of becoming a service provider 166–7 and competitors 165 complex knowledge bases 259 and complex problem handling 145–6, 147 differences between 26 distributed knowledge bases 100–102 external environment of 97, 98–9, 111 firm specific knowledge 27, 99–100 firms and complexity 141, 148 foreign-owned companies (FOCs) 208, 209, 201–12 heterogeneity of firms 26, 27, 151, 175 high tech 12 homogeneity 26 innovation activities 32–3, 184–5, 226–7 and innovation 94, 103 integrators and non-integrators 152–5 integrators 151–3 and knowledge 144–5 knowledge base 100, 257 ‘knowledge businesses’ 133–4 and knowledge creation 98–9 knowledge sharing 133 low tech 12 low-, medium- and high-tech firms 178 relative performance of low-, medium- and high-tech firms 178, 191, 199 nature of 26–7 problems of incumbent firms 31 and problem solving 146–7 and provision of complete solutions to customers 190 and R&D 94, 98, 208 relationship between low-, mediumand high-technology firms and sectors 188–9 relations between 12 relevance of services to competitiveness 182–8

288

Index

research-intensive 12 resource based theory of the firm 260 scale intensive firms 224 and services 160 service innovators 184–5 services as share of turnover 185–8, 189 shifting from goods-focused production to service-orientated 166, 176–7 size and participation in services 182–4 size and productivity 180 supplier-dominated firms 221–2, 224 and systems integration 44–6 tech level and firm size 179 and transformational capabilities 32–4 see also globalization, IEC, industries, mechatronics, problem handling and solving, R&D, service sector, SMEs Flight Data Monitoring (FDM) 126, 130, 131 Flight Operation Quality Assurance (FOQA) 127, 130 Fligstein, Neil 49 Florida, Richard 268, 270–71 food-processing and distributed knowledge 108–11 and packaging 111–12 foreign direct investment (FDI) 204, 206 foreign-owned companies (FOCs) 208, 209, 210, 212 and R&D 210–12 see also firms France 51, 222, 247, 273, 279, 280 Frascati manual 69, 75–6 Fraunhofer Institue for Systems and Innovation Research (ISI) 177 Freeman, C., R. Poignant and I. Svennilson, OECD study 67, 68 generative frameworks 96 German Automobile Manufacturers’ Initiative (AIDA) 53 German Manufacturing Survey 2006 18, 177

Germany 5, 16, 43, 87, 88, 222, 247, 273, 279, 280 industry 262 and innovation 49–50 regulation of labour market 223 and standardization 49–50 Giddens, Anthony 273 ‘going global’ 13–16 globalization 13, 255, 268, 269 goods 161 combining goods and services 163–5 Great Depression 66 Gross Expenditure on R&D (GERD) 204, 248, 249 high- and medium-high-tech industries (HMHT) 222, 225, 227, 228, 229, 239 and EU enlargement 232, 233, 236 interdependences between 236 high-technology definition of 65, 66–9, 97 ‘high tech myopia’ 221 and knowledge creation 98–8 over-emphasis on high-tech 197–8 research intensive industries 67 see also knowledge, R&D high tech (HT) sector of industry 98, 118, 119, 199, 210, 217, 222, 249 focus on 4 ratio of high tech to economic development 6–7 relationship with LMT firms 12–13, 15–16, 118, 221, 236, 241 high throughput experimentation (HTE) 48 higher education and jobs mismatch 272–3 HILAS (Human Integration into the Lifecycle of Aviation) project 120–34, 135 aims 123 Flight Operations 122, 128 innovation 128–32 as knowledge-innovation project 132 and reporting 129 as a R&D consortium 123 tools 125–7, 130–31, 132–3, 135 see also human factors Hoffmeyer, E. 67

Index human error and accidents 121 human factors (HF) concept 120 HF improvements 122–4 HF knowledge 135 HF tools 125–7, 130–31, 132–3, 135 human-factors studies 121 see also HILAS human resource management (HRM) 278–9 IMF 245 incumbent countries 260–61 India 245, 270 industrial complementarities 221 concept of 222–3 between firms 221 between industries 223–4 between sectors 223–6 industrial development 255 institutional 223 Industrial Development Agency (ARP), Poland 238 industries ‘servicizing’ manufacturing industries 176–7,188 high tech (HT) industries 44, 199, 210, 212, 213, 214, 215, 216, 217, 218 high, medium and low tech defined 5, 175, 176, 178–82 interdependencies between different sectors 224–5 intra-sectoral heterogeneity 175 involvement in innovation of different categories 212–16 low tech (LT) 199–200, 210. 213, 214, 215, 216, 217, 217, 218 mature industries 6 medium high tech (MHT) 16, 199, 200, 210, 212, 213, 214, 215, 217, 218 medium low tech (MLT) 199, 212, 213, 214, 215, 216, 217, 218 medium-tech industries (MT) 43 relationship with economic growth 6–7 shift between categories 210 technological basis 43, 44 technologically intensive 68

289

traded and untraded interdependencies 225, 226–32 see also firms, knowledge intensive industries, LMT Information and Communications Technology (ICT) 5, 248–52, 258 ICT complex 102 ICT development 258 information economy 75 information networks 104–5, 131 innovation enabling capabilities (IEC) 17, 26–39, 40, 51, 52, 54 innovations basis of innovative organization 26 comparison of forms 56 contexts of innovation 77 diffusion and innovation 87–8, 93 dynamics of open innovation 45 high performance in innovation 32 high-tech innovation 245 innovation analysis 100 innovation and employment 88–9 innovation in usage 120 innovation quality 87 innovation resources in new EU states 198 innovation studies 77, 93, 278 innovation systems 104 innovation without R&D 76, 221 innovative activity by sector 230–31, 232 input to innovation 85 linear model of innovation studies 25 and LMT industries 93 and ‘newness’ 172 non-incremental 245 non-technological innovative activities 99 open, semi-open and closed forms 54–8 practical problems for innovators 30–31 proximity and innovation 133 science-based innovation 40 spending on innovation 45 stages of innovation 86 technical-scientific basis of 14 types of innovative activities 227 Intel 48

290

Index

intellectual property rights (IPR) 54, 56, 123, 124 inter-agent and inter-industry flows 102–3 disembodied flows 102 embodied flows 102–3 International Ergonomics Association (IEA) Council 120–21 international trade fairs 49 Ireland 233, 248, 272, 275, 279 ISA/ANSI (standards) 51 Italy 43, 222, 233, 248, 273 Japan 5, 60, 69, 108, 140, 243 knowledge ability to utilize knowledge 9 acquisition and use of knowledge 7–8 analytical-synthetic knowledge 256–8 application of generally applicable knowledge 29 barriers to knowledge transmission 104 configuring and reconfiguring distributed knowledge 34–9 defining and failure to define term 96–7 differentiation of 102 diffusion of knowledge 258 distinction between types of knowledge 257–8 distributed knowledge 34, 35–6, 38, 100, 101–2, 103, 112 distributed knowledge base 8, 38, 100–101 empirical evidence of content 106 external knowledge sources 8, 102, 227, 228, 229, 238 firm specific 27 generation of knowledge 38, 93 globally available knowledge 28, 29 global transfer of knowledge 258–60 innovative knowledge 7–8 integration of dispersed knowledge 29–30 knowledge intensity 5 internal knowledge sources 227, 229 knowledge as productive force 4

knowledge base 104, 113 complexity of 256–62 knowledge R&D activities 261 knowledge sharing 123–4 local knowledge 29 management and organization of knowledge 8–10 mapping distributed knowledge 103–6, 113 nature of knowledge 95–7, 267 outsourcing of knowledge acquisition 103 pooling knowledge 37 ‘practical knowledge’ 7, 8 protected knowledge 98–9 relationship between knowledge and occupation 268 repositories of knowledge 36–9 as a resource 4 scientifically generated knowledge 7, 8 specialization in knowledge production 103 synthetic knowledge 256–8 and technology 144 technological knowledge 145 transition between global and local knowledge 29 types of knowledge 7, 28–9 sources of knowledge 94, 104 and trust 133 use of knowledge 282 vocational knowledge 279–80 workplace knowledge 278–9 ‘knowledge society’ 3, 4, 94, 280 knowledge-based society (KBS)/ knowledge-based economy 5, 19, 75, 94–5, 267, 281–2 term defined 95–6, 267 ‘catch up’ 19 centrality of knowledge to 274 knowledge and innovation 221, US 278 social structure of KBS 268–73 and use of knowledge 278 Knowledge Based Economy (OECD document) 98 knowledge-intensive business services (KIBS) 171 knowledge-intensive industry 94, 97–8

Index knowledge workers 269, 270, 272 see also symbolic analysts Korea, South 245, 248 labour force 88 market 10, 11, 279 productivity 180–82 Latvia 233 Lean Flight Initiative 120 ‘Learning Economy’ 4, 281 Levi, L. 64–5 life-cycle concept 167–8 linear model of innovation 25 Lisbon Strategy and Process 5, 253, 280 Lithuania 233 low- and medium-technology (LMT) industries 3, 4, 99, 111 in California 6 capabilities 32 and complementarities 224–5, 229–30 configurational capabilities 34–9 contribution to industrial innovation 16 distributed knowledge and 11, 35–6, 101–2 and economic growth 7 and educational qualifications 9, 10 employees of LMT firms 9 enlargement and dynamics 232–6 expansion of links 13 and EU enlargement 232–6 European LMT 46, 233 firms and vocational education 10–11 and HT firms 12–13, 15–16, 118, 221, 236, 241 and human factors 121 improvements in innovation 198 and industrial regions 221 industries and innovation 7–10, 11, 12, 25, 31, 135, 221 industries and standard setting 47–50 influence on high-tech innovation 12–13, 110 knowledge base of LMT firms 7, 9, 11–12, 14 LMT countries 6

291

LMT enterprises and knowledge management 8–10 LMT firms and their environment 112 LMT innovation 122–4, 134 LMT sectors 95, 102, 118, 140 low level of R&D 7, 95 as market for high-tech products 12, 15–16, 224–5 mature industries 6 mode of innovation 7–10 as part of advanced industrial regions 221 perceived impact 197 place in economic and industrial context 11–12 in Poland 198 productivity 180–81 and policy-regulatory context 11 proximity to markets and customers 14 and regional ties 13–14 relations with developers and manufacturers of production technologies 12 role in economy 101, 176 role of in economic growth 140 role of information sources 228, 229 role of standard setting in 47–53 role of 198 shift of LMT industries in Europe 226, 241 significant relationships 241 societal and institutional conditions 10–14 as supplier-dominated 221–2, 229 working in network 11–12,104 case studies 108–12 see also PILOT, Poland, service sectors and industries ‘low-skill equilibrium’ 272–3 low-tech (LT) industries 76, 199–200, 210, 212 and human factor improvements 122–4 low technology and growth 88 low-tech innovation patterns, debate on 221 relevance of low technology 87–8

292

Index

and traded and untraded interdependences 225 see also Poland Luxembourg 233 Maclaurin, R.W. 70–71 macroeconomic discussions 88 Main Science and Technology Indicators (MSTI) 70, 74 maintenance and after-care services 168 Malaysia 245 managers and career paths 271 Managers/management 10, 93 manufacturers transforming into service providers 161 transition from LMT to high-tech industries 6 mapping knowledge sources 103–6 markets complexity of markets 259 incumbents and challengers 44 market failures 143 markets as fields 44 players in 44 stable markets 49 Marx, Karl 95 Matthew effect 34 ‘mechatronics’ 140, 149–55 and technology fusion 141 metrics 66 Microsoft 48 MIT 70 models of economic growth 253–6 modes of organizing production 142 modularities 103, 148 modularity, product and organizational 143–4 molecular modeling 48 MROs (maintenance and repair organizations) 119 nanotechnology 5, 111–12, 248, 256 National Association of Manufacturers (US) 66 National Research Council (NRC, US) 66 National Science Foundation (NSF, US) 66, 67, 69

‘National Strategy for Regional Development’ (Poland) 237 networks research 110, 112 network-centric perspective 131 New Zealand 111 newly industrialized countries 16 Nicomachean Ethics (Aristotle) 72 non-research-intensive sector 9, 11, 69 innovations in 25, 43 Norway 109, 110, 117 OECD 66, 67–8, 69, 85, 245 analysis of technological development 69–70 concept of technology gap 68 campaign for science policies 67 classification of firms and R&D levels 97–8 classification of industries 3–4 definition of technology levels 69–70 Frascati manual 69, 75–6 knowledge economy 95, 97 OECD classification of industrial sectors 5–6, 175, 178 OECD Council of Ministers 69 OECD countries 6, 68, 74, 87, 250, 260 OECD economies 6 OECD science policies 67 Oslo manual 87, 203, 205, 213 and R&D 74 storytelling functions of 75 and term ‘high technology’ 69, 97–8 OECD NESTI Group 74 OMROM 52 Open-Source General Public License 56 ‘Operational Process Model’ 134 Oracle 48 organizations, layered 54 organizational proximity 133 outsourcing 169–70, 222 strategic aspects of 170 packaging and innovation 111–12 ‘paradox of territories’ 13 patents 246, 259 patent rights 54 patenting 95 Paul, St 73

Index performance improvements, transfusion between firms and industries 102 persistent errors, potential fruitfulness of 25 Personal Digital Assistants (PDAs) 126 PILOT (Policy and Innovation in Lowtech Industries) 17, 26, 31, 35, 118, 281 PISA reading literacy scale 277 Poland 18–19, 197 and low-quality trap 203 changes in sectoral productivity 201–2 comparison to EU15 states 199, 201, 203, 206, 240 competitive performance of manufacturing 199–203 decentralization in 237 determinants of change 198 development programmes 238 economic growth 199 exports to EU 203, 208, 239 firm size and proportion of innovative spending 208 firm size and spending on R&D 208 increased innovation 217 innovation in 202 investment in 204 labour force in 202, 204 level of technology and collaboration 215 licensing of technology 205–6 LMT in 198, 239 regional policies 237–40 reorganisation of firms 202–3 skills barrier 210 technological development 204 unit labour costs and relative unit labour costs 200 variations in innovation activity 212–16 variations in sectoral innovation 207–12 Policy and Innovation in Low-tech Industries see PILOT policy makers 93 Portugal 233 prisoners’ dilemma 119–20, 133

293

problem handling and solving 145–7, 257 product development 143–4, 180 and organisations 144 product services 164 production technologies 10 product decomposability 147–9 product interdependences 148–9 productivity 180–82 and firm size 180 types of 180 PROFIBUS 50, 51, 52–3, 54, 59 PROFIBUS & PROFINET International (PI) 58 PROFIBUS International (PI) 52, 55 PROFIBUS User Organization (PNO) 51, 52, 53, 57–8 Advisory Board 57 PNO Business Office 54 PNO IPR Policy (2006) 54 PNO Board of Directors 56, 57 WG 56–7 PROFINET 52, 54, 59 use in real time 53, 54 proximity – global, cultural and spatial 13–14 public innovation support programmes 11 public research and technology policies 221 ‘qualifications, hybrid’ 10–11 qualification deficits 10–11 qualifications and firms 10–11 vocational training 11 RAND-terms 54 ‘rationalization’ 88 regional ties, erosion of 13–14 Reich, Robert 268, 269, 270–71, 272 R&D (research and development) 3, 4, 11, 13, 74, 95, 96, 118, 142, 144, 203, 246 advanced 49 analysis of level of R&D intensity 177 Chinese R&D 248–52 comparison of regions 44–7 components of 76 definitions 32

294

Index

effects of outsourcing R&D 145 European R&D 45–6 examples 32–4, 37–8 as form of knowledge creation 39–40 government R&D 50, 64 growth in 49 indirect relationship with knowledge generation 97 interaction with innovation 85, 86–7, 118, 119 internal R&D 106–8 intramural R&D 100, 102, 205, 227 investment patterns 45 networks research 110, 112 non-diffusion of R&D 217 OECD definitions 97 organization of R&D activities 147, 155 and other scientific activities 72 in Poland 204–6 and problem decomposability 146–7 and problem solving 147 R&D capacity 26 R&D departments in firms 27 R&D facilities 30 R&D and innovation, interaction between 85–7 R&D intensity 87, 177, 179, 180, 189 and competitiveness 180 and size of firms 190 R&D intensive products 85 R&D spending 45, 75 R&D strategies 43 relationship between R&D intensity and services 186–7 research into 65–6 and research intensity 67, 251 and return on investment ratios 65, 66 and Schumpeter 89 and statistics 74, 75 traditional forms 59 and universities 274 verticalization and de-verticalization 145 see also HILAS, mechatronics, technology research budget and national income 64–5 in aviation industry 119, 123

‘effectiveness’ of research 65 and linear model 254 industrial R&D 49 research intensity 67–8, 89 R&D capacity and innovation 26 research, pure, significance of 275 evaluation of effectiveness 65–6 research-education-development 65 ‘return on investment’ (ROI) ratios 65 risk model 130 risk sharing 120 Romania 233 Rosa, E.B. 74 S&T 245, 246, 261 safety management 121, 124–5, 134 Schumpeter, Joseph 70, 171, 172 concepts 85, 87 and R&D-intensive products 89 Schumpeterian competition 202 Science and Technology Indicators (OECD publication) 69 Science and Technology Studies (STS) 27, 28–9, 30 science and technology 75, 76 Science Research Board (US) 64 science and trust 71 scientific truth 71 service sectors and industries 160, 269 and R&D 87 combining goods and services 163–5 cultural provision of service provision 166 distinction from goods 161 inclusion of services in costs 190–91 industrial services 163–4 innovation and services 162, 165, 171 interrelationship with R&D 187 life cycle service provision 167, 168 models 161–2 nature and components of 161–3, 164 pricing of services 188, 190 product related services 176, 182–8 purchase of services 169–71 relevance of services 176 ‘service concept’ 162 service providers 12

Index service provision by low-, mediumand high-tech firms compared 178 services in manufacturing industries 176–7 see also firms, R&D Shanghai, China 249 Silicon Valley, significance of 271 SIMATIC solutions 57 Singapore 245, 248 skills requirements 168 SMEs (small- and medium-sized enterprises) 76, 99, 112–13 and services 186, 190 and technology level 179, 190 society 77 ‘post-industrial society’ 268 sociology, industrial and organizational 27 solution providers 53, 165 Soviet Union 64 Spain 273 Springer, Jerry 278 standard, defining the global 52 European Standards 50–53 standardization processes 51 standard-setting institutions 49 role of standard-setting 47–53 standard-setting dynamics and innovation process 51–2 standard setting partnership 57 statistics 74, 75 statistical agenda 77 STEP (Science Technology Economic Policy) group, Norway 109–10 substitution effects 88–9 suppliers and knowledge provision 8 supply chains 15 Sweden 44, 262, 278 Switzerland 43 ‘symbolic analysts’ 269–71, 272 systems integration theory 145 coupling of systems 148–9 system standards 47–50 Taiwan 245 Tayloristic forms of work organization 9, 270 technical and engineering societies and standards 49

295

technology adapting new technology 46, 48 advanced knowledge-based technologies 102–3 attendant /partial effects of technical progress 86 concept of high, medium and low technology 70–71 defining high technology 72–3 domestic technological expertise 36 embodied and disembodied technology 203, 205 and employment 88–9, 234 high and low technology, interaction between 72, 88 high/low technology dichotomy 72–3 ‘high technology’ 69–71, 73, 74, 85, 87. 89 integration of different technologies 141 integration of technologies 141 intensity 68–9 knowledge based technologies 103 labelling of technologies 72–3 low technology and growth 88 low technology 87 mixing technologies 103 moral economy of high technology 71–5, 76, 88 nature of technical change 149 novel technology 87 patterns of technological change 221–2 quantifying high technology 74 and R&D 75 technological competition and diffusion of technology 102, 222, 224 technology gap 68 technological process 86 technological progress and innovation 25–6, 70 technology fusion 141 technology standards 48 technology and trust 71, 119 terminology and classification 72–3 types of technologies 69, 70–71 technological complementarities 141 patterns of technological change 221

296

Index

Technological Gap Study Program (US) 68 technological intensity 68–70 Thailand 245 Threat and Error Management (TEM) reports 127, 131 trade, global, increase in 246 in high tech products 246 transaction-cost theory 142–3, 144 Truman, President Harry 64 trust 71 and knowledge sharing 119, 123 TU Munich, Institute for Information Technology (ITM) 58 UK 247, 272, 273, 279 call centres in 270 LMT industries 233 and R&D 67 research budget 64, 275 UN 245 ‘unit export value’ (UEV) and ‘relative unit export value’ (RUEV) 201, 203 unit labour costs (ULC) 200, 201 relative unit labour costs (RULC) 200, 201, 207 universities, research and access 274–7 US Bureau of Standards 65 US Century Fund 277

US 3, 46, 48, 66, 108, 273 balance of payments deficit 67 development of rhetoric of high technology 66 education in 267, 274, 276–7 educational deficiencies 277 government policy 269 growing inequality in 276–7 influence of US universities 275–6 market fields in 44 as most ‘knowledgeable’ society 277–8 protectionism 3 and R&D 67, 68, 274 research budget 64 research in 268 and standard setting 51 and technology gap 68–9 see also knowledge based society value chains 15, 38 Vietnam 245, 246 vocational education and training 278–80 WBRD 245 World Bank 280 WTO 245 Xian, (Shaanxi) China 249–50