Sectoral Systems of Innovation and Production in Developing Countries Actors, Structure and Evolution
Edited by
Franco Malerba Professor of Industrial Economics, KITeS-CESPRI, Bocconi University, Milan, Italy
Sunil Mani Professor, Planning Commission Chair, Centre for Development Studies, Trivandrum, Kerala, India
Edward Elgar Cheltenham, UK • Northampton, MA, USA
© Franco Malerba and Sunil Mani 2009 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: 2009930889
ISBN 978 1 84844 656 4 Printed and bound by MPG Books Group, UK
Contents Contributors I. 1.
II.
2.
3.
4.
5.
6.
vii
INTRODUCTION Sectoral systems of innovation and production in developing countries: an introduction Franco Malerba and Sunil Mani
3
ACTORS AND STRUCTURE OF SECTORAL SYSTEMS IN DEVELOPING COUNTRIES Why is the Indian pharmaceutical industry more innovative than its telecommunications equipment industry? Contrasts between the sectoral systems of innovation of the Indian pharmaceutical and telecommunications industries Sunil Mani From innovation projects to knowledge networks: knowledge as contingency in the sectoral organization of innovation Fernando Perini Learning, innovation and public policy: the emergence of the Brazilian pulp and paper industry Hannes Toivanen and Maria Barbosa Lima-Toivanen The software sector in Uruguay: a sectoral systems of innovation perspective Marjolein Caniëls, Effie Kesidou and Henny Romijn Sectoral system of innovation in Brazil: reflections about the accumulation of technological capabilities in the aeronautic sector (1990–2002) Rosane Argou Marques and L. Guilherme de Oliveira
v
27
57
99
131
156
vi
Sectoral systems of innovation and production
III. DYNAMICS AND EVOLUTION OF SECTORAL SYSTEMS 7.
8.
9.
10.
11.
China’s threat and opportunity for the Thai and Vietnamese motorcycle industries: a sectoral innovation system analysis Patarapong Intarakumnerd and Mai Fujita ‘Low-tech’ industry: a new path for development? The case of the salmon farming industry in Chile Michiko Iizuka Making a technological catch-up in the capital goods industry: barriers and opportunities in the Korean case Yoon-Zi Kim and Keun Lee From ‘nuts and bolts’ to ‘bits and bytes’: the evolution of Taiwan ICT in a global knowledge-based economy Ting-Lin Lee Prospects for Jatropha biofuels in Tanzania: an analysis with strategic niche management Janske van Eijck and Henny Romijn
Index
207
232
259
287
335
367
Contributors Dr Franco Malerba, Professor of Industrial Economics, Director of KITeS, Bocconi University, Milan, Italy Dr Sunil Mani, Planning Commission Chair, Centre for Development Studies, Trivandrum, Kerala, India Dr Patarapong Intarakumnerd, College of Innovation, Thammasat University, Bangkok, Thailand Mai Fujita, Institute of Developing Economies, Japan External Trade Organization, Chiba, Japan Dr Marjolein Caniëls, Faculty of Management Sciences, Open University of the Netherlands, The Netherlands Dr Effie Kesidou, Nottingham University Business School, University of Nottingham, UK Dr Henny Romijn, Faculty of Industrial Engineering and Innovation Sciences, Eindhoven University of Technology, The Netherlands Dr Ting-Lin Lee, Department of Asia Pacific Industrial and Business Management, National University of Kaohsiung, Kaohsiung, Taiwan Dr Michiko Iizuka, Researcher, United Nations University-MERIT, Maastricht, The Netherlands and Visiting Fellow, SPRU – Science and Technology Policy Research, University of Sussex, UK Dr Keun Lee, Professor, Economics Department, Seoul National University, Director of Center for Economic Catch-up, Seoul, Korea Dr Yoon-Zi Kim, Senior Researcher, Overseas Economic Research Institute, The Export-Import Bank of Korea, Seoul Dr Fernando Perini, Program Officer, IPRU – International Development Research Centre and Visiting Fellow, SPRU – Science and Technology Policy Research, University of Sussex, UK vii
viii
Sectoral systems of innovation and production
Mr Janske van Eijck, MSc, General Manager, Diligent, Tanzania Dr Hannes Toivanen, VTT – Technical Research Center of Finland Dr Maria Barbosa Lima-Toivanen, Helsinki School of Economics, Finland Dr Rosane Argou Marques, Senior Advisor, Brazilian Agency for Industrial Development, Brasília, Brazil Luiz Guilherme de Oliveira, University of Brasília, Brazil
PART I
Introduction
1.
Sectoral systems of innovation and production in developing countries: an introduction Franco Malerba and Sunil Mani
1.
THE REASON FOR THIS BOOK
Sectoral systems of innovation and production have been a growing new area of research in industrial economics and the economics of innovation. This growth is due to two basic reasons. First, a sectoral system approach considers a wide range of factors that affect innovation and production in a sector. It places firms and the related capabilities and learning processes as the major drivers of innovation and production. At the same time it pays central attention to other relevant factors that affect innovation and production in a sector: the variety of actors, networks, demand and institutions. In particular, a sectoral system approach examines innovation as the result of both firms’ specific variables (such as firms’ learning and capabilities, R&D and production investments, strategies and organizational structure) and the type of knowledge and technologies that characterize a sector, the links and interdependencies with other related sectors, the role of actors (such as competitors, suppliers, users, universities, financial organizations, public agencies and the government), the characteristics of demand and the type of institutions (such as standards, regulations and norms). A second reason is that a sectoral system approach has a dynamic perspective and takes a process view. Thus it pays a lot of attention to exchange, competition, and cooperation in a coevolutionary setting. A major conclusion of the sectoral system approach is that all these factors and processes often differ from sector to sector and consequently have to be understood in their effects on innovation, diffusion and production. It must be noted that the dimensions of sectoral systems are not necessarily national: they may be also local or global. Therefore the approach calls for a deep understanding of the interplay between national systems and sectoral systems. A sectoral system approach for the study of innovation and production 3
4
Sectoral systems of innovation and production
in sectors, however, is not a straitjacket, but a broad, flexible and adaptable tool. It points to some key variables and fundamental relationships. Only the goals of the analysis will decide which levels of aggregation should be used, depending on the purpose of the analysis. This approach enables quantitative and qualitative comparative analyses across industries, countries and regions because it allows a focusing on the same set of variables. It also provides a framework for policy. Up to now, the work on sectoral systems has concerned mainly developed countries (see, for example, Malerba, 2004). There has been, however, an emerging interest in analyses of sectoral systems of innovation and production in developing countries. There are several reasons for that. Innovation and diffusion have become relevant in most developing countries. Processes of fast growth have been associated with some sectors such as automobiles, electronics and software as well as with the transformation of traditional sectors such as agriculture or food. But the differences across all these sectors in terms of structure and dynamics have been so great that a full understanding of these differences is necessary if innovation is to be encouraged and growth sustained. Therefore this book aims to answer questions such as the following: What are the main features of sectoral systems of innovation and production in developing countries? How do they change and evolve? What are the main policy lessons that one can draw from the analysis of sectoral systems? This book aims at answering these questions by examining a wide range of sectoral systems, from traditional to high technology ones. It does it for a variety of countries. The book originates from the contributions initially presented at the Globelics India Conference at Trivandrum in 2006. After the conferences the papers were completely revised and rewritten for this book. The book is composed of three parts. After an introduction to the volume by Franco Malerba and Sunil Mani, Part II examines the main actors and structure of some key sectoral systems in developing countries. Sunil Mani (Chapter 2) shows that in the same country (India) two different sectors may have a quite different performance because of the features of the specific sectoral systems. Then the book moves to examine key actors and features of sectoral systems in various sectors: Fernando Perini (Chapter 3) discusses networks and knowledge flows in ICT in Brazil; Hannes Toivanen and Maria Barbosa Lima Toivanen (Chapter 4) emphasize the role of private sector firms in the Brazilian pulp and paper industry; Marjolein Caniëls, Effie Kesidou and Henny Romijn (Chapter 5) point to the role of skills, entrepreneurship and clusters in software in Uruguay; Rosane Argou Marques and L. Guilherme de Oliveira (Chapter
Introduction
5
6) discuss the geographical boundaries of the aeronautical sector in Brazil. Part III of the book examines key aspects of the dynamics and evolution of sectoral systems. Patarapong Intarakumnerd and Mai Fujita (Chapter 7) point out that the same sector may evolve quite differently and examine the case of the motorcycle in Thailand and Vietnam; Michiko Iizuka (Chapter 8) illustrates how ‘low tech’ sectors can be highly dynamic in their path to development and examine salmon farming in Chile; Yoon-Zi Kim and Keun Lee (Chapter 9) discuss the role of interdependencies and demand in a key capital good industry such as machine tools in Korea; Ting-Lin Lee (Chapter 10) identifies the key role of two public actors in the evolution of the ICT industry in Taiwan, and finally Janske van Eijck and Henny Romijn (Chapter 11) examine the creation of a new sectoral system in a rural area in Tanzania. This introductory chapter is organized in the following way. Section 2 contains a general discussion of sectoral systems. In section 3 the major themes and points of the book are presented. Finally section 4 draws some conclusions emerging from the findings of this book.
2.
SECTORAL SYSTEMS: AN INTRODUCTION
A sectoral system framework focuses on the nature, structure, organization and dynamics of innovation and production in sectors. A sector can be broadly defined as a set of activities that are unified by some linked product groups for a given or emerging demand and that share some common knowledge. Firms in a sector have some commonalities and at the same time are heterogeneous in terms of learning processes and capabilities. A sectoral system has the following elements: (a) firms in the sector; (b) other actors (in addition to firms); (c) networks; (d) demand; (e) institutions; (f) knowledge; and (g) the basic processes of interaction, variety generation, selection and coevolution. (For a more general discussion see Malerba, 2002 and 2004.) The notion of sectoral systems has the evolutionary theory and the innovation system approach as building blocks. Evolutionary theory places a key emphasis on dynamics, innovation processes and economic transformation. Learning and knowledge are key elements in the change of the economic system. “Boundedly rational” agents act, learn and search in uncertain and changing environments. Agents know how to do things in different ways. Thus learning, knowledge and behaviour entail agents’ heterogeneity in experience and organization; and different competences affect persistent differential performance. In addition, evolutionary theory places emphasis on cognitive aspects such as beliefs,
6
Sectoral systems of innovation and production
objectives and expectations, which are in turn affected by previous learning and experience and by the environment in which agents act. A central place in the evolutionary approach is occupied by the processes of variety creation (in technologies, products, firms and organizations), replication (which generates inertia and continuity in the system) and selection (which reduces variety in the economic system and discourages the inefficient or ineffective utilization of resources). Finally, aggregate phenomena are emergent properties of far-from-equilibrium interactions and have a metastable nature (Nelson, 1995; Dosi, 1997; Metcalfe, 1998). For evolutionary theory the environment and conditions in which agents operate may drastically differ. Evolutionary theory stresses major sectoral differences in opportunities related to science and technologies. The same holds for the knowledge base underpinning innovative activities, as well as for the institutional context. Thus the learning, behaviour and capabilities of agents are constrained and “bounded” by the technology, knowledge base and institutional context. Heterogeneous firms facing similar technologies, searching around similar knowledge bases, undertaking similar production activities, and “embedded” in the same institutional setting share some common behavioural and organizational traits and develop a similar range of learning patterns. The notion of the sectoral system of innovation and production is also linked to the innovation system literature (Edquist, 1997) in that it focuses on learning and interaction among agents. It complements concepts such as national systems of innovation, which are delimited by national boundaries and focused on the role of non-firm organizations and institutions (Freeman, 1987; Lundvall, 1993; Nelson, 1993), regional/local innovation systems, in which the boundary is the region (Cooke et al., 1997), technological systems, in which the focus is on technologies and not on sectors (Hughes, 1987; Callon, 1992; Carlsson and Stankiewitz, 1995), and distributed innovation systems, in which the focus is on specific innovations (Andersen et al., 2002). As an introduction, let’s briefly discuss the main elements of a sectoral system in a general way: a.
Firms in the sector. Firms are the key actors in innovation and production in a sectoral system. They are characterized by specific learning processes, capabilites and organizational structures, as well as by beliefs, expectations and goals (Nelson and Winter, 1982; Teece and Pisano, 1994; Dosi et al., 2000). b. Other actors. In addition to firms, a sector is composed of other agents, which are organizations or individuals. Organizations may be suppliers, users, universities, financial institutions, government agencies, trade unions or technical associations. Individuals may be
Introduction
7
consumers, entrepreneurs or scientists. These agents are also characterized by specific learning processes, competencies, beliefs, objectives, organizational structures and behaviours. Agents interact through processes of communication, exchange, cooperation, competition and command. c. Networks. Within any sectoral system, firms are connected in various ways through market and non-market relationships. Traditional analyses of industrial organizations have examined agents as involved in processes of exchange, competition and command (such as vertical integration). In more recent analyses, processes of formal cooperation or informal interaction among firms or among firms and non-firm organizations have been examined in depth (as one may see from the literature on tacit or explicit collusion, hybrid governance forms, or formal R&D cooperation). This literature has analysed firms with certain market power, suppliers, users facing opportunistic behaviour or asset specificities in transaction, or firms with similar knowledge having appropriability and indivisibility problems in R&D. The evolutionary approach has emphasized that in uncertain and changing environments formal as well as informal networks emerge not because agents are similar but because they are different. Thus, networks integrate complementarities in knowledge, capabilities and specialization. Relationships between firms and non-firm organizations (such as universities and public research centres) have been a source of innovation and change in several sectoral systems: pharmaceuticals and biotechnology, information technology, and telecommunications (Nelson and Rosenberg, 1993). The types and structures of relationships and networks differ greatly from sectoral system to sectoral system, as a consequence of the features of the knowledge base, the relevant learning processes, the basic technologies, the characteristics of demand, the key links and the dynamic complementarities. d. Demand. In a sectoral system, demand may be domestic or international. Demand is not seen as an aggregate set of similar buyers or of atomistic undifferentiated customers, but as composed of heterogeneous agents who interact in various ways with producers. In this way, demand becomes composed of individual consumers, firms and public agencies, which could be part of different countries and national innovation systems, characterized by different size, knowledge, learning processes and competencies, and affected by different social factors and institutions. e. Institutions. Agents’ cognition, actions and interactions are shaped by institutions, which include norms, routines, common habits, established practices, rules, laws, standards and so on. Institutions may
8
f.
Sectoral systems of innovation and production
range from ones that bind or impose enforcements on agents to ones that are created by the interaction among agents (such as contracts), from more binding to less binding, and from formal to informal (such as patent laws or specific regulations as against traditions and conventions). A lot of institutions are national (such as the patent system), while others are specific to sectors (such as sectoral labour markets or sector-specific financial institutions). In all sectoral systems, institutions play a major role in affecting the rate of technological change, the organization of innovative activity, and performance. They may emerge either as a result of deliberate, planned decisions by firms or other organizations, or as the unpredicted consequence of agents’ interaction. Some institutions are sectoral (i.e. specific to a sector), while others are national, and others may be international. The relationship between national institutions and sectoral systems is quite important in most sectors. National institutions have different effects on sectors. For example, the patent system, property rights or antitrust regulations have different effects as a consequence of the different features of the sectoral systems, as surveys and empirical analyses have shown (see, for example, Levin et al., 1987). However, the same institution may take on different features in different countries, and thus may affect the same sectoral system differently. Often, the characteristics of national institutions favour specific sectors that fit better the specificities of the national institutions. Thus, in certain cases, some sectoral systems become predominant in a country because the existing institutions of that country provide an environment more suitable for certain types of sectors and not for others. In other cases, national institutions may constrain the development or innovation in specific sectors, or mismatches between national and sectoral institutions and agents may take place. The examples of the different types of interaction between national institutions and sectoral evolution in various advanced countries in Dosi and Malerba (1996) are cases in point. The relationship between national institutions and sectoral systems is not always one-way, as it is in the case of the effects of national institutions on sectoral variables. Sometimes, the direction is reversed, and goes from the sectoral to the national level. In fact, it may occur that the institutions of a sector, which are extremely important for a country in terms of employment, competitiveness or strategic relevance, end up emerging as national, thus becoming relevant for other sectors. But, in the process of becoming national, they may change some of their original distinctive features. The knowledge base. Any sector is characterized by a specific knowledge base, technologies and inputs. Knowledge plays a central role in
Introduction
9
innovation and affects the types of learning and capabilities of firms. In a dynamic way, the focus on knowledge and the technological domain places at the centre of the analysis the issue of sectoral boundaries, which usually are not fixed, but change over time. Knowledge is highly idiosyncratic at the firm level, does not diffuse automatically and freely among firms, and has to be absorbed by firms through their differential abilities accumulated over time. The evolutionary literature has proposed that sectors and technologies differ greatly in terms of the knowledge base and learning processes related to innovation. Knowledge differs across sectors in terms of domains. One knowledge domain refers to the specific scientific and technological fields at the base of innovative activities in a sector (Dosi, 1988; Nelson and Rosenberg, 1993), while another regards applications, users, and the demand for sectoral products. Recently, a major discontinuity has taken place in the processes of knowledge accumulation and distribution with the emergence of the knowledge-based economy, which has redefined existing sectoral boundaries, affected relationships among actors, reshaped the innovation process, and modified the links among sectors. What do we know about the main dimensions of knowledge? First, knowledge may have different degrees of accessibility (Malerba and Orsenigo, 2000), that is opportunities of gaining knowledge external to firms, which in turn may be internal or external to the sector. In both cases, greater accessibility of knowledge may decrease industrial concentration. Greater accessibility internal to the sector implies lower appropriability: competitors may gain knowledge about new products and processes and, if competent, imitate those new products and processes. Accessibility of knowledge that is external to the sector may be related to the levels and sources of scientific and technological opportunities. Here, the external environment may affect firms through human capital with a certain level and type of knowledge or through scientific and technological knowledge developed in firms or non-firm organizations, such as universities or research laboratories. Knowledge may be more or less cumulative, that is the degree by which the generation of new knowledge builds upon current knowledge. One can identify three different sources of cumulativeness. The first source is cognitive. The learning processes and past knowledge constrain current research, but also generate new questions and new knowledge. The second source is related to the firm and to its organizational capabilities. Organizational capabilities are firm-specific and generate knowledge which is highly path-dependent. They implicitly define what a firm learns and what it can hope to achieve in the future.
10
Sectoral systems of innovation and production
A third source is the feedback from the market, such as in the “success-breeds-success” process. Innovative success yields profits that can be reinvested in R&D, thereby increasing the probability of innovating again. In the case of knowledge spillovers within an industry, however, it is also possible to observe cumulativeness at the sectoral level. Cumulativeness may also be present at the local level. In this case, high cumulativeness within specific locations is more likely to be associated with low appropriability conditions and spatially localized knowledge spillovers. The sources of technological opportunities markedly differ among sectors. As Freeman (1982) and Rosenberg (1982), among others, have shown, in some sectors opportunity conditions are related to major scientific breakthroughs in universities. In other sectors, opportunities to innovate may often come from advancements in R&D, equipment and instrumentation. In still other sectors, external sources of knowledge in terms of suppliers or users may play a crucial role. Not all external knowledge may be easily used and transformed into new artefacts. If external knowledge is easily accessible, transformable into new artefacts and exposed to a lot of actors (such as customers or suppliers), then innovative entry may take place. If advanced integration capabilities are necessary (Cohen and Levinthal, 1989), the industry may be concentrated and formed by large, established firms. Knowledge affects also the types of learning processes and the relevant capabilities that firms have in order to be competitive and innovate. In general, the features and sources of knowledge affect the rate and direction of technological change, the organization of innovative and production activities, and the factors at the base of firms’ successful performance. The boundaries of sectoral systems are affected by the knowledge base and technologies, as well as by the type of demand and links and complementarities among artefacts and activities. These links and complementarities are, first of all, of the static type, as are input–output links. Then there are dynamic complementarities, which take into account interdependencies and feedbacks, both at the demand and at the production levels. Dynamic complementarities among artefacts and activities are major sources of transformation and growth of sectoral systems, and may set in motion virtuous cycles of innovation and change. This could be related to the concept of filière and the notion of development blocks (Dahmen, 1989). Links and complementarities change over time and greatly affect a wide variety of variables of a sectoral system: firms’ strategies, organization and performance, the rate and direction of technological change, the type of competition and the
Introduction
11
networks among agents. Thus the boundaries of sectoral systems may change more or less rapidly over time, as a consequence of dynamic processes related to the transformation of knowledge, the evolution and convergence in demand, changes in competition and learning by firms. g. The main processes and coevolution. The analysis of sectoral systems requires also a careful understanding of the processes of interaction, cooperation and competition. In a sectoral system framework, innovation is considered to be a process that involves systematic interactions among a wide variety of actors for the generation and exchange of knowledge relevant to innovation and its commercialization. Interactions include market and non-market relations that are broader than the market for technological licensing and knowledge, inter-firm alliances, and formal networks of firms. Over time, a sectoral system undergoes processes of change and transformation through the coevolution of its various elements. This process involves technology, demand, knowledge base, learning processes, firms, non-firm organizations and institutions. Nelson (1994) and Metcalfe (1998) have discussed these processes at the general level by focusing on the interaction between technology, industrial structure, institutions and demand. The claim here is that these processes are sector-specific. For example, just looking at three elements such as technology, demand and firms, in sectors characterized by a system product and consumers with a rather homogeneous demand, coevolution leads to the emergence of a dominant design and industrial concentration (Klepper, 1996). However, in sectors with a heterogeneous demand, specialized products and a more fragmented market structure may emerge. Often coevolution is related to path-dependent processes (David, 1985; Arthur, 1989). Here local learning, interactions among agents and networks may generate increasing returns and irreversibilities that may lock sectoral systems into inferior technologies. h. Three last introductory points. Three last points on sectoral systems have to be made here by way of introduction. First, what are the main differences between a sectoral innovation system and a national innovation system perspective? While national innovation systems take innovation systems as delimited more or less clearly by national boundaries, a sectoral system approach would claim that the boundaries of the innovations process in sectors have local, national and/ or global dimensions. Often these three different dimensions coexist in a sector. In addition, national innovation systems result from the different composition of sectors, some of which are so important that they drive the growth of the national economy. For example, Japanese
12
Sectoral systems of innovation and production
growth in the 1970s and 1980s was driven by specific sectors, which were different from the sectors behind the American “resurgence” during the 1990s. As has been pointed out previously, an understanding of the key driving sectors of an economy with their specificities greatly helps in understanding national growth and national patterns of innovative activities. Second, a relevant remark refers to the aggregation issue regarding products, agents or functions. For example, sectoral systems may be examined broadly or narrowly (for example, in terms of a small set of product groups). A broad definition allows us to capture all the interdependencies and linkages in the transformation of sectors, while a narrow definition identifies more clearly specific relationships. Of course, within broad sectoral systems, different innovation systems related to different product groups may coexist. The choice of the level of aggregation depends on the goal of the analysis. Third, a sectoral system perspective should not be seen as a rigid and closed framework, but as a broad, open and flexible framework, able to encompass different elements and variables, according to the focus of the analysis. However, the driving elements of the analysis still have to be knowledge, capabilities, variety of actors, interactions and institutions.
3. THE MAJOR THEMES AND POINTS OF THIS BOOK The chapters in this book identify several relevant aspects of sectoral systems in developing countries that are key for understanding innovation, competitiveness and growth in these countries. They could be grouped in two major parts, which constitute the two parts of this book. 3.1
Understanding the Key Actors and the Main Characteristics of Sectoral Systems and their Effects on Innovation and Developments (PART II)
Some major points emerge from the chapters included in Part II: ●
In the same country two technology-intensive sectors may end up having a quite different innovative and competitive performance owing to the different structure of their sectoral systems. The case of pharmaceuticals and telecommunication equipment in India. This first point is examined by Sunil Mani in “Why is the Indian pharmaceutical industry more innovative than its telecom-
Introduction
13
munications equipment industry?” Mani starts from the remark that in India two sectors which are highly technology-intensive – such as pharmaceuticals and telecommunications equipment – have had quite different innovative and competitive performance owing to the different structure of the two sectoral systems. In both pharmaceuticals and telecom equipment the Indian government intervened early on through essentially the creation of important supporting institutions and instruments. But the innovative performance of both industries has been different: the drug industry has become self-sufficient, has emerged as a net exporter and has a strong patenting record abroad, while the telecommunications industry has increasingly become dependent on MNCs and imports, and the industry does not have many patents to boast of. The differences in outcomes is explained by the differences in the sectoral systems of innovation. In pharmaceuticals, the innovation system of the industry has three strong pillars: private sector enterprises which have invested in innovation, a very proactive government policy regime especially with respect to intellectual property rights, and strong government research institutes. The TRIPS compliance of the intellectual property rights regime making it mandatory for pharmaceutical products to be patented has not reduced the innovation capability of the industry, although it has not made it work on R&D projects that may lead to the discovery of drugs for neglected diseases of the developing world. However, the two main components of the innovation system, namely the enterprises and the government research institutes, do not appear to have all the requisite capabilities to bring a new drug to the market. This is an area where public policy support is still required. In telecommunication equipment, on the contrary, India followed a very rigid policy of indigenous development of domestic technologies by establishing a stand-alone public laboratory that developed state-of-the-art switching technologies. These were then transferred to manufacturing enterprises in both public and private sectors. The domestic enterprises themselves did not have any in-house R&D capability. As a consequence the enterprises were completely dependent on the public laboratory for research. Most of the enterprises constituting the telecom equipment sectoral systems have become mere traders, distributing products manufactured elsewhere. Thus the country, despite possessing good-quality human resources, was unable to keep pace with changes in the technology frontier, and the equipment industry has now become essentially dominated by affiliates of MNCs and by imports.
14
Sectoral systems of innovation and production ●
●
The knowledge base of a sector greatly affects the organization of innovative activity and the type of networks. ICT in Brazil. Fernando Perini, in “From innovation projects to knowledge networks: knowledge as contingency in the sectoral organization of innovation”, examines the relationship between the sectoral knowledge base and the organization of innovative activity in ICT in Brazil. In particular, the type of knowledge base affects the balance between hierarchies and market, the type of governance mechanisms and the inter-organizational channels of knowledge flows. Looking at ICT, strong tie networks are formed in software and middleware, weak tie networks are formed in training, technological services and research, and very limited networks are present in semiconductors, production processes and hardware. In general, early initiatives in product development are associated with higher internalization levels. Different types of organizations did different things. Domestic companies remained focused in hardware and middleware (close to manufacturing activities), while multinational companies connected to private research institutes were important in the emerging software technology. Public research centres and educational institutions became central in training and research activities. The different roles of these organizations reinforce the importance of diversity in governance structures and the different mechanisms for interaction between public and private as well as domestic and multinational stakeholders inside the sectoral systems. The low correlation between the networks in different activities, however, shows that different types of knowledge tended to flow in distinct communities of practice. There were, however, some connections between the types of activities. In particular, three distinct strong connections emerged between different communities: the first took place between production process and laboratorial infrastructure/equipments on one side and training on the other, the second between research and technological services, and the third between product development activities in middleware and software. Private sector enterprises are key actors in a sectoral system of innovation. The pulp and paper industry in Brazil. This is the message that one gets from the chapter by Hannes Toivanen and Maria Barbosa Lima-Toivanen, ‘Learning, innovation and public policy: the emergence of the Brazilian pulp and paper industry’. Brazil occupies an important position among the major world producers of pulp and paper: in 2007 it was the world’s sixth largest pulp producer and the eleventh largest paper producer. In 2006, the industry’s total exports were about US$4.7 billion.
Introduction
●
●
15
It is generally considered to be a successful case. One of the most important agents in the sectoral system is the private sector firms. At present there are 220 companies spread throughout 450 municipalities and located in 17 states and five regions. Public policies did not hamper firms’ competitiveness in the market. Rather, entrepreneurs and business managers enjoyed, and operated under, healthy, internationally competitive incentives for the creation and adoption of scientific, technological and business innovations. Advanced human capital, vibrant entrepreneurship and intense spinoffs are at the heart of the clustering of innovative activities in skillintensive sectors. Software in Uruguay. Marjolein C.J. Caniëls, Effie Kesidou and Henny Romijn, in ‘The software sector in Uruguay: a sectoral systems of innovation perspective’, examine the key role of local skills and clustering of innovative activity in a sector such as software in Uruguay. The sector has undergone a high growth since its inception in the early 1990s. Important factors leading to the emergence of the sector have been favourable demand conditions in Latin America, and the presence of skilled manpower, which can be related to an emphasis on education by the Uruguayan state. The sector is highly clustered geographically, and consists predominantly of a multitude of small and medium-sized firms, indicating that knowledge in the software sector is cumulative mainly at the local cluster level. The sector grew by fulfilling increasing local demand for enterprise resource planning products for small and medium-sized firms, which were overlooked by multinational companies. Uruguayan software companies combined advanced technological knowledge with knowledge about specific markets and applications such as banking, finance, education, health and construction. Successful products were addressed to specific applications or specific customers and incorporated new technological developments. The sectoral system has developed owing to the presence of skilled workers with a good level of education and has grown over time through intense entrepreneurship, spin-offs and labour mobility. Firms learn through internal efforts (R&D) but also access external knowledge and information through networking. In this sectoral system no major role of policy has been present except the one concerning human capital formation and the creation of a good level of general infrastructure. Direct promotion of the sector through public policies and institutional support has not played a major role in the emergence of the sector. Sectoral systems of innovation need not be confined to national borders, but can in fact be global, and therefore the interactions they
16
Sectoral systems of innovation and production
may have with local actors may diminish over time. The aeronautical sector in Brazil. Rosane Argou Marques and L. Guilherme de Oliveira in their chapter on ‘Sectoral system of innovation in Brazil: reflections about the accumulation of technological capabilities in the aeronautical sector (1990–2000)’ advance the point that a sectoral system of innovation has boundaries extending beyond the region and the nation and in fact it very often extends abroad to foreign locations. This is demonstrated with an examination of the Brazilian aircraft manufacturing sector. This sector is dominated by the Brazilian aircraft manufacturer Embraer. The firm was established in 1969 by the Brazilian military government, and it was a state-owned undertaking until 1994 when it was privatized. Although it is successful in achieving international competitiveness in its specific market segment for regional jets, Brazil has not been able to consolidate the supply chain of Embraer within its national borders. There are now only a few Brazilian firms supplying Embraer and some of the foreign first-tier suppliers of Embraer. In fact, the import content increased from about 68 per cent in the 1980s to approximately 95 per cent in the 1990s. Therefore, there is a question about how the local Brazilian suppliers are maintaining themselves in the competitive supply chain of Embraer. The findings of the present analysis show that the local suppliers are improving their innovative capabilities in two directions: by strengthening their basic technological capability regarding production processes and, in a few cases, by upgrading to intermediate and advanced levels of innovative capability. The relationship with Embraer, foreign buyers and Brazilian research institutions are the main sources of knowledge for the technological learning experienced by these Brazilian suppliers that are “surviving” in the supply chain of Embraer. 3.2
The Dynamics and Evolution of Sectoral Systems (PART II)
In this part one can also identify some major points: ●
The same sectors can evolve differently when they are facing similar threats and opportunities. The motorcycle industry in Thailand and Vietnam. Patarapong Intarakumnerd and Mai Fujita in their chapter on ‘China’s threat and opportunity for the Thai and Vietnamese motorcycle industries: a sectoral innovation system analysis’ trace the evolution of the sectoral system of innovation of the motorcycle
Introduction
●
17
industry in both Thailand and Vietnam. The findings illustrate that different sectoral systems of innovation and production evolve differently. The direction and the pace of evolution depend very much on existing absorptive capabilities of agents, strength of their linkages and their process of collective learning to withstand the threats and exploit the opportunities. The industry in both the countries had to face strong import competition from China. The type of response of the sectoral system depended very much on the nature of the economic agents. To illustrate, Thailand can withstand the threats and exploit the opportunities better than Vietnam because, despite still being rather weak and fragmented, its motorcycle sectoral system of innovation and production has relatively more capable agents (i.e. longer-present and more technologically sophisticated TNCs, local champions who are own-brand manufacturers), a government with more vivid and targeted strategies for the automotive sector, more active support of agencies, universities and research institutes, more sophisticated demand conditions, and relatively more interaction (especially knowledge transfer) among agents. The so-called traditional (or low-tech) sectors may change over time quite drastically, and become increasingly knowledge-intensive and innovative. Salmon farming in Chile. Michiko Iizuka, in ‘“Low-tech” industry: a new path for development? The case of the salmon farming industry in Chile’, challenges the view that the ‘low-tech’ sectors – such as food and other natural-resource-based industries – are not dynamic or innovative enough to be the path to development and can only be regarded as a transitional phase to ‘high-tech’ sectors, especially in manufacturing. The chapter clearly shows that the so-called low-tech sectors can be innovative, and have undergone a major transformation that requires advanced capabilities – in particular, of combining existing technological, organizational and market knowledge from different technological domains. In these sectors the innovation process involves wide networks extending beyond national boundaries in order to encourage dynamic interactions in aligning the interests of agents, which subsequently redefine the sectoral boundaries. As an example of a case of transformation of “low-tech sectors”, the Chilean salmon farming industry has shown successful development and reached world leadership in a premium natural-resource-based product. Here, this sector evolved over time by successfully using and combining advanced knowledge. The majority of firms engaged in cumulative and architectural innovation owing to the structural changes that took place in the sector. This made them interdependent
18
Sectoral systems of innovation and production
●
on one another and forced them to perform as a cluster. Some key institutions shaped the agents and interactions among them. In particular an intermediate institution – the Association of the Salmon Industry – expanded its network to broaden its knowledge base so as to enhance negotiating power at the sectoral level. Vertical interdependencies and local demand may impair the full development of a sector. Machine tools in Korea. Yoon-Zi Kim and Keun Lee, in ‘Making a technological catchup in the capital goods industry: barriers and opportunities in the Korean case’, discuss the role of interdependencies and demand in a key capital good industry such as machine tools. The point of the chapter is that growth and catch-up are difficult in capital goods industries usually characterized by small or middle-sized companies, even in countries with a high growth rate such as Korea. This is so because the vertical interaction and exchange of knowledge between advanced final producers and customers’ firms are very important and may play against the development of the sector. The reason is that small firms in the capital goods industry are usually specialized suppliers to big final goods assembly firms in the consumer goods industry or other industries. Local client firms are reluctant to use locally made capital goods owing to their poor quality and low precision levels, so that dynamic processes of learning and capability formation by local capital goods producers cannot be set in motion. In addition, two other reasons may be present. First, while a successful catch-up first requires the ability to produce goods of better quality and lower prices than those produced by incumbent firms from advanced countries, incumbent foreign firms often react by charging predatory prices upon possible local development of capital goods by latecomers. In addition, if the catch-up firms overcome this barrier, then the next strategy used by incumbent firms is to charge latecomers with legal actions for patent violations. Despite these difficulties, the Korean economy has achieved a slow but gradual catch-up in the capital goods industry. The chapter attributes such incremental achievements to several factors, including the strenuous efforts of the government to support joint companies with a foreign partner, the creation of a certification of credibility by the government regarding the quality of the product R&D support, the profiting from the possibility of catching up given by the introduction and adoption of IT or digital technologies in machine tools, and the focus on niche markets in general-purpose machine tools and emerging economies in which limited interaction between producers and users is required.
Introduction ●
19
Sectoral systems of innovation are embedded in the national system of innovation, and their evolution is both nurtured and hampered by the government. ICT in Taiwan. Ting-Lin Lee, in her chapter ‘From “nuts and bolts” to “bits and bytes”: the evolution of Taiwan ICT in a global knowledge-based economy’, traces the evolution of the ICT industry in Taiwan. The ICT sectoral system was composed of two dominant agents, the government and foreign companies. However, the sectoral system of innovation of the Taiwanese semiconductor industry is largely shaped by two government research institutes, namely the Industrial Technology Research Institute (ITRI) and the Institute for Information Industry (III). The ITRI was established in 1973 by the Ministry of Economic Affairs (MoEA) as a national laboratory with the responsibility of developing technical capabilities for high-tech industries that the government had marked as strategic for Taiwan’s development. As the ITRI undertook applied research, it accelerated industrial development by working closely with the private sector to ensure technology transfer from developed nations to domestic industries. With sponsorships from the MoEA, the ITRI conducted research that was shared with private firms. Also, the Institute conducted specific research for individual firms on a contract basis and formed R&D collaborations to update Taiwanese firms on best practices from technologies around the globe. Equally, since 1979, the III has been a key technology contributor to Taiwan’s ICT industry. Its founding and continuing mission has been to increase Taiwan’s global competitiveness through the development of its IT infrastructure and industry. In order to support leading-edge research and speed up the pace of innovative breakthroughs, Taiwan has established a series of opentype national laboratories. The Taiwanese government has also promoted cooperation between industries and universities in recent years. Despite the significant role of government in the creation and nurturing of the semiconductor industry, the government realized that real prosperity is created by the private sector and so promoted a good balance of competition and collaboration among private firms to increase the competitiveness and productivity of the cluster. Also, financial organizations, universities and supporting industries contributed meaningfully to the development of the ICT industry. Taiwan’s “miraculous growth” shows that government can play a leading role in the development of a specific sector but it must understand its role and should not overstep its boundaries to hinder other agents from playing their roles effectively.
20
Sectoral systems of innovation and production ●
Creating and nurturing sectoral systems of innovation in new technology-based industries in the rural areas of the developing world can be both difficult and complex. Biofuels in Tanzania. This is the main message of the chapter ‘Prospects for Jatropha biofuels in Tanzania: an analysis with strategic niche management’ by Janske van Eijck and Henny Romijn. The high petroleum prices have forced the search for more renewable forms of energy. One such alternative source of diesel is from a plant called Jatropha. It will help meet the world’s demand for fuel, without crowding out the world’s supply of food, and it will regenerate dry and denuded soils and create jobs for impoverished farmers. Tanzania is one among a few developing countries that is cultivating Jatropha and then extracting bio-diesel out of it. In this context the authors raise the following questions. To what extent does Tanzania have the potential to develop Jatropha-based energy supply sources, and to what extent could these successfully substitute for fossil-based energy supplies? What has been its progress in that direction? And what are the major obstacles in this respect? These questions are then answered by mapping out the emerging national innovation system for Jatropha biofuels. The authors do this with an approach called strategic niche management (SNM), which is rooted in evolutionary economics. SNM has been designed specifically to study the introduction of radically new technologies in society that promise to contribute to more sustainable development patterns. Essentially, SNM conceptualizes the introduction of such technologies as the start of a broad and long-term transition process, in which widely used technologies with unsustainable characteristics are gradually replaced by new, cleaner technologies. SNM posits that this process encompasses a coevolution of technology and societal factors such as culture, institutions, consumption patterns and preferences, economic regulation, and political governance systems. The main conclusion of this chapter is that the innovation system for Jatropha biofuels in Tanzania is still in its infancy, and that its future is far from being clear. Despite the favourable constellation of many contextual ‘landscape’ factors, there remain prominent barriers within Tanzania’s existing energy regime and agricultural regime. These include structural, infrastructural and logistical problems; technical skill and knowledge gaps; a limited local research infrastructure; the vested interests of powerful actors in the extant energy regime; cultural barriers associated with traditional uses of Jatropha; psychological obstacles emanating from known poisonous qualities of the crop; and a considerable price disadvantage for Jatropha oil except in remote locations.
Introduction
4.
21
SOME CONCLUSIONS
Some conclusions on examining industries in a sectoral perspective in developing countries and on the public policy of support for innovation emerge from this book: a.
Understanding the specificities of the relevant sectoral systems is fundamental in order to identify the sources of innovation and development. This book has pointed to factors such as the knowledge base underpinning innovative and production activities in a sector, the vertical and horizontal interdependencies across sectors, the various actors involved in innovative and production activities, the networks that they form, the characteristics of demand and the relevant institutions. In a dynamic fashion, this requires also the identification of the specific coevolutionary process in which the sector is involved. Most chapters in this book clearly illustrate this point. b. The awareness of the key differences existing across sectoral systems allows an understanding of why some factors affect innovation in some sectors and not in others and why some policies have a big impact in some sectors and a weak one in others. Again, most chapters in this book provide good examples. c. The separation of research from development and production capabilities can be very harmful for innovation and development. The strong separation of research capabilities from production capabilities may lead to companies without any competence in production on one hand or to companies that rely too much on external research but do not have any internal capabilities on the other. As Chapter 2 shows, the two have to be integrated for successful innovation. d. The type of networks that emerge in a sector is strongly associated with the specific knowledge base of that sector. Simply associating public knowledge with decentralized networks and private knowledge with hierarchical networks is too simplistic a statement. As Chapter 3 shows, an understanding of the type of knowledge involved is necessary for an understanding of why networks of certain types are present in the development of a sector. e. The formation of networks and knowledge systems in developing countries may require complex alignment in a multi-level governance structure, including multinational corporation networks and domestic networks. As shown in Chapters 3 and 6, multinational companies may play a fundamental role in the accumulation of technological capabilities in new technologies, but their integration with existing networks and the diffusion of knowledge into existing social
22
Sectoral systems of innovation and production
structures are not a homogeneous process. Specific institutional frameworks may allow for organizational learning and the decentralized interaction between stakeholders with very different interests. f. In knowledge-intensive sectors, skills and human capital formation are particularly relevant for growth, as Chapter 5 shows. g. A vibrant entrepreneurship and intense spin-offs are extremely important for innovation in sectors characterized by small and medium enterprises. As Chapter 5 illustrates, intense dynamic clustering is the result of these processes. h. Public policy has to pay attention to the positive feedbacks but also to the blocking role that links and interdependencies between different sectors have on innovation and development. While these interdependencies often work in a positive way owing to local user–producer interactions, at other times the toughness of international competition at the final product level may block the development of an indigenous advanced capital goods industry because local final producers may decide to import advanced equipment in order to be internationally competitive, as Chapter 9 shows. i. Traditional sectors are not necessarily low-tech or do not necessarily have low knowledge intensity; often they are innovative and they increasingly require the use and integration of advanced and differentiated knowledge. As Chapter 8 illustrates, some of these sectors have been radically transformed recently, are now highly knowledge-intensive and drive development. j. Key institutions such as industrial associations may play a key role in coordination and networking mechanisms. The salmon farming case in Chapter 8 is a clear example. Chapter 4’s pulp and paper story is still another example. k. Depending on the instruments, timing and sectoral context, government can play the role of both a facilitator and an obstacle to the creation of a sectoral system of innovation. The former comes out of the analysis by Chapter 10’s study of the Taiwanese ICT industry, and the latter follows very clearly from Chapter 11.
REFERENCES Andersen, B., Metcalfe, J.S. and Tether, B.S. (2002), “Distributed innovation systems and instituted economic processes”, Working Paper ESSY (http://www. cespri.unibocconi.it/essy/wp/metcalfeetal.pdf). Arthur, B. (1989), “Competing technologies, increasing returns and lock-ins by historical events”, Economic Journal 99(394), pp. 116–131.
Introduction
23
Callon, M. (1992), “The dynamics of techno-economic networks”, in Coombs, A.R.R., Saviotti, P. and Walsh V. (eds), Technological Change and Company Strategies: Economic and Sociological Perspectives, Academic Press, London. Carlsson, B. and Stankiewitz, R. (1995), “On the nature, function and composition of technological systems”, in Carlsson, B. (ed.), Technological Systems and Economic Performance: The Case of Factory Automation, Kluwer Academic Publishers, Dordrecht. Cohen, W. and Levinthal, D. (1989), “Innovation and learning: the two faces of R&D”, Economic Journal 99(397), pp. 569–596. Cooke, P., Urange, M.G. and Extebarria, G. (1997), “Regional innovation systems: institutional and organizational dimensions”, Research Policy 26, pp. 475–491. Dahmen, E. (1989), “Development blocks in industrial economics”, in B. Carlsson (ed.), Industrial Dynamics, Kluwer Academic Publishers, Dordrecht. David, P. (1985), Clio and the economics of QWERTY, American Economic Review 75(2), pp. 332–337. Dosi, G. (1988), “Sources, procedures and microeconomic effects of innovation”, Journal of Economic Literature 26(3), pp. 1120–1171. Dosi, G. (1997), “Opportunities, incentives and the collective patterns of technological change”, Economic Journal 107(444), pp. 1530–1547. Dosi, G. and Malerba, F. (1996), Organization and Strategy in the Evolution of the Enterprise, Macmillan, London. Dosi, G., Nelson, R.R. and Winter, S.G. (2000), The Nature and Dynamics of Organizational Capabilities, Oxford University Press, New York. Edquist, C. (ed.) (1997), Systems of Innovation: Technologies, Institutions and Organisations, Frances Pinter, London. Freeman, C. (1982), The Economics of Industrial Innovation, Frances Pinter, London. Freeman, C. (1987), Technology Policy and Economic Performance: Lessons from Japan, Frances Pinter, London. Hughes, T.P. (1987), “The evolution of large technological systems”, in Bijker, W.E., Hughes, T.P. and Pinch, T.J. (ed.), The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, MIT Press, Cambridge, MA. Klepper, S. (1996), “Entry, exit, growth and innovation over the product life cycle”, American Economic Review 86(3), pp. 562–583. Levin, R.C., Klevorick, A.K., Nelson, R.R. and Winter, S.G. (1987), “Appropriating the returns from industrial R&D”, Brookings Papers on Economic Activity 3, pp. 783–831. Lundvall, B.A. (1993), National Systems of Innovation, Frances Pinter, London. Malerba, F. (2002), “Sectoral systems of innovation and production”, Research Policy 31(2), pp. 247–264. Malerba, F. (2004), Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe, Cambridge University Press, Cambridge. Malerba, F. and Orsenigo, L. (2000), “Knowledge, innovative activities and industry evolution”, Industrial and Corporate Change 9(2), pp. 289–314. Metcalfe, S. (1998), Evolutionary Economics and Creative Destruction, Routledge, London. Nelson, R.R. (1993), National Innovation Systems: A Comparative Study, Oxford University Press, Oxford.
24
Sectoral systems of innovation and production
Nelson, R.R. (1994), “The co-evolution of technology, industrial structure and supporting institutions”, Industrial and Corporate Change 3(1), pp. 47–64. Nelson, R.R. (1995), “Recent evolutionary theorizing about economic change”, Journal of Economic Literature 33(1), pp. 48–90. Nelson, R.R. and Rosenberg, N. (1993), “Technical innovation and national systems”, in Nelson, R.R. (ed.), National Innovation Systems, Oxford University Press, Oxford. Nelson, R.R. and Winter, S.G. (1982), An Evolutionary Theory of Economic Change, Belknap Press of Harvard University Press, Cambridge, MA. Rosenberg, N. (1982), Inside the Black Box, Cambridge University Press, Cambridge. Teece, D. and Pisano, G. (1994), “The dynamic capabilities of firms: an introduction”, Industrial and Corporate Change 3(3), pp. 537–556.
PART II
Actors and structure of sectoral systems in developing countries
2.
Why is the Indian pharmaceutical industry more innovative than its telecommunications equipment industry? Contrasts between the sectoral systems of innovation of the Indian pharmaceutical and telecommunications industries Sunil Mani
The pharmaceutical and telecom equipment industries are two hightechnology industries. The government has intervened in the establishment of both these industries through essentially the creation of important supporting institutions and instruments. The innovative performance of both the industries has been different. In very specific terms the pharmaceutical industry is more innovative. An explanation for this differential performance in innovation is to be found in the nature and composition of the sectoral system of innovation of the two industries. The chapter thus highlights this important explanatory power of the framework of analysis for designing public policy to promote innovations.
INTRODUCTION India pursued a policy of economic growth with technological self-reliance right through the 1950s when it embarked on a planned form of development. Two industries that were targeted especially were the manufacturing of drugs and telecommunications equipment. However, the final outcomes have been very different. The drug industry has become self-sufficient, has emerged as a net exporter and has a strong patenting record abroad, while the telecommunications industry has increasingly become dependent on MNCs and imports and the industry does not have many patents to boast of. I argue that, although the broad external environments pertaining to 27
28
Sectoral systems of innovation and production
both industries are roughly similar, the differences in outcomes could be explained by the differences in their sectoral systems of innovation (SSI). The SSI of the pharmaceutical industry presents an ideal picture where private sector business enterprises occupy the central stage. They have been supported very well by a conducive intellectual property regime, which enabled them to reverse-engineer known technologies and thereby emerge as incremental innovators. On the contrary, a public laboratory dominates the SSI of the telecoms industry, and the production enterprises in the system did not have much innovation capability: the enterprises were completely dependent on the public laboratory. The government too did not support the laboratory adequately, and very often the public technology procurement – the main instrument pursued by the state supposedly to support domestic technology generation through the laboratory – was actually against it. Consequently most of the enterprises constituting the SSI of the telecoms industry have become mere traders, distributing products manufactured elsewhere. The analysis thus brings out important but practical policy prescriptions. The chapter is structured into four sections. Section 1 will outline the framework for analysis, which is essentially the sectoral system of innovation propounded by Malerba (2004). The second section provides some numbers on the innovative performance of the two industries. The third section undertakes a detailed mapping out of the sectoral system of innovation of two high-tech industries in the Indian context, namely the pharmaceutical and telecom equipment industries. India’s pharmaceutical industry is one of the most innovative industries in the country’s manufacturing sector. The innovation system of the industry has three strong pillars: a very proactive government policy regime especially with respect to intellectual property rights, strong government research institutes, and private sector enterprises which have invested in innovation. The TRIPS compliance of the intellectual property right regime making it mandatory for pharmaceutical products to be patented has not reduced the innovation capability of the industry, although it has not made it work on R&D projects that may lead to the discovery of drugs for the neglected diseases of the developing world. Although the innovation system has the capability to develop new chemical entities, the two main components of the innovation system, namely the enterprises and the government research institutes, do not appear to have all the requisite capabilities to bring a new drug to the market. Although the state has been very proactive with respect to this industry, this is an area where public policy support is still required. On the contrary in the case of the telecom equipment industry, India followed a very rigid policy of indigenous development of domestic technologies by establishing a stand-alone public laboratory that
Contrasts between Indian sectoral systems of innovation
29
developed state-of-the-art switching technologies. These were then transferred to manufacturing enterprises in both public and private sectors. The enterprises themselves did not have any in-house R&D capability. The public laboratory was also not given any strategic direction, even though it was, technologically speaking, very competent. Consequently the country, despite possessing good-quality human resources, was unable to keep pace with changes in the technology frontier, and the equipment industry has now become essentially dominated by affiliates of MNCs and by imports. The fourth and last section of the chapter contrasts the two SSIs within the same national system of innovation and draws out the policy implications.
1.
THE FRAMEWORK OF ANALYSIS
The paper adopts a sectoral system of innovation (SSI) perspective introduced by Malerba (2004). The framework involves mapping out the boundaries of the innovation system in terms of the specific agencies of the government dealing with telecommunications development, the policy framework, the equipment suppliers, the service providers and the regulatory agency and tracking the knowledge flows between these various actors within the system. According to Malerba (2004), every sectoral system of innovation has at least three blocks: (i) knowledge, technological domain, and boundaries; (ii) actors, relationships and networks; and (iii) institutions. These three blocks may be elaborated as follows. First, knowledge plays a central role in innovation. It has to be absorbed by firms through their differential abilities accumulated over time. Knowledge differs across sectors in terms of domains. One knowledge domain refers to the specific scientific and technological fields at the base of innovative activities in a sector. The boundaries of sectoral systems are affected by the knowledge base and technologies. Second, sectoral systems are composed of heterogeneous actors. Firms are the key actors in the generation, adoption and use of new technologies. Actors also include users and suppliers, who have different types of relationships with the innovating, producing or selling firms. Other types of agents in a sectoral system are non-firm organizations, government agencies, local authorities and so on. In various ways, they support innovation, technological diffusion, and production by firms, but again their role greatly differs among sectoral systems. Third, in all sectoral systems, institutions play a major role in affecting the rate of technological change, the organization of innovative activity and performance. Innovation greatly differs across sectors in terms of sources, actors, features, boundaries and organization.
30
Sectoral systems of innovation and production
2. THE INNOVATIVE PERFORMANCE OF INDIA’S PHARMACEUTICAL AND TELECOM INDUSTRIES The pharmaceutical industry is more successful from the innovation point of view (Table 2.1). Three indicators are used: (i) trade balance; (ii) R&D expenditure; and (iii) number of US patents granted. On all the three indicators, the pharmaceutical industry has done much better even though both the industries received more or less equal amounts of state support. Note that, while all the pharmaceutical patents have been secured by Indian companies, almost all the telecom patents have been secured by affiliates of MNCs operating from India, based of course on R&D projects performed in India. My argument is that the relative better performance of the pharmaceutical industry may to a large extent be explained in terms of its sectoral system of innovation. This is because if research is done by universities and other government research institutes and if production is Table 2.1
Innovative performance of India’s pharmaceutical and telecommunications equipment industries Pharmaceutical industry Trade balance (millions of US$)
1990–91 1991–92 1992–93 1993–94 1994–95 1995–96 1996–97 1997–98 1998–99 1999–2000 2000–01 2001–02 2002–03 2003–04 2004–05 2005–06 Note: Source:
304.1 402.2 248.5 382.8 501.6 613.0 916.2 1068.8 1103.1 1295.5 1542.3 1637.2 2058.3 2473.4
R&D intensity (percentage)
Telecommunications equipment industry Number of US patents granted
1.78 2.42 2.98 2.80 3.06 3.10 3.17 3.88 3.86 4.72 5.79 7.83 8.79
56* 36 46 48 72 44 54 70
Trade balance (millions of US$) –71.77 –130.38 –97.38 –167.06 –212.70 –117.128 –217.796 –257.116 –302.845 –412.924 –669.806 –1635.404 –2573.272 –3518.318 –5241.180 –5986.310
R&D Number of intensity US patents (percentage) granted Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant Insignificant
* Cumulative number of patents granted during 1995–1999. Mani (2005, 2006); Chaudhuri (2007).
1
2 1 1 2 2 4 5 9 13 29
Contrasts between Indian sectoral systems of innovation
31
done by business enterprises then building a bridge between the two is always a problem. As can be seen below the SSI of the pharmaceutical industry had at its centre stage actors like business enterprises which had strong research and innovative capability. On the contrary the SSI of the telecom equipment industry had at its centre stage a public laboratory devoid of production. Consequently the firms could implement the technologies that they developed (perhaps through reverse engineering) more effectively. This hypothesis is tested by examining the details of the SSI of the two industries.
3. THE SSI OF TWO INDIAN HIGH-TECH INDUSTRIES 3.1.
The Pharmaceutical Industry
Figure 2.1 maps out the sectoral system of innovation of the pharma industry. There are essentially five components to the sectoral system. In broad terms they are: (i) policy and strategic direction; (ii) the intellectual property rights regime; (iii) human resource development or the supply of scientists and engineers;1 (iv) the technology generating sector; and (v) the manufacturing sector. Of the five, the three most important components of the SSI, according to me, are: (i) the public policy support; (ii) the manufacturing enterprises primarily in the private sector; and (iii) government research institutes (GRIs). We deal with each of these components in some detail below. 3.1.1. The public policy support The market conduct or behaviour of the pharmaceutical industry in the country is subject to the following policy framework. This could be classified as: ● ● ● ●
overall policy framework towards the development of the pharmaceutical industry; intellectual property rights or patent regulations; price regulations; and product and quality regulations.
Overall policy framework The overall policy framework governing the industry up to this time has been the Indian Pharmaceutical Policy of 1994. This is because the new drugs policy formulated by the government in 2002 could not be implemented owing to litigation involving it; hence the policy of 1994 continues to be in force. The present Policy, known as the Draft National Pharmaceuticals Policy, 2006,2 has been necessitated owing to
32
Sectoral systems of innovation and production Policy and Strategic Direction Managed by the Department of Chemicals and Petrochemicals and the Department of Science and Technology (through the Pharmaceuticals Research and Development Support Fund) and the Department of Biotechnology. Licensing of firms for permission to manufacture: Drug Controller General.
Price Controls by Drug Prices Control Order 95, administered by the National Pharmaceutical Pricing Authority. Overall policy framework: Pharmaceutical policy 2002. A new policy is on the anvil.
Human Resource Development Apart from the universities and pharmacy colleges, the National Institute of Pharmaceutical Education and Research (NIPER) has been set up by the Government of India as an institute of ‘national importance’ to achieve excellence in pharmaceutical science and technologies, education and training. Through this institute, government endeavour will be to upgrade the standards of pharmacy education and R&D.
Technology Generating Sector • Government Research Institutes Central Drug Research Institute (CDRI) Indian Institute of Chemical Technology (IICT) Indian Institute of Chemical Biology (IICB) Central Institute of Medicinal and Aromatic Plants • In-house R&D centres of leading private sector drugs companies • Contract research organizations primarily in the private sector.
The Intellectual Property Rights Regime The TRIPS-compliant Indian Patents Act 2005
Manufacturing Sector • Public (5) and private sector enterprises (about 5000) • Affiliates of MNCs
Source:
Own compilaton.
Figure 2.1
Sectoral system of innovation of the Indian pharmaceutical industry
several developments that have taken place during the course of the last few years as well as to address some of the major concerns as highlighted above. Price regulation of the essential medicines is an important component of this policy. However, several other matters having a close bearing on the pharmaceuticals sector have also been included. Since the purpose of the present chapter is to analyse the sectoral system of innovation of the Indian pharmaceutical industry, we will focus our attention only on those aspects of the policy that explicitly deal with the promotion of innovation. The major policy initiatives in this area are summarized below:
Contrasts between Indian sectoral systems of innovation
1. 2. 3. 4.
33
promotion of pharmaceutical R&D through the provision of fiscal incentives; promotion of R&D-intensive companies; establishment of a pharmaceutical Research and Development Support Fund (PRDSF); and development of orphaned drugs. In the following I discuss the details of each of these four policy initiatives:
1.
2.
3.
4.
Fiscal incentives for R&D: One of the most important fiscal incentives is the benefit of a 150 per cent weighted tax deduction of the amount spent on R&D per year until 31 March 2015. R&D-intensive companies (gold standard companies): The Pharmaceutical Research and Development Committee headed by Dr R.A. Mashelkar in its report submitted to government in November 1999 recommended that R&D-intensive companies fulfilling certain conditions should be given price benefits for the drugs under the Drug Price Control Order (DPCO). It specified certain norms in this regard and termed these as the gold standards. Firms fulfilling these norms can enjoy a 200 per cent weighted tax deduction. The Pharmaceutical Research and Development Support Fund: At present, the PRDSF has a corpus of Rs 1500 million (where only interest income is available for spending) utilized for funding R&D projects of research institutions and industry in the country. It is not adequate to meet the present-day and emerging requirements of this sector, and it needs to be sufficiently augmented over the next five years. It has been decided to convert it into an annual grant of Rs 1500 million, and thereafter it would be suitably increased further in a phased manner over a period of five years. Priority would be given for R&D in the case of diseases which are endemic to India, like malaria, tuberculosis, hepatitis B, leishmaniasis (kala-azar), HIV/AIDS and so on. Development of orphaned drugs: The Central Drug Research Institute (CDRI) has over time developed a number of drug technologies which could not be commercially produced and marketed. Efforts will be made to identify such technologies with a view to enabling them to reach the market.
Further, the following two initiatives implied in the new draft policy have also further implications for promotion of innovation in the industry. They are: (i) abolition of industrial licensing for bulk drugs, intermediates and formulations; and (ii) automatic approval for foreign technology agreements through RBI.
34
Sectoral systems of innovation and production
The patent regime It is now fairly well accepted that it is the provisions of the Indian Patents Act of 1970, and especially the fact that this Act did not recognize product patents but only process patents, that allowed Indian pharmaceutical companies to reverse-engineer and manufacture at significantly lower costs. But, with the country becoming a member of the WTO in 1995, the patent regime has been made TRIPS-compliant. This TRIPS compliance in very specific terms has led to the introduction of the following set of measures: ●
●
● ● ●
● ● ●
The EMR (Exclusive Marketing Rights) provision was introduced with retrospective effect from 1 January 1995 (a self-expunging provision which became void on 1 January 2005). This transitional arrangement entailed that India should provide for a mailbox mechanism for accepting product patent applications and for examining and granting EMRs till the time it accords recognition to product patents. The minimum patent term was increased from 14 to 20 years. There was a reversal of the burden of proof from patent holder to alleged infringer. The provisions relating to compulsory licensing have been modified to suit the public health requirements and also to comply with TRIPS. Product patents relating to chemicals, drugs, medicines and food products have been introduced. Provision for pre-grant objection to patents has been diluted. There is a grace period in the case of publication of inventions.
The potential effect of these amendments on the innovative behaviour of the domestic industry is now hotly debated. One of the most important consequences is about the availability and prices of many essential drugs. Henceforth some of these drugs can only be manufactured under an explicit licence. According to Ramani et al. (2005), the Indian pharmaceutical firms have three choices open to them in a post-TRIPS-compliant regime. These are: 1. 2.
3.
They can focus on products that are off patent (essentially the generics market). They can collaborate with Western MNCs and biotech companies (two areas that are likely to witness an increase in collaborations are clinical trials and R&D outsourcing). They can focus on innovations that the MNCs will not be interested in, that is mainly ‘tropical’ or developing world diseases.
Contrasts between Indian sectoral systems of innovation
35
Although a bit too early to clearly measure whether the three possibilities are actually happening, there is enough evidence to show that (1) and (2) are indeed happening. We will discuss this in some more detail in the subsequent sections. In the present section we analyse, albeit briefly, the efforts undertaken by Indian pharmaceutical companies towards R&D in neglected but tropical diseases. This discussion is very largely based on Chaudhuri (2005). The Indian private sector started investing in R&D for developing new drugs in the mid-1990s when TRIPS came into effect. According to current estimates there are about 15 domestic pharmaceutical companies that are active in drug research, and they have or are in the process of establishing new research centres with new drug discovery research (NDDR) as the major objective. The total R&D expenditure for the development of new drugs by Indian companies has increased from Rs 6.73 billion in 2002–03 to Rs 10.02 billion in 2003–04, and a number of new chemical entities (NCEs) have been developed which are at different stages of development. Since they do not have all the skills or the financial wherewithal required to engage in the entire process of drug development, they have adopted a strategy to develop new molecules and license out the molecules to the MNCs at early stages of clinical development. Consequent to this the Indian companies are effectively not targeting neglected diseases but only those which interest the MNCs. At this point, it is necessary to mention that the government has taken some initiatives for collaborative research to synergize the strengths of publicly funded R&D institutions and the Indian pharmaceutical industry. The one area where some progress has been made is in the development of an anti-TB molecule (Lupin’s development of the NCE LL 4858 is a case in point). However, no special efforts have been made for the development of new drugs for most of the neglected diseases (such as malaria, HIV/ AIDS, Chagas disease, dengue fever, leishmaniasis and leprosy). Price regulations Drug prices in India are among the lowest in the world (and imports are therefore negligible). This is for several reasons. The first is that only product patents and not process patents (for pharmaceuticals) are so far recognized under Indian law. Therefore Indian manufacturers can make bulk drugs and formulations by “reverse engineering” of the overseas patented medicines, reducing R&D expenses and also avoiding royalty payments. Further, Indian labour costs are low compared to overseas levels. India also has a large pool of technical and managerial personnel and does not need management skills from overseas. Most of the plant and equipment required is made locally. Most importantly, a measure of statutory price control for bulk drugs and formulations operates in India. The DPCO was introduced in 1970, but has since been modified three times, the latest one being in 1995. Over time the number of drugs
36
Sectoral systems of innovation and production
under price control has been significantly reduced from 370 in 1979 to just about 25 in 2005. The National Pharmaceutical Pricing Authority (NPPA) administers the price control regime.3 The government can exempt certain products from price control if they are new drugs discovered in India or bulk drugs produced from the basic stage by a new process discovered in India or drugs manufactured by small-scale industries (capital investment below a certain level) and sold under their own brand names. The most important problem with respect to price monitoring is the absence of an appropriate price index. The government has been depending on IMS Health–ACNielsen (formerly ORG) for tracking data on retail sales both in volume and in value terms. Therefore, having a pharmaceutical price index on the lines of the Consumer Price Index or Wholesale Price Index is being considered. Though details of the proposed index were not available, it is said that the government could create an index by having a basket of drugs whose prices would be benchmarked to a base year. It could then monitor any changes in their prices in relation to the index. However, the therapeutic segments that would form the basket would have to be decided. Also, whether the index would monitor prices of only generic drugs or include patented drugs as well would have to be finalized. Product and quality regulations The Drugs and Cosmetics Act of 1940 and its subordinate legislation Drugs and Cosmetics Rules (DCR), 1945 govern this aspect. The conduct of clinical trials – a growing area of importance – is actually governed by this legislation. The government has decided to amend the DCR and has emphasized the incorporation of good clinical practices (GCP) protocols and to make it legally binding to stress the safety aspect of the patients and strict accordance to ethics. Towards this direction the Department of Science and Technology (Government of India) established the national Good Laboratory Practice (GLP) Compliance Monitoring Authority, with the approval of the Union Cabinet on 24 April 2002. GLP-compliance certification is voluntary in nature and as such it is not binding. The GLP in India are compliant with OECD norms and principles. Industries, test facilities and laboratories looking for approval from regulatory authorities before marketing may apply to the National GLP Compliance Monitoring Authority to obtain GLP certification. So far there are only five Indian laboratories that have received this certification (Table 2.2). 3.1.2. The manufacturing enterprises There has been confusion on the total number of pharmaceutical units in the country. This has been variously estimated to be about 19 203 licensees. Citing the arguments and data provided in the Mashelkar Committee
Contrasts between Indian sectoral systems of innovation
Table 2.2 Sl No
Profile of Indian laboratories with GLP certification
Test facility
Areas of expertise
1
International Institute of Biotechnology and Toxicology (IIBAT)
2
GLP Laboratory, Gharda Chemicals Limited Dr. Reddy’s Laboratories Limited, Discovery Research
Physical-chemical testing Toxicity studies Mutagenicity studies Environmental toxicity studies on aquatic and terrestrial organisms Physical-chemical testing
3
4
Jai Research Foundation
5
Orchid Research Laboratories Limited
6
Advinus Therapeutics Private Limited
7
Zydus Research Centre, Cadila Healthcare Limited
8
Intox Private Limited, Pune
9
Laboratory Animal Research Services (L.A.R.S.), Reliance Life Sciences Private Limited Torrent Research Centre, Torrent Pharmaceuticals Limited
10
37
Physical-chemical testing Toxicity studies Mutagenicity studies Analytical and clinical chemistry testing Physical-chemical testing Toxicity studies Mutagenicity studies Environmental toxicity Studies on aquatic and terrestrial organisms Physical-chemical testing Safety pharmacology and pharmacokinetic studies Toxicity studies Mutagenicity studies Physical-chemical testing Environmental toxicity studies on aquatic and terrestrial organisms Toxicity studies Mutagenicity Studies Toxicity studies Mutagenicity studies Analytical and clinical chemistry testing Toxicity studies Mutagenicity studies Environmental Toxicity Studies on Aquatic and Terrestrial Organisms Toxicity studies
Toxicity studies
38
Sectoral systems of innovation and production
Table 2.2
(continued)
Sl No
Test facility
Areas of expertise
11
GLP Test Facility, Ranbaxy Research Laboratories, Gurgaon
12
SGS Inida Private Limited, Life Science Services
13
Toxicology Centre, Shriram Institute for Industrial Research
Physical-chemical tresting Toxicity studies Mutagenicity studies Analytical and clinical chemistry testing Toxicity Studies Mutagenicity Studies Analytical and Clinical Chemistry Testing Toxicity Studies Analytical and Clinical Chemistry Testing
Source: National Good Laboratory Practice Monitoring Authority, http://indiaglp.gov. in/TestFacility.htm accessed on (accessed at 25 January 2006).
Table 2.3
Structure of India’s pharmaceutical industry
Type of enterprise 1. Bulk drugs 2. Formulations 3. Large volume parenterals 4. Vaccines Total Source:
Number of enterprises 1333 4354 134 56 5877
Mashelkar Committee (2003), p. 49.
on drug regulatory issues, Chaudhuri (2005) argues that there are about 5877 pharmaceutical units in the country. This is because the number of pharmaceutical companies would be less than the number of licensees because manufacturing licences are given to specific units and many companies have multiple manufacturing units. The structure of the drugs manufacturing sector in India is presented in Table 2.3. According to Chaudhuri (2005), the bulk drug industry resembles that of a perfectly competitive industry with no one firm accounting for a significant share. Most of the units in this sector belong to the small-scale sector. Large private sector companies, on the contrary, dominate the formulations industry. See Table 2.4. One of the most important features of the industry is the fact that it is largely dominated by domestic private sector enterprises. In fact there
Contrasts between Indian sectoral systems of innovation
Table 2.4
Top 20 companies in the retail pharmaceutical market in India, 2004
Rank
Sector
Company
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Indian MNC Indian Indian Indian Indian Indian Indian MNC Indian MNC Indian Indian Indian Indian Indian Indian Indian Indian MNC
Cipla GlaxoSmithKline Ranbaxy Nicholas Piramal Sun Pharma Dr Reddy’s Zydus-Cadila Aristo Pharma Abbott India Alkem Labs Aventis Lupin Micro Labs Wockhardt Torrent Novartis India Alembic Unichem Labs USV Pfizer
Source:
39
No. of products 707 205 437 449 350 183 330 175 87 310 44 274 461 238 150 127 169 189 86 29
Annual sale in Rs million 11 285 11 143 9 190 8 720 6 738 4 988 4 959 4 760 4 735 4 477 4 367 4 165 3 903 3 776 3 747 3 725 3 432 3 430 3 390 3 274
Market share (%) 2004 5.51 5.44 4.48 4.25 3.29 2.43 2.42 2.32 2.31 2.18 2.13 2.03 1.9 1.84 1.83 1.82 1.67 1.67 1.65 1.6
Chaudhuri (2005), p. 17.
are only five MNCs in the top 20 and not a single public sector enterprise figures in the list. Two public sector enterprises, Hindustan Antibiotics Limited (HAL) established in 1954 and Indian Drugs and Pharmaceuticals Limited (IDPL) established in 1961, played an important role in creating a domestic private sector pharmaceutical industry (Chaudhuri, 2005, p. 34). This is best summed up by Smith (2000, p. 33): Before HAL opened its doors, the domestic pharmaceutical industry was all but nonexistent. Furthermore, India’s universities had no provisions for the type of specialized training required by pharmaceutical companies. HAL’s founders took the initiative and laid a considerable part of the foundation that supports today’s local and MNC subsidiary drug companies. HAL created a demand for inputs in the form of skilled labor, specialized capital, and relevant services, and provided the critical mass for local pharmaceutical production, created jobs for tens of thousands, spurred innovation, and sparked industrial development in up and
40
Sectoral systems of innovation and production 80 70 60 50 40 30 20 10 0
MNC (%) Domestic Companies (%)
Source:
1952 38
1970 68
1978 60
1980 50
1991 40
1998 32
2004 23
62
32
40
50
60
68
77
Chaudhuri (2005), p. 18.
Figure 2.2
Market shares of foreign and Indian companies in the Indian pharmaceutical industry, 1952–2004
downstream businesses. These contributions eventually rendered India a favorable environment for pharmaceutical production, research, and distribution.
However currently both these units are declared as “sick” or financially distressed companies by the Board for Industrial and Financial Reconstruction (BIFR) and have been virtually closed down. The amended patent law (1972) and the policy of positive discrimination towards indigenous companies vis-à-vis MNCs ensured that domestic companies in 2004 accounted for over three-quarters of the pharmaceutical market (Figure 2.2). Although the data on market shares provided in Table 2.4 appears to give an indication that the market is fairly competitive, this is really not the case, the reason being that the pharmaceutical industry is not a homogeneous one but fragmented into different therapeutic segments such as tranquillizers, analgesics, antibiotics, vitamins and so on. Each of these segments is not a substitute for another. In fact the concentration ratios are much higher within a specific therapeutic group. For instance, Chaudhuri (2005) shows that, if one takes the various sub-groups within antibiotics, the degree of concentration is much higher. Another important structural aspect has been the increased number of mergers and acquisitions in the industry. In the period from January 2004 – when Ranbaxy formalized its purchase of RPG (Aventis) for $80 million,
Contrasts between Indian sectoral systems of innovation
41
making it the fifth largest generics supplier in France – until October 2005, Indian firms made 18 international acquisitions (KPMG, 2006). Glenmark, Jubilant Organosys, Nicholas Piramal and Ranbaxy each acquired two overseas businesses during this time, but the biggest Indian buy was Matrix Labs’ acquisition of Belgium’s Docpharma for $263 million in June 2005. It is generally held that the pharmaceutical enterprises are currently the most aggressive overseas investors of all Indian industries. Several reasons4 could be attributed to this mergers and acquisitions spree. They are the need to: ● ● ● ● ●
●
improve global competitiveness; move up the value chain; create and enter new markets; increase the product offering; acquire assets (including research and contract manufacturing firms, in order to further boost outsourcing capabilities) and new products; and consolidate market shares.
3.1.3. Government research institutes According to Chaudhuri (2005), of the total pharmaceutical R&D expended in the country, nearly two-thirds is contributed by the industry and the remaining by the GRIs primarily under the management of the Council of Scientific and Industrial Research (CSIR). Of the small number of new drugs that were developed by Indian inventors, a lion’s share was the products of research done at the Central Drug Research Institute (CDRI). CDRI is considered to be one of the few public sector organizations in the world which have their own drug development infrastructure. Over the years it has developed and licensed to other private sector enterprises ten new drugs. Unfortunately most of the drugs according to Chaudhuri (2005) did not survive in the market owing to strong competition from MNCs. Apart from the CDRI, which is directly connected with drug research, the CSIR system has 20 other laboratories that are engaged in some form of pharmaceutical research or other. Four of these led by the CDRI have been very active in drug research, as indicated by the fact that they together account for a quarter of both Indian and foreign patents secured by the CSIR system (Table 2.5). 3.2.
The Telecom Equipment Industry
The SSI of the telecom equipment industry is mapped out in Figure 2.3. On the whole, India’s sectoral system of innovation of the telecom
42
Sectoral systems of innovation and production
Table 2.5
Foreign and Indian patents granted to CSIR labs engaged in drugs research, 2003–04
CDRI CIMAP IICB IICT Total for the above Total for CSIR Source:
India
Foreign
7 7 4 24 42 275
5 29 5 39 78 212
Computed from CSIR website.
equipment industry is very weak and fragmented. While the research segment, especially the dedicated public laboratory C-DOT, is very strong in terms of its capability to do successful R&D projects, there have been several attempts in the past to weaken its functioning (Mani, 1995, 2009, forthcoming). Compared to the Chinese one, the strategic direction from the state has been virtually absent.5 Given the fact that the country was demonstrating a growing capability in computer software, efforts should have been made to have a strong presence in telecoms software. This was, of course, accomplished subsequently in some measure by the private sector enterprises, but with little or no state support. The most distinguishing aspect of India’s sectoral system for innovation is the central role that it assigned to the public laboratory C-DOT. While the lab was successful in not just generating technologies that were quite suited to Indian conditions, it was able to effectively transfer the generated technology to a host of public and private sector enterprises. At the very same time it assiduously built up a growing number of component suppliers. In short, the laboratory is credited with establishing a modern telecommunications equipment industry in the country (Mani, 2008, forthcoming). The drawback of this strategy was that the firms did not have their own in-house R&D centres and were dependent entirely on the technologies that they received from the public laboratory. The lab, as mentioned earlier, continued to focus on fixed telephony and that too on circuit switching technology, when packet switching was becoming the state of the art. Further, it failed to take cognizance of the future in mobile communications (just like its counterpart in Brazil, the CPqD, but unlike its Korean counterpart, the ETRI6). The net result is that the licensing firms have become too complacent with respect to their own capability building. This is unlike the Chinese strategy, where the firms have built up considerable innovation capability on their own through their in-house
Contrasts between Indian sectoral systems of innovation Policy and strategic direction • Managed by the Department of Telecommunications (DOT) within the Ministry of Communication and Information Technology. Governed by the National Telecoms Policy of 1999 – wants to make India a major manufacturing base for telecom equipments. Telecom Commission and the DOT are responsible for policy formulation, licensing, wireless spectrum management, administrative monitoring of public sector enterprises. • Regulated by the Telecommunications Regulatory Authority of India (TRAI). • Disputes settlement: Telecom Dispute Settlement and Appellate Tribunal (TDSAT) is to adjudicate any dispute between licensor and a licensee, between two or more service providers and between a provider and a group of consumers, and to hear and dispose of appeals against any decision or orders of TRAI.
Structure of the research segment • Centre for Development of Telematics (C-DOT) • Telecommunication Engineering Centre (TEC) • Indian Institute of Technology Madras • In-house R&D centres of ITI and Hindustan Cables • Contract research organizations, which do R&D outsourcing deals for MNCs like Wipro Technologies • Telecom software manufacturing
Structure of the equipment manufacturing segment • A total of nearly 150 large, small and medium domestic and foreign companies • State-owned undertakings such as ITI and Hindustan Cables • Private sector C-DOT technology licensees • Affiliates of MNCs: Alcatel, Ericsson, Fujitsu, Huawei, Lucent, Motorola, NEC, Nokia, Nortel
Source:
Structure of the telecom service providers Continue to be largely under state ownership: • Fixed telephones: 91 per cent of the fixed lines are still under the public sector entities BSNL and MTNL, but the share of the private sector has been increasing. • Mobile telephony: is largely in the private sector, but the share of public sector enterprises has been increasing and is now about 21 percent.
Own compilation.
Figure 2.3
Sectoral system of innovation of the Indian telecommunications equipment industry (c. 2003)
43
44
Sectoral systems of innovation and production
R&D centres and have in addition acquired considerable production and marketing capabilities and have within a short span of about ten years emerged as internationally competitive. During the period up to and including the 1990s, domestic Indian companies dominated India’s telecom equipment industry. For instance, despite having a public technology procurement policy which did not favour domestic equipment manufacturers, the share of indigenously designed and manufactured equipment accounted for over 50 per cent of the market (Mani, 2009 forthcoming). However, the country just did not have a strategy in place to make its leading state-owned equipment manufacturer, ITI, a national champion, in sharp contrast to the strategy pursued by the Chinese. This will be evident when one makes a comparison of two leading telecommunications equipment manufacturers from China and India. Despite the leading Indian firm being an early starter, on every single indicator the leading Chinese firm outperforms the leading Indian firm. India has been a recipient of substantial FDI in telecommunications, although much of it is in the distribution of mobile communications services. Many MNCs, including two of the leading Chinese telecom equipment manufacturers, Huawei and ZTE, have established or are in the process of establishing manufacturing ventures in the country. During the period April 2000 through January 2009, the telecommunications sector has attracted FDI to the tune of US$6.22 billion, thus working out to about 8 per cent of total FDI inflows to the country. Cumulatively over the period 1991 to 2004, the country attracted FDI in telecommunications to the tune of US$7.14 billion, and this works out at about 18 per cent of the total approved FDI the country received as a whole. As a result of these high foreign investments, the complexion of India’s telecom equipment industry is fast undergoing a change, with foreign affiliates and imports accounting or going to account for a significant share of the domestic market for telecom equipment. In fact a study in 2004 by the Department of Telecommunications found that at that time most of the domestic telecom equipment manufacturers and even the state-owned undertaking ITI, which till recently had been the major equipment manufacturer, had merely become “traders” by importing the equipment and supplying it to the service providers.7 The deregulation of India’s telecom equipment industry had an extremely destabilizing effect on the operations of ITI (Subramanian, 2004), and its very existence was now in danger.8 An important development in the country’s sectoral system of innovation is the growth of R&D outsourcing deals between foreign MNCs and Indian contract research organizations in the area of telecom R&D.
Contrasts between Indian sectoral systems of innovation
45
As this is a growing phenomenon, there are no precise estimates.9 Even C-DOT, the nerve centre of the sector’s innovation system, has entered into a contractual agreement with Alcatel to set up a global R&D centre for broadband wireless products.10 I consider in some detail the three major components of the innovation system, namely: (1) public policy support; (2) government research institutes; and (3) manufacturing enterprises. 3.2.1. Public policy support The most important instrument of public policy support that was unique to this industry was public technology procurement. The government could practise this during the 1980s and even up to the early 1990s because the main consumer, the main telecom service provider, the Department of Telecommunications (DoT) and later on its corporatized version, the BSNL, too were state-owned undertakings. However, in the latter half of the 1990s, the deregulation of the telecom service provision market and the emergence of mobile communication have reduced the state’s ability to support domestic technology generation through public technology support. However, in the best of times, the way the public technology support was actually practised was not very beneficial to domestically developed technology by the public laboratory. This could be explained as follows. The DoT purchases switching equipments only from local manufacturers and does not allow imports of finished switching products.11 This really does not afford any protection to domestically assembled switches, but in fact quite the contrary. There are two types of evidence for this. First, imported equipments attract a customs duty of 15 per cent ad valorem (2002–03). At the same time the imported components for domestically assembled switches also attract customs duties and, given the nearly 50 per cent import content of domestically assembled switches, the procurement policy does not afford any specific protection to domestically assembled or manufactured switches. Second, as noted in Figure 2.6, the fall in price realization of domestically manufactured equipments signal increased competitive pressures. Further in the past the rate of rejection of indigenously manufactured switching equipments by the DoT had been as high as 25 per cent in the early 1980s.12 It has a decentralized telecom switches procurement policy. In order to simplify the procurement process, the department receives tenders and sets a fixed rate through a tendering process commonly known as “rate contract”,13 by which the chief general managers of the various telecom circles are authorized to purchase their requirements from approved vendors. The Telecommunications Engineering Centre (TEC) within the
46
Sectoral systems of innovation and production
department sets the technical standards of all telecom products including switches. Thirty per cent of the total requirements of switching equipments are reserved for the public sector enterprise. However, the price at which this 30 per cent is procured is at the lowest price quotation received for the remaining 70 per cent, for which an open tender is invited. This reservation price is referred to as the L1 price. It is thus seen that the public sector producer of switching equipments has actually to bid for 30 per cent of the switching requirements without actually knowing the price at which the bid is going to be made. Thus it is clear that the public procurement process followed in the case of switching equipments does not afford any protection to the public sector producer, which in this case is ITI Ltd at Bangalore. The price–performance ratio is thus the main criterion for selection and not other non-technical considerations such as deferred credit facilities. At least for several more years, given the near monopoly position of the government carriers, public procurement policy will be an effective instrument for stimulating local R&D activities. However, with the growth of private service providers this will be less effective, especially when the private sector providers that are in the initial years of establishment would also take into account deferred credit facilities which only the MNC vendors can offer. 3.2.2. Government research institute The government research institute, Centre for Development of Telematics (C-DOT), occupies the core of the SSI of the industry. C-DOT was established as a stand-alone public R&D organization by the central government in 1984. It was charged with the responsibility of developing a family of digital switching systems that were suitable for the Indian usage pattern and conditions. Its scope has now been broadened to include transmission and access products as well. Over time, C-DOT has developed a wide range of switching and transmission products for both rural and urban applications. It is claimed that, while the C-DOT main exchange can also function as mobile switching centre for GSM cellular service, the small rural automatic exchanges developed for the rural environment can work without air-conditioning. They come complete with SS7 Intelligent Network signalling systems.14 In addition ISDN facilities are available; what is unique is that these switches have been designed to operate without air-conditioning in harsh environments. About 45 000 exchanges totalling about 23 million telephone lines had been installed in India as at 31 December 2001. This means that approximately 50 per cent of the equipped capacity in the country is based on C-DOT designed switches. Exports in bulk have been to about 22 countries, such as Vietnam, Bangladesh, Nepal, Ethiopia, Nepal, Ghana and Uganda. And this systematic vendor development
Contrasts between Indian sectoral systems of innovation
47
shows that there have been considerable technology spillovers to downstream industries as well. It has an R&D centre in Bangalore with complete test equipment such as microprocessor development systems, CASE tools, object-oriented methodologies and software metrics, along with V5.1, V5.2 interface, and SS7 signalling systems complete with the SSP, SCP and SMP systems. In fact C-DOT, with a total manpower of 1300 employees, has one of the fastest development cycles for digital switching systems anywhere in the world (Mani, 1995). C-DOT also claims to have retrofitting capabilities, that is it has the capability to redesign some of the older switches that it has already developed and that are currently being used in the network to comply with more recent technological changes. This is achieved by making the necessary software changes. I have not been able to secure an independent confirmation of this capability. I now turn my attention to measure quantitatively the innovation capability of C-DOT. To do this, I consider two separate indicators of this capability. First is a summary measure of innovation capability based on production of C-DOT designed switches. Second is a series of evidence to show the spillover effects of the technologies developed at C-DOT. Indicator of innovation capability There are two variants of this index. The first variant of this index is based on the relative market share of domestically designed (namely C-DOT-designed and ITI-manufactured Main Automatic Local Exchanges, in terms of number of lines) and foreigndesigned but domestically manufactured (namely Alcatel-designed and ITI-manufactured).15 On technical grounds, both the technologies are considered to be equal. In very specific terms the index is defined as follows: Index of innovation capability 5 Production of C-DOT-designed exchanges at ITI 3 100 Production of Alcatel-designed OCB-283 exchanges at ITI If the index is greater than 100 and increasing over time, one can say that the innovation capability of the domestic research sector is increasing over time. A major limitation of the index is that it is rather difficult to interpret short-term movements in the index. A second limitation is the fact that the index is defined in terms of production figures and not in terms of number of working connections. But I argue that this will only affect the level of the index and not its direction of movement. This is because the share of C-DOT-designed exchanges has been rising over time. Based on the data during the period 1995 to 2001, the index has been computed, and it is presented in Figure 2.4. Excepting for the initial year, the index shows a continuous rise over time, implying
48
Sectoral systems of innovation and production 250.00
Index of R&D Capability
200.00
150.00
100.00
50.00
0.00 Index of R&D Capability
Source:
1995
1996
1997
1998
1999
2000
2001
175.99
73.54
109.80
84.36
104.80
145.48
234.83
ITI (various issues).
Figure 2.4 Index of innovation capability in switching equipment, 1995–2001 a rising capability. This is quite significant, as this has been happening at a time when the industry was going through a flux: the carrier industry was getting deregulated and MNCs were entering the equipment industry. So despite these factors, which can militate against the usage of domestically designed switches, one sees a systematic and continuous increase in its market share. As seen before this could be largely explained by the public procurement policy of the main consumer, the DoT. The second variant is based on the number of lines of a switching technology actually commissioned within the network of the two main public telecoms service providers, namely within the DoT and the MTNL networks during a year. This variant is thus more of an index of market share of the various technologies, and it is measured by the share of C-DOTdesigned switches in the total number of lines commissioned each year (Table 2.6). This variant thus captures the effect of liberalization. In fact the index shows that, despite public technology procurement, the share of C-DOT-designed switches has continuously fallen all through the period. This of course proves that public technology procurement in India does not afford any protection to domestically designed switches. This proposition could be further explained as follows:
Contrasts between Indian sectoral systems of innovation
Table 2.6
49
Share of C-DOT-designed switches in the total number of lines commissioned in DoT and MTNL networks, 1994–97
Type of switching technology
1993–94
1. AXE-10 74 000 2. EWSD 107 000 3. 5ESS 10 000 4. FETEX-150 160 000 5. NEAX-61 Nil 6. E-10B 766 327 7. OCB-283 79 500 Total foreign technology 1 196 827 (1+2+3+4+5+6+7) 8. Domestic technology 966 583 Of which: ● C-DOT: large 148 500 exchange* ● C-DOT: extra large Nil exchange** ● C-DOT: small 818 083 exchange*** 44.68 Share of domestic technology in the totalnumber of linescommissioned (in percentage)
1994–95
1995–96
1996–97
169 704 203 500 4 000 93 000 10 000 957 330 311 000 1 748 534
128 300 297 328 132 000 113 200 Nil 1 119 994 405 720 2 196 542
113 060 249 544 40 648 93 280 Nil 523 854 490 578 1 510 964
1 198 516
1 121 519
661 905
186 020
328 625
206 166
18 800
11 200
774 094
444 539
Nil 1 012 496 40.67
33.80
30.46
Note: * Small and medium exchanges are those having up to 3000 lines; ** large exchanges are those between 3000 and 10 000; and *** extra large are those having more than 10 000 lines. Source: Rajya Sabha, Unstarred question no. 1171, http://164.100.24.219/rsq/quest. asp?qref=32199 (accessed on 26 September 2006).
1.
2.
Table 2.6 tracks only the share of foreign and domestic technologies in the total annual flow of exchange lines commissioned. C-DOT’s share in the total stock of exchange lines in the country is high at about 50 per cent, with the remaining 50 per cent shared by the other eight technologies. C-DOT is specializing in small and medium exchanges, while the imported technologies are used essentially in large and extra large exchanges. It is also a fact that C-DOT’s capability is largely in small and medium exchanges, though it also has claims to capabilities in designing large and extra large switches.
50
3.
4.
5. 6.
Sectoral systems of innovation and production
It is also clear from an answer to an unstarred question in the upper house of the Indian parliament that the DoT procured almost five times the tendered quantity of switching equipment during the same period, supposedly for modernizing the network with an ISDN facility.16 But the number of subscribers using ISDN in the whole country was just 309.17 So it is clear that the DoT appears to have purchased this ‘overspecified’ equipment far in excess of its actual requirement, and this ‘excess purchase’ appears to have eroded the market share of C-DOT. Further the Comptroller and Auditor General of India in 2000 found a number of other irregularities with this tendering process. For instance, although the suppliers imported most of the components of these switching equipments, the DoT assumed an import content as low as 23 per cent while working out the reduction in rates on account of the fall in customs duty in the 1995–96 budget. This inaccurate assumption by the DoT led to excess payment of Rs 405 million to the suppliers, with a corresponding loss to the government exchequer. The DoT also had to make an avoidable expenditure of Rs 639 million in the procurement of these exchanges against the 1997–99 tender owing to the failure of the Tender Evaluation Committee (TEC) to submit its report within the bid validity period. The TEC took 190 days to finalize its report, against the prescribed limit of 42 days. Despite this fall in market share, C-DOT-designed switches continue to hold the single largest share. In the light of the above comments, it would not be correct to interpret the fall in the overall market share of C-DOT-designed switches to mean a fall in its innovation capability.
Spillover effects of C-DOT Over the last two decades of its existence, C-DOT has made a number of important contributions both in money and in giving a fillip to domestic technology development in this area of high technology. These could be enumerated as follows: 1.
Since its inception in 1984, C-DOT has recouped approximately 25 per cent of the amount that it has received in the form of parliamentary grants through the sale of its generated technologies. This rate of self-generation has increased significantly to more than two-thirds in the recent past (Figure 2.5). This is a remarkable achievement, as elsewhere in India the network of laboratories coming under the purview of the Council of Scientific and Industrial Research (CSIR) has a record of generating only about 10 per cent of its total income through self-generation (Mani, 2002).
51 0.800
1200
0.700
1000
0.600 Rs in Millions
800
0.500 0.400
600
0.300
400
0.200 200
0 External Income (Rs in Million) Parlimentary Grants (Rs in Million) Ratio self generation to parliamentary grants
Source:
3.
4.
2001–02 725.5
1997–98 74.1
1998–99 328 .6
1999–00 190.4
2000–01 817.3
585
550
747.5
709.7
1106.6
1080
0.090
0.135
0.440
0.268
0.739
0.672
1996–97 52.7
0.000
C-DOT (various issues).
Figure 2.5
2.
0.100
Ratio of self generation to parliamentary grants
Contrasts between Indian sectoral systems of innovation
Ratio of self-generation through sale of technology to total parliamentary grants
C-DOT’s technological innovations have contributed to a substantial reduction in the price of switching equipments sold in the country. In fact over the last decade prices have fallen by as much as 75 per cent. This fall in prices has enabled the country to increase the supply of direct exchange lines. C-DOT has transferred eight different types of technologies to very nearly 74 manufacturers in the country (Table 2.7). These 74 companies have their own suppliers of components and spare parts numbering over 600 enterprises. C-DOT has thus effectively contributed to the creation of an indigenous telecommunications equipment industry in the country. More details of the industry are analysed in section 3.2.3. C-DOT has pioneered the telecoms software industry in the country. Every year approximately 80 engineers (out of a total of about 1200) leave the centre. Since the development of modern digital switches is largely software based, this has given these engineers a strong background in the development of telecom software. The growth of the telecoms software sector is analysed separately.
While C-DOT has considerable capability in the design of fixed-line switches, there is some doubt about its ability to design mobile switches. In fact my discussions at the DoT, the TEC and C-DOT revealed that C-DOT does not have, as of now, any real capability in the design of mobile
52
Sectoral systems of innovation and production
Table 2.7
Technology transfer by C-DOT (completed technologies)
Type of technology 256 P RAX SBM RAX DSS MAX IVRS DMX-8 DMX-34 TDMA-PMP OLTE-8 Total Source:
Number of manufacturers 14 14 13 10 9 3 6 5 74
C-DOT (various issues).
switches. In fact no state agencies in India (including the Technology Information and Forecasting Assessment Council) have done a detailed technology foresight exercise for the telecoms sector, and C-DOT, despite possessing the potential, has been totally unprepared for this changeover. This is going to be a serious shortcoming for C-DOT in the future, as the growth of mobile phones is likely to be faster than that of fixed lines. MNCs such as Lucent, Motorola and Siemens have already established themselves as suppliers of state-of-the-art mobile switching centres to the cellular service providers. The problem is so severe that, according the CEO of ITI Ltd (the largest telecoms equipment manufacturer in India), it had no orders for switching equipments for fixed lines from the largest telecom service provider in the country (namely BSNL) for the year 2003. Already there are reports of most of the licensees (switching equipment) of C-DOT having to close their manufacturing activities or scaling them down for lack of sufficient orders. 3.2.3. The manufacturing enterprises There are about 150 enterprises of all sizes and ownership operating in the industry. They can be broadly divided into three: a large but financially distressed state-owned undertaking, ITI, a number of small and mediumsized domestic private sector enterprises and a number of well-known MNCs or their affiliates. The latter have entered the industry only during the period since 1985 when the manufacturing of telecom equipments was opened up to the private sector and indeed even foreign participation. The equipment industry itself can broadly be classified into three: switching, transmission and terminal equipments. The switching equipment sector is largely dominated by the MNCs. The most distinguishing aspect of the manufacturing sector when compared to that of the pharmaceutical
Contrasts between Indian sectoral systems of innovation
53
2000–01
1999–00
1997–98
1996–97
1990–91
0
1000
2000
3000
4000
5000
6000
7000
8000
Average price (Rs per line) 1990–91
1996–97
1997–98
1999–00
2000–01
8000
5411.96
5568.21
4389.41
1978.84
Average price of switching equipment (Rs per line)
Source:
Panning Commission (2002)
Figure 2.6 Trends in average price per line of switching equipments, 1990– 91 through 2000–01 industry is that none of the domestic private sector manufacturers have any in-house research capability and the enterprises were completely dependent on external sources of technology. This, I believe, is an important structural weakness of the industry’s SSI. This weakness is clearly reflected in the innovative performance of the industry (see Table 2.1).
4.
CONCLUSIONS
The pharmaceutical and telecom equipment industries are two high-technology manufacturing industries that India has sought to promote. One can easily discern a sectoral system of innovation in both the industries, although there are important differences between the SSI of both the industries. Table 2.8 summarizes these important contrasts between the two SSIs. Employing three standard indicators of innovativeness, one sees that the pharmaceutical industry is more successful. Our explanation for this
54
Sectoral systems of innovation and production
Table 2.8
Contrast between the SSI of the Indian pharmaceutical and telecom equipment industries
1. Degree of complexity in technology 2. The main actor
3. Main instrument of state support 4. Networks 5. Innovative performance Source:
Pharmaceutical industry
Telecom equipment industry
Complex
Complex
Domestic private sector manufacturing enterprises with strong in-house R&D capabilities The intellectual property right regime (Indian Patents Act of 1970) Interaction between equal partners High
Government research institute with strong innovation capability Public technology procurement Interaction between unequal partners Low
Own compilation.
difference was largely in terms of the constitution of the respective SSI. Thus the present study is yet another instance18 to show that a dichotomous relationship between research and production is not very desirable for promoting innovations in the industrial sector.
NOTES 1. 2. 3.
4. 5.
6.
The areas are medicinal chemistry, combinatorial chemistry, bioinformatics and structure-based molecular modelling, genomics and proteomics, clinical pharmacology, and regulatory toxicology. Department of Chemicals and Petrochemicals, http://chemicals.nic.in/npp_circulation_latest.pdf (accessed on 11 August 2006). The functions of the NPPA, inter alia, are to: (i) implement and enforce the provisions of the Drugs (Prices Control) Order in accordance with the powers delegated to it; (ii) monitor the availability of drugs, identify shortages, if any, and take remedial steps; and (iii) collect/maintain data on production, exports and imports, market share of individual companies, profitability of companies, etc., for bulk drugs and formulations. See KPMG (2006), p. 25. The TIFAC did a major technology foresight exercise covering nearly 17 different areas including the telecommunications sector. Known as the ‘Vision 2020’ reports, these were published in 1996. Going through the list of seven major recommendations of the report on telecommunications one finds that the study did not anticipate at all the phenomenal growth of mobile telecommunications in the country. Mani (2009, forthcoming) has the details.
Contrasts between Indian sectoral systems of innovation 7.
8.
9.
10.
11. 12. 13. 14.
15.
16.
17. 18.
55
The study even states that, ‘in order to take advantage of lower customs duty, a separate procedure of “high-sea sale” is being followed. Even reservation quotas of PSUs are being used for trading goods manufactured abroad and without any commitment of transfer of technology.’ See Department of Telecommunications (2004), p. 4. The prime minister, Manmohan Singh, sanctioned Rs 10.32 billion for the revival of ITI, and ITI entered into a technical tie-up with Alcatel for the manufacture of 3 million telephone lines. An announcement to this effect was made in the upper house of the Indian parliament on 24 March 2005. See Economic Times, http://economictimes.indiatimes.com/articleshow/1061839.cms (accessed on 27 September 2006). Further recent press reports indicate that the company is going to take up the production of mobile communications equipment under foreign technology licensing. According to one of the leading consultancy organizations, the R&D outsourcing market for IT in India is estimated to grow more than $8 billion by 2010 from $1.3 billion in 2003, at a CAGR of 30 per cent. There have been a number of high-profile R&D outsourcing deals between Western MNCs and Indian enterprises, for instance the WIPRO–Ericsson deal and the Sasken–Nortel deal are two of three high-profile deals in this area. The project is to develop WiMAX (Worldwide Interoperability for Microwave Access) broadband technology. WiMax is a lot like WiFi, the short-range wireless technology that allows web surfers to connect to the internet at so-called hot spots. But unlike WiFi’s 50-metre range, WiMax’ reach is 1 to 10 miles, offering a way to bring the internet to entire communities without having to invest billions of dollars to install phone or cable networks. It must of course be added that the new private entrants are not governed by this stipulation and are free to import switching equipments. See Mani (1992), p. 97. The DoT receives and evaluates bids from domestic firms (including affiliates of MNCs) and awards rate contracts based on price and performance. These are the systems that are used to find out if a number is busy or available and involve a separate system that checks up the databases of phone numbers; they also provide toll-free services; in this way the main telephone network does not get overloaded; these systems are also used to interconnect mobile and land-based telephone numbers. Currently the Indian telecoms carrier industry employs eight different types of switching technologies, like C-DOT, E-10B and OCB-283 (Alcatel), 5ESS (Lucent technologies), EWSD (Siemens), FETEX-150L (Fujitsu) and NEAX-61E (NEC). Of all these eight technologies, C-DOT is the single largest, with a market share of 50 per cent. See Rajya Sabha, Unstarred question no. 4125, http://164.100.24.219/rsq/quest. asp?qref=21560 (accessed on 27 September 06). According to the answer given by the Ministry of Communications, the DoT has actually ordered 0.91 million lines of digital switching equipments in response to a tender for just 0.2 million lines. See the response to the same question, no. 4125. From the same response it is also clear that the number of ISDN subscribers even in developed countries ranges from 0.5 million in the USA to just 40 000 in Italy. China’s innovation system was transformed all through the 1990s. In 1988 the Torch Programme was launched to encourage something like spin-off enterprises, called NTEs (New Technology Enterprises), from existing R&D institutes and universities. Local governments contributed to investment in infrastructure and supporting institutions for the New and High-Tech Industry Zones that became incubation bases for the NTE start-ups. Scientists and engineers, often with support from their parent institutions, went into commercial application of their inventions and expertise by means of the creation of NTEs. And, by the early 1990s, reform policy included another solution to change individual R&D institutes into production entities. This, as well, was an adaptation to an actual evolution already realized by many industrial R&D institutes.
56
Sectoral systems of innovation and production
REFERENCES Chaudhuri, Sudip (2005), The WTO and India’s Pharmaceuticals Industry Patent Protection, TRIPS, and Developing Countries, Oxford University Press, New Delhi. Chaudhuri, Sudip (2007), Is Product Patent Protection Necessary in Developing Countries for Innovation? R&D by Indian Pharmaceutical Companies after TRIPS, Working Paper Series 614, Indian Institute of Management, Calcutta. Department of Telecommunications (2004), Annual Report, Government of India, New Delhi. KPMG (2006), The Indian Pharmaceutical Industry: Collaboration for Growth, KPMG Consulting Private Ltd, Mumbai. Malerba, Franco (2004), Sectoral Systems of Innovation, Concepts, Issues and Analyses of Six Major Sectors in Europe, Cambridge University Press, Cambridge. Mani, Sunil (1992), Foreign Technology in Public Enterprises, Oxford and IBH, New Delhi. Mani, Sunil (1995), ‘Technology Import and Skill Development in a Microelectronic based industry: Case of India’s electronic switching systems’, in Amiya Kumar Bagzhi (ed.), New Technology and the Workers’ Response, New Delhi: Sage, pp. 98–122. Mani, Sunil (2002), Government, Innovation and Technology Policy: An International Comparative Analysis, Edward Elgar, Cheltenham, UK and Northampton, MA, USA. Mani, Sunil (2005), ‘Innovation capability in India’s telecommunications equipment industry’, in Saith, Ashwani and Vijayabaskar, M. (eds), ICTs and Indian Economic Development, Sage Publications, New Delhi, pp. 265–322. Mani, Sunil (2006), The Sectoral System of Innovation of Indian Pharmaceutical Industry, Working Paper Series 382, Centre for Development Studies, Trivandrum. Mani, Sunil (2009, forthcoming), Innovation Capability in Developing Countries: A study of the Telecommunications Industry, Edward Elgar, Cheltenham, UK and Northampton, MA, USA. Mashelkar Committee (2003), Report of the Expert Committee on a Comprehensive Examination of Drug Regulatory Issues, including the Problem of Spurious Drugs, Central Drugs Standard Control Organization, New Delhi. Planning Commission (2002), Working Group on the Telecommunications Industry, New Delhi: Planning Commission. Ramani, Shyama V., P. Pradhan and M. Ravi (2005), ‘Biotech in post TRIPS India’, Nature Biotechnology, 23(1), 18–19. Smith, Sean Eric (2000), Opening up to the World: India’s Pharmaceutical Companies Prepare for 2005, Asia Pacific Research Centre, Institute for International Studies, Stanford University, Stanford, CA. Subramanian, Dilip (2004), ‘Impact of Deregulation on a Public Sector Firm’, Economic and Political Weekly, 39(49), 5233–45.
3.
From innovation projects to knowledge networks: knowledge as contingency in the sectoral organization of innovation Fernando Perini
1.
INTRODUCTION
Identifying relationships between the knowledge base of sectors, the role of different actors and networks, and the channels of the knowledge flows within and across sectors is a fundamental challenge for the dynamic understanding of sectoral innovation systems (Tidd, 1997; Malerba, 2002; Acha and Cusmano, 2005; Giuliani and Bell, 2005; Owen-Smith and Powell, 2005). Scholars have used joint ventures, surveys on university– industry links, patents citation and co-authorship in publications as ways to measure knowledge and knowledge flows and examine the process of specialization and formation of comparative advantages among firms, sectors and countries.1 However, despite their key role in the organization of innovative activities, detailed studies on networks formed around innovation projects in sectors remain particularly rare. This chapter uses a large database of innovation projects inside the Brazilian ICT sector to examine contingencies between the knowledge base of sectors, the specialization of governance mechanisms and the patterns of horizontal collaboration between firms and research and educational institutes. Although there are many mechanisms of knowledge flow in sectors (i.e. labour mobility, informal relationships, interaction with users, etc.), increasingly innovation projects are recognized as a fundamental organizational format for the active creation of knowledge within and among organizations. Given their dynamic nature, many innovative organizations make use of them to balance between exploitation of technological niches and exploration of new opportunities (Davies and Hobday 2005). Innovation projects became central in evolutionary theories of industrial dynamics as they are essentially problem-solving activities
57
58
Sectoral systems of innovation and production
adapting and localizing knowledge to specific organizational needs (von Hippel 1994; Dosi et al., 2000). This chapter suggests that examination of innovation projects in different knowledge-related activities can provide in-depth insights into the nature of the knowledge governance in sectors and a way to investigate evolutionary mechanisms of change inside the industrial organization. This research is based on a contingency approach to investigate the coevolution between knowledge base and the organizational network emerging between firms and technological partners. The knowledge base has been acknowledged as a fundamental contingency in the organization of firms and the industrial structure (Woodward, 1965; Tidd, 2001; Birkinshaw et al., 2002; Thompson, 2003; Foss and Klein, 2005). However there are still many weaknesses in the measurement of knowledge and knowledge flows in sectors (Patel and Pavitt, 1995; Meyer, 2002). Empirical studies are hindered by the lack of reliable databases. The lack of standardization in contractual relationships and the private and confidential nature of its content make codifying transactions in different projects within and among organizations in sectors a very complex task. However, the increasing use of IT systems and reporting standards provides new ways to scrutinize the decentralized innovation dynamic occurring in sectors. Therefore, secondary data on innovation projects can contribute significantly to the direct exploration of the knowledge flows in sectors. This chapter draws upon an exclusive dataset containing details of the innovation projects conducted by national and multinational manufacturing companies as well as educational and research institutes in the Brazilian ICT sector and declared under the Brazilian ICT Law between 1997 and 2003. After decades of important substitution policies, the market was opened to foreign investment during the early 1990s. The previous regulation in the Brazilian ICT sector was substituted by tax incentives for the commercialization of a set of industrialized products in the internal market conditioned to local manufacturing and investments in R&D. An R&D offset scheme implemented in the sector promoted an overall private investment of more than USD 2 billion in innovation during the last decade and involved more than 200 companies as well as 200 universities and research institutes. The ICT Law became one of the pioneering projects for the development of sectoral innovation systems in Latin America following its liberalization policies. Therefore, from the empirical perspective, the chapter sheds new light upon the structure that co-evolved from the institutional changes in the Brazilian ICT sector and the interaction between multinational and national innovation systems during the period in different technological trajectories. This research uses 10 088 innovation projects in the Brazilian ICT sector
From innovation projects to knowledge networks
59
to operationalize the elusive concept of knowledge and more than 35 000 economic transactions inside innovation projects to observe the even more ambiguous concept of inter-organizational knowledge flows inside sectors. Although using secondary data on the economic transactions in innovation projects as a proxy for knowledge flows in sectors is certainly a simplification, it nevertheless offers many advantages and complementarities in relation to its traditional counterparts such as patents, citations and surveys. The use of innovation project-level data to analyse the knowledge in sectors is useful, as it avoids any assumptions that knowledge flow (or leakage) among firms may happen “in the air” (Marshall, 1891), that it is costless (Teece, 1977) or that it may happen as a by-product of commercial relations (Bell and Pavitt, 1993). It reinforces the idea that intentional flow of codified types of information, as well as tacit knowledge embedded in people and constructed in organizational routines, is a requirement for organizational learning in firms and sectors (Teece, 1988; Nelson, 1994). In fact, innovation projects seem to be a particularly adequate unit of analysis for the investigation of the sectoral innovation systems in the developing context. Innovation projects are a key mechanism in which interaction among organizations takes place and relevant knowledge for the various parties is constructed and transferred. Innovation projects provide a way to discuss ‘relevant knowledge’ according to the needs of the parties involved rather than any assumption that the knowledge developed should be new to the world (i.e. patentable knowledge). Therefore, they provide new ways of measuring the longitudinal evolution of innovation systems in developing countries (Bell, 2005) and comparing the governance structures driving change in specific directions as well as systemic characteristics that may speed up, detain or reverse the formation of these sectoral networks. This chapter proposes and tests different ways in which the type of innovative activities influences (i) the boundaries between firms and technological partners in different knowledge-related activities, (ii) the process of specialization in different governance mechanisms and (iii) the formation of channels for knowledge flows between different sets of actors. This chapter distinguishes between ten different innovation-related types of project in the ICT sector: laboratory and equipment infrastructure, technological training, technological services, R&D quality systems, process technology, product development in software, middleware, hardware and semiconductors as well as research activities. These three propositions are tested using, respectively, (i) a longitudinal examination of the boundaries between firms and technological partners in different types of innovation project, (ii) a project-based index of revealed technological advantage (PRTA) and (iii) a social network correlation technique (QAP).
60
Sectoral systems of innovation and production
The structure of the chapter is as follows. The second section briefly discusses how the knowledge network is defined and used in this chapter. Section 3 describes the dimensions used to examine the interaction between the knowledge base involved in different types of innovation projects and the organizational characteristics of the knowledge network. Section 4 describes the characteristics of the database of innovation projects in the Brazilian ICT sector, and the main characteristics of the sectoral knowledge networks. This is followed by the details of the methods used in the investigation as well as consideration of some of the limitations of the research design (section 5). Section 6 details the results of the empirical analysis of the knowledge network in the Brazilian ICT sector empirically examining the relationship between the knowledge base of projects and the organization of the innovation in sectors. The last section summarizes the empirical findings and outlines some preliminary implications for firm strategy and institutional design of sectoral networks.
2.
ON PROJECT-BASED KNOWLEDGE NETWORKS
The term ‘knowledge network’ can be used as a metaphor to represent the complexity of the innovation process (DeBresson and Amesse, 1991; von Tunzelmann, 2004) or a middle way between market and hierarchy (Powell, 1990), or even to describe the fundamental nature of the firm and all its economic activities (Coase, 1937; Williamson, 1985). In contrast, the same term, ‘knowledge network’, has also been used as an attempt to employ new methodological tools for the analysis of the interactions between agents (Wasserman and Faust, 1994; Powell et al., 1996; Pyka and Küppers, 2002; Malerba, 2005). Social network analysis has emerged recently as one of the most promising tools for the analysis of the knowledge flows in innovation studies where specific rules/norms or institutions would allow definition of the boundaries of the observable network, its participants and the scope of their activities. Given the widespread use of the term ‘knowledge network’ to discuss different interactions between actors, the natural first step is to clarify what is meant by the term and its underlying assumptions. First of all, as previously discussed, the term as used in this chapter is limited to the quantitative examination of networks formed by innovation projects. Naturally, the evolution of industrial sectors under evolutionary principles embraces a wider range of activities, basing its foundations on the concept of capabilities. By focusing on innovation projects, the analysis emphasizes tacit knowledge creation and flows rather than technology transfer by simple acquisition of equipments or traditional commercial transactions in direct acquisition of products and services (Bell and Pavitt,
From innovation projects to knowledge networks
61
1993; Giuliani and Bell, 2005). By delimiting the analysis to innovation projects, the analysis focuses on relatively dynamic capabilities, as projects are by definition characterized by their relative uniqueness and defined time span. A second aspect is the necessary distinction of the level of aggregation inside the observed networks. Even the term ‘project-based network’ could refer to different levels of aggregation, such as: individuals in the labour market (Granovetter, 1973); among groups inside an organisation; as well as in vertical (Hardstone, 2004) or horizontal relations (Acha and Cusmano, 2005) within an industry.2 This chapter focuses on the latter and, in particular, the innovation activities occurring inside and among multinational and national manufacturing firms and technological counterparts such as educational and research institutes in the sector.3 However, as in any study conducted within the social sciences, any bounded social network is influenced by the behaviour of the agents and interacts with other social networks in higher and lower levels (Giddens, 1979; Malerba, 2005). The interaction with foreign organizations and institutions, clients and a wider range of stakeholders would influence the behaviour and decision making of agents in a specific sector. Therefore, the analysis of the behaviour of the bounded network, in terms of locus, sector technology and/or institution, should not ignore the possible influence of the unobserved networks over the network examined. This is especially important in decentralized networks, as the usual coherence provided by a leading organization (a company or other form of association) is missing or tenuous. Third, it is useful to distinguish between normative and operational aims inside a knowledge network. While in most instances a knowledge network would involve certain formal aims that would define the network (e.g. an institutional set associated with regional/sectoral development, etc.), individual agents may operate under a different set of strategic aims and under a different set of organizational principles. Therefore, knowledge networks are not necessarily formed by relatively homogeneous entities that are coordinated together. The underlying complexity of interactions would mean that individual organizations would pursue specific operative aims that are not necessarily aligned with wider formal sectoral aims proposed for the network’4 and heterogeneity will be a key characteristic of a complex sectoral network, resulting in a wider diversity of aims in individual organizations (Dosi et al., 2000; Malerba, 2005). At the same time, rather than profit maximizing, organizational development is influenced by uncertainty and path dependence (Simon, 1979) – a result of the endogenous formation of the bounded networks defined by previous interactions in innovation projects as well as the wider range of organizational and social structures.
62
Sectoral systems of innovation and production
However, despite accepting historical and sectoral heterogeneity as a fact, this chapter does assume that there is a possibility of identifying some general contingencies in the evolutionary process taking place within knowledge networks. Careful consideration should allow the extraction of common configurations emerging in these project-based knowledge networks (Meyer et al., 1993), and further research into these configurations should allow examination of how these configurations are contingent on the characteristics of the knowledge base (Tidd, 1997).
3.
PROPOSITIONS
This section proposes three relationships between the knowledge base in innovation projects, the role of different types of governance mechanisms and the channels of inter-organizational knowledge flow. A brief examination of the theoretical basis for each one of the propositions is presented in turn. Proposition 1 The balance between markets and hierarchies in innovative activities is influenced by the knowledge base and the availability of disperse resources in the knowledge network. Companies develop internally new knowledge or collaborate with partners, becoming active actors in different knowledge networks inside the sectoral innovation system. However, it is important to note that networks do not emerge in the same way in different types of activities. The decision to integrate vertically specific types of knowledge or use partners in the network would fundamentally determine the characteristics of the knowledge network formed. Therefore, an investigation of the type of activities internalized and outsourced in innovation projects provides a necessary first step in the examination of the bottom-up evolution of the project-based knowledge network and its characteristics. The first proposition argues that two key elements would significantly influence the boundaries found between firms and possible technological partners in project-based knowledge networks: the type of knowledge activity and the availability of external resources (i.e. dispersed resources inside the knowledge network). According to the transaction cost theories of the firm (Coase, 1937; Williamson, 1985), companies will be especially interested in developing certain types of innovative activities in-house, where the costs of searching for and identifying appropriate partners and developing and enforcing
From innovation projects to knowledge networks
63
appropriate contracts are high. In contrast, companies tend to use external sources of knowledge in innovative activities when they cannot fully appropriate from their own investments. Particularly, companies would underinvest in some activities such as long-term research, training and other infrastructure. In such cases, the social benefits derived from these activities would provide a fundamental rationale for government intervention (Arrow, 1962). The formation of public goods would allow individual companies to access the qualified human resources using the labour market and infrastructure or information services provided by universities and research institutes. Companies, on the other hand, would target investment at industrial R&D where they would be able to appropriate directly from their investments as long as an adequate intellectual property rights regulatory framework was in place. It is essential however to recognize that the boundaries of the firm inside sectors are more complex than simple black-and-white distinctions between public and private knowledge (Nelson, 1989). Given the increasingly interactive nature of knowledge creation within sectors the distinctions between producers and users of knowledge are increasingly fuzzy (Geuna et al. 2003). Resource-based theorists complement this discussion, arguing that the internal importance of the accumulation of technological capabilities inside organizations is a necessary way of identifying possible technological opportunities and possible technological sources (Teece, 1994; Penrose, 1995). Industrial networks are a result of the companies’ needs to grow, balancing their internal growth with the resources available outside the firm (Pavitt, 2001). The industrial product development networks would not be driven only by exogenous factors (e.g. technology created in/absorbed from universities and research institutes), but primarily by the endogenous differentiation of the capabilities accumulated by firms in the industrial structure (Nelson, 1994; Gulati, 1999; Gulati and Gargiulo, 1999). Capabilities outside the firm would not be a substitute for internal capabilities, as organizational learning would allow for economies of repetition and the formation of comparative advantages (Brusoni et al., 2001). Balancing accumulation of internal capabilities and exploitation of external sources is at the centre of the firm’s technological renewal and diversification. In many sectors, and in particular ICT, the boundaries between different agents are blurring, given the extensive need for inter-organizational linkages (Antonelli et al., 2000). The governance of innovative activities becomes increasingly diverse and complex in order to coordinate the knowledge flows between different public and private agents. Projectbased knowledge networks would be fundamental to combine the ability to accumulate internal capabilities within firms with governance structures
64
Sectoral systems of innovation and production
that would allow a wider reconfiguration of the capabilities inside a network of organizations. An analysis of the project-based knowledge can provide a way to avoid the dichotomy between market and hierarchy in the analysis of industrial organization. Although companies will tend to integrate vertically the activities that provide them with comparative technological advantage, they will also need to integrate with external sources as new opportunities arise, interacting with other groups and firms. Proposition 2 Different types of knowledge base require specific governance mechanisms resulting in long-term specialization in the knowledge network. Organizations specialize in order to develop comparative advantages resulting from a specific match between knowledge base and governance structure (Pavitt, 1998). In contrast to a normative definition of the functions/technological areas of individual organizations inside the sector, we consider that different configurations would emerge mainly as the result of interaction among actors that compete as well as collaborate in different forms (Miller, 1986). Innovation projects became fundamental organizational forms both as a way to solve specific scientific problems in specialized functional departments and as a way to integrate knowledge from different sources in dedicated project-based organizations. The literature on organizational contingencies describes many advantages and disadvantages of specific organizational forms: ●
●
●
●
Strong hierarchical ties within firms are required to develop scale and scope in technological capabilities, allowing diversification into new market opportunities (Chandler, 1990; Gann and Salter, 2000). The project-based organizations are necessary to cope with uncertainty in production, fast-changing client needs and integrating dispersed knowledge into complex products and systems. However, they may be inefficient ways to accumulate knowledge that would be dispersed across projects (Hobday, 2000). Subsidiaries of multinational companies would make use of different matrix structures to become flagships in emerging economies and tap into new knowledge while diffusing knowledge inside the organization (Rugman, 1997; Ernst and Kim, 2002). Public investments in large high-tech projects may develop centres of excellence in specific areas and are considered a source of dynamic comparative advantages in sectors and countries, fundamental to retain competitiveness in a world with low trade barriers
From innovation projects to knowledge networks
65
and an increasing division of labour (Dunning, 1998; Patel and Vega, 1999). Given the diversity of organizational mechanisms observed in different sectors and cases, identifying empirically how the knowledge base determines the distributed governance structures within sectors is an important research question. Following a technology deterministic perspective (Woodward, 1965; Perrow, 1970; Thompson, 2003), we could expect that different types of knowledge should be associated with specific types of stakeholders, organizational structures and behaviours inside the knowledge network formed by firms and technological partners. There is no best practice, as different organizational forms inherited specific advantages and disadvantages related to their internal dynamics. A wider range of governance mechanisms connecting the innovation process within firms with knowledge in universities and research institutes needs to be considered in order to support the identifying and exploiting of technological opportunities within distributed networks (Fombrun, 1986; Rothwell, 1994; Powell et al., 1996). Therefore, quantitative analysis on the process of specialization in sectors is fundamental, as it might unveil differences and complementarities between the role played by national and multinational firms, universities and research institutes in the decentralized innovation process. It can shed light on the source of leadership in specific activities and help in the identification of look-ins and constraints in the complex decentralized knowledge governance in sectors. Proposition 3 Different types of knowledge base require different types of inter-organizational channels, limiting the possible knowledge flow to specific communities of practice. The third proposition investigates the process of knowledge flow inside the project-based knowledge network. The process of knowledge flow in sectors is sometimes simplified as a simple matter of diffusion or knowledge spillover. However, knowledge is not diffused evenly throughout the knowledge network. Given the variety of findings from econometric evidence in different sectors, it is necessary to recognize that the underlying channels developed among actors inside the sectoral network are fundamental determinants of the direction and speed of the knowledge transfer in sectors (Breschi and Lissoni, 2001; Narula and Zanfei, 2003; Bell and Marin, 2004). The knowledge flow is influenced by the costs of codification and the diffusion of knowledge among partners as well as the conflict of
66
Sectoral systems of innovation and production
interests in the process of searching for and following technological opportunities (Teece, 1989; Cummings and Teng, 2003). The social engagement in communities of practice has been recognized as the fundamental process by which individuals and organizations get to know what they know and by which they become who they are (Walker et al., 1997; Wenger, 1999). The organization of specialized communities of practice inside sectors is a fundamental dimension in understanding the diffusion of tacit knowledge (Brown and Duguid, 1991). The reuse of knowledge will be influenced by specific professional norms and practices developed in previous relations as well as by the power of receptors to bend rules to their advantage along the dynamic development of the network (Dyer and Singh, 1998; Lane and Lubatkin, 1998; Knight, 2002). Ultimately this will result in different inter-organizational channels in which specific types of knowledge are diffused (Brown and Duguid, 2001; Swan et al., 2002). The position of the individuals and organizations in different networks would provide them with a strategic advantage. As technological opportunities change over time, organizations need to recombine the accumulated knowledge or ‘transfer’ new types of knowledge inside and outside the organizational boundary (Hansen et al., 2005). Therefore, identifying the overlapping structures of specific communities of practice in different types of innovative activities allows detection of the number of actors participating in multiple networks (i.e. gatekeepers/boundary spanners) (Tushman and Katz, 1980) and which kind of organisational channels are used for the transfer of similar types of knowledge. The role played by the key actors involved in different networks can significantly determine the knowledge flows in the sector. The involvement of specific types of organizations in different activities and the isolation of these same organizations from different types of knowledge can provide insight into the direction and possible misalignments in the diffusion of knowledge (von Tunzelmann, 2004). The behaviour of the key organization in the network becomes fundamental in explaining the evolution of the network and in the translation of knowledge along the different steps involved in the decentralized innovation process.
4.
GENERAL CHARACTERISTICS OF THE DATABASE
A useful way to define the network is in terms of the broad institutions, the actors, the ties and the content of the interactions involved in it (Malerba, 2005). This chapter delimits its analysis to the network formed
From innovation projects to knowledge networks
67
by innovation projects declared under the tax scheme developed in the Brazilian ICT sector, called ‘ICT Law’. The tax scheme defined R&D obligations proportional to sales in the national market in exchange for different types of tax exemptions/waivers for manufacturing companies’ products. In order to be entitled to the tax scheme the companies were obliged to invest approximately 5 per cent of their national turnover in innovative activities.5 Ex post, the activities conducted should be described in structured project-type forms and in turn audited by the regulatory governmental agency (SEPIN) connected to the Brazilian Ministry of Science and Technology. Through a collaboration agreement with the Brazilian Ministry of Science and Technology, the database of projects used for administrative purposes was codified for this research. While adhering to the confidentiality requirements of the contract, this research uses the normalized procedure for collecting data from the companies as a way of exploring the relation between the types of project and the organization of the knowledge network.6 This unique database of projects contains information from the executor of individual projects in order to identify the process of knowledge creation in different organizations and transactions among firms and technological partners to identify the process of inter-organizational knowledge flow. In terms of projects, the dataset contains 10 088 projects executed under the Brazilian ICT Law between 1997 and 2003 (an average of 1261 per year). The costs of projects expanding beyond one specific year needed to be declared separately for the different years. The projects total an amount of R$1.6 billion executed internally by the companies and R$1.1 billion executed in partnership with universities and technological institutes (annual average of R$358.1 million) (see Table 3.1). In terms of actors, the dataset involves 211 companies and 181 educational and research institutes operating under the Brazilian ICT Law for the period 1997–2003. These actors are located throughout the entire Brazilian territory with the exception of the Manaus Free-Trade Zone, which receives specific incentives to manufacture and for R&D activities. The nodes of the network are companies and their ‘technological partners’. The companies could be subdivided into national and multinational companies with local manufacturing of products operating under the incentives (usually products that integrate advanced electronics, such as computers, mobiles and telecommunication equipment). In turn, the technological partners could be subdivided into organizations that would fit the definition of educational and/or research institutes from either public or private ownership. The regulation defined that a specific part of the investments (approximately 40 per cent) should be conducted
68
Table 3.1 Total
Sectoral systems of innovation and production
Longitudinal distribution of the projects 1997
1998
1999
2000
2001
2002
2003
Average
Total
Investments 304.3 346.8 389.5 560.4 249.6 349.4 306.3 358.1 2864.4 (million reais) Number of 1194 1381 1439 1741 783 1235 1055 1261 10 088 projects Average 2421.5 2738.8 2907.5 3868.1 4555.0 4818.0 8799.7 3665.5 33 774.1 project size (thousand reais) Equiv. 2637.2 2823.0 2666.2 3582.1 1535.3 2090.1 1563.6 2355.2 19 252.6 staff/FT * Notes: * Estimated number of full-time staff (direct + indirect HR costs)/(Average cost man/ Hour × 2000).
with technological partners in an explicit attempt to promote university– industry linkages. These partners were especially important in the regulation that aimed to reinforce these organizations as the key nodes in the sector. The database of projects contains details on the costs of innovative activities both inside companies and with technological partners. As the regulation does not define the type of activities that should be conducted inside the firm boundaries or with partners, this database provides a useful source for investigating the firm decision making between integrating vertically and using the network of partners to conduct specific types of activities. In terms of ties, the knowledge flows are developed on the basis of more than 35 000 transactions within the projects between firms and educational/technological organizations, creating 948 ties between these 392 nodes. These transactions are used to operationalize the flow of knowledge among the organizations in the network. There were also transactions with other companies, creating a wider, open network (commercial software companies, suppliers of equipment and training abroad, and other organizations not classified as ‘technological partners’ inside the network). However, the analysis of these transactions would add another layer of complexity and is therefore beyond the scope of this chapter. Table 3.2 summarizes the boundaries of the knowledge network under examination following the elements discussed in Section 2. The definition of the type of activities is connected to the definition used in the standard procedures, namely investments in laboratory and
From innovation projects to knowledge networks
Table 3.2
69
The knowledge network under the Brazilian ICT Law
Dimensions
Description
Institutions
The Brazilian ICT Law – Manufacturing companies operating under the ICT were required to invest approximately 5% of the national sales in innovative activities (2.3% needed to involve a research and/or educational institute) in order to benefit from tax incentives. It resulted in more than $2 billion invested in innovation projects between 1997 and 2003 (period under analysis) 214 manufacturing firms of products under the incentives (51 foreign companies and 163 domestic companies) 182 technological partners that met the regulation requirements (47 private research institutes, 20 public research institutes, 75 private educational institutes and 40 public educational institutes) Innovation projects allowed under the incentives were classified using the following categories: laboratory and infrastructure for S&T, quality systems for R&D, training in S&T, technological services, development of products in hardware, software, semiconductors, middleware and production processes, as well as research activities. More than 35 000 transactions within innovation projects conducted under collaborative agreements between firms and technological partners
Actors
Knowledge
Knowledge flows
infrastructure for S&T, quality systems for R&D, training in S&T, technological services, development of products in hardware, software, semiconductors, middleware,7 and production processes, as well as research activities. This categorization at project level represents an advantage in terms of defining the knowledge base independently from the final product classification (e.g. Pavitt taxonomy, most of the sectoral system studies), as it allows the existence of multi-technology firms (Granstrand and Sjolander, 1990). Table 3.3 shows the distribution of ties according to the type of activities. It also summarizes some basic statistics about the network in terms of investments in projects, the number of firms, the number of ties, and the strength and density of the network divided by the different activities. Table 3.3 also contains some details about the density and concentration in the different networks. Foreign companies represented 72 per cent of the total investments, while domestic companies were 28 per cent. The concentration is especially
70
Sum of investments (Million reais) (with partners) Number of firms (with partners) Number of partners Number of ties (>Million reais) Tie strength (Thousand reais) Average Tie strength (Thousand reais) Maximum Density Concentration (10-firm ratio) Concentration (5-firm ratio) Concentration (3-firm ratio) Concentration (1-firm ratio)
Dimension
Infrastructure 169.7 103.7 142 64 96 174 18 570 11 584 407 73% 42% 26% 9%
Quality 118.2 27 170 67 52 120 5 174 3349 105 53% 36% 26% 10%
Technological services 84.7 65.8 104 76 71 162 12 387 20 957 258 72% 51% 40% 25%
Training in S&T 159.5 100.4 177 87 117 240 20 388 28 565 394 70% 51% 41% 25%
Semiconductors 44.7 4 30 15 18 22 1 189 1427 17 99% 97% 95% 72%
108.9 13.5 140 44 54 90 3 145 1818 53 63% 47% 38% 23%
Production processes
Table 3.3 Descriptive statistics about the ‘ICT Law’ knowledge network, 1997–2003
Hardware 203.4 46.3 191 81 71 141 8 304 7300 181 63% 48% 34% 13%
System 621.7 212.4 234 127 92 230 31 799 28 188 834 64% 49% 36% 16%
Software
838.3 121 385 97.2 271 195 157 111 140 121 425 304 56 23 830 309 58 622 9229 1512 381 70% 65% 45% 48% 29% 37% 12% 18%
Research
From innovation projects to knowledge networks
71
high among the top 20 companies. The top 20 represent 73 per cent of the total investments. Of these 20 companies, 16 are subsidiaries of foreign multinational companies, accounting for 64 per cent. A similar concentration can be observed among the receivers of investments. In relation to the proportion of the resources allocated to technological partners by the firms, approximately 60 per cent of the total investments went to private research institutes, followed by private educational institutes (18 per cent), public research institutes (12 per cent) and public educational institutes (9 per cent). Figure 3.1 provides a visual representation of the knowledge networks divided by the different activities. Companies are represented as circles and technological partners as squares. Domestic companies are represented in white, foreign companies are represented in blue, educational institutes are in red and research institutes in black. The diameter is proportional to the sum of innovation projects conducted by the specific organization during the period between 1997 and 2003. Before entering into the characteristics of these networks, it is useful to observe the relevance of the innovation projects database compared to the total investments in innovation in the Brazilian telecommunications and computers sector. One way to proceed is to compare the results with an external measurement of the total investments in R&D conducted by these two sectors. The total investments in R&D in the telecommunications sector and the computer sector by private companies as assessed by the PINTEC (Brazilian innovation survey) were R$627 million in 2000 and R$637 million in 2003 according to the two innovations surveys conducted in the Brazilian ICT sector (Brasil, 2006). In addition, the innovation survey estimated that the total outsourcing of R&D was R$153.9 million in 2000 and R$184.2 million in 2003. From these figures in this section, it is possible to estimate that the SEPIN database contains on average more than 55 per cent of the investments in R&D in the computer and telecommunications sector (the average annual investment under the ICT Law was R$386 million for the entire period). In addition, more than 85 per cent of the innovation projects outsourced occurred inside the regulatory framework.8 Although there are some differences in the concept used to classify R&D in the two databases, the overall number obtained via these two different databases illustrates two general observations about the dataset: (i) There is possibly more R&D activity within companies in the sector, as it contains a much larger sample, such as software companies and services that do not have a manufacturing production system with products/minimum standards required by the regulations. Anyway, the number of projects in the dataset is indeed a significant proportion. (ii) Almost the totality of
72
Sectoral systems of innovation and production Research Networks
Laboratorial Infrastructure Network
Training in S&T Network
Quality Systems Network
Technological Services Network
Source:
Multinational Companies
Educational Institutes
Domestic Companies
Research Institutes
Based on MCT/SEPIN data using NetDraw 2.37 (Borgatti, 2002).
Figure 3.1
Knowledge networks in the Brazilian ICT sector divided by type of activity, 1997–2003 – complete
From innovation projects to knowledge networks Semiconductors Network
Production Process Network
Hardware Network
Middleware Network
Software Network
Figure 3.1
Multinational Companies
Educational Institutes
Domestic Companies
Research Institutes
(continued)
73
74
Sectoral systems of innovation and production
the outsourced R&D in the computer and telecommunications sector was conducted under the regulation. Therefore, in general, we assume that the project and the ties pointed to here do provide an important measurement of the investments that the companies would make inside the limits of the sector under analysis.
5.
RESEARCH METHODS
The methods used for the investigation of each one of the propositions are presented in this section. Proposition 1 The balance between markets and hierarchies in innovative activities is influenced by the knowledge base and the availability of dispersed resources in the knowledge network. The project-based networks in different activities were analysed using the trend (two years’ average) for the amount of knowledge created in specific innovative activity and the locus of execution of the projects (firms or technological partners). The significance of the differences between the vertical integration in different activities was verified using an ANOVA test (Appendix). The type of knowledge-related activity was defined as the total investments in specific innovative activity in relation to the total investment. Managers classified individual projects ranging from infrastructure to R&D, technological services, training, hardware, middleware, software, semiconductors, process technology, other types of product development, or research. Vertical integration in innovative activities was defined as the percentage of the total costs in innovation managed directly by the company in contrast to the sum of the costs of the projects whose management was outsourced to an external organization. In each project, managers assigned the organization responsible for the project management. The availability of dispersed knowledge was measured based on the percentage of total investments made in the particular type of innovative activities both by companies and by technological partners. This provides a way to estimate the relative amount of “knowledge” generated inside the network. Proposition 2 Different types of knowledge base require specific governance mechanisms, resulting in long-term specialization in the knowledge network.
From innovation projects to knowledge networks
75
A specialization index was adapted from the revealed technology advantage (RTA) index.9 In our case, we use the value of projects conducted by the organization to arrive at the project-based revealed technological advantage (PRTA) index calculated for the different types of agents (“types of governance mechanism”) for the different types of knowledge activities. The project-based specialization index (PRTA) could be defined as: Pij b Si Pij PRTAij 5 Sj Pij a b Sij Pij a
The governance mechanisms were divided among foreign companies, domestic companies, public and private research institutes, and public and private educational institutes. Pij was the costs of the project executed by organizational type i in knowledge related activity j. As in the traditional RTA, values greater than one suggest that an organizational type is comparatively specialized in the innovative activity in question relative to other organizational types (as it conducted more projects in this activity than the general average for the group), while values less than one are indicative of a position of comparative disadvantage. This procedure would allow one to control for the general concentration of specific organizations as well as the rules that define broader proportions that should be spent in companies and technological partners. Proposition 3 Different types of knowledge base require specific types of inter-organizational linkages, limiting the possible knowledge flow to specific communities of practice. In order to test this hypothesis, a correlation was used to analyse the interdependence between the structures of the different networks. In particular, the Quadratic Assignment Procedure (QAP)10 is used for all the possible combinations between the ten knowledge networks. Each network structure was represented by a valued matrix (Aikk), where i is the type of activity and k is the number of organizations in the network. In this case, k was constant and equal to 392, as there are 211 firms and 181 technological partners in the network. The values of these networks were the sum of the transactions among partners (i.e. the valued network). These values were normalized using the natural logarithm in order to obtain a normal distribution among the technological partners (Appendix). The mathematical procedure could be defined as:
76
Sectoral systems of innovation and production
Corr (A1kk 0 A1kk) Xii 5 ° ( Corr (Aikk 0 A1kk)
c Corr (A1kk 0 Aikk) ¢ 11 f ( c Corr (Aikk 0 Aikk)
The result of the correlation was a matrix (Xkk) containing the strength of the overlapping between each pair of networks.
6.
RESULTS
This section describes the results obtained from the empirical test of the propositions. The aim is not just a validation of the theoretical propositions but also an exploration of the specific dynamic occurring in the Brazilian ICT sector during the period and the role played by different organizations driving change in the sectoral innovation system. Proposition 1 The balance between markets and hierarchies in innovative activities is influenced by the knowledge base and the availability of dispersed resources in the knowledge network. The first proposition is related to the balance between hierarchies and markets in innovation activities inside the network. In order to explore the dynamic of the formation and interaction among the knowledge networks, Figure 3.2 shows the trends in the accumulated technological capabilities in the different technologies (estimated based on the percentage of the total investments) and the degree of vertical integration of the innovative activities (based on the number of projects controlled by the company compared to those outsourced to technological partners). Firstly, Figure 3.2 reinforces the visual inspection of the networks represented in Figure 3.1. (i) There are basically no or very incipient networks related to semiconductors, the production process and hardware. (ii) Wider networks with relatively weak ties were formed via activities such as training, technological services and research. (iii) There are strong-tie networks in middleware and, most of all, software, where considerable governance mechanisms could be expected through the technological partners. Secondly, these three groups are significantly different in terms of boundaries of the firm in innovative activities and accumulated capabilities, as supported by the ANOVA test (Appendix). The different characteristics of these groups would allow the proposition of the following taxonomy:
77
90%
80%
50%
40%
30%
Vertical Integration of innovative activities
70%
20%
0% 0% Outsourced
10%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Figure 3.2 The size and boundaries of the knowledge networks in the Brazilian ICT sector
In-house
100%
PP, 2002.2003 SE, 1997 QS, 2002.2003
QS, 1998.1999 QS, 2000.2001 PP, 2000.2001 Tr, 1997 PP, 1998.1999 PP, 1997 SE, 2002.2003
Tr, 2000.2001 IL, 2000.2001 IL, 2002.2003 Tr, 1998.1999 IL, 1998.1999 Re, 2000.2001 TS, 1998.1999 Re, 2002.2003 Tr, 2002.2003 Re, 1997 Re, 1998.1999 IL, 1997 TS, 2002.2003 TS, 1997 TS, 2000.2001 SE, 2000.2001 SE, 1998.1999
60%
Sy, 2002.2003
HW, 2000.2001
HW, 2002.2003
HW, 1998.1999 QS, 1997 HW, 1997
SW, 2000.2001
SW, 1998.1999
Sy, 2000.2001
Sy, 1998.1999 SW, 1997
Sy, 1997
SW, 2002.2003
Peercentage of Total Investments
Sy System (Dev) SW Software (Dev) IL Infra& Labs Re Research TS Tech. Services PP Production HW Hardware (Dev) SE Semiconductors( Dev) Tr Tech Training (Dev) QS Quality System (Dev) 2 per. Mov. Avg. (Sy) 2 per. Mov. Avg. (SW) 2 per. Mov. Avg. (IL) 2 per. Mov. Avg. (Re) 2 per. Mov. Avg. (TS) 2 per. Mov. Avg. (PP) 2 per. Mov. Avg. (HW) 2 per. Mov. Avg. (SE) 2 per. Mov. Avg. (Tr) 2 per. Mov. Avg. (QS)
78
Sectoral systems of innovation and production
Number of ties per million in R&D 0 1 2 3 Infrastructure Technological Services Training S&T Semiconductors Production process Hardware System Software Research Figure 3.3 ●
Number of ties per million reais in different technologies
Enabling networks (low levels of investment, low vertical integration). Companies tended to use the market in these innovative activities with relatively lower investments. The group of points at the bottom right refers to activities such as training in science and technology, technological services (e.g. metrology, certification) and research activities. Investments within innovation projects tend to be smaller and almost entirely outsourced. Only a smaller part of the investments in technological services (36 per cent), training in S&T (45 per cent) and research activities (22 per cent) were conducted internally. To some extent, investments in infrastructure and laboratories could also be associated with this group, although they have a less significant difference in terms of firm boundaries (46 per cent). These ties with external organizations were also weaker. In fact, when considering number of ties in relation to the total investments, there are just 1 tie/million reais in infrastructure and laboratories projects, 1.6 ties/million reais in training projects, 2.1 for technological services and 2.5 for research activities. These numbers contrast significantly with averages of 0.4 to 0.6 tie per million reais invested in the other ‘product development’ networks (Figure 3.3).
From innovation projects to knowledge networks
●
●
79
Although weak ties are usually assumed to be related to research projects (as companies would look for technologies opportunities based on these types of activities), it is clear that these other networks also have a fundamental connection between firms and a large number of different supporting organizations. These ties would be important for the development of human resources, technological information and so on. Some of these supporting organizations (providers of training, technological services, infrastructure, etc.), which are sometimes neglected within innovation studies, need to be understood in more detail, as they cannot be assumed to be available in most developing countries. Developing networks (low levels of investment, high vertical integration). Individual companies tended to conduct most of their product development projects in-house whenever there were limited total investments in specific technologies. The points at the bottom left were mainly composed of three groups of innovation activities during different time periods: product development using hardware, semiconductors and production process technology as well as quality systems. An important aspect of these three arrows is the strong internalization of the technological investments inside firms, and although some linkages could exist with international partners or other stakeholders there was very limited horizontal collaboration with partner technological institutions. This indicates that in these networks the companies resisted using external sources of technology. This analysis reinforces the visual analysis (Figure 3.1) that the formation of dispersed governance mechanisms has been limited in these activities. There are also very different trends, as shown by the arrows. The arrow related to semiconductors shows incipient, but increasing, initiatives to accumulate technological capabilities inside the companies. An opposite trend is observed in relation to production technology that has decreased and outsourced activities. The arrow and the dots related to hardware show that there is an upward movement, although it has been turbulent throughout the period, probably as a result of the instability in the initiatives undertaken by different companies in this type of technology. Developed networks (high levels of investment, intermediate vertical integration of innovative activities). A different portrait could be developed around the dynamic involving the two largest networks: the networks formed by product development projects using middleware and software technology. They are both characterized by higher levels of investment and an intermediate level of desegregation of the activities between hierarchies and partnerships.
80
Sectoral systems of innovation and production
The analysis of these trends over time shows that the development of the network evolved in opposite directions. From this trend, we can infer that the established and newcomer companies have shifted their investments from middleware to software during this period. In the middleware network, while the investments in middleware technology were reducing, companies tended to retain internal projects rather than consolidate. At the same time, the companies that were increasing their investments in software identified existing capabilities available in partners, and the general vertical integration decreased. In the two large areas of investment (software and middleware), the data implies that there was considerable scope for governance structures with strong ties among partners. These lower levels of vertical integration in relation to other types of product development knowledge networks support the hypothesis that, as there are increasing resources available in the network, governance mechanisms would tend to emerge and integrate disperse resources. When these capabilities decrease – as in the case of middleware – the level of vertical integration tends to increase simultaneously. This supports a resource-based view, where the evolution of the knowledge networks is mainly connected with endogenous differentiation among companies in product development activities. The number of ties per total investment is significantly lower in product development when compared to training, technological services and research activities. It suggests that firms have fewer, but strong, ties in product development, while companies will also tend to have more, but weaker, ties in relation to technological services, training and research activities. These findings therefore tend to support the first proposition that both the type of activities and the need to integrate disperse resources influence the boundaries between firms and technological partners in innovative activities. These findings also contain a dynamic portrayal of the dynamic occurring inside the sector during the period. In the development of new products, the technological opportunities identified by the companies changed considerably during the period, mainly from middleware to software. Clearly companies inside the framework identified limited opportunities in microelectronics, hardware and production processes, showing that the same institutional framework can result in very different investment behaviours. Therefore the understanding of the technological trajectories opened to specific industries cannot be overstated when examining firm behaviour.
From innovation projects to knowledge networks
81
Proposition 2 Different types of knowledge base require different governance mechanisms, resulting in long-term specialization in the knowledge network. A next step is to expand the analysis from the simple bilateral relation in terms of vertical integration, to the analysis of the emerging role played by different actors inside the knowledge network. Table 3.4 shows the measurement of the project-based revealed technological advantage (PRTA) for the different governance mechanisms. In Table 3.4 it is possible to observe patterns of specialization in the different nodes, thereby identifying how the knowledge base was associated with different governance mechanisms in the network (i.e PRTA>1). Analysing the results, some patterns of specialization emerged in the knowledge network. While domestic companies focused their investments in middleware and hardware (as well as relatively higher investments in quality systems), foreign companies were predominant in emerging software network (1.08 for multinational against 0.69 in domestic firms). The latter also undertook important initiatives in the smaller semiconductors (1.77) and production process (1.79) activities. The results provide a strong indication that, while domestic companies tend to be more connected to their manufacturing base in hardware (2.18), multinational companies tend to be more capable of diversifying into distinct competences in middleware (1.38) and software projects. The organizational characteristics of the MNCs may have allowed subsidiaries to develop capabilities in niches in the international division of labour inside the corporation operating in global projects and disconnecting themselves from the manufacturing basis and local market. Among the technological partners, private research institutes became key governance structures spanning many activities such as research activities (1.33), software (1.17), training (1.46), technological services (2.39) and development of labs and technological infrastructure (1.31), while public research institutes became highly specialized in research (3.08) and technological services (1.57). It is possible to speculate on the organizational characteristics that define these differences. The indication is that public funds tended to complement private investments in long-term infrastructure and the research personnel required for these activities, creating a relative comparative advantage for these organizations. Meanwhile, the governance of these organizations and their policies could be too rigid to adapt to the short-term requirements of companies, as private research institutes became fundamental inter-organizational linkages in the software project-based networks. Further qualitative research should help corroborate these conclusions.
82
Foreign companies Domestic companies Private research institutes Public research institutes Private educational institutes Public educational institutes
Specialization index
Table 3.4
2.15
4.98
75
40
1.33
47
3.08
0.58
163
20
0.25
51
1.85
0.17
0.49
0.41
2.18
0.96
0.51
1.02
0.89
1.17
0.69
1.08
Count Research Hardware Software
0.56
0.02
0.29
0.14
0.97
1.77
0.35
0.97
0.73
0.77
–
1.06
SemiMiddleconductors ware
0.31
0.20
0.45
0.22
0.77
1.79
Process
3.39
1.25
0.65
1.46
0.61
0.72
0.32
0.98
1.57
2.39
0.49
0.42
0.20
0.67
0.21
0.43
1.64
1.26
1.55
2.26
0.51
1.31
0.53
0.83
Training Technological Quality Infrastructure services
Revealed technology advantage of the different organizational mechanisms
From innovation projects to knowledge networks
83
Finally, in terms of educational institutes, both private and public organizations specialized in similar areas such as research, training and infrastructure (as possibly expected). Public educational institutes, a group composed mainly of federal and state universities, were particularly specialized in the research and training areas (4.98 and 3.39 respectively). Clearly, the public organizations developed their comparative advantage from their traditional role inside the structured national educational system financed with public resources. Interestingly, these traditional organizations did not tend to diversify into collaborative activities in the new technological areas, with the exception of hardware (1.85), where domestic companies also had their relative advantage. Although the public educational institutes kept their relative strength in hardware, they clearly lag behind in the emerging software trajectory. This pattern of specialization sheds new light on the distributed innovation process following the liberalization of the sector. Recent academic discussions in the sector have been heated as authors investigate different patterns, for instance that the process of liberalization resulted in decreasing capabilities in the cluster previously concentrated in domestic firms in Campinas (Szapiro and Cassiolato, 2003), the active role of policy and multinational equipment manufacturers in the sector (Mani, 2004), and the dependence of the innovation system in software on multinational companies in Brazil (Stefanuto, 2004). The pattern of specialization described above shows how these different governance structures co-evolved as a result of the mixture of technical change, foreign direct investment and sectoral policies. It could be observed from Table 3.4 that foreign companies, private research institutes and (to a small degree) private educational institutes could be considered key nodes integrating disperse capabilities inside the fast-expanding software project network in Brazil. The following proposition examines how these organizations may in fact result in distinct communities within the sector, restricting knowledge flows among specific agents. Proposition 3 Different types of knowledge base require specific types of inter-organizational channels (knowledge flows are limited to specific communities of practice). The next level of our analysis is to examine how the different knowledge flows occur inside and among different project-based knowledge networks. Using the payments within innovation projects as a proxy for knowledge flows among actors in different activities (see Figure 3.1 for the visualization of the networks), Table 3.5 shows the result of correlation among the different valued networks.
84
Sectoral systems of innovation and production
Table 3.5
QAP correlation among the knowledge networks developed in different activities 1
1 Infrastructure 2 Technological services 3 Training 4 Quality systems 5 Semiconductors 6 Production technology 7 Hardware 8 Middleware 9 Software 10 Others 11 Research
2
3
4
5
6
7
8
9
10
11
0.283 0.580 0.173 0.276 0.197 0.204 0.035 0.064 0.033 0.057 0.403 0.132 0.606 0.181 0.081 0.155 0.262 0.309 0.473 0.334
0.035 0.210 0.330 0.034 0.617
0.101 0.265 0.449 0.392 0.183
0.207 0.416 0.577 0.410 0.358
0.097 0.049 0.047 0.012 0.209
0.217 0.145 0.274 0.274 0.181
0.137 0.158 0.600 0.314 0.173 0.319 0.242 0.293 0.368 0.217
Notes: *All the correlations are significant at 0.01. QAP procedure developed in UCINET 6 (Borgatti et al., 2002)
The first clear result of the correlation is that the knowledge flows inside the project-based networks are not homogeneous, therefore supporting the proposition that different types of knowledge base require specific types of inter-organizational channels. Although the relationship between the different activities does exist, most of the networks presented in Figure 3.1 are significantly different from each other, as demonstrated by the relatively small correlation between the different networks in most of the cases. Different knowledge activities would create significantly different communities of practice that would co-evolve in the sector. The second set of results with empirical relevance refers to those networks that do have a relatively strong correlation. Establishing 0.5 as a threshold to a strong relationship, we could distinguish just five intertwined networks. These intertwined networks could be further grouped into three distinct communities of practice: ●
Training and infrastructure/training and production technology. The analysis suggests companies are connected with the same partners for the improvement of the infrastructure and training in new technologies. In addition, production technology was also particularly related to training in new technologies.
From innovation projects to knowledge networks ●
●
85
Research and technological services. Other channels became specialized in providing research activities and technological services (metrology) for the companies. It is interesting to note that, in general, research and technological services (possibly centres of excellence in different technologies) were not strongly related with the linkages involved in product and process development. Product development in software and quality systems/software and middleware. Specific channels became related to the improvement of quality systems in R&D (e.g. CMM certification) and the development of products in software. Here, it is also possible to observe a strong relation between the formation of the capabilities in middleware and software. Although this test does not allow us to attribute causality, the dynamic changes shown in Figure 3.1 reinforce the interrelation between the decreasing middleware project network and the growing software project network. While the newcomers (especially multinational companies) shifted their investments towards opportunities in software, private research institutes became key integrators between ‘old’ and ‘new’ capabilities.
The quantitative analysis of the structure of these communities of practice provides an interesting insight into the complex and multi-dimensional nature of the accumulation of knowledge inside organizations and the knowledge flows in sectors. Naturally, innovation projects involve many other types of interactions occurring inside the sector, for instance through the interaction with customers, suppliers and other functions of the company inside and outside the sector. However, even focusing on a subset of the relationships and activities inside the innovation process, this inductive approach shows how different governance mechanisms specialized in specific types of activities (proposition 2), requiring particular channels for the diffusion of knowledge (proposition 3). Given that a number of organizations may take the lead along the different activities and possible conflict of interests between these different organizations, the decentralized innovation projects in sectors remain strongly susceptible to network failures. There are also constant opportunities for entrepreneurs inside and outside existing organizations to align the interests of different networks.
7.
ANALYSIS AND IMPLICATIONS
A dynamic understanding of the relationships between the knowledge base of sectors, the main actors and the knowledge flows in sectors is a key
86
Sectoral systems of innovation and production
challenge for innovation management and policy. This chapter contributes to the empirical literature on the organization of sectoral innovation systems using innovation projects to validate three contingencies between the knowledge base and the structure and dynamics of the knowledge networks in sectors. This chapter uses longitudinal analysis of the boundaries between firms and technological partners in different innovation activities, a project-based index of revealed technological advantage for different governance structures (PRTA) and social network correlation techniques (QAP) to identify a number of principles of the co-evolution between innovation projects and knowledge networks. Three distinct propositions show that, rather than being cumulative (the same actors going from basic training, to product development, to research activities), the development of the system is multi-dimensional, where different governance structures are involved in specific activities in a process of specialization and differentiation. The first proposition states that the knowledge base influences significantly the boundaries of the innovative activities as well as the endogenous availability of resources inside the network. The results support the proposition that project-based networks would emerge more naturally from activities such as research, training, technological services and infrastructure. In these activities, the natural trend would be the formation of market-mediated ties with a plethora of organizations, accumulating capabilities in universities and public research institutes (enabling networks). Companies would be willing to maintain sporadic channels of interaction with many organizations in these activities. On the other hand, early initiatives in product development (developing networks) were associated with higher internalization levels. A mixture of vertical integration and strong interaction with external partners was only observed in middleware and software (developed networks) product development where a larger number of actors invested in the technology, suggesting an endogenous reconfiguration of the disperse capabilities inside the sectoral networks. The second proposition examines the process of specialization of governance structures inside the project-based knowledge network in different activities. A great variety of organizations coordinated available resources in new opportunities, forming complex inter-organizational sectoral governance structures. Domestic companies remained focused in hardware and middleware (close to manufacturing activities), while multinational companies (matrix R&D units) connected to private research institutes (project-based organizations) were important in the emerging software technology. Public research centres and educational institutions (usually functional structures) became central in training and research activities. The different roles of these organizations reinforce the importance of
From innovation projects to knowledge networks
87
diversity of governance structures and the different mechanisms for interaction between public and private as well as domestic and multinational stakeholders inside the sectoral systems. The third proposition examines the patterns in the knowledge flows occurring between companies and technological partners. The low correlation between the networks in different activities shows that the different types of knowledge tended to flow in distinct communities of practice. There were, however, some connections between the types of activities developed between companies and technological partners in specific activities. In particular, three distinct strong correlations emerged between different communities of practice: (i) production process and laboratorial infrastructure/equipment were connected with training; (ii) research was correlated with the same partners involved in technological services; and (iii) product development activities in middleware and software were strongly correlated. The correlation between the different networks provides an interesting insight into the multi-dimensional governance of knowledge in sectors and the complexity involved in aligning interests dispersed in specific communities of practice. These findings have important policy implications for those involved in organizational studies and institutional design of sectoral policies. First, the results show that the breadth and depth of the definition for innovation projects influence significantly the kind of knowledge networks that will emerge within sectors. There are clear differences between innovation activities that would promote decentralized networks or vertical hierarchies inside the sector. However, simply associating these different activities with public and private knowledge may not be a helpful distinction, particularly if this is followed by an attempt to jump into recommendations about the type of activities that should be supported by public investments. In the initial moments of the formation of these networks, copying specific institutions from developed countries into the developing economy may be insufficient or even distort and hinder the cohesion and alignment of the knowledge networks in sectors. Balancing short- and long-run sustainability between public and private investments in these networks would certainly involve a constant process of institutional and organizational learning. The incentives that would promote the most adequate level of learning should consider the characteristics of the existing capabilities in the knowledge and the technological opportunities open to the industry. The analysis also shows the diversity and multi-level nature of the governance mechanisms needed in order to integrate multinational and national networks in high-tech sectors. While most of the justification for trade and science and technology policy relies heavily on econometric
88
Sectoral systems of innovation and production
measures to understand knowledge spillovers, measuring them in terms of the possible economic outcomes, it is necessary to recognize that knowledge spillovers are ultimately a result of the underlying organizational and political structures. In this sense, the organizational characteristics of the networks between multinational and national innovation systems may determine the occurrence and direction of the possible knowledge flows. Hence, there are important lessons to be derived from this experience in the Brazilian ICT sector. Multinational companies played a fundamental role in the accumulation of technological capabilities in new technologies, but their integration with existing networks and the diffusion of knowledge into existing social structures are not a homogeneous process. The institutional framework allowed space for organizational learning and the decentralized interaction between stakeholders with very different interests. However, a more precise identification of the ways multinational companies support or obstruct the development of knowledge networks in developing countries should be the focus of more detailed research. Finally, new methods for measuring the process of accumulation and the diffusion of knowledge could help the “management” of these decentralized knowledge networks. Along the way, compromise and adjustment between different interests should allow the identification of endogenous growth opportunities. However, different groups will naturally have disagreements about their relative past and potential contribution to the general performance of the sector. Indeed, the regulation of those organizations that will be involved in participation in the knowledge network is susceptible to political influence. The political strength of specific stakeholders may block the participation of other groups, resulting in a reinforcement of specific path dependencies. Certainly, an analysis of these project-based networks could help in developing more adequate sectoral strategies in the different developmental aims. Although the network was deeply influenced by the context of the tax regime that was in place between 1997 and 2003, there is no reason to constrain the methods used to this unique source of funding. A project-level analysis of the knowledge networks could encompass a wider diversity of institutional frameworks and organizational forms organized by projects. Although this level of analysis needs to be considered in terms of the costs of accountability, the analysis of the structure emerging from projects certainly could have a long-run pay-off as long as they guide organizational and institutional learning. In addition, a project-level analysis of sectoral systems raises theoretical and practical lines of enquiry that remain largely open. For instance, how do individual actors contribute to the vitality of these networks? How could changes in the rules promote a better allocation of resources in the decentralized network? Would a rule-based allocation, a discretionary
From innovation projects to knowledge networks
89
allocation or a combination of both improve the long-term processes of variety creation and selection inside the decentralized network? Which kind of interventions (or non- interventions) should be carried out in different stages of the development of the knowledge networks? Which mechanisms should be combined in order to promote wider knowledge spillovers, stronger growth, and sustainability in the networks? Further detailed analysis of the behavioural aspects of these networks as well as the developments in methodological tools needs to be considered in order to attempt to answer these questions. Finally, if we expect to understand the formation of knowledge systems in developing countries, and especially the interaction between the knowledge networks in multinational and domestic networks within sectors, we must go beyond the unproductive debate between nationalistically charged statements and ideologically influenced liberalizing reforms. We must take into account that the integration between networks requires complex alignment in multi-level governance structures (Kim and von Tunzelmann, 1998; von Tunzelmann, 2004) and at the same time that the organizational changes, learning and adjustments between conflicting interests are negotiated on a project-by-project basis (Nonaka and Takeuchi, 1995; Lam, 2000). Thus, investigating the project-based knowledge networks not just provides useful insights into the nature of the sectoral system, but could also provide a framework to discuss institutional interventions and organizational strategies to promote the sustainability of this fragile, complex and dynamic process.
NOTES 1. 2. 3.
4. 5. 6.
For a review of different uses of data to map knowledge flows see Meyer (2002). For some examples of individuals, groups, and industrial vertical and horizontal relations, see Hobday (2000), DeFillippi and Arthur (1998), Acha and Cusmano (2005) and Manning (2005). Although these specific linkages, sometimes identified in the literature as ‘university– industry linkages’, have a considerable amount of literature of their own, they rarely allow the examination of the networks and governance structures formed by the aggregation of individual ties. This is supported by the idea that the normative definition of functions within the innovation in developing countries is usually relatively weak, as has been shown by many empirical qualitative studies, e.g. Bell and Albu (1999). This percentage decreased slightly during the last three years of the analysis. See www. mct.gov.br/sepin for more details about the regulatory framework. Further details about the database will be provided in the author’s forthcoming thesis. Three datasets were accessed in the Brazilian Ministry of Science and Technology in Brasilia for three different periods under a non-disclosure agreement and for academic purposes only. The dataset was cleaned and integrated into the different levels of analysis. The consolidated data about the network was based on the dataset of the
90
7. 8. 9. 10.
Sectoral systems of innovation and production innovative projects developed by companies for the period 1997–2003, declared under the Brazilian ICT policy. The original classification was ‘System (hardware + software)’ characterizing projects in the interface. The term ‘system’ was substituted here for ‘middleware’ to avoid confusion with sectoral systems. It is supported by anecdotal information that highlights that outsourced R&D projects and the projects under the regulation are usually considered to be synonyms by the people interviewed. The index is usually used with patent and scientific publications. QAP correlation (number of permutations: 5000; random seed: 24 322).
REFERENCES Acha, V. and Cusmano, L. (2005), ‘Governance and co-ordination of distributed innovation processes: patterns of R&D co-operation in the upstream petroleum industry’, Economics of Innovation and New Technology, 14(1–2), pp. 1–21. Antonelli, C., Geuna, A. and Steinmueller, W.E. (2000), ‘Information and communication technologies and the production, distribution and use of knowledge’, International Journal of Technology Management 20(1), pp. 72–94. Arrow, K.J. (1962), ‘The economic implications of learning by doing’, Review of Economic Studies 29(3), pp. 155–173. Bell, M. (2005), ‘Towards the dynamic analysis of sectoral innovation systems in industrialising countries’, paper presented to Globelics 2005, St Petersburg. Bell, M. and Albu, M. (1999), ‘Knowledge systems and technological dynamism in industrial clusters in developing countries’, World Development 27(9), pp. 1715–1734. Bell, M. and Marin, A. (2004), ‘Where do FDI-related technology spillovers come from in emerging economies? An exploration in Argentina in the 1990s’, European Journal of Development Research 16(3). Bell, M. and Pavitt, K. (1993), ‘Technological accumulation and industrial growth: contrasts between developed and developing countries’, Industrial and Corporate Change 2(2), pp. 157–210. Birkinshaw, J., Nobel, R. and Ridderstråle, J. (2002), ‘Knowledge as a contingency variable: do the characteristics of knowledge predict organization structure?’ Organization Science 13(3), pp. 274–289. Borgatti, S.P. (2002), NetDraw: Graph Visualization Software, Analytic Technologies, Harvard, MA. Borgatti, S.P., Everett, M.G. and Freeman, L.C. (2002), Ucinet 6 for Windows: Software for Social Network Analysis, Analytic Technologies, Harvard, MA. Brasil (2006), Pesquisa Industrial Invoção Technológica (PINTEC) 2000/2003, Portal de Ministério de Ciência e Technologia. Breschi, S. and Lissoni, F. (2001), ‘Knowledge spillovers and local innovation systems: a critical survey’, Industrial and Corporate Change, 10(4), 975– 1005. Brown, J.S. and Duguid, P. (1991), ‘Organizational learning and communitiesof-practice: toward a unified view of working, learning, and innovation’, Organization Science 2(1), pp. 40–57. Brown, J.S. and Duguid, P. (2001), ‘Knowledge and organization: a social-practice perspective’, Organization Science 12(2), p. 198.
From innovation projects to knowledge networks
91
Brusoni, S., Prencipe, A. and Pavitt, K. (2001), ‘Knowledge specialization, organizational coupling, and the boundaries of the firm: why do firms know more than they make?’, Administrative Science Quarterly 46(4), pp. 597–621. Chandler, A.D. (1990), Scale and Scope: The Dynamics of Industrial Capitalism, Belknap Press of Harvard University, Cambridge, MA. Coase, R.H. (1937), ‘The nature of the firm’, in Williamson, O.E. and Winter, S.G. (eds), The Nature of the Firm, 1993, Oxford University Press, Oxford. Cummings, J.L. and Teng, B.S. (2003), ‘Transferring R&D knowledge: the key factors affecting knowledge transfer success’, Journal of Engineering and Technology Management 20(1–2), pp. 39–68. Davies, A. and Hobday, M. (2005), The Business of Projects: Managing Innovation in Complex Products and Systems, Cambridge University Press, Cambridge, New York. DeBresson, C. and Amesse, F. (1991), ‘Networks of innovators: introduction’, Research Policy 20, pp. 363–379. DeFillippi, R.J. and Arthur, M.B. (1998), ‘Paradox in project-based enterprise: the case of film making’, California Management Review 40(2), pp. 125–139. Dosi, G., Winter, S.G. and Nelson, R.R. (2000), The Nature and Dynamics of Organizational Capabilities, Oxford University Press, Oxford. Dunning, J.H. (1998), ‘Globalization, technological change and the spatial organization of economic activity’, in Chandler, A.D., Hagström, P. and Sölvell, Ö. (eds), The Dynamic Firm: The Role of Technology, Strategy, Organization and Regions, Oxford University Press, Oxford, New York, pp. xvii, 469. Dyer, J.H. and Singh, H. (1998), ‘The relational view: cooperative strategy and sources of interorganizational competitive advantage’, Academy of Management Review 23(4), pp. 660–679. Ernst, D. and Kim, L. (2002), ‘Global production networks, knowledge diffusion, and local capability formation’, Research Policy 31(8–9), pp. 1417–1429. Fombrun, C.J. (1986), ‘Structural dynamics within and between organizations’, Administrative Science Quarterly 31(3), pp. 403–421. Foss, N.J. and Klein, P.G. (2005), The Theory of the Firm and Its Critics: A Stocktaking and Assessment, 05-03, DRUID, Copenhagen. Gann, D.M. and Salter, A.J. (2000), ‘Innovation in project-based, serviceenhanced firms: the construction of complex products and systems’, Research Policy 29(7), pp. 955–972. Geuna, A., Salter, A.J. and Steinmueller, W.E. (2003), ‘Science and innovation: rethinking the rationales for funding and governance’ Northampton, MA: Edward Elgar Publishing. Giddens, A. (1979), Central Problems in Social Theory: Action, Structure, and Contradiction in Social Analysis, University of California Press, Berkeley. Giuliani, E. and Bell, M. (2005), ‘The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster’, Research Policy 34, pp. 47–68. Granovetter, M.S. (1973), ‘The strength of weak ties’, American Journal of Sociology, 78(6), 1360. Granstrand, O. and Sjolander, S. (1990), ‘Managing innovation in multi-technology corporations’, Research Policy 19(1), pp. 35–60. Gulati, R. (1999), ‘Network location and learning: the influence of network resources and firm capabilities on alliance formation’, Strategic Management Journal 20(5), pp. 397–420.
92
Sectoral systems of innovation and production
Gulati, R. and Gargiulo, M. (1999), ‘Where do interorganizational networks come from?’, American Journal of Sociology, March, pp. 177–231. Hansen, M.T., Mors, M.L. and Løvås, B. (2005), ‘Knowledge sharing in organisations: multiple networks, multiple phases’, Academy of Management Journal 48(5), pp. 776–793. Hardstone, G.A.P. (2004), ‘Capabilities, structures and strategies re-examined: incumbent firms and the emergence of complex product systems (cops) in mature industries’, Technology Analysis and Strategic Management, 16(2), 173–96. Hobday, M. (2000), ‘The project-based organisation: an ideal form for managing complex products and systems?’, Research Policy 29(7, 8), p. 871. Kim, S.-R. and von Tunzelmann, N. (1998), Aligning Internal and External Networks: Taiwan’s Specialization in IT, SPRU, Brighton. Knight, L. (2002), ‘Network learning: exploring learning by interorganizational networks’, Human Relations 55(4), p. 427. Lam, A. (2000), ‘Tacit knowledge, organizational learning and societal institutions: an integrated framework’, Organization Studies 21(3), p. 487. Lane, P.J. and Lubatkin, M. (1998), ‘Relative absorptive capacity and interorganizational learning’, Strategic Management Journal 19(5), pp. 461–477. Malerba, F. (2002), ‘Sectoral systems of innovation and production’, Research Policy 31(2). Malerba, F. (2005), ‘Sectoral systems of innovation: a framework for linking innovation to knowledge base, structure and dynamics of sectors’, Economics of Innovation and New Technology, 14(1–2), pp. 63–82. Mani, S. (2004), Coping with Globalisation: An Analysis of Innovation Capabilities in Brazilian Telecommunications Equipment Industry Intech, Maastricht. Manning, S. (2005), ‘Managing project networks as dynamic organizational forms: learning from the TV movie industry’, International Journal of Project Management 23, pp. 410–414. Marshall, A. (1891), Principles of Economics, Macmillan, London. Meyer, A.D., Tsui, A.S. and Hinings, C.R. (1993), ‘Configurational approaches to organizational analysis’, Academy of Management Journal 36(6), pp. 1175–1195. Meyer, M. (2002), ‘Tracing knowledge flows in innovation systems’, Scientometrics 54(2), pp. 193–212. Miller, D. (1986), ‘Configurations of strategy and structure: towards a synthesis’, Strategic Management Journal 7(3), pp. 233–249. Narula, R. and Zanfei, A. (2003), The International Dimension of Innovation, 9, MERIT-Infonomics, Maastricht. Nelson, R.R. (1989), ‘What is private and what is public about technology?’, Science, Technology and Human Values 14(3), p. 229. Nelson, R.R. (1994), ‘The co-evolution of technology, industrial structure, and supporting institutions’, Industrial and Corporate Change 3(1), pp. 47–63. Nonaka, I. and Takeuchi, H. (1995), The Knowledge-Creating Company, Oxford University Press, Oxford. Owen-Smith, J. and Powell, W.W. (2005), ‘Knowledge networks as channels and conduits: the effects of spillovers in the Boston biotechnology community’, Organisation Science 15(1), 5–21. Patel, P. and Pavitt, K. (1995), ‘Patterns of technological activity: their measurement and interpretation’, in Stoneman, P. (ed.), Handbook of the Economics of Innovation and Technological Change Oxford: Blackwell, pp. 14–51.
From innovation projects to knowledge networks
93
Patel, P. and Vega, M. (1999), ‘Patterns of internationalisation of corporate technology: location vs. home country advantages’, Research Policy 28(2–3), pp. 145–155. Pavitt, K. (1998), ‘Technologies, products and organization in the innovating firm: what Adam Smith tells us and Joseph Schumpeter doesn’t’ Industrial and Corporate Change 7(3), pp. 433–451. Pavitt, K. (2001), Can the Large Penrosian Firm Cope with the Dynamics of Technology?, 68, SPRU, Brighton. Penrose, E.T. (1995), The Theory of the Growth of the Firm, 3rd edn, Oxford University Press, Oxford, New York. Perrow, C. (1970), Organizational Analysis: A Sociological View, Tavistock Publications, London. Powell, W.W. (1990), ‘Neither market nor hierarchy: network forms of organization’, in Staw, B.M. and Cummings, L.L. (eds), Research in Organization Behavior, vol. 12, JAI Press, Greenwich, CT, pp. 295–336. Powell, W.W., Koput, K.W. and Smith-Doerr, L. (1996), ‘Interorganizational collaboration and the locus of innovation: networks of learning in biotechnology’, Administrative Science Quarterly 41(1). Pyka, A. and Küppers, G. (2002), Innovation Networks: Theory and Practice, New horizons in the economics of innovation, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Rothwell, R. (1994), ‘Towards the fifth-generation innovation process’, International Marketing Review 11(1), pp. 7–31. Rugman, A.M. (1997), ‘Canada’, in Dunning, J.H. (ed.), Governments, Globalization, and International Business, Oxford University Press, Oxford, New York, pp. xviii, 518. Simon, H.A. (1979), ‘Rational decision making in business organizations’, American Economic Review, 69(4), pp. 493–513. Stefanuto, G.N. (2004), O Programa Softex e a indústria de Software no Brasil, Unicamp, Campinas. Swan, J.A., Scarbrough, H. and Robertson, M. (2002), ‘The construction of “communities of practice” in the management of innovation’, Management Learning, 33(4), p. 477. Szapiro, M. and Cassiolato, J. (2003), ‘Telecommunications system of innovation in Brazil: development and recent challenges’, paper presented to the First Globelics Conference on Innovation Systems and Development Strategies for the Third Millennium, Rio de Janeiro. Teece, D. (1988), ‘Technological change and the nature of the firm’, in Dosi, G., Freeman, C., Nelson, R., Silverberg, G. and Soete, L. (eds), Technical Change and Economic Theory, Pinter, London, pp. 256–281. Teece, D. (1989), ‘Inter-organizational requirements of the innovation process’, Managerial and Decision Economics 10, pp. 35–42. Teece, D. (1994), ‘The dynamic capabilities of firms: an introduction’, Industrial and Corporate Change 3, pp. 537–556. Teece, D.J. (1977), ‘Technology transfer by multinational firms: the resource cost of transferring technological know-how’, Economic Journal 87(346), pp. 242–261. Thompson, J.D. (2003), Organizations in Action: Social Science Bases of Administrative Theory, Transaction Publishers, Piscataway, NJ. Tidd, J. (1997), ‘Complexity, networks and learning: integrative themes for research
94
Sectoral systems of innovation and production
of innovation management’, International Journal of Innovation Management 1(1). Tidd, J. (2001), ‘Innovation management in context: environment, organization and performance’, International Journal of Management Reviews 3(3), pp. 169–183. Tushman, M.L. and Katz, R. (1980), ‘External communication and project performance: an investigation into the role of gatekeepers’, Management Science 26(11), pp. 1071–1085. von Hippel, E. (1994), ‘“Sticky information” and the locus of problem solving: implications for innovation’, Management Science 40(4), pp. 429–439. von Tunzelmann, N. (2004), ‘Network alignment in the catching-up economies of Europe’, in McGowan, F., Radosevic, S. and von Tunzelmann, N. (eds), The Emerging Industrial Structure of the Wider Europe, Routledge, London, New York. Walker, G., Kogut, B. and Shan, W. (1997), ‘Social capital, structural holes and the formation of an industry network’, Organization Science 8(2), pp. 109–125. Wasserman, S. and Faust, K. (1994), Social Network Analysis: Methods and Applications, Structural analysis in the social sciences, 8, Cambridge University Press, Cambridge, New York. Wenger, E. (1999), Communities of Practice, Cambridge University Press, Cambridge. Williamson, O.E. (1985), The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting, Free Press, Collier Macmillan, New York, London. Woodward, J. (1965), Industrial Organization: Theory and Practice, Oxford University Press, New York.
From innovation projects to knowledge networks
95
APPENDIX DESCRIPTIVE STATISTICS, TESTS AND COMPLEMENTARY TABLES Vertical Integration 0%
20%
40%
60%
80% 100%
Infrastructure Technological Services Training in S&T Hardware Production Semiconductors Software System Others Research Internal costs
Figure A3.1
Vertical integration in different innovative activities conducted by companies under the Brazilian ICT Law, 1997–2003
0.4 Size (percentage of investments)
0.9
Vertical Integration
0.8 0.7 0.6 0.5 0.4 0.3 1.00
Figure A3.2
External Activities
2.00 Group
3.00
0.3
0.2
0.1
0.0 1.00
2.00 Group
Mean for different technological groups defined in H1
3.00
96
Sectoral systems of innovation and production
Table A3.1
INT
IMP
ANOVA test for different technological groups defined in H1 ANOVA
Between groups Within groups Total Between groups Within groups Total
Sum of squares
df
Mean square
F
Sig.
1.557 .520 2.077 .398 .055 .453
2 37 39 2 37 39
.779 .014
55.394
.000
.199 .001
134.455
.000
97
15.44
14.67
16.52
100.00
5.01
6.83
3.32
100.00
100.00
6.15
1.14
2.48
9.86
100.00
1.69
6.97
4.49
28.29
100.00
1.87
0.11
1.48
3.48
17.40
100.00
1.18
6.60
3.66
18.71
24.74
100.00
1.04
1.35
2.28
5.28
13.78
76.27
100.00
11.25
8.53
3.27
35.38
10.94
30.63
100.00
1.06
6.67
7.89
57.87
8.78
17.73
100.00
0.67
4.54
1.05
10.49
29.49
53.76
100.00
5.16
15.42
2.56
31.77
9.56
35.54
Quality Infrastructure (%) (%)
32.23
12.45
45.12
Tech. Services (%)
24.18
39.21
75.67
Training (%)
10.34
46.12
Process (%)
17.95
41.16
Middleware (%)
10.81
Semiconductors (%)
42.71
Software (%)
Foreign companies Domestic companies Private research institutes Public research institutes Private educational institutes Public educational institutes Grand total
Hardware (%)
Total (%)
Row Labels
Research (%)
Distribution of investments per technological activity and type of organization (used in H2 for calculating PRTA)
Table A3.2
98
Sectoral systems of innovation and production
Table A3.3
Normality test for ties in the knowledge network (H3) Tests of normality
KNOW
Semiconductors Hardware Infrastructure Others Process technology Quality system Research Services Software System Training
Kolmogorov– Smimovaa
Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln) Tie strength (ln)
Statistic
df
Sig.
Statistic
df
Sig.
.096 .045 .047 .100 .092 .065 .025 .078 .043 .049 .043
22 141 174 60 90 120 304 162 425 230 240
.200* .200* .200* .200* .056 .200* .200* .018 .055 .200* .200*
.973 .995 .990 .974 .946 .986 .997 .983 .991 .990 .996
22 141 174 60 90 120 304 162 425 230 240
.782 .892 .239 .221 .001 .229 .847 .043 .010 .094 .757
Notes: * a
Shapiro–Wilk
This is a lower bound of the true significance. Lilliefors Significance Correction.
4.
Learning, innovation and public policy: the emergence of the Brazilian pulp and paper industry* Hannes Toivanen and Maria Barbosa Lima-Toivanen
INTRODUCTION It is conventional wisdom that responsive and adaptive innovation policy requires the interaction and feedback between different elements of the national innovation system. Nevertheless, this simple prescription is difficult to accomplish, in particular when policy objectives include socially and environmentally sustainable development in addition to international industrial competitiveness (Lundvall and Borrás, 2005). Lack of responsiveness increases the risk of structural and allocative inefficiencies as policy outcomes, and appears to be pronounced when innovation policies prioritize science and technology-based infant industries (McKendrick et al., 2000). This chapter develops perspectives on responsive and adaptive innovation policy, and examines in particular the role of population-level technological learning, global industry dynamics, and entrepreneurship for the performance of sectoral innovation systems. The vehicle of this discussion is the emergence of the Brazilian pulp and paper industry (PPI) and its sectoral science, technology, and innovation policies over the last hundred years. Over the last few decades, the industry has expanded rapidly and ranks today as one of the leading centres of production and technology in the world, and stands out also as an exceptional success story of Brazilian sectoral innovation policies. Though favourable natural conditions and cheap labour have contributed to the industry’s ascent, its international competitive advantage has been created through sustained and momentous national effort in scientific and technological innovation. Introduction of non-native pulp wood tree species, notably eucalyptus, and their subsequent improvement through silviculture and biotechnology, as well as the development of new chemical pulp processes, have been the most important scientific and technological 99
100
Sectoral systems of innovation and production
accomplishments that established Brazil as world leader in quality and price in short fibre pulp production (Dahlman and Frishtak, 1993; Leão, 2000; Mora and Garcia, 2000; Hilgemberg and Bacha, 2001; Juvenal and Mattos, 2002; ABTCP, 2004). We explore industry’s development as a network learning phenomenon, and highlight how the dynamics of technological learning shifted as the organization of industry and its sectoral innovation system changed in tandem with the advance of strategic knowledge frontiers. Following Wright (1999), we offer a stylized history of the Brazilian pulp and paper industry that identifies the agents of learning that amounted to strategic technological progress, and how technological knowledge was acquired, accumulated and implemented over time. The point of this exercise is to demonstrate the need for improved understanding of the dynamics of technological learning for industry evolution, and consider the role of entrepreneurship and corporate social responsibility for successful leverage of innovation policies in the context of emerging economies and globalization.
ESTABLISHMENT OF THE LEARNING NETWORK, 1900–1955 The foundation of the sectoral innovation system of the Brazilian forest products industry was laid in the late nineteenth century, when the railways introduced eucalyptus into the country. A fast-growing hardwood tree, eucalyptus establishes the raw material base and foundation of the Brazilian pulp and paper industry today. Eucalyptus is not indigenous to Brazil, and its development as a raw material base for the paper industry was not a mechanical replication of existing production processes, but a long interactive learning process that involved selection and adaptation of the eucalyptus species into the Brazilian biological environment and innovation in hardwood pulping technologies. Long duration, vast scale of operations, scientific and technological complexity, and the need to adapt nascent industry to the new biological environment made this learning process similar to the territorial expansion of American agriculture and the emergence of the pulp industry in the North American South (Hunter, 1955; Olmstead and Rhode, 1993; Stanturf et al., 2003; Toivanen, 2004). Eucalyptus was introduced in Brazil in 1864, and the first plantations produced cross ties for railroads and coal for locomotives. Railroad companies spearheaded modern forestry in Brazil, as they responded to allegations of deforestation along newly laid railway tracks, and attempted to secure a steady supply of quality fuel and timber for their own operations.
The emergence of the Brazilian pulp and paper industry
101
Systematic scientific work to discover the most suitable eucalyptus tree for the Brazilian biological environment and launch its industrial exploitation started in the early 1900s with Edmundo Navarro de Andrade. Known as the “father of eucalyptus” in Brazil, Navarro de Andrade had received his agricultural education at the University of Coimbra in Portugal and returned to Brazil with selected eucalyptus seedlings. Employed by a leading railway company, Navarro de Andrade ran an experimental silvicultural station in Jundiaí, São Paulo, between 1904 and 1909, where he attempted to select and breed improved Brazilian eucalyptus species. Later he obtained seedlings of 144 different eucalyptus species from Australia and established another experimental research station in Rio Claro, São Paulo. Navarro developed 12 eucalyptus species that were particularly adapted to large-scale industrial forestry in Brazil and suitable for the production of different lumber products. Director of forest operations of a large railway company for 37 years, Navarro de Andrade was instrumental in the initial promotion of eucalyptus plantations and nascent Brazilian forest products industries. He established altogether 18 eucalyptus plantations for the Companhia Paulista de Estradas de Ferro in the state of São Paulo, yet his legacy lies in the cultivation of a culture of environmentally sound forestry and its scientifically advanced exploitation in Brazil, perhaps best captured in the establishment of the Serviço Florestal do Brasil, in 1921 (Doughty, 2000; Mora and Garcia, 2000; ABTCP, 2004; Martini, 2004). The needs of Brazilian railways encompassed the sectoral innovation system that selected, adapted and improved eucalyptus species for industrial production in the early twentieth century. Consequently, the Brazilian industrial forestry and lumber industries boomed in the early twentieth century, but the country’s pulp and paper industry was practically nonexistent and lagged far behind leading North American and Scandinavian regions (Toivanen, 2004; Lamberg et al., 2005; Lilja, 2005). Though one can speculate that the lack of demand for South American pulp and paper products contributed too, technology constituted the critical bottleneck. Available chemical pulp processes did not lend themselves to either largescale or economic pulping of eucalyptus. The rapidly expanding paper business in North America and Europe was based on a sulphite pulp process and exploitation of soft wood and long-fibre spruce that produced best-quality paper. In order to search for new raw materials and chemical pulp processes, North American and European firms and research centres undertook systematic research, yet attempts to apply economically the sulphite process to pine or hardwoods were unsuccessful, as were experiments to improve the sulphate process (Reed, 1995; Boyd, 2001; Stanturf et al., 2003; Toivanen, 2004).
102
Sectoral systems of innovation and production
Although some imaginative people, the foremost of whom was Navarro de Andrade, suggested the development of eucalyptus pulp processes in the hope of creating a nascent Brazilian pulp and paper industry, the lack of pre-existing organizational capabilities, such as experienced paper firms or a specialized technological community, undermined the few resulting research efforts. Scattered and modest attempts to make pulp and paper from eucalyptus were not sufficient to tackle the scientific and technological challenges involved. In 1925, Navarro de Andrade’s short visit to the US Forest Products Laboratory in Madison, Wisconsin, produced the first sulphite newsprint made from eucalyptus, and another firm, Gordinho Braune & Cia, in São Paulo, started the production of bleached sulphite pulp from eucalyptus two years later (Martini, 2004). Nevertheless, these and some other experiments did not amount to much industrial activity, did not really improve eucalyptus into an economically viable raw material for papermaking, and did not yield enough technological results to generate substantial interest for further research in the opportunities of hardwood and short-fibre eucalyptus. Instead, Brazilian pulp and paper firms focused on long-fibre softwoods, especially Brazilian pine and araucaria, and imported substantially sulphite pulp, mainly from Scandinavia and North America, for papermaking (Doughty, 2000; Mora and Garcia, 2000; ABTCP, 2004). Eucalyptus pulp remained a curiosity as an industrial product and research subject until the Second World War, when raw material shortages and dependency on pulp imports prompted the government to introduce incentives for research on new fibre sources. The incentives included tax breaks for firms undertaking research, new reforestation policy, and the establishment of sectoral research institutes, such as the National Pine Institute, and culminated in 1940 in a decree that provided low-interest finance for the establishment of new pulp mills (ABTCP, 2004). Another government incentive for scientific and industrial forestry was the establishment of large government-sponsored forestry plantations, most importantly in Minas Gerais in 1950 (Bacha, 2003). The emphasis on national self-sufficiency was a global phenomenon, and particularly strong in North America, where the war and disruption of international trade created severe pulp shortages throughout the 1940s, and in this sense Brazilian public policy followed the international model (Toivanen, 2004). Scarcity of market pulp and new political incentives accelerated the industry’s research and development work on chemical pulping of eucalyptus and its use for papermaking. Working systematically on the subject from 1942, Hasso Weiszflog, at the Companhia Melhoramentos, in São Paulo introduced a critically improved process a few years later. Weiszflog’s process made eucalyptus pulp an increasingly viable substitute
The emergence of the Brazilian pulp and paper industry
103
for other pulps used for papermaking, and important experiments with bleached eucalyptus pulp improved its suitability for white papers. Raw material substitution was decisive also for the launch of a long-term research programme on eucalyptus fibre at Leon Feffer & Co. in 1951, as it had depended on imported pulp since the installation of its first paper machine in 1942 (Mora and Garcia, 2000; ABTCP, 2004). Another landmark was the successful experiments of S/A Indústrias Reunidas Francisco Matarazzo with the sulphite pulp process in 1953, which produced paper made solely from eucalyptus pulp (Bacha, 2003). The introduction of eucalyptus in Brazil occurred without the explicit aim of fostering new industries. Instead, the incumbent railways provided the impetus for an early sectoral innovation system in forestry, as they considered advanced forestry strategically important for their existing operations. Scientific enthusiasm, the increased size of the research system, and the exceptional personality of Navarro de Andrade explain why the sectoral innovation system gained enough independence to define research trajectories of little value to railways, in particular to examine the possibilities for making pulp and paper from eucalyptus. Though successful and important, early experiments with the sulphite pulp process did not yield an immediate technological or economic breakthrough in eucalyptus pulp. However, they did define eucalyptus pulp as a central issue for the emerging sectoral innovation system of the Brazilian pulp and paper industry. From the 1940s, the refocused Brazilian political economy supported such a strategy too, and removed rapidly many of the factors that had arrested research on eucalyptus pulp before, as domestic production of pulp and paper ranked among the central industry policy goals.
BEGINNINGS OF THE COMPREHENSIVE GOVERNMENT INNOVATION POLICY, 1955–1970 Between 1955 and 1970, the sectoral innovation system of the Brazilian forest products industry was significantly expanded and augmented. The government strengthened and expanded knowledge creation and transfer institutions, such as research institutes and universities. It also expanded innovation policy and created new policy instruments, which focused on implementation of new knowledge and technology. These instruments included state- and federal-level forestry initiatives, government subsidies as incentives for investments in new pulp and paper capacity, and different regulatory and legal initiatives. The new policies created a system that can be counted as an innovation system, as it encompassed broadly different policy sectors and actors, mobilized the private industry in implementation
104
Sectoral systems of innovation and production
of policy goals through several initiatives, and amounted to a real innovation policy. Finally, these initiatives allowed the Brazilian system to integrate more tightly into the emerging global sectoral innovation initiatives, which were launched after the Second World War. From the 1950s, three factors prompted the Brazilian paper industry to invest increasingly in research and development of the eucalyptus forestry and pulp process. First, the ascendancy of Juscelino Kubitschek to the presidency in 1955 invigorated ambitious industrial policies. Applying the slogan “50 years in 5”, Kubitschek envisaged a crash programme of national industrialization that could replace imports of durable and intermediary goods, and looked to make Brazil self-sufficient in wood pulp and paper by 1960 (Leão, 2000; ABTCP, 2004), though, especially in the case of the pulp and paper industry, Kubitschek’s policies continued and strengthened incentives and institutions that had been created in the 1940s and early 1950s. Second, the worsening global shortage of pulp, which resulted in an over 150 per cent market price increase, paralysed overseas imports into Brazil and prompted interested in new fibre sources. The third and decisive factor was the maturation of sulphate pulp process technology after decades of a sustained, global wave of innovation, and its emergence as the dominant mass production technology of pulp in the 1950s. The sulphate process allowed the pulping of pine and several other tree species, whereas their chemical and mechanical constituency could not be dealt with adequately in the sulphite process, the chosen technology of the majority of earlier experiments with eucalyptus pulp. In the US alone, production of sulphate pulp increased from 0.5 million metric tons in 1939 to 1 million in 1947, and to 4 million in 1958, totalling 60 per cent of the country’s pulp production. Most of this growth resulted from the emergence of the Southern pulp and paper industry that processed pine (Toivanen, 2004). Maturation of the sulphate pulp process as a mass production technology boosted the rise of the pulp and paper industry in regions where traditional wood species could not be effectively processed with the sulphite process. From the early 1950s, new eucalyptus mills went increasingly online in Portugal and directly facilitated Brazilian learning in the technology (Nunes, 2002). Similarly, the adoption of sulphate and kraft pulp processes allowed New Zealand firms to process local wood species economically for the first time, and the local industry grew rapidly (Baker, 2004), as occurred also in Australia (Doughty, 2000). In Scandinavia, leading firms increasingly began to use pine for papermaking. Subsequently, adaptation of the sulphate process into practically all major pulp wood species, including eucalyptus, has marginalized sulphite and other pulp processes (Toivanen, 2004). The introduction of the mass
The emergence of the Brazilian pulp and paper industry
105
production of sulphate pulp presented great opportunities for Brazilian firms, yet its successful application required intensive industry-level technological learning in the new process and its adaptation to Brazilian virgin fibre sources, araucaria and, most notably, eucalyptus. Industry-wide learning in new pulp technologies during the 1950s catapulted the share of short-fibre, consisting practically only of eucalyptus, production within total Brazilian pulp production from a minuscule 4 per cent in 1950 to 60 per cent by 1960 (Appendix). Central vehicles of such industry-level learning and innovation were the new industrial policies that introduced new institutions in the early 1950s. A critical tool of new policy was the Banco Nacional de Desenvolvimento Econômico e Social (BNDES), which extended much of the government financial aid and coordinated extensively national industrial investments after its establishment in 1952. Until 1967, BNDES extended only loan guarantees to pulp and paper firms. Nevertheless, its projects proved strategically important for the emerging eucalyptus pulp industry, beginning with the pioneering long-term research programmes of Cia Suzano de Papel and Papel Simão. Another milestone project was begun in 1957, when the bank launched its third project in the pulp and paper industry and provided support to the Panamericana Têxtil. The company built in Mogi-Guaçu, São Paulo, a small sulphate pulp and paper mill that used a mix of eucalyptus and pine, and quickly constituted a strategic site of technological learning. Upon its entry into Brazilian markets and the eucalyptus pulp industry, US-based Champion Paper Company acquired the mill in 1961, and pioneered there a new strategy that focused upon the advantages of eucalyptus fibre. Champion developed the mill into a large-scale research and development site in eucalyptus silviculture, genetics, and chemical pulping (Comércio, 1994; Juvenal and Mattos, 2002). The demonstrated viability and new technological opportunities of the eucalyptus pulp process prompted BNDES to revise its financial aid strategy to the pulp and paper sector, and it approved direct finance schemes to the sector in 1967. Moreover, BNDES devised a financing strategy that addressed central issues for the development of the nascent eucalyptus pulp industry into a viable large-scale operation. In 1968, the bank specified minimum production requirements for mills that applied for its financial aid to build new or expand existing mills. Printing paper mills were required to have daily production of at least 250 tons, other paper mills at least 50, and pulp mills 100. Critically, BNDES also requested that companies procured at least half of all pulp wood from forests owned by companies themselves, undertook systemic research and development of eucalyptus process technologies, and improved paper grades. In addition,
106
Sectoral systems of innovation and production
the bank incentivated the specification of technical standards, through the Associação Brasileira de Normas Técnicas (hereafter ABNT) (Juvenal and Mattos, 2002). Industrial policies were also decisive for the development of eucalyptus forestry and its institutional advancement in the 1950s and 1960s. New educational institutions, foremost at universities, government and private sectoral research institutes, and finally a set of legislative measures established a three-pronged development strategy of eucalyptus forestry during the period. A pioneering initiative in education was the inauguration of the first Brazilian school of forestry at the Federal University of Viçosa in Minas Gerais in 1960, which in addition to education developed quickly into an important research and technology transfer centre. Several other universities followed suit and launched courses on forestry engineering, silviculture and other aspects of eucalyptus forestry. The schools educated a scientifically and technically advanced work force, accelerated the diffusion of knowledge and technology, and improved the international contacts for Brazilians (Bacha, 2003). The development of a sectoral research system demonstrated in particular public–private partnership in the 1960s, as government and industry established a number of new important sectoral forestry research institutes that augmented existing ones. A Brazilian flagship institution of forestry research, Instituto de Pesquisas e Estudos Florestais (hereafter IPEF), was established in 1968 by Professor Helládio do Amaral Mello, from the University of São Paulo, who emerged as the leading advocate of scientific and industrial forestry in Brazil in the 1960s (Leão, 2000). IPEF was a joint venture of the university and pioneering eucalyptus forestry and pulp firms, such as Champion, Duratex, Madeirit, Rigesa, and Suzano. IPEF provided a template for public–private partnerships in the pulp and paper industry R&D, and other forestry schools followed its lead (Mora and Garcia, 2000; Bacha, 2003). In addition, several regional initiatives were established, which further augmented the industry’s educational and research infrastructure in the 1960s. Amaral Mello also assumed a central role in the revision of Brazilian forestry regulation in the 1960s, as a series of new laws removed obstacles to eucalyptus and pine plantations and established new fiscal incentives for industrial reforestation. Much of this work culminated in the new forestry law, Código Florestal (Law No. 4 771, 15 September 1965), which was passed a year after the military dictatorship assumed political power in Brazil. The new forestry law allowed the large-scale use of eucalyptus and pine for reforestation and the subsequent transformation of the Brazilian landscape. The old law had required that tree species used for reforestation had to be local and correspond with the logged forest, thereby arresting
The emergence of the Brazilian pulp and paper industry
107
industry’s opportunities to introduce more productive tree species plantations. The new Forestry Code allowed firms to tax-deduct reforestation activities up to 50 per cent of their full value, and created a government fund to enhance national reforestation activities, though this was realized first in 1970 (Leão, 2000). Development of the Brazilian sectoral innovation system of forest industries occurred in tandem with a globally concerted effort to improve the organization of eucalyptus research and accelerate it. The Food and Agriculture Organization of the United Nations (hereafter FAO), founded in 1946, was the single most important international institution that advanced research and accelerated the circulation of scientific and technological knowledge. It organized seminars, funded research and development projects, and published manuals that advocated the benefits of industrial eucalyptus forestry, provided technical blueprints for implementation, and diffused the latest research. FAO also networked the global eucalyptus research community and sponsored the first Eucalyptus World Congress in Rome in 1956 with almost 100 participants from 26 countries. The second one took place in São Paulo in 1961 with 240 delegates from 19 countries, and focused on industrial opportunities for eucalyptus. Besides networking the community of specialists and coordinating international research efforts, the conferences staged the launch of practical joint projects, such as the international seed bank, which was realized by the Commonwealth Scientific and Industrial Research Organization, a project hosted by IPEF in Brazil. At the time of the third world conference, titled “Man-Made Forests” and organized in Canberra in 1967, the international eucalyptus community had developed into a well-funded and organized scientific field. Naturally, improved Brazilian organizational capabilities allowed the country to benefit to a greater extent from the rapidly expanding knowledge frontier and absorb more efficiently stateof-the art knowledge (Doughty, 2000).
INNOVATION, INDUSTRIAL GROWTH AND THE CULTURE OF ENTREPRENEURSHIP, 1970–1985 By the late 1960s, the necessary elements for the rapid growth of the Brazilian pulp and paper industry were in place, including mass production sulphate pulp technology, forestry plantations of selected eucalyptus species, pools of a scientifically and technologically advanced work force, a comprehensive sectoral innovation system, capital, and an advantageous political economy. Domestic short-fibre pulp production also increased tremendously, and the paper industry was able to practically eliminate its
108
Sectoral systems of innovation and production
Table 4.1
Cellulose pulp production: imports, exports and consumption in Brazil in 1957, 1963, 1968, and 1973 (metric tons)
Year
1957
1963
1968
1973
Production Imports Exports Consumption
165 137
448 54 3 499
624 35 12 647
1.130 123 194 1.059
Source:
302
Juvenal and Mattos (2002).
previous dependency on pulp imports (Table 4.1). Whereas the share of short-fibre production in total Brazilian pulp production had remained around 60 per cent throughout the 1960s, and the total production of pulp in Brazil did not increase significantly, a dramatic take-off in the production of eucalyptus started in the late 1970s. Production of short fibre pulp increased over 330 per cent between 1970 and 1980, and 18 per cent between 1980 and 1985. The share of short-fibre in total Brazilian pulp production increased from 60 to 74 per cent, though this declined in the context of modest growth to 69 per cent in 1985 (Appendix). While Brazilian industrial and innovation policies should evidently be credited for this development, the ascendancy of the Brazilian pulp and paper industry between 1970 and 1990 enveloped significant organizational and cultural changes as new entrants and entrepreneurs emerged as the leading firms. Juvenal and Mattos (2002) have noted that a few new entrants, notably Aracruz, Cenibra and Jari, leveraged most of the change in eucalyptus pulp production until 1984. Though the military dictatorship and its key institutions exercised heavy-handed coordination of industrial development, the government was also keen to develop private capitalism and create the most favourable conditions for private investment and entrepreneurship since the 1960s (Goldstein, 1999). It is beyond the scope of this chapter to assess how these measures fared, yet it is evident that classical Schumpeterian entrepreneurship contributed significantly to the Brazilian pulp and paper industry from the late 1960s. New firms and entrepreneurs defined completely new corporate strategies that departed radically from those of incumbent Brazilian firms. Entrepreneurial firms focused on the export of eucalyptus pulp and pursued a strategy based on innovative eucalyptus forestry, state-of-theart sulphate pulp processes, export markets, and economies of scale. They disregarded the industry’s traditional emphasis on vertical integration of pulp and paper production. Nothing embodied the culture of entrepreneurship better than the founder of Aracruz Celulose, Erling Lorentzen,
The emergence of the Brazilian pulp and paper industry
109
a Harvard Business School graduate and native Norwegian. A shipping magnate in the 1960s, Lorentzen entered the pulp industry as he recognized Japanese demand for pulp and wood cords, and launched Aracruz in 1967. A pioneer of the eucalyptus pulp business in Brazil, Lorentz also defined the role of environmental and social responsibility for the long-term viability of forest industries in Latin America. Local low wages contributed importantly to Aracruz’s early success, but in addition the company embarked early on a substantial social-investment strategy, financing local hospitals, schools, housing, and training programmes. Though not completely successful, the local social contract did reduce illegal logging from Aracruz’s eucalyptus plantations and labour conflicts, and turned the company into a business school blueprint of social responsibility (Hart, 1997). Another important entrant was Celulose Nipo-Brasileira (hereafter Cenibra), which was a joint venture between Companhia Vale do Rio Doce and a consortium of Japanese papermakers. With the aim of establishing a eucalyptus pulp mill in Belo Oriente, Minas Gerais, and exporting pulp primarily to Japan and North America, the partners launched a feasibility study in 1970. The company benefited directly from pioneering government eucalyptus plantations in Minas Gerais, and in this sense provided also a private extension of earlier industrial policy (Guess, 1982). Cenibra was incorporated in 1973 and, after securing a wood supply from local eucalyptus forests, development of new industrial eucalyptus plantations, and logistics infrastructure, the company built a pulp mill. In many ways, Cenibra also followed Aracruz’s example of embedding environmental and social responsibility into its business plan, and nurtured close social and cultural alliances with the surrounding local society and state (Young, 1995; ABTCP, 2004). Aracruz and Cenibra launched massive forestry operations with the intention to establish subsequently large-scale pulp mills, and leveraged rapid industrial change in Brazil. Both firms also received important support from the government, which had designed special support measures for the pulp and paper industry in its second National Plan for Development between 1975 and 1979. In this framework, BNDES revised its policies and targeted explicitly greenfield large-scale projects. The bank financed about half of Aracruz’s US$400-million pulp mill that went online in 1978 with record capacity of over one thousand tons per day, and similarly extended support to Cenibra’s 750-ton-per-day mill that went online in 1977 (Juvenal and Mattos, 2002). The two mills alone produced over 650 000 tons per year once fully in operation, which took a couple of years, sufficient to increase Brazilian short-fibre pulp production by 75 per cent from its 1976 level (Appendix).
110
Sectoral systems of innovation and production
The extent to which government coordinated and supported entrepreneurs is best embodied in American financier Daniel Ludwig’s Jari project, and the case also highlights the importance of social and cultural strategies in which Aracruz and Cenibra excelled. Ludwig envisioned a business plan to build extensive gmelina-tree plantations at the remote Jari river basin in Amazon. Owing to the lack of local infrastructure, Ludwig built an integrated pulp and paper mill on a floating barge, financed in part with government loan guarantees. The mill went online in 1978, and soon the project hit burgeoning problems. South-east Asia-originated gmelina did not succeed in the biological environment of Jari, and the Brazilian government refused to extend loan guarantees for another barge mill, since it was to be built completely in Japan. The most serious problems arose because of Ludwig’s failure in environmental and social issues. Environmentalists heavily criticized the project, and Ludwig’s project suffered from serious labour conflicts, severed relations with local landless people, and mushrooming social unrest. By 1981, the billion-dollar project was announced a failure by the Brazilian government, which invited Azevedo Antunes, a Brazilian manganese magnate, to form a Brazilian investor group that could take over the company and qualify for government support, for which the non-Brazilian Ludwig was ineligible. The private sector assumed 75 per cent ownership of the faulted project, and the government the rest. The new owners and management changed Jari’s course, and embarked on a highly successful corporate strategy based on eucalyptus forestry, export markets, and environmental and social responsibility appropriate to Brazilian culture (Hoge, 1982; Reier, 1990). It would be an exaggeration to fault incumbent pulp and paper firms for not acting on opportunities in eucalyptus pulp between 1970 and 1985, yet Aracruz, Cenibra and Jari did spearhead a new business strategy and model in the Brazilian and global perspective. Their exclusive pursuit of eucalyptus forestry, economies of scale and the global export trade established a stark contrast to the incumbent pulp and paper firms that produced pulp mainly to supply their own paper production. A culmination of this strategy was the transformation of Aracruz’s export port (Portocel) in Espírito Santo into a technologically advanced and dedicated pulp export port in 1985. A joint venture of Cenibra and Aracruz and financed substantially through BNDES, Portocel featured advanced railway, road and waterway logistics to the pulp mills of Cenibra and Aracruz, and created new economies of scale in pulp exports (Juvenal and Mattos, 2002). In addition, the entrepreneurial new entrants defined environmental and social responsibility important for the long-term viability of their business in Brazilian society, and implemented comprehensive programmes to ensure credibility in the issues and foster alliances.
The emergence of the Brazilian pulp and paper industry
111
SHIFTING LEARNING DYNAMICS OF THE SECTORAL INNOVATION SYSTEM: PRIVATE INITIATIVE AND GOVERNMENT RESPONSE, 1967–1990 The take-off of the entrepreneurial eucalyptus pulp industry gave rise to new learning dynamics in the sectoral innovation system of the Brazilian pulp and paper industry. Whereas government initiated frontiers of forestry research, nurtured scientific and technological capabilities and extended other incentives in order to generate private interest in the infant industry, new entrepreneurial eucalyptus firms turned the tables. They pioneered new biotechnological research and innovations in Brazil that translated directly and immediately into new business strategies and industrial operations. The government, incumbent firms, and the industry’s sectoral innovation system followed their pioneering suit only after the success of the new business and innovation strategy was evident. In the late 1960s, Aracruz and Cenibra recognized that biotechnology allowed improved control of the eucalyptus stock and thereby increased productivity. In particular, novel techniques of asexual reproduction, that is cloning of existing trees through cuttings and avoiding the use of seedlings, marked an important breakthrough in the production of standardized and controlled eucalyptus forests, and contributed to tremendous productivity improvements in eucalyptus forestry since 1970 (Ferreira, 1992; Doughty, 2000). In pioneering private experiments and research programmes, the Brazilian government and incumbent paper firms embraced new technological opportunities and launched a series of initiatives which turned the country into a leader in forestry biotechnologies (Doughty, 2000). Investments in eucalyptus R&D quickly yielded impressive returns, first movers in large-scale eucalyptus plantations learned after a few years of experimentation. Standardized stock improved disease resistance, improved economies of scale in forestry and wood handling, and made the management of the pulp digesting process easier. In addition, biotechnological innovation improved growth yields. First Aracruz eucalyptus trees required 12 years to grow to logging size, but the company’s systematic research and development programme diminished the average growth cycle to seven years by the mid-1980s. IPEF has estimated that the average annual production of Brazilian planted forests increased from 15 cubic metres per hectare in 1970 to almost 35 cubic metres in 1985 (Leão, 2000). Though government bodies provided assistance, private firms, and in particular new pulp firms, were responsible for the implementation of the latest forestry innovations. Aracruz started silvicultural forestry research in 1967, when it created its first industrial eucalyptus plantations using
112
Sectoral systems of innovation and production
species imported from Rio Grande do Sul and derived from seedlings imported to Brazil in the early twentieth century. Six years later Aracruz started systematic genetic improvement of the eucalyptus programme, and sent silviculturist and eucalyptus researcher Edgard Campinhos to Australia and Asia to collect seedlings and do research. Using seedlings for 50 different species, Aracruz created over 1000 experiment stations to try out different characteristics and sort out the best examples. Subsequently, the company acquired French and Australian biotechnology experts to reproduce quality stock from cuttings rather than from seeds, and probably to learn the cloning technique. The cloning programme yielded rapid improvements in disease resistance and pulping qualities, and allowed enhanced standardization of fibre raw material (Doughty, 2000). In Brazil, IPEF and the University of São Paulo in Piracicaba experimented with the cloning technology from 1975, and also launched a systematic research programme on the subject (Leão, 2000). Other companies, such as Cia Vale do Rio Doce, Cia Suzano, Duratex Florestal and Klabin, among others, followed the example of Aracruz from 1976, and launched programmes of cloning. They too sent research teams to collect seedlings from Australia and Indonesia, and cloned the best examples for forest plantations (Ferreira, 1992; Doughty, 2000). In most companies, research-intensive forestry was delegated to specialized forestry divisions, such as the 1983-established Cenibra Florestal. Cenibra acquired its seedlings from Southern Africa, and established a cloning program in Guanhães, Minas Gerais, in 1984, and another one in Ipatinga, also in Minas Gerais, in the following year. Cenibra centralized its cloning operations on the Ipatinga plantation, which exceeded 700 hectares by 1989 and 2,500 in 1994 (Rocha and Campos, 1994). Following the firms’ race to acquire cloning techniques and launch systematic research programmes, the industry’s central associations assumed national leadership in the field in the late 1980s. Empresa Brasileira de Pesquisa Agropecuária (hereafter EMBRAPA) had established the Centro Nacional de Pesquisa Florestal in 1978 (hereafter CNPF) in order to address the increased interest in eucalyptus forestry. IPEF focused in particular on diffusing and commercializing the best available eucalyptus seedlings in Brazil, and acquired actively the assistance of the world’s leading experts, evolving into the leading producer of eucalyptus seedlings in Latin America (Leão, 2000). Though the firms appear to have spearheaded the cloning technology in Brazil, in 1987 IPEF initiated a programme to facilitate industry-wide technological learning in the subject. The programme diffused new innovations and the latest know-how, coordinated national research efforts, created centralized databanks, and advocated for the possibilities of biotechnology and genetic technologies
The emergence of the Brazilian pulp and paper industry
113
in the forest products industries. The last function shielded industry also from social, ethical and environmental criticism, since it provided credibility and national legitimacy for forest biotechnologies (Santos, 1994; Juvenal and Mattos, 2002). By the industry’s standards, the cloning programme was a success in Brazil, as companies created ever larger plantations of cloned eucalyptus. The Cenibra programme yielded adapted clones for its plantations and the company soon utilized practically exclusively cloned trees for reforestation. According to Doughty (2000), all Aracruz plantations consisted of cloned eucalyptus by 1989 (Ferreira, 1992). In total, eucalyptus plantations increased from 1 million hectares in 1970 to 3.6 million by 1990, and the majority of the growth came from plantations of biotechnologically improved examples of species of E. saligna and E. grandis (Doughty, 2000). Growth of the plantation area also fuelled environmental criticism, such as an alleged spread of monoculture at the cost of biodiversity, and most forestry firms initiated plantation schemes that created pathways of natural forests inside vast eucalyptus fields (Rocha and Campos, 1994).
CATCH-UP LEARNING DYNAMICS AND THE SECOND-GENERATION INNOVATION SYSTEM Since 1985, the Brazilian pulp and paper industry’s evolution has been characterized by incumbent firms’ catch-up with pioneering eucalyptus pulp firms and consolidation. Latecomers into the eucalyptus business benefited from the industry’s advanced sectoral innovation system, which rapidly diffused the latest innovations and knowledge and government policies, though their success cannot be credited only to these factors. Established large-scale paper firms had exceptional organizational capabilities and political leverage to enter the eucalyptus business at a moment when the industry’s operations and size had reached an unprecedented scale in Brazil. Between 1985 and 2005, the eucalyptus pulp and paper industry expanded steadily, and Brazilian annual production of shortfibre pulp increased 60 per cent between 1985 and 1995, and 81 per cent between 1995 and 2005. Total annual production of pulp leaped from 3.7 million tons in 1985 to over 10 million tons in 2005 (Appendix). This massive expansion enveloped deep changes in the industry’s organizational structure, as incumbent Brazilian firms began to emulate the strategy of the pioneering entrepreneurial firms and caught up with them. Government initiatives facilitated this development, as incumbent firms benefited from the national initiatives to develop and diffuse eucalyptus cloning technologies and eucalyptus species, and from the government’s
114
Sectoral systems of innovation and production
Second Pulp and Paper National Plan (hereafter II PNPC), which was introduced in 1986. It defined an ambitious expansion plan for the Brazilian pulp and paper industry, and called for the increase of production of pulp from 3.4 million tons per year to 6.6 million tons in 1995, and the doubling of annual production of 4 million tons of paper. The policy continued and amplified earlier incentives in place for firms that invested in eucalyptus forestry, pulp and paper mills, but also allowed BNDES and other financial institutions to assume a more active role in the development of the industry, in particular as direct owners (ABTCP, 2004). Government incentives prompted several incumbent firms to strengthen their forestry operations and begin production of eucalyptus pulp, and also gave rise to important new ventures. One of the biggest ventures was the incorporation of Bahia Sul in Mucuri, Bahia, by Cia Suzano de Papel, Cia Vale do Rio Doce, BNDES and its arm BNDESPAR, and International Finance Corporation, a World Bank organization. Following the blueprint strategy of pioneering firms, Bahia Sul established eucalyptus plantations and built a pulp mill with 250 thousand tons per year capacity in 1992. Half of its pulp production was supplied to owner companies’ paper mills. Similarly, Votorantim Celulose e Papel, part of a diversified industrial conglomerate, acquired small forests and pulp firms in the early 1990s, in part with the assistance of BNDESPAR. The company built new pulp mills with 300 000 tons per year capacity in 1992 and new vertically integrated paper machines, as well as expanded production capacity through acquisitions. Throughout the 1990s, the company expanded its production capacity, producing 1 million tons of pulp and some 700 000 tons of paper in 2005 (Mattos and Valença, 1999). Votorantim and other latecomers to the eucalyptus pulp business benefited from the advanced sectoral innovation system, which widely diffused knowledge, technology, and eucalyptus seedlings. The company’s rapid expansion was in part based on accelerated learning in advanced forestry technologies, as it had little pre-existing capability in eucalyptus cloning. It first acquired such a capability in the 1990s, yet Votorantim succeeded in producing equally impressive productivity gains in forestry to the pioneering firms. In the early 1970s, the Votorantim eucalyptus plantation produced an average of 20 cubic metres per hectare a year, and it took eight trees to produce one cubic metre of wood. In the 2000s, the respective figures were 45 cubic meters and 3.3 trees (Knight, 2002; Valle, 1994). Globally leading pulp and paper firms have exemplified related catch-up strategies. In 2000, Finnish–Swedish StoraEnso, then the world’s second largest pulp and paper enterprise, entered a joint venture with Aracruz. The two companies jointly own Veracel Celulose, whose sulphate eucalyptus pulp mill in Bahia went online in 2005, and it is one of the world’s
The emergence of the Brazilian pulp and paper industry
115
largest one, with 900 000 tons’ annual production capacity. The joint venture marked StoraEnso’s entry to Latin America and secured access to state-of-the-art eucalyptus know-how. The two companies share production of Veracel and, whereas Aracruz sells pulp in the market, StoraEnso uses its entire share to substitute birch pulp at its European paper mills. In this sense the joint venture has been an important learning experience for the Scandinavian firm, as it has been able to experiment with a new pulp source and adjust the European printing paper mills to new pulp. The forceful entry of incumbent firms into the eucalyptus forestry and pulp business, and the ensuing merger wave, changed the industry’s organizational structure. In the 1990s, a consolidation wave took place in the Brazilian pulp and paper industry, as over 30 major mergers occurred between 1992 and 2001 (Fonseca and Zeidan, 2002). Aracruz, Cenibra, Celmar, Veracel and Jari remained focused upon pulp production and export markets, and dominated 71 per cent of market pulp production in 2002. Incumbent firms adapted eucalyptus operations into their existing strategy and structure that emphasize vertical integration of pulp and paper production, and produce relatively little market pulp. Votorantim, Klabin, Suzano and Ripasa were integrated forward in papermaking and supplied a variety of markets, and used most of the short-fibre pulp in their own mills, though, upon learning the eucalyptus pulp processes and scaling up forestry operations, they too produced significant amounts of market pulp, as did Bahia Sul and Votorantim. Champion, a pioneer of eucalyptus pulp absorbed by US-based International Paper Company in 1998, followed the strategy of its North American parent company and used practically all pulp for papermaking in Brazil (Juvenal and Mattos, 2002). The response of the Brazilian sectoral innovation system to the recent rise of genomic research and the improvement of eucalyptus has followed largely historical precedents. Firms, industry associations and state and federal governments have launched cooperative initiatives that coordinate national research efforts and transfer technology. Indeed, it appears that Brazil’s sectoral innovation system is renewing itself at an amazing pace and is poised to be world leader in eucalyptus genomic research. The cooperative nature and extent of the Brazilian genomic research platform, which extends to the regulation of biosafety and other legislative initiatives, represents a departure from the previous structure of the sectoral innovation systems. Genomic eucalyptus research took hold in Brazil when some of the leading research institutes, such as IPEF, advocated possibilities of gene technology for forestry in the mid-1980s. A real turning point occurred in the early 1990s with global advances in genomic research, however. Since 1994, the industry advocated heavier government participation and
116
Sectoral systems of innovation and production
initiation of a national eucalyptus genomic mapping project, eventually launched at the turn of the millennium. Genomic eucalyptus research is supported with particularly strong investment in genomic research in Brazil. In 1997, the State of São Paulo Research Foundation (hereafter FAPESP) created the Organization for Nucleotide Sequencing and Analysis Network, which encompasses 30 laboratories across the state. FAPESP coordinates the Forest – Eucalyptus Genome Sequencing Project Consortium, which includes also Votorantim, Ripasa, Suzano, and Duratex, and aims to produce genetically improved eucalyptus (Silveira and Borges, 2005). In addition, the National Council for Scientific and Technological Development (CNPq) has funded several research projects, and the federal Ministry of Science and Technology (MCT) launched the Brazilian Genome Project in 2000. Its many dedicated genome mapping initiatives include the Genolyptus Project – The Brazilian Eucalyptus Genome Network – launched in 2002 and a major cooperative project that involves all the central trade associations, universities, research institutes, and government bodies. The Genolyptus Project reflects a larger government ambition to create industrial competitiveness through biotechnological research and innovation programmes, and it may eventually produce the first transgenic eucalyptus species (Sedjo, 2001; ABTCP, 2004; Grattapaglia, 2004a, 2004b). Expansion of forestry-related biotechnologies has been accompanied by increased environmental, social and ethical criticism of clone and genomic research and development work. This has prompted the government to introduce clear environmental and bioethical limitations. A specific biosafety law regulating several aspects of agricultural biotechnologies and innovation was introduced in the mid-1990s, but its implementation suffered from criticism and opposition. In 2004, the Law of Biosafety eventually introduced a comprehensive regulatory framework for biotechnology in Brazil, and provides an important framework for innovation in forestry at public research institutions and industry (Silveira and Borges, 2005). Its passage marked also an important extension of innovation policies into ethics and safety laws, and provides one of the latest examples of network learning and expansion in the sectoral innovation system of the Brazilian forest products industry.
THE SECTORAL INNOVATION SYSTEM OF THE BRAZILIAN PULP AND PAPER INDUSTRY TODAY The growth and evolution of the sectoral innovation system of the Brazilian pulp and paper industry has been punctuated by the needs of
The emergence of the Brazilian pulp and paper industry
117
firms and economic and industrial policies, as well as the global advances in science, technology and world trade. The result is a unique sectoral innovation system that addresses exclusive Brazilian knowledge and innovation needs, as well as maintains a division of labour between Brazilian and foreign actors. While large and somewhat diffuse, the system should be characterized as highly focused on the core issues for the competitiveness of the Brazilian pulp and paper industry. A typology of the sectoral innovation system of the Brazilian pulp and paper industry is provided in Table 4.2. Even at the risk of simplifying too much, one could argue that the system invests in basic research and fundamental innovation only when it comes to exploit further the advantages offered by eucalyptus. In the case of research and innovation in other scientific and technological areas, such as chemical processing, energy, equipment and machinery, the system creates national capacities to use the globally best available practices and technologies. Brazil as a nation is a late entrant into the pulp and paper industry, and thus much of the system is geared towards catching up. Evidently, the capacity to exploit and absorb knowledge and innovations from abroad is certainly one of the great strengths of the Brazilian sectoral innovation system. The training and educational system in Brazil turns out a body of skilful labour and a scientifically and technologically advanced work force, which can take advantage of the best technologies and practices developed elsewhere. Most firms train blue-collar workers’ and there exist some vocational schools. Critical for the supply of a scientifically and technologically advanced work force are the federal and state universities, which have created special curriculums in pulp and paper science and engineering, forestry engineering and management. The most important ones are at the Federal University of Viçosa (hereafter UFV), the Federal Rural University of Rio de Janeiro (hereafter UFRRJ) and the São Paulo State University (hereafter UNESP). Industry associations and professional societies, of which the Brazilian Pulp and Paper Association (BRACELPA) and the Brazilian Pulp and Paper Technical Association (hereafter ABTCP) are the two most important ones, also provide continuous education and circulate the latest knowledge that is highly relevant for the industry. Knowledge and technology transfer is also catalysed by several sectoral research institutes and international scientific and technical organizations and especially by the large presence of foreign firms in Brazil (IPEF, 2002). Apart from eucalyptus science and technology, foreign firms are a key vehicle of technological learning and a source of innovations in the Brazilian pulp and paper industry. Engineering service firms, such as the Finnish Pöyry Group, provide the latest scientific and technological knowledge and deliver state-of-the-art pulp and paper mills. Equipment
118
Central actors
MCT MDIC MMA MAPA CNPq
Federal ministries Federal agencies
Federal government
Various
State governments in: Minas Gerais São Paulo Paraná
State governments
ABTCP BRACELPA
Industry associations, technical associations, professional societies
Industry
Firms e.g. Aracruz e.g. Votorantim Strategic alliances Veracel Cenibra
Pulp and paper firms Strategic alliances
Brazilian firms
Firms Engineering services Machinery and equipment suppliers Strategic alliances
Pulp and paper firms Engineering services Machinery and equipment suppliers Strategic alliances
Foreign firms
Overview of the sectoral innovation system of Brazilian pulp and paper industry
Type of actors
Table 4.2
FAO The World Bank Latin American Development Bank, technical and scientific associations
International organizations Scientific organizations, scientific and technical initiatives
International
119
Basic research
Training
Universities UFV UFRGS PUC Dedicated research programmes and funding EMBRAPA
Universities FAO UFV UFRGS UFRRJ UFPR
Universities USP
Universities USP
Vocational schools Special programmes International exchange programmes ABTCP
Special programmes at local level Firm-specific international exchange and study programmes
Special programmes International exchange programmes
FAO Eucalyptus World Congress Scientific associations
120
Universities Sectoral research institutes EMBRAPA CNPF
CNPq MCT
BNDES BNDESPAR
Forestry laws Biosafety Law
Research funding
Industrial development and finance
Regulation and standardization
Federal government
(continued)
Applied research
Table 4.2 Industry
Brazilian firms
State governments
FAPESP State governments
Sectoral research institutes SIF
ABNT ABTCP
Sectoral research institutes IPEF Special programmes PPI Firm-specific R&D Special programmes PPI
Firm-specific R&D PPI
Special PPI programmes Organization for Nucleotide Sequencing and Analysis Network, SP Forest Eucalyptus Genome Sequencing Project Consortium, SP Brazilian Genome Project – Genolyptus Project
State governments
Firm-specific R&D PPI
Foreign firms
IFC Latin American Development Bank
FAO Eucalyptus World Congress Scientific associations
International
121
Notes: ABNT = Associação Brasileira de Normas Técnicas (Brazilian Association of Technical Norms) ABTCP = Associação Brasileira Técnica de Celulosa e Papel (Brazilian Pulp and Paper Technical Association) BNDES = Banco Nacional de Desenvolvimento Econômico e Social (National Bank of Economic and Social Development) BRACELPA = Associação Brasileira de Celulose e Papel (Brazilian Pulp and Paper Association) Cenibra = Celulose Nipo-Brasileira CNPF = Centro Nacional de Pesquisa Florestal (National Centre of Forestry Research) CNPq = National Council for Scientific and Technological Development EMBRAPA = Empresa Brasileira de Pesquisa Agropecuária FAO = The Food and Agriculture Organization of the United Nations FAPESP = São Paulo Research Foundations IFC = International Finance Corporation IPEF = Instituto de Pesquisas e Estudos Florestais (Forestry Science and Research Institute) MAPA = Ministério da Agricultura, Pecúria, e Abastecimento (Ministry of Agriculture, Livestock and Supply) MCT = Ministério de Ciência e Tecnologia (Ministry of Science and Technology) MDIC = Ministério do Desenvolvimento, Indústria e Comercio Extérior (Ministry of Development and Foreign Trade) MG = Minas Gerais MMA = Ministério do Meio Ambiente (Ministry of Environment) PPI = Public–private initiatives PUC = Pontífica Universidade Católíca (Catholic University) SIF = Sociedade de Investigaçoes Florestais at the UFV SP = São Paulo UCB = Catholic University of Brasilia UESC = State University of Santa Cruz UFFRJ = Universidade Federal Rural de Rio de Janeiro (Federal Agricultural University of Rio de Janeiro) UFG = Federal University of Goiás UFPR = Universidade Federal do Paraná (Federal University of Parana) UFRGS = Universidade Federal de Rio Grande do Sul (Federal University of Rio Grande do Sul) UFV = Universidade Federal de Viçosa (Federal University of Viçosa) UNESP = São Paulo State University UNICAMP = University of Campinas USP = University of São Paulo.
122
Sectoral systems of innovation and production
and machinery suppliers, such as the Finnish Metso and German Voith, do the same in pulp and paper equipment. Firms operating in smaller markets than mill designs or paper machines have located complete manufacturing plants to Brazil in order to benefit from the rapidly expanding industry. An illustrating example is the Finnish harvesting and logging machine company Ponsse, which has substantial manufacturing in Brazil, as do several foreign suppliers of chemicals. Lastly, strategic alliances between Brazilian and foreign firms also serve this purpose, as the abovediscussed cases of Veracruz and Cenibra attest. The role of foreign firms for knowledge and technology transfer is also facilitated by the industry and trade associations, which often network and liaise with their international counterparts. Eucalyptus occupies most of the attention of the Brazilian research efforts. The sectoral innovation system maintains a rather clear distinction between applied and basic research, although the recent advances in genomic and biotechnology blur this distinction. Basic research is carried out mainly at the universities, and three of them stand out as central hubs of research: the Federal University of Rio Grande do Sul (hereafter UFRGS), the UFV and the UNESP. Eucalyptus genomic research has been organized in three major initiatives that involve, in varying mixes, the federal and state governments, universities and the industry (see Table 4.2). The eucalyptus genomic research projects are best characterized as a new type of public–private innovation iniatives. The above-discussed “Genolyptus Project – The Brazilian Eucalyptus Genome Network” is based on a strong partnership among the Brazilian federal government through the MCT, the academic and research sector through seven universities and EMBRAPA, and finally 12 companies in the pulp and paper sector (Grattapaglia, 2004b). Basic research is also importantly supported by the international organizations and scientific associations. Applied public research is undertaken in sectoral research institutes, and their research agenda is also quite dominated by forestry and eucalyptus, as is evident in the review of the sector done by the IPEF and the MCT in 2002 (IPEF, 2002). One could even argue that there might be currently too many Brazilian research institutes on forestry and eucalyptus. IPEF (2002) lists 54 research institutes active in forestry and 16 private ones. Although these institutes serve all kinds of knowledge needs in the area of forestry, they also constitute the backbone of the Brazilian knowledge base for industrial forestry. These institutes maintain the basic knowledge on eucalyptus, such as the seed banks, growth data, genomic information libraries and databases, maintain global knowledge networks and diffuse nationally the latest knowledge and innovations (Grattapaglia, 2004b). In addition to IPEF, the most important ones are
The emergence of the Brazilian pulp and paper industry
123
the EMBRAPA and the Sociedade de Investigaçoes Florestaisat (hereafter SIF) at the UFV. The sectoral innovation system is adapted to the somewhat complicated Brazilian federal system of government. Science, technology, innovation, industrial, and environmental policies are on the agenda of the federal government, as well as the most active state governments. Thus the need for coordination is evident at the level of governance structures and policy instruments. This is most obvious in the public funding for research and innovation, as well as industrial policies, which both are often important in state-level politics. At the federal level, three ministries and their agencies bear the majority of public responsibility for the sectoral innovation system of the Brazilian pulp and paper industry: the MCT, the Ministry of Environment (hereafter MMA), and the Ministry of Development, Industry and Foreign Trade (hereafter MDIC). The MCT plays a critical role in the creation of a sound research infrastructure, networking, and funding for advanced eucalyptus research, and does so in great part through its research funding arm, the CNPq. It also cooperates with the MMA in matters that relate to forestry and forestry research. The Ministry of Agriculture, Livestock and Supply (MAPA) has a major role in financing biotechnology and genomic research, as well as maintaining major sectoral research institutes. MDIC provides a general framework for industrial policy, and its finance arms give specific support actions for industrial activities. Of these the most important one is the above-discussed BNDES and its subsidiaries. We can summarize the existing sectoral innovation system of Brazilian pulp and paper with short, simplifying characterizations on research policy and the role of firms, its two key features in our judgement. From the point of view of research policy, the fundamental and most serious scientific and technological research goal is to improve the productivity of Brazilian forests and foremost the eucalyptus. This strategy has manifested itself with the ambition to emerge as the global leader on the subject, and indeed Brazilian scholars, research institutions and firms have accomplished this. The scientific and technological ambition level in other research areas is considerably less, and more attention is put on capacity building and technology transfer, which allow the quick adoption of innovations developed elsewhere. This same dual strategy also underpins educational policies and institutions. Private Brazilian and foreign firms alone bear the responsibility to put the innovations to work in industrial activity and achieve a return on heavy investments in new pulp and paper mills. They compete in a relatively open world market against pulp and paper producers from Asia, Australasia and the Northern Hemisphere, and operate therefore with the
124
Sectoral systems of innovation and production
need for continuous improvement. Firms have responded to this with the same dual strategy as the public sector: focusing Brazilian cutting-edge research, development and innovation efforts on the improvement of eucalyptus and in other science and technology areas transferring the best available solutions from abroad to Brazil.
CONCLUSION This chapter has explored the role of the dynamics of technological learning for efficient innovation policy design, and argued that the exceptional success of public policy in fostering a thriving Brazilian pulp and paper industry was based on an interactive, responsive and adaptive sectoral innovation system. Although the policy experience did include some casualties and notorious failures, the long-term outcomes of investments in innovation have paid off handsomely in the Brazilian pulp and paper industry. The sectoral innovation system grew up over a long period of time as a network phenomenon, which gradually linked different actors and included new policy measures. Though government advocated policy priorities, most notably national industrial competitiveness, it did not coordinate extensively the system or its actors, and any attempts at ambitious coordination would have been thwarted by the scale and complexity of the actors involved, as they ranged from regional, to national, to global actors. Instead, the system allowed for a healthy amount of competition and diversity. This is best illustrated by the critical role of the new entrepreneurial firms in the eventual take-off of the Brazilian pulp and paper industry in the 1970s. While the incumbent papermakers remained largely locked into old strategies, the new entrants pioneered strategies focused on eucalyptus pulp and international export trade, and also emphasized the significance of social and environmental responsibility for the long-term viability of their business. Eventually, the success of entrepreneurs and new entrants prompted the incumbent pulp and paper firms to revisit their competitive strategies. Since the late 1980s, incumbent Brazilian papermakers have caught up with the pioneers of the eucalyptus pulp and paper business, and added to the momentum of the industry. The sectoral innovation system has importantly facilitated this catch-up process, as it had developed advanced mechanisms to transfer knowledge and diffuse innovations, and diminished potential first-mover advantages enjoyed by the technological pioneers. Lately, the leading global pulp and paper enterprises have also begun to utilize these services and revise their traditional production strategies.
The emergence of the Brazilian pulp and paper industry
125
If there is any central reason why we must consider the development of the sectoral innovation system of the Brazilian pulp and paper industry a success worthy of study, it is the critical role played by private firms. This case is unlike so many other cases of sectoral innovation policies in that public policies did not create allocative inefficiencies fatal to the competitiveness of the market. Thus entrepreneurs and business managers enjoyed, and operated under, healthy, internationally competitive incentives for the creation and adoption of scientific, technological and business innovations.
NOTE *
This paper was prepared within the ‘Game Global Project’ at Lappeenranta University of Technology, Finland.
REFERENCES ABTCP (2004), The History of the Pulp and Paper Industry in Brazil, Associação Brasileira Técnica de Celulose e Papel (ABTCP), São Paulo. Bacha, C.J.C. (2003), ‘The evolution of wood-based industries in Brazil and their means of securing wood’, Oxford Development Studies 31(2), pp. 197–217. Baker, A. (2004), ‘Governments, firms, and national wealth: a new pulp and paper industry in postwar New Zealand’, Enterprise and Society 5(4), pp. 669–690. Boyd, W. (2001), ‘The forest is the future? Industrial forestry and the Southern pulp and paper complex’, in Scranton, Philip (ed.), The Second Wave: Southern Industrialization from the 1940s to the 1970s, University of Georgia Press, Athens, pp. 168–218. BRACELPA (2006), Statistics of the Sector: Pulp, Brazilian Association of the Pulp and Paper Industry (BRACELPA), São Paulo, http://www.bracelpa.org. br/en/numeros/pasta/01cel.pdf. Comércio, J. (1994), ‘Silvicultura Clonal na Champion’, in Anais: I Reunião Regional sobre Clonagem Intensiva em Eucalyptus, 19–24 Junho, IPEF, Aracruz. Dahlman, C. and Frishtak, C. (1993), ‘National systems supporting technical advance in industry: the Brazilian experience’, in Nelson, R. (ed.), National Innovation Systems: A Comparative Analysis, Oxford University Press, New York. Doughty, R.W. (2000), The Eucalyptus: A Natural and Commercial History of the Gum Tree, Johns Hopkins University Press, Baltimore, MD. Ferreira, M. (1992), ‘Melhoramento e a silvicultura intensiva clonal’, IPEF 45, pp. 22–29. Fonseca, M. da G. de and Zeidan, R. M. (2002), ‘Situação da indústria de papel e celulose no Brasil: perspectivas de integração (EU e ALCA)’, in Estudo da Competitividade por cadeias Integradas: Um Esforço de Coordenação de
126
Sectoral systems of innovation and production
Criação de Estratégias Compartilhadas, Convênio MDIC/FECAMP/NEIT-IEUNICAMP. Goldstein, A. (1999), ‘Brazilian privatization in international perspective: the rocky path from state capitalism to regulatory capitalism’, Industrial and Corporate Change 8(4), pp. 673–690. Grattapaglia, D. (2004a), ‘Integrating genomics into Eucalyptus breeding’, Genetics and Molecular Research 3(3), pp. 369–379. Grattapaglia, D. (2004b), ‘Genomics applied to eucalyptus: the Genolyptus Project’, in Kellison, R., McCord, S. and Gartland, K.M.A. (eds), Forest Biotechnology in Latin America: Proceedings from the Workshop Biotecnología Forestal. Global Biotechnology Forum March 2–5, 2004, Raleigh: Institute for Forest Biotechnology 57–62. Guess, G.M. (1982), ‘Institution-building for development forestry in Latin America’, Public Administration and Development 2, pp. 209–324. Hart, S. (1997), ‘Beyond greening: strategies for a sustainable world’, Harvard Business Review, January–February, pp. 66–76. Hilgemberg, E.M. and Bacha, C.J.C. (2001), ‘A evolução da indústria brasileira de celulose e sua atuação no mercado mundial’, Revista Análise Econômica 19(36), pp. 145–164. Hoge, W. (1982), ‘An amazon buy out by Brazil’, New York Times, 15 January. Hunter, H. (1955), ‘Innovation, competition, and locational changes in the pulp and paper industry, 1880–1950’, Land Economics 31(4), November, pp. 313–327. IPEF (2002), Ciência e Technologia no Setor Florestal Brasileiro: Diagnóstico, Prioridades e Modelo de Financiamento, Instituto de Pesquisas e Estudos Florestais (IPEF) and Ministério de Ciência e Tecnologia, Piracicaba. Juvenal, T.L. and Mattos, R.L.G. (2002), ‘O setor de celulose e papel’, in BNDES 50 Anos: Histórias Setoriais, BNDES, São Paulo. Knight, P. (2002), ‘Jacarei is set to join the million tonnes’, Pulp and Paper International, October, pp. 29–32. Lamberg, J.-A., Näsi, J., Ojala, J. and Sajasalo, P. (eds) (2005), The Evolution of Competitive Strategies in Global Forestry Industries, Springer, Dordrecht. Leão, R.M. (ed.) (2000), A Floresta e o Homen, EdUSP/IPEF, São Paulo. Lilja, K. (ed.) (2005), The National Business System in Finland: Structure, Actors, and Change, Helsinki School of Economics, Helsinki, B:60. Lundvall, B.-Å. and Borrás, S. (2005), ‘Science, technology, and innovation policy’, in: Fagerberg, J., Mowery, D.C. and Nelson, R. (eds), The Oxford Handbook of Innovation, Oxford University Press, Oxford, pp. 599–631. McKendrick, D.G., Doner, F. and Haggard, S. (2000), From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry, California University Press, Stanford. Martini, A.J. (2004), ‘O Plantador de Eucaliptos: A Questão da preservação florestal no Brasil e resgate documental do legado de Edmundo de Navarro de Andrade’, MS thesis, Faculdade de Filosofia, Letras e Ciências Humanas, Universidade de São Paulo, São Paulo, http://www.ipef.br/servicos/teses/?id=39. Mattos, R.L. and Valença, A.C. de V. (1999), ‘A Reestruturação do setor de papel e celulose’, BNDES Setorial 10, pp. 253–268. Mora, A.L. and Garcia, C.H. (2000), A Cultura do Eucalipto no Brazil [Eucalyptus cultivation in Brazil], Sociedade Brasileira de Silvicultura/SBS, São Paulo.
The emergence of the Brazilian pulp and paper industry
127
Nunes, P. (2002), ‘Processos Químicos em Portugal: O Motor desconhecido da economia portuguesa no século XX’, in Brito, J.M.B. de, Heitor, M. and Rollo, M.F. (eds), Engenho e Obra: Uma abordagem à História da Engenharia em Portugal no Século XX, Dom Quixote, Lisboa. Olmstead, A.L. and Rhode, P. (1993), ‘Induced innovation in American agriculture: a reconsideration’, Journal of Political Economy, 101, February, pp. 100–118. Reed, G.M. (1995), Crusading for Chemistry: The Professional Career of James Holmes Herty, University of Georgia, Athens. Reier, S. (1990), ‘Enchanted forest: how Daniel Ludwig’s disastrous pulp project in the Amazon became increasingly profitable’, Financial World, 20 February. Rocha, M.G. de B. and Campos, W. de O. (1994), ‘Clonagem Intensiva em Eucalyptus grandi na Cenibra’, in Anais: I Reunião Regional sobre Clonagem Intensiva em Eucalyptus, 19–24 Junho, IPEF, Aracruz. Santos, P.T. dos (1994), ‘Contribuição dos grupos de trabalho de propagação vegetativa do IPEF na clonagem de Eucalyptus’, in Anais: I Reunião Regional sobre Clonagem Intensiva em Eucalyptus, 19–24 Junho, IPEF, Aracruz. Sedjo, R. (2001), Biotechnology’s Potential Contribution to Global Wood Supply and Forest Conservation, Discussion Paper 01–51, Resources for the Future, Washington, DC. Silveira, J.M. da and Borges, I. de C. (2005), ‘An overview of the current state of agricultural biotechnology in Brazil’, Proceedings Workshop on Agricultural Biotechnology for Development: Institutional Challenges and Socio-economic Issues, Bellagio, Italy, 30 May – 1 June, http://bcsia.ksg.harvard.edu/BCSIA_ content/documents/BrazilSilveira.pdf. Stanturf, J.A., Kellison, R., Broerman, F. and Jones, S. (2003), ‘Innovation and forest industry: domesticating the pine forests of the southern United States, 1920–1999’, Forest Policy and Economics 5, pp. 407–419. Toivanen, H. (2004), ‘Learning and corporate strategy: the dynamic evolution of North American pulp and paper industry, 1860–1960’, Unpublished Ph.D. dissertation, Georgia Institute of Technology, Atlanta. Valle, C.F. do (1994), ‘Clonagen na Votorantim Celulose e Papel’, in Anais: I Reunião Regional sobre Clonagem Intensiva em Eucalyptus, 19–24 Junho, IPEF, Aracruz. Wright, G. (1999), “Can a nation learn? American technology as a network phenomenon”, in Lamoreaux, R., Raff, D. and Temin, P. (eds), Learning by Doing in Markets, Firms, and Countries, University of Chicago Press, Chicago, pp. 295–320. Young, J. (1995), ‘Cenibra’s $800-million expansion will double production capacity’, Pulp and Paper, February, pp. 98–104.
128
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
Year
13 416 15 433 19 621 24 222 23 021 23 028 25 128 27 798 26 969 31 868 32 074 34 958 32 680 28 885 27 375 32 519 31 780 27 473 23 947
Bleached
Table A4.1
24 951 27 089 25 703 23 325 23 949 27 154 26 834 28 032 40 014 47 663 48 255 60 567 83 508 107 506 121 331 133 692 181 872 169 051 185 942
Unbleached
Softwood
38 367 42 522 45 324 47 547 46 970 50 182 51 962 55 830 66 983 79 531 80 329 95 525 116 188 136 391 148 706 166 211 213 652 196 524 209 889
Total
1 131 1 420 1 977 3 055 4 337 6 875 8 377 12 174 23 388 29 843 61 745 75 796 98 629 120 787 132 829 146 721 181 729 200 785 240 850
Bleached 461 1 008 7 680 5 381 12 652 16 111 17 471 17 969 29 061 35 307 58 163 57 914 63 330 62 321 62 254 57 141 56 244 77 914 66 387
Unbleached
Hardwood
Chemical and Semi-chemical
1 592 2 428 9 657 8 436 16 989 22 986 25 848 30 143 52 449 65 150 119 908 133 710 161 959 183 108 195 083 203 862 237 973 278 699 307 237
Total 39 959 44 950 54 981 55 983 63 959 73 168 77 810 85 973 119 432 144 681 200 237 229 235 278 147 319 499 343 789 370 073 451 625 475 223 517 126
Total
55 400 62 900 65 900 68 400 64 900 72 900 75 900 79 400 86 000 84 600 86 200 94 000 103 500 128 400 161 400 201 500 210 000 123 586 106 507
Highyield pulp
95 359 107 850 120 881 124 383 128 859 146 068 153 710 165 373 205 432 229 281 286 437 323 235 381 647 447 899 505 189 571 573 661 625 598 809 623 633
Total
13.10 12.08 2.90 3.60 13.35 5.23 7.59 24.22 11.61 24.93 12.85 18.07 17.36 12.79 13.14 15.76 –9.49 4.15
53 298 –13 101 35 12 17 74 24 84 12 21 13 7 5 17 17 10
5 18 15 27 31 33 35 44 45 60 58 58 57 57 55 53 59 59
11 7 5 –1 7 4 7 20 19 1 19 22 17 9 12 29 –8 7
Annual Increase in Short-fibre Increase in evolution short-fibre to total long-fibre (%) production production production p.y. (%) p.y. (%) p.y. (%)
PRODUCTION OF LONG- AND SHORT-FIBRE PULP IN BRAZIL,
Historical evolution of pulp production (in metric tons)
APPENDIX 1950–2005
129
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
24 287 44 635 52 593 48 946 48 235 56 259 70 344 73 854 78 020 75 925 95 663 106 858 132 387 172 166 188 456 195 375 202 972 207 794 193 436 191 378 203 151 216 703 224 820 239 486 301 090 289 032 261 849 221 520 122 410 95 278
202 671 233 521 239 549 259 690 281 593 322 910 288 424 376 648 431 085 463 587 511 319 648 714 609 619 627 255 703 275 742 268 855 338 911 974 970 619 1 051 240 1 022 860 957 753 987 644 1 022 833 1 056 322 1 074 205 1 149 656 1 123 827 1 159 668 1 151 502
226 958 278 156 292 142 308 636 329 828 379 169 358 768 450 502 509 105 539 512 606 982 755 572 742 006 799 421 891 731 937 643 1 058 310 1 119 768 1 164 055 1 242 618 1 226 011 1 174 456 1 212 464 1 262 319 1 357 412 1 363 237 1 411 505 1 345 347 1 282 078 1 246 780
259 852 291 216 325 416 346 772 383 554 442 941 475 447 508 517 649 470 847 220 1 363 079 1 678 136 1 632 072 1 712 823 1 921 885 2 083 508 2 078 285 2 168 858 2 201 216 2 246 070 2 369 582 2 377 540 2 794 642 3 246 655 3 351 528 3 729 049 3 760 118 4 098 038 4 332 950 4 739 250
80 501 94 691 103 942 242 932 258 305 307 416 355 393 294 765 343 695 427 262 477 690 438 988 421 712 382 526 244 157 343 234 266 869 266 781 299 190 304 180 348 286 362 692 339 414 361 593 301 248 283 985 271 319 292 793 289 097 245 363
340 353 385 907 429 358 589 704 641 859 750 357 830 840 803 282 993 165 1 274 482 1 840 769 2 117 124 2 053 784 2 095 349 2 166 042 2 426 742 2 345 154 2 435 639 2 500 406 2 550 250 2 717 868 2 740 232 3 134 056 3 608 248 3 652 776 4 013 034 4 031 437 4 390 831 4 622 047 4 984 613
567 311 664 063 721 500 898 340 971 687 1 129 526 1 189 608 1 253 784 1 502 270 1 813 994 2 447 751 2 872 696 2 795 790 2 894 770 3 057 773 3 364 385 3 403 464 3 555 407 3 664 461 3 792 868 3 943 879 3 914 688 4 346 520 4 870 567 5 010 188 5 376 271 5 442 942 5 736 178 5 904 125 6 231 393
116 722 113 206 136 863 140 136 158 783 164 888 162 578 182 994 170 177 190 343 220 138 223 569 196 535 216 386 207 916 237 366 312 513 358 330 390 471 397 688 426 421 436 455 431 596 431 777 460 742 452 599 492 965 465 257 427 037 455 513
684 033 777 269 858 363 1 038 476 1 130 470 1 294 414 1 352 186 1 436 778 1 672 447 2 004 337 2 667 889 3 096 265 2 992 325 3 111 156 3 265 689 3 601 751 3 715 977 3 913 737 4 054 932 4 190 556 4 370 300 4 351 143 4 778 116 5 302 344 5 470 930 5 828 870 5 935 907 6 201 435 6 331 162 6 686 906
9.69 13.63 10.43 20.98 8.86 14.50 4.46 6.26 16.40 19.84 33.11 16.06 –3.36 3.97 4.97 10.29 3.17 5.32 3.61 3.34 4.29 –0.44 9.81 10.97 3.18 6.54 1.84 4.47 2.09 5.62
11 13 11 37 9 17 11 –3 24 28 44 15 –3 2 3 12 –3 4 3 2 7 1 14 15 1 10 0 9 5 8
60 58 60 66 66 66 70 64 66 70 75 74 73 72 71 72 69 69 68 67 69 70 72 74 73 75 74 77 78 80
8 23 5 6 7 15 –5 26 13 6 13 24 –2 8 12 5 13 6 4 7 –1 –4 3 4 8 0 4 –5 –5 –3
130
1 317 833 1 349 877 1 367 510 1 420 520 1 426 311 1 440 799 1 449 701
Unbleached
Softwood
1 405 298 1 422 205 1 438 495 1 508 728 1 511 866 1 537 586 1 536 328
Total
5 091 948 5 295 451 5 292 351 5 751 391 6 812 205 7 311 794 8 011 474
Bleached 267 577 243 814 212 620 265 578 286 134 300 632 304 660
Unbleached
Total 5 359 525 5 539 265 5 504 971 6 016 969 7 098 339 7 612 426 8 316 134
Hardwood
Chemical and Semi-chemical
(continued)
BRACELPA (2006).
87 465 72 328 70 985 88 208 85 555 96 787 86 627
Bleached
Source:
1999 2000 2001 2002 2003 2004 2005
Year
Table A4.1
6 764 823 6 961 470 6 943 466 7 525 697 8 610 205 9 150 012 9 852 462
Total
444 309 501 796 468 561 495 398 459 042 470 131 499 651
Highyield pulp
7 209 132 7 463 266 7 412 027 8 021 095 9 069 247 9 620 143 10 352 113
Total
7,81 3,53 –0,69 8,22 13,07 6,07 7,61
8 3 –1 9 18 7 9
79 80 79 80 82 83 84
13 1 1 5 0 2 0
Annual Increase in Short-fibre Increase in evolution short-fibre to total long-fibre (%) production production production p.y. (%) p.y. (%) p.y. (%)
5.
The software sector in Uruguay: a sectoral systems of innovation perspective Marjolein Caniëls, Effie Kesidou and Henny Romijn
1.
INTRODUCTION
Recently, a number of developing countries have witnessed the emergence of a local software industry. This phenomenon is not limited to large emerging economies like Brazil, China and India, but also extends to a range of smaller countries in the South, including at least seven Latin American countries, South Africa, South Korea, Malaysia, Thailand and Vietnam. This chapter focuses on the rising software industry in Montevideo, Uruguay. The fast growth of the sector since the early 1990s is by all means remarkable. Uruguay is a traditional agricultural economy, as demonstrated by its main exports, meat, leather, wool and rice. The blossoming of the Uruguayan software industry came as a surprise to many local and international observers. As with many other Latin American countries, Uruguay’s economy was performing very badly during the 1990s. While the industrial value added declined by 1 per cent per annum during 1991–2001, the software sector grew by 4 per cent per annum during the same period.1 This growth mainly emanated from the export market. At the end of the 1990s, exports constituted 39 per cent of the value added of Uruguay’s software sector (Snoeck et al., 1992; CUTI, 2004), further increasing to 47 per cent in 2004, representing a total value of US$79.6 million (CUTI, 2004).2 The most interesting feature of the Uruguayan software sector is that it has emerged without substantial direct support from the state and that it has managed to create internationally competitive sophisticated products and services. In this chapter we try to explain the success of the sector by investigating its sectoral system of innovation, which creates insight into its growth dynamics and the underlying driving forces of its competitiveness and innovative prowess. We use a slightly adapted form of Malerba’s (2002) 131
132
Sectoral systems of innovation and production
sectoral systems of innovation (SSI) framework, in which we give prominence to the central role played by interactive innovation for the sector’s competitiveness. In section 2 we explain how we use Malerba’s framework for the analysis in this chapter, and we briefly describe the sample. Basic information about the nature of the knowledge base in the software sector, and how this affects opportunities for entry and development of developing country producers, is discussed in section 3. In sections 4 to 7 we discuss the findings from our survey structured according to Malerba’s sectoral system framework (see next section for details). Section 8 concludes.
2.
THEORY AND METHODOLOGY
Malerba’s (2002) article defines a sectoral system as a “set of new and established products for specific uses and the set of agents carrying out market and non-market interactions for the creation, production and sale of those products” (p. 248). This definition includes production along with innovation processes and characteristics, but for the purpose of this chapter we will focus primarily on the innovation aspects; hence we speak of sectoral systems of innovation (SSI). Malerba’s broader system of innovation and production consists of five main building blocks (p. 251): 1. 2. 3. 4. 5.
the sector’s knowledge base and learning processes; basic technologies, inputs and demand, with key links and dynamic complementarities; type and structure of interactions among firms and non-firm organizations; institutions, defined as standards, norms, routines, common rules, established practices and habits, and laws and regulations; processes of variety generation and selection.
For the purpose of our analysis, the first three building blocks take centre stage. We start with the second building block, because this is especially important for the explanation of the emergence and early development of the industry, as a result of a combination of various demand and supply conditions. The industry’s emergence was also made possible by the possibility of adopting an incremental development path, something which is intricately bound up with peculiar characteristics of software production. These issues are discussed in section 4. Learning (the first block) constitutes a key feature of the success story. Moreover, in the software sector, learning is known to be a highly interactive process (the third block). The sector is commonly organized in tight
The software sector in Uruguay
133
geographical clusters, of which Silicon Valley and Bangalore are only among the most well known. This clustering is known to be commonly associated with exchange of information and knowledge among firms and non-firm organizations, as for example documented by Saxenian in respect of the Silicon Valley cluster (1994). Software firms in Uruguay are likewise highly clustered in and around the capital city of Montevideo; 98 per cent of the firms are located in Montevideo itself, while 2 per cent are located in Zonamerica, the business and technology park on the city’s outskirts. This implicates specific importance for Malerba’s third SSI building block about the type and structure of inter-agent interactions. Since learning and local interactions are so closely bound up in our sector, these two building blocks are analysed together in section 5. The fourth building block plays a supporting role in our analysis. We limit ourselves to the more formal, tractable types of sector institutions relating to policies and functioning of supporting industry organizations. The role of these institutions has been quite unimportant for the emergence and early growth of the sector, although they are now beginning to assume a more prominent role in its further expansion. Public policy is examined in section 6, and the role played by institutional entities in section 7. Information concerning the fifth building block is largely lacking, because our field study is based on a one-shot survey. A proper appreciation of the importance of this aspect would necessitate longitudinal research. Moreover, our sector in Montevideo is still too new for variety generation and selection to have had a significant impact on the industry. The primary data for the SSI analysis of the software sector in Uruguay were collected by means of personal interviews among 97 software firms and all relevant support institutions, research institutes and universities in Montevideo in 2004. The software development and consulting services sector consists of an estimated 150 firms, excluding 1600 one-person companies (Stolovich, 2003). The estimate is based on a list from the Uruguayan Business Association of Information Technologies (CUTI), direct verification by means of contacting companies, and a search in the telephone directory. These 150 firms were approached and asked to participate in the actual survey. Eventually, 98 firms were willing to take part, of which 97 provided full information. This represents a 65 per cent response rate. This sample includes all the important and biggest software developing and consulting firms in the cluster. The non-participating firms consisted mostly of very small ventures. We administered a structured questionnaire by means of face-to-face interviews with the director and/or the chief engineer of the R&D department in each of the sample companies. The design of the questions was partly based on the EU Community Innovation Survey.
134
Sectoral systems of innovation and production
3.
THE KNOWLEDGE BASE OF THE SOFTWARE SECTOR3
Software development includes so many different activities and skills that it is hard to treat software as a single sector of the economy (Mansell and Wehn, 1998). In this research, the software sector will be defined as all companies that develop and sell software, as well as firms that offer software consulting services. This definition captures producers of computer software, while it excludes hardware producers, as well as companies that develop software for direct use in their own products or processes (embedded software producers). Computer software is sold separately from hardware, and its production is often outsourced (Bitzer, 1997). The software sector in Uruguay predominantly produces computer software. The software sector in general is still relatively new. It emerged in the 1970s and has grown to a market of over US$370 billion. Annual growth of the sector has been over 15 per cent per year since the 1990s. The market is mainly controlled by developed nations, with about half of all software sales originating in the United States and a third in Western Europe. Software, both packaged software and software services, is one of the most rapidly growing sectors in the OECD countries. Software accounts for about 10 per cent of the global ICT market and is an innovative sector. New products succeed each other quickly, and the organization of software production also changes rapidly (Commander, 2005). An important distinction is between standardized and individual software. Standardized software consists of a product sold to many customers, whereas individual software is tailor-made for one particular client. In the latter case the offerings are high in service content, particularly when the software is sold to business clients. In that case, the company sells the hours its employees have worked on the product, which is much less risky and less demanding in terms of breadth of skills than developing, programming and marketing a standardized software product for many users. We will see that many Uruguayan firms started out by doing such jobbing work. However, selling standardized software products tends to be financially more attractive. A standardized software product requires large up-front development investments, but can then be reproduced infinitely without any extra costs to the developer. An hour of work, on the other hand, can only be sold once (Cusumano, 2003). One may thus expect some jobbing firms to move into development and sale of standard products over time, after they have become well established. As we shall see, this pattern is visible in Uruguay. Since entry into the sector can be done incrementally, starting with less capital-intensive, service-focused activities and gradually deepening the
The software sector in Uruguay
135
skill range, there are opportunities for countries that do not have much experience in the sector to enter the market (UNCTAD, 2002). Global demand for software is growing rapidly, and is chronically short of cheap manpower. This explains why several developing countries are beginning to make significant inroads into the global software market. As Heeks (1999) puts it: “Computer science graduates need arm themselves with just a PC and a couple of user contacts to become part of the local information economy. Add a modem and they are global “infopreneurs”.’
4.
EMERGENCE AND DEVELOPMENT OF THE SOFTWARE SECTOR IN URUGUAY
The emergence of the software sector in Uruguay can be traced back to the early 1990s. A major impetus was the fast-growing demand for software in Latin America during the 1990s. Table 5.1 shows the expenditure on software products and services in selected Latin American countries. The software market grew by a whopping 13 per cent per annum in these countries from 1992 until 1999. These favourable conditions of demand for software products and services stimulated the emergence of a local software sector in the Latin American region. Research by the Inter-American Development Bank has shown that several software clusters developed in Mexico during this period. These clusters catered primarily to growing local demand for software products and services (Rabellotti, 2003). In contrast to the Mexican situation, the Uruguayan software sector confronted a small local market, which induced budding software entrepreneurs to look further afield. The cluster grew by taking advantage of the increasing demand in neighbouring Table 5.1
Market for software products and services (US$ millions)
Brazil Mexico Argentina Colombia Total (selected Latin American Countries) India Israel Source:
1992
1999
Annual growth (%)
2 373 1 120 527 162 4 182
5 984 2 097 1 423 562 10 066
14.1 9.3 15.2 19.4 13.3
425 598
1 080 1 974
14.2 18.6
Authors’ calculations based on data from WITSA (2000).
136
Sectoral systems of innovation and production
Table 5.2
Participation in education in Latin American countries Gross enrolment ratio (% of relevant age group)
Argentina Bolivia Brazil Colombia Paraguay Peru Chile Uruguay Source:
Primary (1990–91)
Secondary (1990–91)
Tertiary (1990–91)
106 95 106 102 105 118 100 109
71 37 38 50 31 67 73 81
39 21 11 13 8 30 21 30
World Bank (2004, Table 2.11).
countries such as Argentina and Brazil. In Brazil, for instance, demand for software products by the banking and the telecoms sectors has been very great since the mid-1990s and was served by a mixture of local and foreign firms (Veloso et al., 2003, p. 2). A notable reason why Uruguay specifically was able to build a successful local software sector under these circumstances has to be sought in its comparatively good educational background. The Uruguayan state has been consistently emphasizing education in its policies. Around the time of take-off of the software sector, in 1990–91, the country’s enrolment ratios compared favourably to those of other major Latin American countries, particularly for secondary and tertiary education (see Table 5.2). Adult literacy rates in Uruguay in 1990 were among the highest in the Latin American region, along with those of Argentina and Chile (see Table 5.3). However, there were also push factors at work in the labour market owing to the bad state of the Uruguayan economy at this time, driving underemployed people to seek out promising new activities. There was a sizeable surplus of well-qualified professionals skilled in diverse fields like informatics, finance and accountancy who could not be absorbed easily in a badly performing economy. During 1980–85, real GDP contracted by 5 per cent per annum. Although growth began to pick up somewhat after 1985, in 1990 real GDP had only regained its level of 1980.4 In this situation people with complementary skills found each other to set up joint software businesses in order to fulfil the increasing local and Latin American demand for enterprise resource planning (ERP) products
The software sector in Uruguay
Table 5.3
137
Adult literacy rates in major Latin American countries % ages 15 and older
Argentina Bolivia Brazil Colombia Paraguay Peru Chile Uruguay Source:
Male
Female
96 87 83 89 92 92 94 96
96 70 81 88 88 79 94 97
World Bank (2004, Table 2.13).
for small or medium enterprises. This niche market was overlooked by big TNCs. Hence, initially Uruguayan software firms faced modest competition, which facilitated their entry to foreign markets. These mostly small and medium-sized software firms combined detailed technological knowledge with knowledge of a specific market, or application knowledge, for example in areas like banking and finance, education, health and construction. The key to success in supplying this market lies in having hybrid knowledge of informatics on the one hand and specialized knowledge of specific user sectors on the other. Innovations were usually applicationbased and tailor-made, i.e. satisfying a new need of the customer, while at the same time incorporating new technical trends in informatics in new releases. Under the influence of this constellation of demand and supply forces, the development trajectory of the software sector in Uruguay closely reflects the incremental path described in section 3. A typical firm in the sector starting off in the early 1990s would have been working towards meeting the requirements for a specific product and, in turn, would be fulfilling the needs of specific business clients. For instance, the managing director of a local firm told us that at that time ‘the business model resembled the duties of a scrivener’. At the same time, customers functioned as ‘sponsor godfathers’ to start-up companies, financing the development of their first product. Once the product was ready, it could be sold to another customer through the recommendation of the first customer. Thus, during the first stage of the development of the industry, most of the firms were developing custom-made software applications for user sectors, an activity which we classified above as software services. Many researchers have noted that the sector has diversified considerably
138
Sectoral systems of innovation and production
since then, involving considerable technological upgrading (Mejía and Rieiro, 2002; Stolovich, 2003; Failache et al., 2004). Many firms have come up that make highly standardized application software packages for sectors such as banking, health, education and transport. Another group of firms continues to focus on consulting services, but has moved into more skill-intensive activities. These include, for example, management solutions for small and medium enterprises. The first type of firm derives profits from economies of scale, while the second type enjoys economies of scope. However, there remain a number of – primarily small and mediumsized – software enterprises in the sector that continue to rely on low-skill services. Some basic statistics from our own recent survey among Uruguayan software firms (Table 5.4) confirm this picture. Our sample clearly bears out the increasing importance of software development, with 47 companies exclusively being involved in this activity, and another 26 firms relying on a mix of development and service consulting. The fast expansion of the sector in the 1990s is also reflected in Table 5.4, with 41 companies having begun operations during this decade. The expansion of the sector primarily manifested itself in a fast growth of the number of companies, rather than in growth in firm size. The great majority of the sample firms are small or very small. Fifty sample companies have between 1 and 10 employees, and another 35 employ between 11 and 50. Another noteworthy characteristic of the sample is that the great majority (73) are national firms, while another 20 are multinationals with a Uruguayan home base. Just three companies are foreign multinationals, and these are entirely export-oriented. The rest of the sample have a mixed foreign–domestic market orientation. Since the sample includes all the major firms in the industry, we can conclude that the sector is predominantly home-grown.
5.
LEARNING PROCESSES AND INNOVATION OUTCOMES
In this section we delve into technological learning, one of the main drivers of the technological upgrading that has characterized the Uruguayan software cluster since its establishment in the early 1990s. We use data from our survey. For conceptualization purposes we draw on studies about technological change in developing countries, which suggest that firms can increase their technological capabilities and competitiveness by learning from internal and external sources (e.g. Dahlman et al., 1987; Katz, 1987; Romijn, 1999). We concentrate on innovation capabilities in this
The software sector in Uruguay
Table 5.4
Key characteristics of the sample
Characteristics
Principal activity of the firm Software development Consultancy and services Both (software development and consultancy and services) Other related business activity Type of firm National Domestic multinational Foreign multinational Non-profit organization Year firm started 1970 or earlier 1971–1980 1981– 1990 1991–1999 2000–2004 Firm size (number of employees) 1–10 11–50 51–250 > 250
Source:
139
Number of firms
Average number of employees
Average total sales in US$ (000)
Average export sales in US$ (000)
47
19
2186
785
19
45
3070
1939
26
18
435
74
5
21
967
640
73 20
15 44
1155 1852
355 1059
3
113
10 900
10 580
1
7
n.a.
0
4 6 28 41 18
40 77 24 15 23
3517 2691 3475 428 1784
975 2032 971 95 1759
50 35 10 2
6 25 67 256
115 2629 4140 19 000
15 578 2290 17 800
Kesidou (2007), p. 49.
study, since the main characteristic of the software sector is its continuous product innovation effort. We use Lall’s definition of innovation capability, namely the skills and knowledge which are necessary in order for a firm to be able to improve and change products, processes and production organization (Lall, 1992). Since the firm’s innovation capabilities are
140
Sectoral systems of innovation and production
not directly observable, in our survey we obtained information about the firms’ innovation performance outcomes. The extent and quality of firms’ internal learning would be contingent upon the absorptive capacity already developed (Cohen and Levinthal, 1990). The primary data about absorptive capacity in the firms in our survey cover the R&D that the firm has undertaken, the experience and education of its employees, and the age of the firm. Firms may also use a variety of mechanisms in order to learn from external sources. These sources can either be located within the Montevideo cluster or outside it. Two sources which we expected to be important are knowledge spillovers from firm spin-offs, and knowledge spillovers through mobility in the labour market. Hence our questionnaire included questions about the origin of the firms, and about the extent of their recent labour turnover. A third potentially important source of new knowledge for firms is through their interactions with other parties in the industry in the course of doing business. To capture these effects, we introduce the concept of “knowledge flows”. The concept encompasses both market-based knowledge transactions resulting from formal collaborations between actors, and free and direct knowledge flows arising from informal contacts, i.e. knowledge spillovers.5 We collected separate information about local (cluster-based) and non-local knowledge flows. There are no significant business activities in Uruguay outside Montevideo that could act as sources of new knowledge for our sample firms; hence ‘non-local’ in this case always refers to international knowledge sources. We constructed four knowledge-flow variables designed to capture interaction effects, as follows: Respondents were asked to rate the importance of various sources of information, advice or assistance to their upgrading or innovation efforts on a Likert scale running from 0 (nonexistent) to 4 (crucial for innovation). They were asked to do this for 13 different potential sources of knowledge: the business group; temporary personnel; customers; suppliers; competitors; alliance partners; consultants; research institutes; universities; innovation centres; sector institutes; exhibitions; and electronic information. Moreover, firms were requested to indicate the geographical location of each of these sources of knowledge (local versus international). Finally, firms were asked to clarify the nature of the relationship between their firm and each of the knowledge sources, i.e. whether it constituted a formal relationship involving knowledge transactions (KT), or an informal linkage or contact in which knowledge is transferred spontaneously in a purely informal manner without any compensatory payments, i.e. knowledge spillovers (KS). Using these three attributes – importance, location and type of the relationship – we were
The software sector in Uruguay
141
able to group the flows into LKS, LKT, IKS or IKT-type flows. Then, for each of these four categories of knowledge flow, we added up the 13 Likert scores, ending up with one aggregate importance score for each flow category. LKS, LKT, IKS and IKT are thus ordinal variables, which can assume a minimum value of 0 and a maximum of 52. Despite the rather modest size of the firms, their innovation performance was found to be impressive. Over half of them (52 per cent) had been first in introducing a completely new product to the market, while 70 per cent had upgraded existing products and/or introduced new custom-based services during the five years preceding the interview. The average total number of realized innovations in this period was 4, and the average percentage of turnover deriving from those product innovations was as high as 45 in 2004. A sizeable minority of firms (38 per cent) had also obtained an international quality certification. Information about the internal learning activities and absorptive capacity from the individual sample firms shows that the accumulated average number of R&D man-years was 10.4, and the average percentage of R&D manpower in relation to the firms’ total workforce was 36 per cent. This signifies the importance of continuous innovation activities and considerable innovation capacity in the industry. The great majority of employees have undergone vocational or technical training, and several have a graduate degree. Twenty-seven per cent of the companies employed someone with an M.Sc. degree or higher. Eight per cent employed someone with a foreign university degree. Employee working experience is a little over five years on average, and these employees worked in 1.7 companies before joining the current firm. We now turn to the external learning activities. Remarkably, almost half of the sample firms (48 per cent) had come into existence as a result of a spin-off. Spin-offs apparently are a very commonly used mechanism of new firm formation in this industry. Table 5.5 shows the main characteristics of the spin-offs found in the software cluster in Montevideo. The majority of the spin-offs were created during the last decade. Most of these firms are small with no more than 50 employees. The private sector is the dominant source of spin-offs, particularly the TNC segment. Eleven out of 47 spin-offs in the sample derive from domestic software TNCs, while no fewer than 7 originate from the few foreign software MNCs present in the cluster. Another seven derive from national software firms. The percentage of spin-offs from the universities and the incubator programme represent just a small fraction of the total spin-offs in the sample, despite the fact that such spin-offs were encouraged by the Uruguayan state and the universities (see section 7 about the incubator programme). With the exception of one firm, all the spin-offs hailed from entities located in
142
Sectoral systems of innovation and production
Table 5.5
Characteristics of spin-offs
Characteristics (a) Year firm started Before 1990 1990–94 1995–99 2000–2004 (b) Size of the firm 1–10 employees 11–50 employees > 51 employees (c) Type parent organization Domestic software multinational Foreign software multinational National software firm Other national firm Other multinational (foreign) University Incubator (d) Location of parent organization Local National International Source:
Organisations (n = 47) 8 8 14 17 23 20 4 11 7 7 10 4 3 5 46 0 1
Kesidou (2007), p. 131.
Montevideo. This points to the importance of spin-offs as a conduit for local knowledge spillovers and technological capability building in the sector. A second form of knowledge spillovers seems to be associated with labour mobility, with over one-third (35 per cent) of the workforce having joined the firm during the five years preceding the interview. A third major type of knowledge spillovers is visible in the contacts with other parties that firms cultivate. The average score across the sample firms for the LKS variable is 6.1. The figure suggests that firms on average benefit freely from, say, two to three local knowledge sources and the same number of international ones, and that the linkages to these sources are of moderate intensity or that they have one single free source of knowledge that is crucial to them. However, transactions-based local knowledge flows (LKT) get an even higher score of 9.1. Thus, while some of the local external knowledge is perceived by the respondents as being ‘in the air’, a substantial part of it does not come for nothing. Overall, these results thus suggest that
The software sector in Uruguay
143
both local knowledge spillovers and local knowledge transactions are important for technological development and growth of the industry. Translated in terms of Malerba’s (2002) SSI framework, our findings point towards high cumulativeness of knowledge at the local level. This is indicated by significant local knowledge spillovers and low appropriability, and high local knowledge transactions with which firms supply themselves with inputs complementary to their internal knowledge generation processes. Some more detailed insight into the nature and importance of the different types of interaction-based local knowledge flows is provided in Table 5.6. For each mechanism of local knowledge flow, the table ranks the number of sample firms that reportedly utilized it. The most-used local knowledge flow is a commercial one, not a spillover. This is knowledge acquired in the course of doing business with, and interacting with, customers. No fewer than 85 out of 97 firms (87.6 per cent of the sample) reportedly benefited from this knowledge flow mechanism. In addition, interactions with customers also induce free knowledge spillovers. These are found under the heading of backward and forward linkages in the left-hand column of the table, mentioned by 55 firms (56.7 per cent). The second most important type of local knowledge flow emanates from visiting exhibitions and conferences, mentioned by 66 firms (68.0 per cent). These, too, constitute a mix of pure knowledge spillovers (in the case of free events, mentioned by 30 firms) and knowledge transactions (in the case of fee-paying events, listed by 36 firms). A third important type of local knowledge flow consists of collaborative relations with actors at the same stage of the value chain, which are reported by 54 firms (55.7 per cent). These relations consist entirely of knowledge spillovers. Some of the domestic TNCs play a major role in generating these types of knowledge flows. Artech and Ideasoft are two prominent software development firms, which have created software engineering tools and are strongly involved in R&D. The main product of Artech – Genexus – is a generating code or platform upon which other software firms are able to develop a variety of software solutions. This product is used by many national software firms, which have formed a unique user community. Besides the advantage afforded by being located near to their main technology supplier, these Uruguayan firms have formed an informal user network among themselves. Hence, knowledge spills over among software firms (users) and facilitates problem solving and innovation. The relationship between TNCs and local software firms becomes particularly close when the latter take the role of B-testers. Selected software firms test the technological tools, detect errors and offer ideas for further improvement. These collaborations facilitate the sharing of technological knowledge and
144
Sectoral systems of innovation and production
Table 5.6
Importance of different local knowledge flows in the Uruguayan software cluster
Local Knowledge Spillovers
Local Knowledge Transactions
No. of Importance firms attached to reporting mechanism mechanism1 (avg)2 Horizontal interactions with other firms and technological collaborations Vertical interactions: backward and forward linkages Interactions in exhibitions and conferences Interactions with universities and research institutes Interactions with support institutes
No. of Importance firms attached to reporting mechanism mechanism1 (avg)2
54
1.6
Interactions with customers
85
3.4
55
1.8
39
1.4
30
1.0
36
1.1
20
0.7
Purchase of knowledge products and services from suppliers Interactions in exhibitions and conferences Purchase of knowledge services from consultants
25
1.4
19
0.6
Notes: 1. To construct the data in this column, we have given the value of 1 if a local source of knowledge is used by a firm (irrespective of its importance) and 0 if the source is not used. 2. Firms were asked to rank the importance of each mechanism of local knowledge flow on a scale ranging from 0 (unimportant) to 4 (crucial). The figures in this column of the table represent the average rankings per knowledge flow across the sample firms.
the creation of new knowledge. The marketing manager of a software firm within the Montevideo cluster said that “face-to-face informal communication with the employees of Artech constitutes one of the most important ways of acquisition of technological knowledge and learning from one of the most successful companies in the cluster”. The remaining knowledge mechanisms listed in Table 5.6 are of lesser importance. They emanate primarily from knowledge transactions with consultants (25.8 per cent) and equipment suppliers (40.2 per cent). The
The software sector in Uruguay
145
least-used knowledge links in the sample concern the knowledge spillovers arising from interactions with support institutes, reported by just 19 firms (19.6 per cent), and linkages with universities and research institutes, which are pure knowledge spillovers used by 20 firms (20.6 per cent). The linkages with universities are mainly informal. They usually involve research projects by students, who are placed in a software firm for a year. The limited importance of the institutions for knowledge accumulation in the private sector is a common pattern in developing countries, where the great majority of firms – especially small and medium ones – are not in contact with the state, universities or collective organizations. Table 5.6 also shows average importance scores for the different knowledge flow mechanisms, as reported by the sample firms. Each firm was asked to rate the importance of each local knowledge link for its innovation processes on the Likert scale explained earlier. Table 5.6 shows that only one catego.ry has an average value above the mid-point in the scale (2), namely local knowledge transactions with customers. With a rating of 3.4, this knowledge flow appears to be by far the most important mechanism for the innovation processes undertaken by the firms. In fact, it is the only truly important category. The scores assigned to the other categories are all below 2, with the forward and backward linkages and horizontal interactions with other firms attracting the highest scores (1.8 and 1.6, respectively). The highest average ranking for interactions involving efforts by public, collective and other non-profit sector-promoting institutions is a mere 1.1, for fee-paying conferences. In this respect, the Montevideo software cluster is still no different from other developing country clusters, which tend to be characterized by tenuous and ill-developed linkages with promoting organizations and weak regional innovation systems in general (Arocena and Sutz, 2001). The overall picture emerging from Table 5.6 is one in which local transactions, rather than spillovers, are the dominant interaction-based knowledge transfer mechanism. We conclude that geographical proximity clearly matters, but the idea that all the crucial knowledge is ‘in the air’, in the way that Marshall (1920) suggested and many others have adopted, is not clearly confirmed by our findings. Local knowledge acquisition for innovation is also strongly bound up with local market transactions. Finally, we should take note of the fact that international sources of knowledge have also been important for the innovative performance of the cluster. The sample average score for the international knowledge transactions (IKT) of 5.4 suggests that international knowledge transfer contracts obviously do occur in the industry. In addition, international knowledge spillovers are significant, with a sample average score of 5.9. The high number of spin-offs that originated from TNCs with a foreign
146
Sectoral systems of innovation and production
home base could also be significant (indirect) sources of international knowledge spillovers. In sum, our results point to the importance of both local and international knowledge linkages for fostering the firms’ innovation processes, a finding that is not unique to our study (see, for example, Simmie, 2003). Quite possibly, the local and non-local knowledge flows are complementary. International knowledge flows not only matter for the directly receiving firms in the cluster; they also subsequently diffuse more widely through the operation of all kinds of local knowledge transmission mechanisms.
6.
THE ROLE OF PUBLIC POLICY
Although public policy cannot be credited with actively promoting the emergence of the sector, over time it has become a factor of some importance for the sector’s more recent expansion and its future growth potential. Public policy, and in association with this the role played by public institutions (see section 7), is therefore an important contextual factor influencing the performance of the sector. A direct policy encouraging the development of a local software sector was not in place in Uruguay in the early 1990s. The impact of the state on the emergence of the software industry was indirect and primarily entailed the fostering of human capital accumulation through investments in education. There are public as well as private university programmes in computer engineering and informatics in Uruguay. The majority of the graduates come from the Universidad de la República (47 per cent), while 45 per cent have studied in the University ORT of Uruguay.6 The Catholic University provides 5 per cent of the graduates, and the remaining 3 per cent come from the Instituto Universitario Autónomo del Sur and the Taller de Informática (Mejía and Rieiro, 2002). The number of graduates from computer study programmes in Uruguay is somehow ambiguous. The main problem arises from the fact that at the Universidad de la República computer engineering graduates also take an analyst programmers Licenciado degree as a part of their curriculum. Thus, they are counted twice (first as engineers and then as analysts/programmers) in the graduate statistics. A better way to calculate the number of graduates would be to consider only the analyst and programmer graduates in order to avoid the double counting problem. Using this methodology, we find that an average of 250 graduates in computer studies received their degree every year for 1990–99. Aside from promoting relevant education, the government’s role has
The software sector in Uruguay
147
not been positive. The state possibly even discouraged the strengthening of the capabilities of the budding software industry by offering important contracts for government projects to foreign firms (Kesidou, 2007). Fiscal incentives started only in 1999. In particular, software products are exempted from taxes imposed on the revenues of industrial and commercial activities (Impuesto a la Renta de Industria y Comercio: IRIC). In addition, the exports of software services are exempted from value-added tax (Failache et al., 2004; Pérez Casas, 2004). In 2001, the Laboratorio Tecnológico del Uruguay (LATU) in collaboration with the University ORT of Uruguay created an incubation programme: the Ingenio, an initiative financed by the Inter-American Development Bank (IDB). It hosted approximately 30 firms and offered infrastructure and training in order to strengthen the business and/or technical capabilities of the firms. However, the participation was limited to firms that had their own resources: usually professionals with previous experience who took the decision to start a software business. In this respect, the programme did not succeed in attracting a large participation, because it did not offer grants or other types of financing to the participants. This is because software development requires approximately two years of R&D before product commercialization can begin. Although the president of LATU acknowledged the shortcomings of the incubation programme, he also stated that, overall, the incubation programme was a good experience for Uruguay. Since December 2005 the incubation programme has given small grants to the participants in the form of a salary. Simultaneously, the Uruguayan government initiated the Programa de Desarrollo Tecnológico (PDT). The Ministry of Education and Culture was in charge of executing the project, which was financed 80 per cent by IDB and 20 per cent by the Uruguayan government itself. The main objective of the programme was to promote innovation by SMEs in Uruguay and to stimulate the mobility of local researchers. As far as the first objective is concerned, firms were called to submit innovative projects for financing. However, this programme addressed all fields of research and not solely the software sector. Competition for finance was prominent, and grants were difficult to obtain. The programme has already been discontinued.
7.
SUPPORTING INSTITUTIONS AND PRIVATE NETWORKS
In the course of time, sectoral support organizations have sprung up to aid the development of the software sector. These institutions did not play
148
Sectoral systems of innovation and production
a notable role in the development of the sector, however; more appropriately, their emergence should be seen as a response to the remarkable growth of the software sector, aiding its further development. Zonamérica Business and Technology Park is a key location for a number of the most important software enterprises in Uruguay. This is a privately owned, tax-free zone founded in 2002, located on the outskirts of Montevideo. Zonamérica created a technological park, the so-called Silicon Plaza, using the most advanced infrastructure. Besides software firms, other high-tech firms such as biotechnology firms are located in Zonamérica. The rest of the firms are usually financial business, logistic and distribution centres. Firms located in Zonamérica are exempted from customs duties and taxes (Whitelaw, 2004). These are unique conditions, since in most free-trade zones in the world only customs benefits are granted while tax exceptions are partially given. Consequently, Zonamérica Business and Technology Park offers many advantages to foreign firms that wish to enter the Latin American market. One of the largest Indian software firms, Tata Consultancy Services (TCS), decided to settle down in Zonamérica. According to the vice-president of TCS-Iberoamérica the decision to invest in Uruguay, and in particular in Zonamérica, was taken after careful examination of other alternative locations such as Costa Rica, Chile, Brazil and Argentina. Among other reasons the VP of Tata stressed that ‘Zonamérica is considered to be the most advanced technology park in the Latin American region and, in combination with the qualified personnel available at a reasonable price, and the political stability of Uruguay, it constitutes the best location for the objective of our business’. Currently the sector has two main institutions that serve the interests of local parties, CUTI and Integro. In total 2 216 companies (including many single-person firms) are registered with CUTI. The institution assists firms to develop their business capabilities and reinforces common action for the promotion of the Uruguayan software products in foreign markets. It functions as a link between the software firms and national and multinational funding organizations. In addition, CUTI organizes trade fairs and exhibitions in order to promote Uruguayan software abroad and to improve the image of the country as a technology centre. CUTI has administered a Programme for the Support of the Software Sector (Programa de Apoyo al Sector del Software) since 2002. Fiftyfive per cent of the financing of the programme comes from the IDB, the remainder being contributed by CUTI itself. The main objective of this programme has been to stimulate the export capability of the firms. The campaign is aimed at the Latin American markets. According to one participant of this export mission, the main strategy to enter a foreign market involved a series of calculated steps. ‘Usually a local consultant, with an
The software sector in Uruguay
149
extended network of contacts and knowledge of the local market, is hired. In addition, we inform the potential customers, previous to our visit, about the software industry of Uruguay.’ The simplest strategy for exports is the search for distributors in the foreign market. A more advanced strategy which involves more transfer of knowledge is the creation of technology alliances between foreign and local software firms, which would assimilate and then commercialize the technology. Larger and dynamic firms would also open a branch of their firms in the foreign market, from which they would be able to offer the necessary services. The director of a local software firm that actively participates in CUTI explains its role: CUTI is a political actor that gains its relevance by representing common needs of the sector. CUTI is the interlocutor which searches, and connects the different political segments nationally. CUTI puts strategic goals above the individual goals of the firms. The concept of the mission of CUTI is to take overall action that the firms cannot take individually. The benefits of these actions will be seen in two or three years. However, most of the firms in Montevideo do not perceive the role of CUTI in the same way. They fail to see the importance of CUTI for the development of the software business.
Integro is an IT business alliance. It represents the effort of local firms to share costs and acquire training in relation to quality certifications and marketing strategies for exports. Integro is perceived to be more significant than CUTI for knowledge spillovers between the software firms in the cluster. The director of a local software firm explains the reasons for its participation in Integro: “Integro is an alliance which enables us to better commercialize our product abroad. In addition, we receive training in marketing, sales and negotiation all together. It is our intention to work together to achieve synergy.” This opinion is supported by the directors of several other firms. For example, the director of another software firm elucidates the activities of Integro: Eight companies that participate in the Integro group decided to undertake the CMMI [Capability Maturity Model Integration] project together. CMMI evaluates the quality of the process by which each firm produces software. It is difficult for a small firm to go through such a process on its own due to financial and time constraints. Cooperation generates synergies in terms of access to credit and sharing of experiences. Firms may gain easier access to finance because they provide guarantees to one another. They also exchange new knowledge and experiences. There are no earlier experiences with CMMI in Uruguay, so there is no local benchmark against which firms can assess how they are doing. Being part of a group of firms that follow the same trajectory is important because it enables us to compare experiences, avoid reappearance of mistakes and promote the application of best practice, etc.
150
Sectoral systems of innovation and production
This overview of institutional support to Montevideo points to a largely passive attitude among policy makers in the early stages of the industry’s development. The limited importance of the institutions also shows through in the limited knowledge linkages that they have established with the sample firms (see section 5), however, but it seems that they did begin to react constructively once the industry began to show its teeth. One fruitful area is the organization of conferences and exhibitions, which makes possible collaboration and exchange between private parties. However, it will take time before public institutions such as CUTI will be able to play a noteworthy role, particularly because many firms in developing countries, especially small ones, do not have positive expectations about public initiatives. It will take time and effort to win them over. The review also makes it clear that self-organizing initiatives among firms in the cluster (like Integro) are emerging. This is a positive development because it signals an awareness on the part of companies that they can capture external effects and achieve more through coordinated action and a willingness to subordinate their individual interests in favour of broader common goals.
8.
CONCLUSIONS
Drawing together the discussion from the previous sections, we can make the following concluding observations concerning the growth dynamics of the sector, and the underlying driving forces of its competitiveness and innovative prowess, from a sectoral systems of innovation perspective: ●
●
Since its emergence, the Montevideo software cluster has undergone an impressive development, showing clear signs of technological upgrading and diversification. From an exclusive focus on labourintensive software services it has moved into software design and more complex services. The sector also shows considerable signs of innovative activity, particularly in the form of product innovation. The emphasis on product innovation is clearly industry-specific, and has to do with the nature of the knowledge base in the software sector in general. The nature of knowledge and its accumulation in the software sector also influences the size-wise composition of the industry. Entry barriers are low because the activities are knowledge-intensive rather than capital-intensive. This makes it easy for new entrants to set themselves up in business, resulting in a fragmented sector composed of many small and medium-scale companies, and just a few
The software sector in Uruguay
●
●
●
●
151
large players that dominate the high-skill end of the market where entry barriers are higher. Firms’ knowledge acquisition activities were varied and included learning from a variety of internal sources and external sources. Among the internal sources, R&D featured prominently. Further, the firms’ educational base was a major aspect of their absorptive capacity and innovation capability. A more in-depth analysis of the external knowledge accumulation processes of the firms pointed to the importance of knowledge cumulativeness at the local sectoral level. There is a lot of evidence of local knowledge spillovers, involving a range of different local parties. This readily explains the current organization of the industry in terms of one tightly concentrated spatial cluster. Local knowledge flows involving market transactions also seem to be important for the firms’ innovation processes, alongside free spillovers. Firms evidently acquire a lot of bits and pieces of external knowledge and information which is complementary to what they themselves possess, and they fit these into their own knowledge base. The existence of local inter-firm complementarities which manifest themselves through market transactions may very well be an additional explanation for the spatial clustering observed in the sector, all the more so because many knowledge transactions in the sector could also have a spillover element in them. The local spillovers take a number of different forms. The three main types of spillover mechanisms commonly observed in geographical clusters (see, for example, Saxenian, 1994), are recognizable: (a) firm spin-offs; (b) inter-firm labour mobility; and (c) inter-party business interactions. We appear to be looking at a classic example of the sort of innovative cluster documented in the economic geography literature (Jaffe et al., 1993; Audretch and Feldman, 1996; Martin and Sunley, 2003) and the industrial district literature (Storper, 1995). These findings have important policy implications for Uruguay and other developing countries. The first building block of the sectoral systems of innovation approach, namely the sector’s knowledge base and learning processes, plays a dominant role in driving the growth of the software sector in Uruguay. In particular, the presence of knowledge spillovers and knowledge flows based on market transactions taking place at the local level reinforces the idea that geographic proximity may generate advantages related to the fast circulation of knowledge, not only in advanced economies but also in developing countries. A policy ingredient for state and regional agencies would be the identification and support of these geographic spaces.
152 ●
●
●
Sectoral systems of innovation and production
The role of public policy has evidently been modest, with the notable exception of the indirect effect associated with the promotion of human capital formation in the country in general. Aside from this, the sector pulled itself into existence through a combination of favourable demand conditions in the Latin American region at a time when many skilled professionals with knowledge of software requirements in user sectors were looking for productive employment in a badly performing economy. To the extent that sectorspecific policy does exist, its origins are too recent to be able to detect a significant influence. This kind of reactive policy stance is very typical of the experience of developing country clusters of SME in general. Several lessons emerge from the analysis about what a more proactive public policy could look like. First, it should facilitate labour mobility by promoting more flexible and less regulated labour markets, especially for SMEs. Morever, public policy needs to reinforce the networking activities of firms both locally and internationally, invest in higher education and in R&D and provide R&D subsidies. More bridges should be built between the private sector and local universities. The attraction of more innovative foreign firms to the sector could also help to foster more local spillovers and learning through inter-firm collaborations and spin-offs. Awareness of the importance of sectoral systems of innovation is crucial for drawing up policies that enhance accumulation and circulation of knowledge in developing countries. The main difference between the sectoral system that we examined in Uruguay and others located in advanced countries derives from the weak role played by formal institutions. We face a fragmented SSI, which in the examined Uruguayan case was characterized by low levels of interactions between firms and universities and public and private institutions. For SSI to be a potential path of economic development for developing countries and in particular for small countries, public policy needs to nurture ‘interactive learning spaces’ (Arocena and Sutz, 2004) alongside developing the knowledge base (i.e. investments in education) and learning processes (i.e. investments in R&D) of firms and institutions.
NOTES 1. Own calculations, based on data from the Central Bank of Uruguay. 2. The Uruguayan Business Association of Information Technologies (CUTI) carries
The software sector in Uruguay
3. 4. 5.
6.
153
out annual surveys and collects information on a sample of 149 firms with more than one employee and 1600 individuals working in software development and consulting services. This section draws on Teeselink (2006). Based on data from the World Bank World Development Indicators 2005 (database 1980–88), and the Central Bank of Uruguay (database 1988–2004). We follow the definition of knowledge spillovers given by Jaffe as ‘intellectual gains by exchange of information for which no direct compensation is given to the producer of the knowledge or the compensation is less than the value of the knowledge’ (Jaffe, 1996, p. 5). ORT is a world-wide organization for education and training. The acronym stands for Obshestvo Remeslenofo zemledelcheskofo Truda, meaning ‘The Society for Trades and Agricultural Labour’.
REFERENCES Arocena, R. and Sutz, J. (2001), ‘Changing knowledge production and Latin American universities’, Research Policy 30, pp. 1221–1234. Arocena, R. and Sutz, J. (2004), ‘Emerging neoperipheral structures and gardening policies’, Paper presented at the Druid summer conference. Audretsch, D. and Feldman, M.P. (1996), ‘R&D spillovers and the geography of innovation and production’, American Economic Review, 86(3) 630–640. Bitzer, J. (1997), ‘The computer software industry in East and West: do Eastern European countries need a specific science and technology policy?’, http://www. diw.de/deutsch/produkte/publikationen/diskussionspapiere/docs/papers/dp149. pdf (accessed on 23 December 2005). Cohen, W. and Levinthal, D. (1990), ‘Absorptive capacity: a new perspective on learning and innovation’, Administrative Science Quarterly, 35(1), pp. 128–152. Commander, S. (2005), The Software Industry in Emerging Markets, Edward Elgar, Cheltenham, UK and Northampton, MA, USA. Cusumano, M. (2003), ‘Finding your balance in the products and services debate’, Communications of the ACM, 46(3), pp. 15–17. CUTI (2004), The Uruguayan Chamber of Information Technologies, survey data downloaded from www.cuti.org.uy. Dahlman, C.J., Larson-Ross, B. and Westphal, L.E. (1987), ‘Managing technological development: lessons from the newly industrializing countries’, World Development 15(6), pp. 759–775. Failache, C., Muinelo, L. and Hounie, A. (2004), ‘Estudio de competitividad sectorial: technologías de la información en Uruguay’ [Sectoral competitiveness study: information technology in Uruguay], Inter-American Development Bank, Washington, DC. Heeks, R. (1999), ‘International perspectives: software strategies in developing countries’, Communications of the ACM 42(6), pp. 15–20. Jaffe, A. (1996), ‘Economic analysis of research spillovers and implications for the Advanced Technology Program’, Paper presented for the Advanced Technology Program, December. Jaffe, A.B., Trajtenberg, M. and Henderson, R. (1993), ‘Geographic localization of knowledge spillovers as evidence by patent citations’, Quarterly Journal of Economics 63(3), pp. 577–598.
154
Sectoral systems of innovation and production
Katz, J.M. (ed.) (1987), Technology Generation in Latin American Industry, Macmillan, London. Kesidou, E. (2007), ‘Local knowledge spillovers in high tech clusters in developing countries: the case of the Uruguayan software cluster’, Ph.D. thesis, Faculty of Technology Management, Eindhoven University of Technology, Eindhoven. Lall, S. (1992), ‘Technological capabilities and industrialization’, World Development, 20(2), pp. 165–186. Malerba, F. (2002), ‘Sectoral systems of innovation and production’, Research Policy 31, pp. 247–264. Mansell, R. and Wehn, U. (1998), Knowledge Societies: Information Technologies for Sustainable Development, Oxford University Press, Oxford. Marshall, A. (1920), Principles of Economics, Macmillan, London. Martin, R. and Sunley, P. (2003), ‘Deconstructing clusters: chaotic concepts or policy panacea?’, Journal of Economic Geography 3, pp. 5–35. Mejía, T. and Rieiro, M. (2002), ‘Aprendizaje, innovación y cometitividad de la industria del software en Uruguay’ [Learning, innovation and competitiveness of the software industry in Uruguay], Trabajo de investigacion monografica, Universidad de la Republica (UDELAR), Montevideo. Pérez Casas, A. (2004), ‘Las conditiones sociales del florecimiento de la industrial Uruguaya de software’, MA thesis, Faculty of Social Sciences, Universidad de la Republica, Montevideo. Rabellotti, R. (2003), Software Clusters in Mexico: Guadalajara, Monterrey, Distrito Federal, Aguascalientes, Inter-American Development Bank, Washington, DC, http://www.iadb.org/sds/doc/casestudysoftwaremexico.pdf. Romijn, H. (1999), Acquisition of Technological Capability in Small Firms in Developing Countries, Macmillan, London, and St Martin’s Press, New York. Saxenian, A. (1994), Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Harvard University Press, Cambridge, MA. Simmie, J. (2003), ‘Innovation and urban regions as national and international nodes for the transfer and sharing of knowledge’, Regional Studies 37(6/7), pp. 607–620. Snoeck, M., Sutz, J. and Vigorito, A. (1992), Tecnología y transformación: La industria electrónica. Uruguaya como punto de apoyo, CIESU-Trilce, Montevideo. Stolovich, L. (2003), ‘What does the data on Uruguay’s information technology industry indicate’, mimeo, CUTI/PASS, Montevideo. Storper, M. (1995), ‘The resurgence of regional economics, ten years later: the region as a nexus of untraded interdependencies’, European Urban and Regional Studies 2(3), pp. 191–221. Teeselink, M. (2006), ‘The potential of the Vietnamese software sector: Export success or domestic strength?’, M.Sc. thesis, Faculty of Technology Management, Eindhoven University of Technology, Eindhoven. UNCTAD (2002), ‘Changing dynamics of global computer software and services industry: implications for developing countries’, http://www.unctad.org/en/ docs/psitetebd12.en.pdf (accessed on 23 December 2005). Veloso, F., Botelho, A.J., Tschang, T. and Amsden, A. (2003), Slicing the Knowledge-Based Economy in Brazil, China and India: A Tale of 3 Software Industries, http://www.softex.br/media/MIT_final_ing.pdf. Whitelaw, J.A. (2004), ‘The Uruguayan tax free zones’, in Latin American
The software sector in Uruguay
155
Tax Group of PricewaterhouseCoopers (eds), Practical Latin American Tax Strategies 7(2). WITSA (2000), Digital Planet 2000: The Global Information Economy, World Information Technology and Services Alliance, Vienna, VA. World Bank (2004), World Development Indicators 2004, World Bank, Washington, DC.
6.
Sectoral system of innovation in Brazil: reflections about the accumulation of technological capabilities in the aeronautic sector (1990–2002) Rosane Argou Marques1 and L. Guilherme de Oliveira2
1.
INTRODUCTION
Since the end of the 1990s, Brazilian government policies have given incentives to strengthening linkages between firms and the science and technology infrastructure (research institutions) for promoting innovation and, consequently, for improving the country’s competitiveness. The Brazilian government put forward innovation as the focus for policies when it launched in 2004 the “Industrial, Technology and Trade Policies”. Since then, government agencies and research institutions have increased their role of articulating, modelling programmes, financing and diffusing technological knowledge, especially for the segment of small and mediumsized firms (SMEs). This set of industrial and S&T policies has a different focus when compared to those implemented by the military government (1960s to mid-1980s) and even after that phase, during the 1990s, when the government decision in relation to an explicit industrial policy was that “not having an industrial policy is better”. The military government had given a strong role to multinationals and state government firms as the primary sources of technology transfer to the local SMEs’ development. These policies have followed the international path in terms of objective, which was import substitution industrialization, although there are several differences among countries in the way they have been implemented. Some general examples of these differences can be observed when comparing the South Korean policies for industrialization during the 1970s and 1980s and the Brazilian ones. 156
The aeronautic sector in Brazil
157
South Korean policies followed the Japanese by controlling imports and stimulating exports; regulating capital and technology transfer from abroad; investing in increasing the number of engineers; and giving incentives to the development of technological capabilities in Korean firms. Large private Korean corporations had an important role in the Korean policies. Brazilian policies were towards incentives for increasing foreign direct investment, building up industrial sectors in order to be self-sufficient and creating government-owned firms in strategic areas. The government invested in the education system and in research institutions, as well as research and development activities at the government-owned firms. The government-owned firms and multinational corporations played an important role in building up a diversified industrial system. These firms have focused mostly on the Brazilian internal market except, for example, the aeronautic sector. In an attempt to understand the accumulation of technological capabilities experienced by Brazilian suppliers to the aeronautic sector, this chapter explains the differences in activities carried out among them and technological changes implemented from the 1970s to 2002. The chapter makes the connection between activities undertaken and their impact in terms of technological change implemented and level of technological capability built up by the SMEs. Additionally, it distinguishes this connection over time, aiming at identifying if there were any differences in the factors influencing the accumulation of technological capabilities experienced by the SMEs. This issue is particularly addressed in the aeronautic sector because it is one of the most innovative and robust in the country. This sector was responsible for 3.1 per cent of Brazilian total exports in 2005, which was US$118.308 million (FOB). Moreover the Brazilian aeronautic firm Embraer is one of the largest passenger aeroplane manufacturers worldwide, with a focus on commercial, executive and defence aviation. Embraer was established in 1969 by the Brazilian military government and was managed by the military until 1994, when it was privatized. Although it is successful in achieving international competitiveness in its specific market segment for regional jets flying from 45 to 108 passengers, Brazil has not been able to consolidate the supply chain of Embraer within the national borders. There are few Brazilian firms supplying Embraer and some of the foreign first-tier suppliers established in Brazil during the 1990s. In fact, the import content increased from approximately 68 per cent in the 1980s (Dagnino and Proença, 1989) to approximately 95 per cent in the 1990s (Bernardes, 2000a). Therefore, there is a question about how the local Brazilian suppliers are maintaining themselves in the competitive supply
158
Sectoral systems of innovation and production
chain of Embraer. The findings show that they are improving their innovative capabilities in two directions: by strengthening their basic technological capability regarding production processes and, in a few cases, by upgrading to intermediate and advanced levels of innovative capability. The relationships with Embraer, foreign buyers and Brazilian research institutions are the main sources of knowledge for the technological learning experienced by these Brazilian suppliers that are “surviving” in the supply chain of Embraer. The structure of the chapter is as follows. We examine the literature about the system of innovation and technological capability accumulation and build on that the analytical framework in the next section. Following that, the Brazilian aeronautic sector is briefly described in section 3, and the method is explained in section 4. Then reflections about the accumulation of technological capabilities experienced by Brazilian small and medium-sized firms (SMEs) are made in section 5, and final comments are in section 6.
2.
SECTORAL SYSTEM OF INNOVATION AND TECHNOLOGICAL CAPABILITY: ANALYTICAL FRAMEWORK
2.1
The System of Innovation Approach
The discussion underlying the literature about the sectoral system of innovation considers the fact that sectors have specificities that the analytical framework utilized in the national system of innovation can hardly capture. Particularly, one could explain important characteristics of innovation that the macro-environment has not clearly shown by focusing down the camera to look at the actors, connections, modes, nodes and technological dynamics of sectors. In this case, it is useful to utilize the concept of a sectoral system of innovation, which is bounded by a set of market and non-market relations with the purpose of technological development in products and services, particularly for the “creation, production and sales of products” (Malerba, 2002, p. 248). Broadly, systems of innovation are the network of government and non-government agencies, science and technology institutes, educational organizations and firms, among other organizations, whose flows influenced the direction and extent of innovation. The country’s macroeconomic and industrial policies, international regulations, market governance and socio-cultural institutions influence the network dynamism and trajectory. The interaction between the former and the latter has influenced knowledge accumulation and learning
The aeronautic sector in Brazil
159
processes in firms (Nelson and Winter, 1982; Freeman, 1987; Nelson and Rosenberg, 1993; Cooke et al., 1997; Hodgson, 1993). For understanding this issue, many authors have focused on distinct but interrelated areas of systems of innovation. Some of these areas are related to technological, sectoral, national, regional, financial and political, among others, systems of innovation. It is worth mentioning that the analytical framework utilized in this chapter focuses on examining the evolution of a sector, specifically its dynamics and transformation over time, regarding technological development and linkages among actors. This approach refers, thus, to the sectoral system of innovation idea. Although many researches have been done about this theme, less comprehension exists on the relation among actors in a sectoral system and technological capabilities accumulation (Malerba, 2002). Before focusing on the sectoral system of innovation, it is relevant for the debate to introduce a few considerations about technological change and the broad system of innovation literature. First, the theory that technological change is not an isolated process emerged as an attempt to explain innovative behaviour and the consequent technological capabilities accumulation and evolution in firms. Second, technological change is a consequence of the capability of firms in managing and generating innovation as well as in acquiring and diffusing technological knowledge (Freeman, 1987). In fact, the development of such a capability is a process that requires that a given firm interact with other firms, research institutions, universities and funding institutions, among other organizations. Third, government policies have an important role in regulating and coordinating the pace (quantity) and nature (quality) of the development of technological capabilities. This explains partly the differences of industrial development between countries, regions and sectors; other factors that may influence the differences are the specific endowments and characteristics of where firms are located (De Ferranti and Perry, 2002). Differences between the three aspects of technical evolution influence the trajectory of countries for catching up industrialized countries (Freeman, 1987; Viotti, 2002). Generally, researchers focus their analysis in two ways. First, they observe differences among countries and regions (national or regional systems) regarding the type of exported products, investments in R&D, investments in education and training, science and technology capabilities, industrial structure, and patents, among other variables (Freeman, 1987; Lundvall, 1992; Nelson and Rosenberg, 1993; Patel and Pavitt, 1994; Cooke et al, 1997; Cassiolato and Lastres, 1999; Viotti, 2002). Second, they examine technological diffusion and development within industrial
160
Sectoral systems of innovation and production
networks (technological or sectoral systems) (Carlsson and Jacobson, 1994; Carlsson, 1995; Breschi and Malerba, 1997; Malerba, 2002). Viotti (2002) considers that industrializing countries are adopters of technological knowledge from developed countries and their firms may develop incremental innovations according to their capabilities to do so. However, these firms are not developing innovations in the same sense that Lundvall (1992) or Schumpeter (Malerba, 2002) defined. In fact, Viotti refers to innovation in these countries as not really being innovation to the world market but adaptations to meet the specificities of industrializing countries. He compares the case of South Korea and Brazil and concludes that the former is an active learning system and the later is a passive learning system. An active learning system is characterized by the capability to improve and adapt technologies, while a passive learning system is characterized by the capability to adopt technologies. The conclusion that the Brazilian system of innovation, as in other Latin American countries, is not innovative is supported by other researches. Cassiolato and Lastres (1999), Katz (2000, 2001) and Bernardes (2000c) examined the technological behaviour of national and foreign firms and the influences of macroeconomic policies and industrialization strategies defined by governments on this behaviour. Their common argument is that the Brazilian system is fragmented, lacking long-term industrial policies. Moreover, they refer to local firms as lacking innovative capabilities to succeed in competing in the world market. Although there is a lack of explicit long-term industrial policies, some researchers hold that the Brazilian system of innovation is heavily influenced by the government development policies (Dahlman and Frischtak, 1990; Cassiolato and Lastres, 1999; Marques, 2004; Oliveira, 2005). Particularly, industrial structural changes resulting from import substitution industrialization policies maintained and strengthened the role of imported technologies and subsidiaries of foreign firms in the indigenous technological development. Dahlman and Frischtak (1990) observed that by 1960 more than 50 per cent of the total goods manufactured in Brazil were produced by subsidiaries of foreign corporations. They also noted that the government created the research infrastructure to improve technological capabilities and develop a local supply chain to support the production facilities of foreign subsidiaries and national state-owned firms. More recently, Quadros et al. (2001) and Costa and Queiroz (2002) argue that local foreign subsidiaries accounted for the largest share of private R&D activity in Brazil, which concentrates on the adaptation of products and processes to the local endowment. According to them, there has been a ‘moderate’ improvement in Brazilian firms’ technological capability after the Brazilian government shifted from import substitution industrialization
The aeronautic sector in Brazil
161
policies to liberalization policies during the 1990s, although local subsidiaries of foreign firms are still important innovators. 2.2
The Technological Capability Approach
More specifically to the firm level, further relevant literature regards learning and technological capability accumulation experienced by firms in industrialized countries. This literature highlights the trajectory of capabilities accumulation during the industrialization phase, focusing on the idea that technological change is endogenous or internal to the firm. Moreover, learning dynamics does not have only an endogenous own character but also elements of capturing external innovation through technology transfer, among other forms. In this way, efforts at technological adaptation associated with internal learning processes act upon the rhythm of acquisition of competencies, which can occur because of the characteristics of the external technology transferred (Oliveira, 2005; Fransman, 1984). Research about technological capabilities accumulation is based on the evolutionary perspective of innovative efforts undertaken by firms. Since the 1980s this perspective has been developed, considering firms as differing and dynamic organizations as well as stores of knowledge (Nelson and Winter, 1982). The evolutionary perspective also considers that firms evolve over time when they attempt to adapt themselves to their environment. This adaptation process has implications for the path of technological capabilities accumulation, which is related to the main characteristics of the innovative activities within firms, being uncertain and path-dependent on their knowledge base. Following this perspective, technological capabilities refer to the dynamic and competence-building activities firms undertake to generate new products, processes and services. There is a variety of definitions for technological capability. Earlier studies consider technological capability as the systematic efforts for acquiring knowledge to improve production capacity (Katz, 1976). Other studies refer to the “capacity to manage technology and to implement technical change” (Bell, 1984, p. 189). Some others include in the concept the ability of individuals, and infrastructure and activities undertaken to implement changes in production and techniques (Figueiredo, 2003, referring to the studies of Bell in 1982 and Scott-Kemmis in 1988). There exist some concepts that limit it to the ability of individuals and ignore their organizational context (Pack in 1987, in Figueiredo, 2001). Broader than Pack’s definition, Enos’s (1991) definition refers to the technical knowledge necessary to achieve a common organizational objective that is embodied in the mind of engineers and technicians. Both definitions focus heavily on
162
Sectoral systems of innovation and production
individuals as the locus of technological capabilities, neglecting important organizational aspects that these capabilities integrate (Figueiredo, 2001). In fact, based on previous studies, most literature about industrializing countries refers to technological capabilities as the ability to absorb, use, adapt, improve and change existing technologies. This ability involves the effective use of technological knowledge in production, investment and innovation (Westphal et al., 1985). A central role is given to a firm’s in-house technological learning efforts to master new technologies, adapting them to local conditions, diffusing, and exploiting them by exporting (Lall, 1992). At this stage, it is important to make the distinction between domains of technological capabilities that refer to distinct processes of learning. In developing a framework for distinguishing between the forms of technological development experienced by South Korea, Westphal et al. (1985) refer to three domains of technological capabilities: production, investment and innovation. Production capability consists in the ability to operate production processes and adapt them to changing market circumstances. Investment capability refers to the skills for expanding and establishing new production facilities. Innovation capability consists in the ability to carry out activities for creating and implementing changes in techniques and organizational processes. They argue that technological development is costly because it requires stable and long-term investments in skills and technological knowledge as well as improvements in organizational processes for learning to adapt imported technologies. Drawing on Westphal et al. (1985), Lall (1992) developed a framework for explaining firm-level differences in technological capabilities. The framework considers technological capabilities as divided into two domains: investment and production, which innovative activities vary according to the degree of complexity from simple routine to adaptive and innovative. Adding to previous work, Lall argues that production capability is not only the ability to operate and improve imported production techniques but includes the firm’s in-house efforts in engineering for absorbing technologies, as well as linkages with other organizations. Linkages capability refers to the ability to transmit technological information and receive it from other organizations, such as suppliers, consultants, customers, service firms and universities. These linkages are supposed to assist the firm to improve its productive efficiency and also the diffusion of technologies (Lall, 1992). Kim (1997b) also examined the process of technological capabilities accumulation experienced by Korean firms developing a “learning model” of acquisition–assimilation–improving foreign technologies. He considers technological capability as the ability of firms to utilize technological
The aeronautic sector in Brazil
163
knowledge in an efficient manner in order to assimilate, use and adapt existing technologies. It has three main elements: production (management, engineering and repair and maintenance); investments (training, project development and implementation); and innovation (basic and applied research, and development of new products, processes and services). The accumulation of technological capabilities took place in Korean firms from imitation of foreign technologies, such as through reverse engineering and technology transfer, to innovation based on firms’ internal efforts to develop and produce new products to the market. Based on this research, more recent studies examined technological capabilities in a different spectrum: from the analysis of internationalization of innovative capabilities (Ariffin, 2000) to explaining the differences between the processes of accumulation in firms (Figueiredo, 2001). Drawing on Lall (1992) and Bell and Pavitt (1995), among others, Ariffin (2000) and Figueiredo (2001) consider two levels of technological capabilities according to the activities for generating and managing technological change undertaken by firms: routine capability and innovative capability. Routine capability refers to the firm’s ability to utilize knowledge and technologies and undertake activities in distinct functions: product, production and organizational processes. Innovative capability permits the creation, modification or improvement of these functions. In fact, their routine capability refers to the first level of technological capability, defined by Lall (1992) as experienced-based capability. Innovative capability, according to Lall, refers to the second and third levels of complexity, defined as search-based and research-based. This distinction is important for explaining the path of technological capabilities accumulation experienced by firms in industrializing countries, which are building up capabilities from routine to innovative levels. The literature reviewed so far analyses technological capabilities accumulation by basically examining distinct aspects of learning efforts undertaken by large firms, which are national champions, foreign subsidiaries, state-owned firms or recently privatized firms that are acquiring foreign technologies. These firms have to actively invest in the development of skills, knowledge and experience for learning and consequently building up technological capabilities. Their learning efforts change over time according to the technological complexity of products and production. This literature thus highlights the importance of deliberately investing in learning to build up technological capabilities. It does little, however, to discriminate the dynamic process of technological capabilities accumulation in small and medium-sized firms supplying a complex system manufacturer, such as the aeronautic sector located in an industrializing country.
164
2.3
Sectoral systems of innovation and production
The Analytical Framework
Based on the literature reviewed, the broad analytical framework used for organizing the reflections about technological capability of Brazilian firms supplying to the aeronautic sector is shown in Figure 6.1. At the centre of our reflections we consider the relationship between technological change (left side) and technological capability accumulation (right side). In fact, technological change is related to the level of technological capability of firms and to their linkages with other firms and research institutions (Lall, 1992; Fransman, 1984). In fact, we move further the understanding of a sectoral system of innovation by focusing on the co-evolution of technological change, technological capability and linkages between suppliers, buyers and research institutions. We have also attempted to make the connections between those three elements of the sectoral system. We consider, in the chapter, domains and levels that are relevant to the analysis of technological change and technological capability in the Brazilian aeronautic sector based on interviews and visits to firms, and on previous research (Frischtak, 1992, 1994; Bernardes, 2000a, 2000b, 2000c; Bernardes and Oliveira, 2002; Marques, 2004; Oliveira, 2005). The domains considered in the framework are product, production (process and equipment-related), and organization of project management and design procedures. Particularly, technological change consists of the introduction in the Sectoral environment Linkages: National–national firms; national–foreign firms Firms and research institutions
Technological changes – levels: – High impact – Middle impact – Low impact
Product Production Organization
Technological capability – levels: Innovative: – Advanced – Intermediate Routine: – Pre-intermediate – Basic
Type of linkages: Procurement Collaborative agreements (co-design, research) Technology transfer
Figure 6.1
Analytical framework for the aeronautical system of innovation
The aeronautic sector in Brazil
165
firm of technology embodied in the three domains (based on Bell and Pavitt, 1993). It can be related to the introduction of completely new technologies to the firm or to adaptations in the already existent product, production and organizational processes. This research classification is thus based on three levels of impact of the change in the firm (see Appendix 6.1). The incorporation of substantially or completely new technologies in any or all of these three domains belongs to high-level impact. Changes incorporated in product, production and organization for improving and upgrading the already existent technologies refer to middle-level impact. Finally, the duplication of technologies already in utilization by the firm associates with low-level impact. It is important to stress that the changes may be interconnected in the sense that a change in product may influence changes in production, which may call for changes in organizational processes. This interconnection has not been explored in this chapter. Technological capability is defined as the ability of firms to manage and generate technological changes, i.e. their ability to innovate (Bell, 1984). The domains are product-centred, production (process and equipments), and organizational processes (project management and design procedures). Product-centred technological capability consists in the ability of firms to innovate in product design, specifications and quality. Production technological capability is classified in two domains: production processes and production equipment (machinery, equipment and software). Organizational-centred technological capability regards the ability of firms to manage and generate changes in project management and design procedures. Technological capability is classified in two levels: routine technological capability and innovative technological capability. Routine technological capability is the ability of a firm to utilize and adapt knowledge to implement changes in the distinct domains, and has two levels: basic and pre-intermediate (see Appendix 6.2). In fact, this level is related to the production capability of firms, whereas innovative technological capability permits the creating, modifying or improving of technologies. There are two levels for the latter capability: intermediate and advanced. Moreover, we examine some aspects of the Brazilian aeronautic system of innovation which are historical developments of the sector in terms of government support, launch of new products, and linkages among actors in the system. Particularly regarding the classification of linkages, we consider it may be: procurement of services and goods; collaborative agreements for generating technological changes; and technology transfer.
166
3.
Sectoral systems of innovation and production
THE BRAZILIAN AERONAUTIC SYSTEM OF INNOVATION
The Brazilian aeronautic sector is one of the most important high-technology sectors in the country. The evolution of this sectoral system of innovation was examined in Marques (2004), which set out the background for the development of this section. Moreover, the main objective here is to examine how Brazilian aeronautics has been developed since the foundation of Embraer in 1969. Using a historical approach, this section will give an overview of the steps followed by Embraer to become the world third producer of regional jets seating from 45 to 108 passengers. Therefore section 3.1 briefly explores important features in the period pre-foundation of Embraer (before 1969). Then section 3.2 examines the main characteristics of the three phases in which it has been possible to identify more explicit industrial development policies: (i) starting-up (1969–78); (ii) seeking the international market (1979–94); and (iii) post-privatization of Embraer (1995–2002). Finally, the general market characteristics are described in section 3.3. 3.1.
The Brazilian Aeronautic Sector in the Period before 1969
The period before 1969 can be examined taking into consideration three important periods of development of the Brazilian aeronautic sector. The first period was 1910–30. According to Dagnino and Proença (1989), the first remote step for developing an aeronautic sector in Brazil was in the decade to 1910 when Santos Dumont, a Brazilian industrialist, developed the first heavier-than-air flying machine. Although some investments were made in this period, Brazil lacked engineering and technological capabilities, as well as the government policies, that such an industry required. Then, the efforts made did not result in any significant development of the Brazilian aeronautic sector. The second period was during the Second World War (1935–1945). Brazil, as an ally of the USA, functioned as a producer of attack aircrafts. During this time, the American Air Force trained Brazilian pilots and turned out aeronautic engineers to help in the production of aircraft. Brazil produced approximately one aircraft per day for the USA during this time (Dagnino and Proença, 1989). Although more government efforts were made than in the first phase, they were not sufficient for setting up the basis for the sectoral economic growth. The lack of technological and engineering capability to design and produce aeroplanes in Brazil continued. At that time, the production was restricted to light aircraft for utilization in agricultural matters. Therefore, the Brazilian Air Force decided to create an aeronautical
The aeronautic sector in Brazil
167
institute to turn out highly qualified engineers to support the infant aeronautic sector. The Technological Institute of Aeronautics (ITA) was founded by the end of the 1940s, which marked the beginning of the third period. ITA was founded with the support of MIT and NASA. ITA turned out approximately 200 engineers until 1970, but most were contracted to work in other sectors owing to the lack of companies to contract them in the aeronautics sector (Dagnino and Proença, 1989; Dagnino, 1993; Bernardes, 2000a). During the 1950s, the Brazilian Air Force was aware that it was also necessary to create a research centre for applying aeronautic engineering knowledge to the development of a “Brazilian aircraft” that could fly according to the particular endowment and characteristics of the Brazilian territory. They founded the Aerospace Technical Centre (CTA), which absorbed ITA, and developed other institutes for aeronautic research. The main research project at CTA was for the design and production of a 19-seat aircraft. The Brazilian Ministry of Aeronautics contracted an entire research group from Germany that worked with Brazilian engineers from ITA with the aim of developing such an aircraft. The first prototype flew in 1959, and further improvements were necessary. In 1969, the project was then concluded and the first aircraft, called Bandeirante, could be produced. This group of researchers founded the first state-owned aircraft producer in Brazil in 1969, Embraer, with the support of the Brazilian Ministry of Aeronautics (Dagnino and Proença, 1989; Dagnino, 1993; Bernardes, 2000a). 3.2.
The Brazilian Aeronautic Sector from 1969 to 2002
The starting-up phase began in 1969 when the Brazilian Ministry of Aeronautics founded the Empresa Brasileira de Aeronáutica S.A. Embraer. The aeroplane manufacturer was founded as a spin-off of the CTA, with the objective of supplying the Aeronautic Command with parts, components, and training and attack aircraft (Dagnino and Proença, 1989; Coutinho and Ferraz, 1993; Bernardes, 2000a). According to Frischtak (1994, p. 602), ‘although the production of airplanes in Brazil dates back to 1910, when the first monoplane was built in the country, the development of the Brazilian passenger airplane manufacturing can be equated with the development of Embraer’. The first regional aeroplane manufactured in Brazil was during the 1970s and had the capacity for 19 passengers, with twin-engine turbo propeller (Frischtak, 1992; Bernardes, 2000a). Before that, there were approximately six companies that manufactured one- and/or two-seats aeroplanes, and only one was in the market by the time Embraer was founded.
168
Sectoral systems of innovation and production
The Ministry of Defence was the main buyer and also gave strong tax incentives and subsidies to Embraer for developing production and technological capabilities to manufacture the 19-seat aeroplane (Dagnino and Proença, 1989; Coutinho and Ferraz, 1993; Bernardes 2000a). These incentives were oriented for financing (through subsidies and tax exemption), marketing (through procurement and protectionism) and developing technologically (through the creation of special decrees for technology transfer and supporting research). In this first ten years of existence, the main market was national. According to Dagnino and Proença (1989), although the Ministry of Defence had heavily invested in the creation of a national production chain (aeroplane and its systems and components), approximately 68 per cent of parts, components and sub-systems were imported. Nevertheless, some suppliers had developed capacity from the production of parts and components to the production of small aeroplanes (one to ten seats), such as Aeromot (located in Porto Alegre, State of Rio Grande do Sul) and Neiva (Botucatú, State of São Paulo). This development was possible owing to the Ministry of Defence special programmes for ‘nationalization’ of aero parts (systems, structural parts and other components). However, according to interviews at the Institute for Development and Coordination of the Aerospace Industry (IFI), low production scale, high quality, and high development costs of aero parts influenced the production concentration at Embraer. Few local supplier firms had developed the capacity to produce parts and components, and they had relied heavily on technological transfer from CTA through IFI consultancy. Therefore, most aero parts were imported at the end of 1978. The seeking international market phase corresponds to an increase in exports of small-body aeroplanes (10–30 seats), whose main market was the USA. The American aeronautic market deregulation in 1978 was an important factor influencing the success in the export-oriented phase (Coutinho and Ferraz, 1993). Therefore the second phase corresponds to the period when the market changed from national to foreign. A 30-seat advanced twin-engine turbo propeller aeroplane was the main commercialized product during this phase. According to Frischtak (1992, p. 13) “At the end of 1990, Brasilia’s market share in the 20–45 seat category was 25 per cent worldwide, just slightly below that of its major competitor (the SAAB SF340). In the U. S. market, the Brasilia had the dominant position in that year in terms of the total number of aircraft in service, again for the 20–45 seat category.” The launching of the eight-seat twin-engine turbo propeller pressurized aeroplane in 1979 is the starting point of this phase (Frischtak, 1994; Bernardes, 2000a). This aeroplane was the first one entirely developed by
The aeronautic sector in Brazil
169
Embraer, for which the market was not mainly the Brazilian Air Force but American large corporations. It was a business aeroplane called EMB 121 Xingú designed to appeal to the American market. Embraer undertook the whole process from product development, finance, and design and manufacture, including the pressurized system (one of the main innovations in this model) (Frischtak, 1994). It was the first aeroplane developed using the concept of communality or ‘family’. The second in the family was the 30-seat aeroplane called Brasília EMB 120, which was launched in 1981 to supply the US and Latin American market (Frischtak, 1992). The programme for ‘nationalization’ of aero parts stopped almost completely in this phase. The phase of ‘denationalization’ had begun, together with Embraer focusing on technological development at the international frontier in its market segment: small regional aeroplanes. It affected negatively the development of Brazilian firms supplying Embraer, which did not have the strength to compete with large international corporations and/or with companies receiving high subsidies from local governments, such as those located in the USA or Europe. The new market demanded many improvements in digital technologies, new materials and sophisticated software, among other technological developments, that the Brazilian SMEs could not supply. Brazilian suppliers had lacked government incentives and economies of scale, investments in R&D and financial health for this technological upgrading, and the Ministry of Defence reduced the budget to CTA and IFI, which decreased significantly IFI support to Brazilian SMEs. The recession in the international civil aeronautic market and the Brazilian government decreasing procurement and subsidies had been the main factors affecting the financial crisis of Embraer at the beginning of the 1990s (Bernardes, 2000a, 2000c). The company was thus privatized in 1994. Many Brazilian SMEs exited from the market owing to the economic recession in the period 1990–94. The most important products in the post-privatization period have been the ERJ 145, the ERJ 190 and business jets. The most important in terms of market success up to 2007 is the ERJ 145, which has the basic platform of the EMB 120 and incorporates new technologies in avionics, propulsion and aerodynamics. It was launched in 1995 and had sold a thousand up to 2007. Summarizing, the Brazilian government has supported civil aircraft manufacturing in all stages of its development, basically through (Green, 1987; Bernardes, 2000a): (1) funding R&D; (2) joint government–private ownership, corporate funding at main markets loan rates; (3) protection of the home market; and (4) export development policies. However, at least one important question can be asked: what is the extent to which
170
Sectoral systems of innovation and production
the Brazilian SMEs in the aeronautic sector have built up technological capabilities to continue supplying Embraer? Before examining this question, we describe some relevant aspects of the strategies implemented by Embraer that had affected the whole aeronautic sector in Brazil. 3.2.1. Strategies We examine the strategies of Embraer, focusing on the development of its aeroplane models, which are described in Table 6.1. Some characteristics of the models can highlight the understanding about the learning strategy of Embraer. Embraer developed the models by a mix of strategies that combined internal and external learning efforts to adopt, improve and develop an array of technologies and management tools that were mostly utilized by large companies manufacturing aeroplanes and systems for the sector. The strategies influenced a reduction in the time to market of aeroplanes from the 1970s on, particularly a reduction in the development of a new model from ten years to two to three years, at the same time as an increase in the technological complexity in the aeroplane, for example from a turbo propeller seating 30 passengers to a turbo jet seating 100 passengers. Technological and management expertise was developed by a mix of internal efforts and investments, the Brazilian government’s support from its technological centres and universities, and the defence Ministry’s offset policies involving technology transfer and co-development partnership with foreign firms, as explained before. The Brazilian government has invested in knowledge development in the field of aeronautics since the 1940s until nowadays. It means that there were 30 years of government investments in education and R&D until academics were able to manufacture the first aeroplane in 1969 when Embraer was founded to take it over from the CTA. The Brazilian government had many ups and downs regarding investments owing to macroeconomic and financial circumstances. Frischtak (1992, 1994) describes the most important aeroplane models developed by Embraer and the correspondent learning curves. Before 1992, the most important aeroplane model in terms of its impact on Embraer’s technological learning curve was the AMX.3 This model was developed in a co-development international partnership between the governments of Brazil and Italy. After that, the model CBA 123, which was an international partnership with Argentina and Chile, was strongly based on Embraer’s own internal learning efforts and investments. The CBA 123 was a commercial failure and was never sold owing to its high market price. In fact, the learning experience for the development of the CBA 123, the AMX, and the Brasília and Bandeirante had influenced the development of the ERJ 145.
The aeronautic sector in Brazil
171
Since the financial problems that Embraer faced before 1994, the financial situation and market behaviour became the central focus for the company’s executives. This means that in the previous period the main focus was capacity building to develop, manufacture, deliver and give customer support all over the world. For doing that, the Brazilian government was the main financial source. As the government was not investing any more as before, the company needed to find its own way to survive and improve competitiveness in the aeronautic sector. Thus the strategy followed was to implement the concept of strategic partnership as risk sharing and co-development of new models with international suppliers, such as GEAE, Honeywell and Parker Hannifin, among others. This strategy was first implemented for the ERJ 145, for which the partners were Gamesa (Spain), Sonaca (Belgium), Latecoere (France) and ENAER (Chile). At the same time, increasing the complexity of technologies, production and management, as well as the international market and logistics, had a correspondent increment in the import rate of the final aeroplane. A complex worldwide network, reducing time between the first idea and delivery of the first aeroplane, and shrinking costs without problems with certification, safety and quality were major challenges for Embraer. The company sought solutions and achieved an international standard that is considered by its competitors as one of the best management capacities in the sector. Managing logistics and suppliers, and integrating different software tools for managing development, production and customer support were two important things in achieving low costs, quality and increased payload capacity. All this development had two important impacts at least: creation of barriers for new entrants at the first level of the supply chain, and decreasing the number of direct suppliers. In fact, Embraer implemented a risk-sharing/co-development type of contract to complement its own knowledge base, shorten the time to develop and produce aeroplanes, and increase its financial capacity. A risk-sharing partnership is a convention for development in which the supplier funds a percentage of the new aeroplane model correspondent to the system or component it will supply and has a share in the total investment. It means that, for each aeroplane sold, Embraer shares the profits with its risk-sharing partners according to the partner share in the total aeroplane budget. As explained above, partners are suppliers of high-cost systems or components such as flight control systems, propulsion systems, energy systems, or fuselage parts, among others. They cooperate in the project definition and development, manufacture of prototypes, tests and certification. Therefore there is intensive interaction between Embraer’s R&D team and each partner’s R&D team. This type of contract with suppliers created a new dynamic in
172
Sectoral systems of innovation and production
which Brazilian SME suppliers have not managed to participate yet, with the exception of an Embraer subsidiary. 3.3.
General Characteristics of the Brazilian Civil Aircraft Market
According to Dagnino (1993), Donângelo et al. (2000) and Bernardes (2000a) the Brazilian international civil aircraft market is mainly the US and Europe, although there are investments for increasing the participation of China and Asia. During the 1970s and 1980s the main market however was the Brazilian, whose imports were restricted by the “Law of Similars” (Chapter III, Section V of Decree Law No. 37, as implemented by Decree No. 61 574 of 20 October 1967), which was part of the import substitution industrialization policies (Green, 1987). Then Embraer was granted the monopoly for production and commercialization of the aircraft turbo-prop with more than eight seats. Piper was the only foreign competitor of Embraer that was selling in the Brazilian market, owing to a licence agreement signed between them before the Law of Similars was implemented. During this period, the main competitors of Embraer in the US and Europe were De Havilland, Cessna, Fairchild, Piper, Saab, BAe, Dornier, Fokker and Canadair. By the end of the 1980s and during the 1990s, some civil aircraft manufacturers exited from the civil aircraft market, such as BAe, Cessna, Saab and Fokker, among others, while others merged or were acquired, such as Fairchild–Dornier, and Bombardier–Canadair. Fairchild–Dornier filed for bankruptcy in 2002. Since the 1990s, the main international competitor of Embraer has been Bombardier–Canadair, which is called Bombardier Aerospace. Bombardier is the third largest world producer of regional jets, while Embraer is the fourth largest, with a 45 per cent share of the regional jet market in 2000. The fierce competition between the two companies led them to complain at the World Trade Organization about unfair subsidies given by the Brazilian and Canadian governments. Bombardier complains that Embraer’s jets are less technologically advanced than their jets and is doing well in the market owing to the lower labour costs, the cheap Brazilian currency and the Brazilian government subsidies. Embraer complains that Bombardier’s jets are subsidized by the Canadian government low loan rates. The WTO complaints started in approximately 1998 and were still going on in 2002 (Padgett, 2003). The main civil aircraft models manufactured by Embraer are shown in Table 6.1. Embraer’s market segment ranges from small turbo-props seating 8 to 30 passengers, which are the models developed during the 1970s and 1980s, to medium-sized jets seating 35 to 108 passengers. They fly specifically short hauls or regional routes, mainly linking hub routes
The aeronautic sector in Brazil
Table 6.1 Year (first plane flew) 1972 1979 1983 1995 1995 1998 2002 2004 2005
173
The evolution of aircraft models produced in Brazil Model EMB 110 Bandeirante EMB 121 Xingú EMB 120 Brasília ERJ 145 ERJ 140 ERJ 135 ERJ 170 ERJ 190 ERJ 195
Seats Altitude (feet)
Speed Characteristics (km/h)
19
22 500
413
8
26 000
450
30
30 000
50 44 37 70 98 108
37 000 37 000 37 000 37 000 37 000 37 000
(As*Vc)
555
Light twin turboprop Twin turbo-prop pressurized Turbo propeller
7 847
16 650
833 833 833 870 870 870
Twin turbofan (jet) Twin turbofan (jet) Twin turbofan (jet) Jet Jet Jet
41 650 36 652 30 821 60 900 85 260 93 960
3 600
Source: Websites – http://www.embraer.com and http://www.airliners.net/. Information gathered in January 2001.
and small airports. As*Vc is the performance indicator that represents the number of seats (As) multiplied by the speed (Vc). Mowery and Rosenberg (1981) consider it an important indicator of performance development. Table 6.1 shows that there has been a substantial increase in aircraft performance since the launch of the jetliner ERJ 145. By the end of the 1980s, the Brazilian government had started a process of opening up the national market and had changed the macro-policies from import substitution industrialization to liberalization. It reduced substantially the subsidies given to Embraer, and there was no longer an import restriction for small regional aircraft as there had been in the previous period. Embraer, which was controlled by the Ministry of Aeronautics, was sold to private companies in 1994. Nowadays, the main market of the Brazilian aeronautic sector is foreign, which accounted for approximately 72.5 per cent of the turnover in 2000. AIAB (2000) observed an increase in the total exports from US$0.70 billion in 1997 to US$2.50 billion in 2000. As a consequence, there was a rise in the participation of this sector in the Brazilian GDP, measured by the total turnover divided by the total GDP, which jumped from 0.29 per cent in 1997 to 1.06 per cent in 2000. Embraer corresponded to about 80 per cent of the total Brazilian aeronautic sector in 2007. The economic performance of Embraer has been positive since 1997: the turnover increased from US$0.29 billion in 1996 to US$4.6 billion in
174
Sectoral systems of innovation and production
2005. In fact, exports accounted for most of the turnover, which grew from US$0.13 billion in 1996 to US$3.2 billion in 2005. Following that growth, employment has increased from 6737 people in 1998 to 16 953 people in 2005. In 2005 Embraer accounted for approximately 2 per cent of the total Brazilian exports.4 Embraer implemented many changes during the period 1996–2005. As explained before, some important changes were concerned with management of project development and relationships with suppliers, as well as procurement. At the same time, there was a reduction in the production cycle or from starting production to the phase-out from eight months in 1996 to five months in 2000 (Damiani, 2001). In this context, local suppliers, which have been mostly subcontracted for supplying pieces and parts, assembly jigs and tools and engineering projects, may comply with a just-in-time delivery time and high quality standards. To do so, they may comply with Embraer’s technical, quality and financial requirements. The procurement unit audits the supplier once or more a year and monitors compliance with these requirements. Local suppliers then implement and correct the suggested items in the auditor’s report. In 2005 most of Embraer’s suppliers were foreign: imports accounted for US$1.73 billion.
4.
RESEARCH METHODS
Firms were selected using purposeful sampling. The logic and power of purposeful sampling, as opposed to probability sampling, are to select information-rich cases from which it is possible to learn about issues of central importance to the purpose of the research (Patton, 1990; Yin, 1994). The main issue that this chapter is concerned with is accumulation of technological capabilities experienced by Brazilian SMEs in the Brazilian aeronautic sector. SMEs are firms with fewer than 500 employees as defined by the Brazilian Institute for Geography and Statistics (IBGE). The selection of the local SME suppliers was done in two stages: the pilot phase and the main fieldwork. In the pilot phase, a catalogue of firms from the Institute for Development and Coordination of the Aerospace Industry (IFI),5 called CESAER 2001, were surveyed to produce/collect information about the composition of the Brazilian local suppliers in the aeronautic sector. Thirty-one local Brazilian firms out of 98 were primarily selected. After that, the results were compared with the sample survey of a Brazilian research on the aeronautic system of innovation, coordinated by Roberto Bernardes, Jose E. Cassiolato and Helena Lastres, and financed by the Brazilian FINEP,6 and checked with Embraer’s procurement unit and with Roberto Bernardes. Then 12 SME suppliers of Embraer were
The aeronautic sector in Brazil
175
selected for in-house interviews and visits to their plants. Embraer and five subsidiaries of foreign first-tier suppliers of Embraer were also interviewed and visited. Based on the research interviews and double-checked with Embraer, the production chain structure was built as shown in Figure 6.2. Embraer is the main buyer and aircraft system integrator, i.e. a firm that designs the product and integrates its components, which outsources manufacturing and design activities seeking to make its organizational structures leaner. Outsourcing these activities requires close inter-firm interaction and coordination at the knowledge and organizational level (Brusoni and Prencipe, 1999). Few firms compose the first tier in this supply chain, which are foreign firms, with the exception of Embraer, and produce airframe structures. These first-tier firms jointly develop the aircraft models with Embraer and are also system integrators. The only Brazilian supplier in the second tier is Eleb, a producer of landing gear sub-systems, which has participated in joint development projects with Embraer and foreign suppliers of Embraer. Local SME firms supplying in the chain are mostly in the third and fourth tiers, among the ones that are mainly supplying to all tier firms and are shown in the middle boxes in Figure 6.2. They are all owned by Brazilian firms or individuals, with the exception of Eleb, whose shareholders are Embraer (60 per cent) and Liebherr Aerospace, Germany (40 per cent). The five subsidiaries of foreign suppliers located in Brazil are in the first tier, supplying airframe structures and related systems, and propulsion systems. Their activities in Brazil are mostly customer support and assemblage of airframe structural sections to Embraer. They were selected because their plants in Brazil resulted from the second Brazilian government efforts for “nationalization” of the supply chain, for which Embraer is responsible for undertaking efforts to attract foreign suppliers to locate in Brazil. The firms that settled plants in Brazil are first-tier suppliers that jointly develop projects with Embraer. Their contact details were given by Embraer’s technological development unit and double-checked with the procurement unit. 4.1.
Brief Description of the Sample of Suppliers
4.1.1. Local SME supplier firms This section aims at characterizing the local SME suppliers interviewed in terms of their ownership, period of foundation, number of employees, market characteristics and product characteristics. Brazilian firms or individuals own all SME suppliers, although there is one firm that is a joint venture between Embraer and the German Liebherr Aerospace.
176
The supply chain structure of Embraer (2002)
Own elaboration based on the research.
Figure 6.2
Source:
Propulsion parts & pieces
Avionics parts & pieces
Avionics components
LS – Local SME Suppliers FS – Foreign Suppliers BR – Brazil Fuselage, energy system, landing gear, hydraulic system, interiors, environmental system, electric system Navigation, communication, surveillance, flight command
Airframe parts & pieces (LS)
Propulsion components
Turbine Avionics & related sub-systems sub-systems
Avionics & propulsion stems FS
3rd tier
Sub-systems parts & pieces (FS)
Airframe structural components (LS)
Airframe structural section (wings & fuselage sections) Embraer
Machinery & equipment, assembly jigs & tools, materials (LS, FS) Engineering projects, software, technical assistance (LS, FS)
EMBRAER
2nd tier
Sub-systems components FS
Eleb BR FS in BR& abroa
Subsystems
Airframe structure & related systems Embraer FS in BR & abroad
Aircraft manufacturer (main contractor)
1st tier 4th tier
The aeronautic sector in Brazil
177
The majority of local SME firms were supplying the Brazilian market, particularly Embraer and the Brazilian Air Force, which was approximately 80 per cent of their market. There were two firms that entered the export market which were only supplying Embraer until the end of the 1990s. These suppliers were exporting landing gear, motor gliders and spare parts to Embraer aircraft models EMB 120 Brasília, and engineering consultancy in design, production and assemblage of airframe structural sections, particularly fuselage sections. The majority of local SME suppliers were manufacturing fuselage parts and components (third and fourth tiers) using steel, aeronautic aluminium and composite materials, while two firms are producing parts for avionics in aluminium and composite material. The other parts and components supplied vary from nails and pieces to hydraulic components. One supplier firm produced landing gear utilizing new composite materials and electronic components. Assembly jigs and tools are utilized as support to the assembling of fuselage sections as well as to the integration of fuselages to the airframe. Jigs and tools are also made to order for supporting prototype tests. They were made in steel and had electronic components to adjust the jig or tool to the size of the fuselage section and airframe. Engineering projects and consultancy were based on mechanic and aeronautic engineering knowledge. Specifically, local SME suppliers were designing airframe structural sections, parts and components, as well as defining their production specifications. Particularly, engineering firms sold consultancy services to European suppliers of Embraer. The consultancy was related to design definition, production processes and integration of fuselage sections to the airframe. The software firm supplied enterprise resource planning (ERP), that is software utilized in the management of integration among all units of the firm, to a subsidiary of a foreign supplier of propulsion systems to Embraer. The development of the ERP was made in close integration with the customer, as a team of technicians were working inside the customer’s plant. 4.1.2. Characteristics of foreign suppliers located in Brazil The main characteristics of the five subsidiaries of foreign suppliers interviewed are shown in Table 6.2. The large majority were settled in Brazil in the period 1999–2001. They assemble systems and airframe structures. They have also participated in the co-design arrangement with Embraer for developing the projects ERJ 145 and ERJ 170. Technical support to system integration in the aeroplane at Embraer’s plant was another activity undertaken by them. The subsidiary QF has not focused only on
178
* Number of employees in November 2002.
Own elaboration based on interviews.
Source:
2000
2000
1999
Assembly of interior compartments Overhaul, accessories and component repair of power plant – engines (assembly and production of parts for engines before 1992 for Embraer’s military aircraft AMX) Customer support for repairing and assembling hydraulic system, flight controls Assembly and repair passenger window transparencies Assembly structural parts and customer support
Activity in Brazil
Note:
RF (first-tier supplier – hydraulic system; second-tier supplier – flight control) SF (second-tier supplier – airframe structural sections) TF (second-tier supplier – airframe structural sections)
2001
PF (first-tier supplier – airframe structure) QF (first-tier supplier – propulsion system)
1951 (acquired by General Electric Aircraft Engines’ GEAE/USA, in 1992)
Foundation
ERJ 145 ERJ 170
ERJ 145 ERJ 170
ERJ 170
ERJ 145 ERJ 170 AMX ERJ 170
Embraer aircraft model
100% Belgium
99% UK 1% USA
100% USA
99% GEAE USA 1% Brazilian Ministry of Defence
100% USA
Ownership
Characteristics of the subsidiaries of foreign suppliers of Embraer located in Brazil, 2002
Firm
Table 6.2
99
9
12
350
106
Employees (*)
Embraer
Embraer
Embraer
Brazil (including Embraer) Export
Embraer
Market
The aeronautic sector in Brazil
179
supplying Embraer. It has overhaul and maintenance of turbines as its main market focus since it acquired the Brazilian manufacturer of engines Celma. The Brazilian plant is one of the few of GEAE that includes the development of processes for turbine maintenance. The other interesting case is TF, which produces aeroplane structural sections to Airbus aeroplanes in Brazil. This firm transferred part of its European manufacturing activities to Brazil. Although there are non-R&D or design activities locally, it is working for the implementation of minor improvement activities in accordance with the quality and other exigencies of Embraer. The other three cases only assemble parts and components for the fuselage and interiors, as well as hydraulic systems and flight controls, which are manufactured in their home countries. In fact, all the subsidiaries had few Brazilian suppliers, which were mostly related to engineering services for fuselage structural assemblage and compliance with Embraer exigencies.
5.
REFLECTIONS ABOUT THE ACCUMULATION OF TECHNOLOGICAL CAPABILITY EXPERIENCED BY BRAZILIAN SME SUPPLIERS
This section comments briefly on the results of our fieldwork. Table 6.3 classifies the SMEs as shown in Appendix 6.2. The results show that SME suppliers have mostly maintained themselves in the supply chain by using their already existent basic routine technological capability to implement technological changes. Fewer SMEs have upgraded to the preintermediate routine technological capability, and even fewer have moved to innovative intermediate. In fact, only one SME developed advanced innovative technological capability during the period 1970–2002. Furthermore, all firms mastered basic levels of technological capabilities in all domains: product-centred, production (process and equipment-related), and organizational process (project management and design procedures). The majority of firms (Group 4) implemented technological changes utilizing their existent basic routine technological capability, which associates with passive learning efforts. This means that they are able to manage the replication of specifications from customers, basic quality control, routine replacement of components in machinery, equipment and software, basic coordination of project development and basic routine design procedures. From the sample of 12 firms, one SME maintained its pre-intermediate routine technological capability, and three have accumulated technological capability in the period 1995–2002.
180
Source:
–
–
–
–
–
– –
–
– –
1981– 94
–
–
–
1 –
1995– 2002
Own elaboration based on interviews.
Product Production Process Equipmentrelated Organizational processes Project management Design procedures
1970– 80
Advanced
–
–
–
– –
1970– 80
–
–
–
2 –
1981– 94
Intermediate
1
1
1
1 1
1995– 2002
–
–
–
1 –
–
1
1
– 2
1981– 94
–
–
1
1 3
1995– 2002
Pre-intermediate 1970– 80
Level/period
Technological capability: number of firms by domain and level
Technological capability
Table 6.3
3
3
3
2 3
1970– 80
9
8
8
7 7
1981– 94
Basic
11
11
10
9 8
1995– 2002
The aeronautic sector in Brazil
181
The three local SMEs that built up technological capability are classified under the headings of: ● ● ●
Group 1, which includes firms that built up advanced innovative capability; Group 2, which includes firms that built up intermediate innovative capability; Group 3, which includes firms that built up pre-intermediate routine capability.
SMEs classified in these three groups had pro-actively invested in learning in order to manage and generate technological changes in different domains. These firms are compared below, considering the domains of technological capabilities accumulation in the period 1995–2002. The firm considered in Group 1 is Firm 5, which built up advanced technological capability in product-centred innovation. From the mid-1980s up to 1994, this firm acquired intermediate technological capability in products for developing landing gear for Embraer aircraft models. It had also managed tests for improving the production process, moving from basic to preintermediate routine technological capability. During this period, efforts were made to improve the internal coordination of projects through managing team working, allowing the firm to master the pre-intermediate level of technological capability in project management. From 1995 up to 2002, Firm 5 had a pro-active learning behaviour by investing in in-house R&D and partnership with customers and universities/research centres for product development, among other learning activities. The firm had also improved technological capabilities from: ● ● ● ●
pre-intermediate to intermediate in the production process; pre-intermediate to intermediate in the organization of project management; basic to intermediate in production equipment-related; and basic to intermediate in the organization of design procedures.
The local SME considered under the heading Group 2, for example Firm 1, built up intermediate innovative technological capability in product and pre-intermediate routine technological capability in the production process. In reality, Firm 1 developed the two-seat aeroplane AMT 600 by utilizing its intermediate innovative technological capability acquired in the period 1981–94 when it received technology transfer while working on a government project for replacing training aeroplanes. The other SME that built up pre-intermediate technological capability
182
Sectoral systems of innovation and production
Table 6.4
Number of firms that implemented technological changes by domain and level
Technological changes
Level/period High
Middle
Low
1970– 1981– 1995– 1970– 1981– 1995– 1970– 1981– 1995– 80 94 2002 80 94 2002 80 94 2002 Product Production Process Equipmentrelated Organizational processes Project management Design procedures Source:
1 –
3 2
7 9
1 –
3 2
6 7
– –
1 1
1 3
–
2
8
–
–
5
–
–
2
–
1
1
–
–
–
–
–
–
–
–
4
–
–
–
–
–
–
Own elaboration based on interviews.
in the production process was Firm 4, classified in Group 3. This firm undertook efforts for managing team working, training of employees and in-house tests for the production process and product development from its foundation in 1995. It had also managed tests for assuring quality control in production and implemented an ERP computer system. Considering the impact of technological changes implemented by the SME suppliers (Table 6.4), all firms have implemented high-level changes. The technological changes were implemented mostly during the period 1981–94 and 1995–2002. The latter was the period in which SMEs implemented high-level changes in all domains, although most firms implemented changes in product and production processes and equipment. It is important to say that two firms did not mention the implementation of any technological change in the period 1970–80, while one firm implemented change at the high and middle levels in products in this period. The research findings so far show that there are differences in the level and domains of technological changes implemented by the SMEs compared to the level of capability accumulation. Particularly, local SMEs that maintained their basic routine technological capability implemented mostly changes at the high, middle and low levels in production processes in the period 1995–2002. These changes related to the implementation of
The aeronautic sector in Brazil
183
ISO 9000 standards and quality assurance procedures to comply either with Embraer requests or with a government programme for promoting exports. They have generated and managed the implementation of changes by contracting a consultancy firm to help them. They generated and implemented the changes working together with the consultancy firm, utilizing their already existent knowledge about the ISO 9000 procedures. The local SMEs that moved to more innovative levels of technological capability implemented mostly high-level technological changes in products, production and organizational procedures. Particularly, those SMEs that achieved advanced technological capability in products implemented high-level technological changes in project management and design procedures, which was not observed in the other firms. Moreover, the research findings suggest that differences among the four groups of SMEs refer to their market strategy. The group of SMEs that maintained their basic routine technological capability supplied mostly Embraer, following Embraer requests and blueprints, whereas the other groups of SMEs also supplied other customers, such as foreign firms and/or subsidiaries of foreign suppliers of Embraer located in Brazil. In particular, the SME that built up advanced technological capability in products was largely affected by the development of Embraer itself. The accumulation of technological capability experienced by this SME was influenced both by its internal efforts and by its relationship with other firms and universities and research centres, as well as its participation in government programmes. According to the interviews, two firms did not implement any significant technological change in the period 1970–80, while the one firm that did implement change in the period did so in products. In the following period (1981–94), the numbers of firms that did not implement changes were as follows: ● ● ● ●
five firms – products and production processes; seven firms – production equipment-related; eight firms – organization of project management; and nine firms – organization of design procedures.
Regarding the period 1995–2002, all firms implemented at least technological changes in products and production equipment. The numbers of firms that did not implement changes were: ● ●
one firm – products; two firms – production process;
184 ● ●
Sectoral systems of innovation and production
11 firms – project management; and eight firms – design procedures.
Table 6.5 summarizes the extent to which the 12 sampled SMEs accumulated technological capabilities, associating them with the level of technological changes implemented. The shadow boxes show that the majority of the firms implemented changes to the high, middle and low levels utilizing the already existent basic routine technological capability. The four firms that moved to other levels of technological capabilities implemented technological changes at the high and middle levels in products and production processes and equipment. Just one firm improved to intermediate technological capability in the organization of project management and design procedures in the period 1995–2002. Figure 6.3 compares the direction and extent to which the SME suppliers have accumulated technological capability. The direction may be to managing and generating high-level impact technological change by maintaining the existent technological capability at one side, or upgrading the technological capability while managing the implementation of higher levels of technological change. The findings show four groups of SMEs: ●
●
●
●
Group 1: one firm that achieved advanced technological capability at least in one area, i.e. product-centred, production (process, equipment-related), or organizational processes (project management, design procedures), and implemented high-level technological change in at least one of these areas; Group 2: one firm that built up intermediate technological capability in at least one of those areas explained in Group 1, and implemented high-level technological change in at least one of those areas; Group 3: two firms that built up pre-intermediate technological capability in at least one of those areas explained in Group 1, and implemented middle-level technological changes at least in one of those areas; Group 4: eight firms that implemented at least low-level technological changes in one of those areas explained in Group 1, and maintained basic technological capability.
Table 6.6 shows the characteristics of linkages for each group of SME suppliers taking into consideration the sources of external knowledge, and linkage types, activities and impact. In fact, linkages experienced by firms in Group 3 are similar to those of firms in Group 4, whereas the differences arise in the activities and impact of the linkages. First, firms in Group 3 hire experts, undertaking training
185
–
–
–
–
Low Production process High
–
–
–
–
1981– 94
Middle
High
Product
1970– 80
–
–
–
Firm 5
1995– 2002
Advanced
–
–
–
–
1970– 80
–
–
Firm 5
Firm 5
1981– 94
Firm 5
–
–
Firm 1
1995– 2002
Intermediate
–
–
–
Firm 1
1970– 80
Firm 5 Firm 8
–
–
–
1981– 94
–
–
Firm 1 Firm 4 Firm 8
Firm 1
–
1970– 80
–
Firm 4
1995– 2002
Pre-intermediate
Technological capability level/period
–
Firm 1
Firm 1 Firm 8
Firm 10
1981– 94
Basic
Firm 2 Firm 6 Firm 7 Firm 10 Firm 11 Firm 3 Firm 8 Firm 9 Firm 10 Firm 9 Firm 2 Firm 3 Firm 7 Firm 11 Firm 12 Firm 3 Firm 7
1995– 2002
Classification of firms according to technological change and technological capability: level and period
Technological change (domain/level)
Table 6.5
186
–
–
–
Middle
Low
Production Equipmentrelated High
1970– 80
(continued)
Technological change (domain/level)
Table 6.5
–
–
–
1981– 94
–
–
–
1995– 2002
Advanced
–
–
–
1970– 80
–
–
–
1981– 94
–
–
Firm 5
1995– 2002
Intermediate
–
–
–
1970– 80
Firm 8
–
Firm 5
1981– 94
–
–
1995– 2002
Pre-intermediate
Technological capability level/period
–
–
–
1970– 80
Firm 10
Firm 6
1981– 94
Basic
Firm 1 Firm 2 Firm 3 Firm 4 Firm 8 Firm 9 Firm 11 Firm 12
Firm 8 Firm 9 Firm 11 Firm 4 Firm 7 Firm 9
1995– 2002
187
– –
Middle Low
Source:
–
–
– –
–
– – –
–
–
Own elaboration based on interviews.
– –
– – –
–
– – –
–
Low
–
Organizational processes Project management High Middle Low Design procedures High
–
Middle
– –
–
– – –
–
–
– –
–
– – –
–
–
– –
Firm 5
Firm 5 – –
–
Firm 5
– –
–
– – –
–
–
– –
–
Firm 5 – –
–
–
– –
–
– – –
–
– –
–
– – –
–
–
– –
–
– – –
–
–
Firm 7 Firm 9 Firm 12 – –
– – –
Firm 6 Firm 7 Firm 10 Firm 2 Firm 9
188
Sectoral systems of innovation and production
Technological changes
High
Group 4 (fourth tier)
Group 3 (fourth tier and supplier to all tiers)
Group 2 (third tier)
Group 1 (second tier)
Middle
Low
Basic Routine Capabilities
Pre-intermediate Intermediate Routine Capabilities Innovative Capabilities
Advanced Innovative Capabilities
Technological Capabilities Accumulation
Figure 6.3
Groups of SME suppliers in the Brazilian aeronautic sector
outside the firm. Furthermore, their customers and suppliers helped in the implementation and management of technological changes, while specific work in the customer plant was undertaken as well as their technical support at Group 3 plants. Second, SMEs in Group 4 had trained employees in-house on the shop floor since their foundation up to 2002. In some cases, Group 4 SMEs have employees working inside the Embraer plant. Embraer trained these employees in areas related to the products and production quality assurance, among other topics such as safety and confidentiality. In this sense, Embraer was transferring technological information to local SME suppliers to the extent of their involvement working inside the Embraer plant. Another research finding worth mentioning is that local SMEs classified in Group 4 relied more heavily on the technical assistance given by Embraer and CTA for the generation of technological changes than the other groups. Differences between them and Group 1 are that this group was actively learning from Embraer and foreign buyers by participating in joint product development activities, which contributed to the implementation of technological changes mostly in products, project management and design procedures, and to the building up of advanced innovative technological capability. Other important linkages are with Brazilian and
The aeronautic sector in Brazil
Table 6.6
189
Comparison between the groups of SMEs and linkages Group 1
Group 2
Universities/ research centres; Brazilian Air and Navy forces; suppliers; Collaborative Types of Collaborative agreement, linkages agreement, technology technology transfer and transfer and procurement procurement Activities/ Joint product Contracting linkages development university with customers; tests and contracting experiments; research informal tests and tests at at foreign universities and customers’ research centres, plant hiring experts, training outside Blueprints/ Impact of Contributing linkages to investments specifications in research and from the development of engine supplier and tests at product; and implementation universities of technological contributing to the changes implementation in project of technological management changes in and design products procedures; proposals from suppliers contributing to the implementation of changes in production
Sources Universities/ of external research knowledge centres; Embraer and foreign customers
Group 3
Group 4
Brazilian and foreign customers, suppliers, internet
Embraer, consultancy, supplier
Procurement
Procurement
Hiring experts; eventual training outside; informal contacts with university and research centre
Hiring experts, informal contacts with university and research centre
Technical support from customer contributing to the implementation of tests and prototype activities that result in changes in product and process
Blueprints and specifications from customers and suppliers contributing to the implementation of all technological changes
190
Sectoral systems of innovation and production
foreign universities and technological centres for researching and training employees. As in the case of Group 2, it relied more on the acquisition of blueprints, specifications and technical assistance from foreign buyers, engine suppliers and universities than from Embraer. Moreover, it has received technology transfer from foreign firms when participating in the Brazilian military government offset programmes.7 Particularly related to inter-firm linkages, although it is difficult to separate the influence of linkages with Embraer and other firms, the research findings suggested that linkages between Group 1 and Embraer contributed largely to the implementation of technological changes in products, production, and organizational processes. Moreover, its linkages to foreign suppliers of Embraer and foreign buyers contributed to the accumulation of innovative technological capabilities for generating changes in products and organizational processes. It is worth mentioning that Embraer itself particularly influenced collaborative agreements between Firm 5 and foreign firms. Technology transfer from foreign buyers to Firm 5 also had an important impact on the innovative capability accumulation, and it was related to some of the high-level technological changes implemented by this firm. The technology transfer was negotiated by the Brazilian military government at the end of the 1970s up to the mid-1980s in offset programmes. The distinction between Group 1 and Group 2 in terms of their participation in government offset programmes lies in the extent to which the government funded and negotiated the technology transfer. In the case of Group 2, the government negotiated the technology transfer of seat production, which was a field in which the firm already had technological capability, whereas in the case of Firm 5 the government funded and negotiated technology transfer for design and production of parts and components, an area in which this firm lacked capability.
6.
FINAL COMMENTS
Understanding the technological capability accumulation and the role of linkages with national and foreign buyers and research institutions in the Brazilian aeronautic sector involves a difficult exercise for perceiving changes in the innovative environment and in the supply chain. Initially, the development of this sector was tied to the strengthening of CTA. In the period pre-privatization of Embraer, the participation of the national government was very important through its technological policies for developing research capacity in the aeronautic field. The creation of an R&D environment together with tax incentives and government procurement
The aeronautic sector in Brazil
191
made possible the accumulation of technological capabilities experienced by Embraer, Firm 5 in Group 1, and Firm 1 in Group 2. During this period, Embraer based its development of worldwide overhaul and commercialization on an agreement with the American Piper. The other pillar was the Brazilian Embassy worldwide in the commercial efforts of Embraer. It was a period of fierce competition owing to the “deregulation” of the American market: many American small aircraft producers were exiting from the market, merging with large corporations or seeking international alternatives. The organization of Embraer’s supply chain changed drastically after its privatization in 1994. The economic crises and the privatization led to new behaviour in the firm regarding project management and the supply chain organization. Embraer elaborated a complex management architecture for moving up with its projects of jetliners. The management architecture involved reducing the number of first-tier suppliers and increasing collaborative agreements (co-design and risk sharing), with fewer strategic foreign suppliers. The Brazilian SME suppliers have not been incorporated in collaborative agreements, with the exception of the firm in Group 1. The study concludes that the interaction between SMEs and research institutions for implementation of technological changes that were new to the firms drove these SMEs to invest in engineering and testing/searching activities, non-existent activities before the interaction. On the other hand, the activities undertaken with Embraer were more related to technological upgrading making use of the existing knowledge base rather than for accumulating technological capabilities, with the exception of Firm 5. The interaction between SMEs and foreign buyers was also possible owing to the existing knowledge base of the SME. These foreign buyers considered very important the tacit knowledge accumulated by the SMEs in the specific areas in which they were supplying Embraer. The experience as suppliers of Embraer was very important to the foreign buyers, which were supplying systems and sub-systems to Embraer. Therefore, the foreign buyers had little influence on the technological changes the SMEs implemented or on their accumulation of technological capabilities, with the exception of Firm 1 and Firm 5. Finally, it is important to emphasize that the few SME suppliers of Embraer that built up technological capability from basic production to more innovative levels (including learning by searching and by interacting with research institutions) constitute a special group of firms: they have, over time, been strategic to Embraer, engaged in military projects and very competitive in terms of international markets. In fact, the type of activities undertaken by the SMEs differed according to the novelty of the technological change to the SME. The completely new technologies called for interaction with research institutions (universities
192
Sectoral systems of innovation and production
or technological centres) and searching and testing activities in-house. Following that, product, production and organizational changes related to the improvement of existing technologies called for doing and adapting activities through close interaction with Embraer (national buyer) and the foreign buyer. Embraer had also played an important role in training employees of the SME suppliers. The results suggested that there is a strong relationship between the novelty of the technological change to the firm and the type of activities undertaken with research institutions, national buyer and foreign buyer. Compared with other sectors and countries, the Brazilian aeronautic sector has little or no distance from the global innovation frontier. A study by Dahlman and Frischtak in 2005 (in Knight and Marques, 2008) gives a good analysis of the challenges the knowledge economy poses for Brazil in Table 6.7
Brazil – selected economic sectors in relation to their distance from the global innovation frontier
Little or none
Moderate
Significant
Agriculture and tropical forest management
Vehicles: agricultural machinery; buses; trucks; compact cars Principal strength: engineering of new products/processes
Electronics; instruments; IT equipment
Principal agent: Embrapa (Brazilian Agricultural Research Enterprise) Energy: biomass (ethanol), deep water petroleum exploitation Principal agent: Petrobrás Aeronautics: regional jet aircraft Principal agent: Embraer
Electric motors Principal agent: WEG
Source:
Auto parts, white line (stoves, refrigerators, washing machines, etc.) Principal strength: engineering of new products/processes Software: financial; administrative; security Principal strength: cost/quality of software engineers Cosmetics Principal strength: biodiversity
Principal weakness: embedded chips and components Chemical products Principal weakness: fine chemicals
Pharmaceutical products
Principal weakness: research on new molecules Capital goods related to information technology Principal weakness: limited demand
Dahlman and Frischtak, 2005 (in Knight and Marques, 2008, p. 188, Table 8).
The aeronautic sector in Brazil
193
an increasingly competitive world. Table 6.7 is taken from that study and looks at some sectors and sub-sectors of the Brazilian economy in terms of their distance from the global innovation frontier. Brazil’s position in the global economy has some similarities to that of the other members of the BRIC group – Russia, India and China. All four countries have large populations and landmass, a diversified industrial base, and a critical mass of scientific and technological resources and institutions. All four are not members of either the European Union or the North American Free Trade Area, major economic blocs. Brazil’s agricultural, biofuel, aeronautics and electric motors potential and substantial mineral resources are major comparative advantages for trade. Strategic alliances with all three other BRIC countries could help build scientific and technological resources in the areas where Brazil is moderately and significantly distant from the global innovation frontier, and in which it is possible to complement efforts. Brazil’s science and technology policies are an attempt under way to develop a shared vision of a desired future in which Brazil will be a major player in the world economy and also make intensive use of technology to solve its centuries-old problems of deep socio-economic inequalities, and some case studies in closing the technology gap. Finally, we suggest that government policies for upgrading Brazilian SME suppliers to the aeronautic sector should take into consideration instruments for strengthening their technological capabilities. This is particularly relevant to the characteristics of the strategies of Embraer for managing the relation with suppliers. In fact, the strategies moved forward for managing suppliers increased the internationalization of procurement and participation in product development, as observed in section 3. Furthermore, foreign suppliers are sharing product development activities as well as the necessary investments and returns. This behaviour poses another question referring to the lack of financial capabilities of Brazilian SME suppliers for sharing risks and product development. Following that, we also suggest that policy instruments should focus on stimulating the upgrading in the SMEs to completely new and more advanced technologies while forcing them to strengthen linkages with research institutions and implement engineering, project management and design procedures. Concluding, the Brazilian aeronautic system of innovation generally follows the loose characteristic of the national system of innovation, as observed in section 2.1, although there are a few exceptions of successful firm specific cases of tight linkages with research institutions, national and foreign buyers, participation in government programmes and innovative capability accumulation. Government policies, thus, should also focus on
194
Sectoral systems of innovation and production
the diffusion of the innovation culture in the sector through the successful cases of SMEs that have accumulated innovative capability and not focus only on Embraer or on foreign buyers.
NOTES 1. 2. 3. 4.
Email:
[email protected]. Email:
[email protected]. The AMX is a surveillance and light attack aircraft serving military purposes. Information gathered from Bernardes (2000a) and Embraer reports to investors at www. embraer.com.br in February 2008. 5 IFI is an institute of the Technical Aerospace Centre (CTA). 6. FINEP – Financiadora de Estudos e Projetos (Finance of Studies and Projects) – is a foundation of the Ministry of Science and Technology (www.finep.org.br). 7. The offset programmes of the military government allow foreign firms supplying the government to transfer technology to Brazilian firms in accordance with the specific procurement agreement.
BIBLIOGRAPHY Associaçâo das Industrias Aerospaciais do Brasil (AIAB) (2000), ‘O setor Aerospacial: Corte em 2000’ (The Brazilian Aerospace Sector: Cutting in 2000), Mimeo, São José dos Campos: Association of the Brazilian Aerospace Industry. Albuquerque, E.M. (1996), “Sistema nacional de inovação no Brasil: uma análise introdutória a partir de dados disponíveis sobre a ciência e a tecnologia”, Revista de Economia Política 16(3), pp. 56–72. Altenburg, T. and Meyer-Stammer, J. (1999), ‘How to promote clusters: policy experiences from Latin America’, World Development, 27(9), pp. 1693–1713. Alvarez, R. (2004), ‘Sources of export success in small and medium-sized enterprises: the impact of public programs’, International Business Review 13, pp. 383–400. Amsden, A.H. and Hikino, T. (1993), ‘Borrowing technology or innovating: an exploration of the two paths to industrial development’, in Thomson, R. (ed.), Learning and Technological Change, St Martin’s Press, New York, pp. 243–266. Ariffin, N. (2000), ‘The internationalization of innovative capabilities: the Malaysian electronics industry’, D.Phil. thesis, University of Sussex/SPRU, Brighton. Ariffin, N. and Figueiredo, P.N. (2003a), ‘Internationalization of innovative capabilities: counter-evidence from electronic industry in Malaysia and Brazil’, Paper presented at the DRUIDS Conference, June, Copenhagen. Ariffin, N. and Figueiredo, P.N. (2003b), Internacionalização de Competências Tecnológicas: Implicações para Estratégias Governamentais e Empresariais de Inovação e Competitividade na Indústria Eletrônica no Brasil, Editora FGV, Rio de Janeiro. Arnold, E., Allinson, R., Muscio, A. and Sowden, P. (2004), Making the Best Use of Technological Knowledge: A Study of the Absorptive Capacity of Irish SMEs, Report, Technopolis, Brighton.
The aeronautic sector in Brazil
195
Arocena, R. and Sutz, J. (1999), ‘Looking at national systems of innovation from the South’, Mimeo, Universidad de la Republica, Montevideo. Bell, M. (1984), ‘Learning and the accumulation of industrial technological capacity in developing countries’, in Fransman, K. and King, M. (eds), Technological Capability in the Third World, Macmillan Press, London, pp. 187–209. Bell, M. and Albu, M. (1999), ‘Knowledge systems and technological dynamism in industrial clusters in developing countries’, World Development 27(9), pp. 1715–1734. Bell, M. and Pavitt, K. (1993), ‘Technological accumulation and industrial growth: contrasts between developed and developing countries’, Industrial and Corporate Change 2(2), pp. 157–210. Bell, M. and Pavitt, K. (1995), ‘The development of technological capabilities’, in Haque, I. (ed.), Trade, Technology and International Competitiveness, World Bank, Washington, DC, pp. 69–101. Benkard, C.L. (1999), Learning and Forgetting: The Dynamics of Aircraft Production, Working Paper Series, May, National Bureau of Economic Research, Cambridge. Bernardes, R. (2000a), O Arranjo Produtivo da Embraer na Regiao de Sao Jose dos Campos [The production organization of Embraer in the region of Sao Jose dos Campos], Report, Federal University of Rio de Janeiro – Institute of Economics, Rio de Janeiro. Bernardes, R. (2000b), Embraer: Elos entre o Estado e o Mercado [Embraer: links between the state and the market], Editora Hucitec, São Paulo. Bernardes, R. (2000c), Oportunidades de Mercado, Produção e Acesso ao Conhecimento: Linhas de Ação para o Fortalecimento da Performance Technológica do Setor Aeronáutico, Report, FINEP, Rio de Janeiro. Bernardes, R. and Oliveira, L.G. (2002), ‘O Arranjo Produtivo da Rede Embraer de Fornecedores”, Nota Técnica, BNDES, Rio de Janeiro, http://www.finep. org.br (accessed in November 2002). Braga, C.A.P. and Cabral, A.S. (1986), O Estado e o Desenvolvimento Tecnológico da Indústria Aeronáutica Brasileira [The state and the technological development of the Brazilian aeronautic industry], Discussion Paper 23/86, University of São Paulo/IPE, São Paulo. Breschi, S. and Malerba, F. (1997), ‘Sectoral innovation systems: technological regimes, Schumpeterian dynamics, and spatial boundaries’, in Edquist, C. and Mackelvey, M. (eds), System of Innovation: Growth, Competitiveness and Employment, Edward Elgar Publishing, Cheltenham, UK and Lyme, NH, USA, pp. 261–287. Brusoni, S. and Prencipe, A. (1999), ‘Unpacking the black box of modularity: technologies, products, and organisations’, Mimeo, University of Sussex /SPRU/ CoPS Innovation Centre, Brighton. Caniëls, M. and Romijn, H. (2001), Small-Industry Clusters, Accumulation of Technological Capabilities and Development: A Conceptual Framework, Eindhoven University of Technology, Eindhoven. Capdevielle, M., Cimoli, M. and Dutrénit, G. (1997), Specialization and Technology in Mexico: a Virtual Pattern of Development and Competitiveness?, Interim report, May, International Institute for Applied Systems Analysis, Laxenburg. Carlsson, B. (1995), ‘The technological system for factory automation: an international comparison’, in Carlsson, B. (ed.), Technological Systems and Economic
196
Sectoral systems of innovation and production
Performance: The Case of Factory Automation, Kluwer Academic Publishing, Dordrecht, pp. 441–475. Carlsson, B. and Jacobsson, S. (1994), ‘Technological systems and economic policy: the diffusion of factory automation in Sweden’, Research Policy 23, pp. 235–248. Cassiolato, J. and Lastres, H. (1999), ‘Local, national and regional systems of innovation in the Mercosur’, Paper presented at DRUID’s Summer Conference on National Innovation Systems, Industrial Dynamics and Innovation Policies. Cooke, P., Uranga, M. and Etxebarria, G. (1997), ‘Regional innovation systems: institutional and organisational dimensions’, Research Policy 26(4/5), pp. 475–491. Costa, I. and Queiroz, S.R.R. (2002), ‘Foreign direct investment and technological capabilities in Brazilian industry’, Research Policy 31(8–9), pp. 1431–1443. Coutinho, L. and Ferraz, J.C. (1993), ‘Estudo da Competitividade da Indústria Brasileria: Competitividade da Indústria Aeronáutica’, Campinas: Ministério da Ciência e Tecnologia (MCT), Financiadora de Estudos e Projetos (FINEP), Programa de Apoio ao Desenvolviments Cientifico e Tecnológico (PADCT). Dagnino, R. (1993), “Competitividade da indústria aeronáutica”, Nota Técnica do Complexo Metal-Mecânico, Estudo da Competitividade da Indústria Brasileira, http://www.mct.gov.br (accessed in January 2001). Dagnino, R. and Proença, D.J. (1989), The Brazilian Aeronautics Industry International Labour Office/International Labour Organization, Geneva. Dahlman, C. (1984), ‘Foreign technology and indigenous technological capability in Brazil’, in Fransman, M. and King, K. (eds), Technological Capability in the Third World, Macmillan, London. Dahlman, C. and Frischtak, C. (1990), National Systems Supporting Technical Advance in Industry: The Brazilian Experience, University of Sussex/SPRU, Brighton. Damiani, J.H.S. (2001), ‘A nova face da gerência de projetos: o projeto do avião ERJ 170 na Embraer’, Revista ESPM, March–April, ESPM, São Paulo. De Ferranti, D. and Perry, G.E. (2002), Closing the Gap in Education and Technology, World Bank Latin American and Caribbean Studies, http://web. worldbank.org/WBSITE/EXTERNAL/NEWS/0,,contentMDK:20070794~m enuPK:34457~pagePK:34370~piPK:34424~theSitePK:4607,00.html (accessed in October 2002). Douâgelo, A., Coelho, F. and Ichimura, J. (2000), Gerência do marketing mix Internacional de cima Empresa Basileira: o caso da Embraer (Management of International Marketing Mix in a Brazilian Company: the case study of Embraer), Cadernos Discentes COPPEAD 2: 39–55, Rio de Janeiro: COPPEAD/ UFRJ. Dutrénit, G.B. (1998), ‘From knowledge accumulation to strategic capabilities: knowledge management in a Mexican glass firm’, D.Phil. thesis, University of Sussex/SPRU, Brighton. Dutrénit, G.B. (2001), Learning and Knowledge Management in the Firm: From Knowledge Accumulation to Strategic Capabilities, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Enos, J. (1991), The Creation of Technological Capabilities in Developing Countries’, London and New York: Pinter. Figueiredo, P.N. (1999), ‘Technological capability: accumulation paths and the underlying learning processes in the latecomer context: a comparative analysis
The aeronautic sector in Brazil
197
of two large steel companies in Brazil’, D.Phil. thesis, University of Sussex/ SPRU, Brighton. Figueiredo, P.N. (2001), Technological Learning and Competitive Performance, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Figueiredo, P.N. (2003), ‘Learning, capability accumulation and firms differences: evidence from latecomer steel’, Industrial and Corporate Change 12(3), pp. 607–643. Frasman, M. (1984), Indigenous Technological Capability in the Third World, London: Macmillan. Freeman, C. (1987), Technology and Economic Performance: Lessons from Japan, Pinter, London. Frischtak, C. (1992), Learning, Technical Progress and Competitiveness in the Commuter Aircraft Industry: An Analysis of Embraer, Industry Series Paper, World Bank Industry and Energy Department, OSP, Washington, DC. Frischtak, C.R. (1994), ‘Learning and technical progress in the commuter aircraft industry: an analysis of Embraer’s experience’, Research Policy 23(5), pp. 301–312. Galimberti, I. (1993), ‘Large chemical firms in biotechnology: case studies on learning in radically new technologies’, D.Phil. thesis, University of Sussex/ SPRU, Brighton. Garvin, D.A. (2000), Learning in Action: A Guide to Putting the Learning Organization to Work, Harvard Business School Press, Boston, MA. Green, R.D. (1987), Brazilian Government Support for the Aerospace Industry, Report, US Department of Commerce/International Trade Administration, Washington, DC. Hobday, M. (1995), ‘East Asian latecomer firms: learning the technology of electronics’, World Development 23(7), pp.1171–1193. Hodgson, G.M. (1993), ‘Economics and Evolution: bringing life back into economics’, Cambridge, UK: Polity Press; Ann Arbor, MI: University of Michigan Press. Hobday, M. (2000), ‘Innovation in Southeast Asia: lessons for Europe?’, in Haley, Usha C.V. (ed.), Strategic Management in the Asia Pacific: harnessing regional and organizational change for competitive advantage, London: ButterworthHeineman, pp. 87–107. Iansiti, M. and Clark, K.B. (1994), ‘Integration and dynamic capability: evidence from product development in automobiles and mainframe computers’, Industrial and Corporate Change 3(3), pp. 557–605. Information Society Technologies/European Commission (2002), SME Participation in the FP5/IST Programme, Report, European Commission, Brussels, http://www.europe.eu.int (accessed in November 2003). Katz, J. (1976), Importación de Tecnología, Aprendizaje y Industrialización Dependiente, Fundo de Cultura Económica, Mexico, pp. 52–75. Katz, J. (2000), Passado y Presente del Comportamiento Tecnologico de America Latina, Serie Desarrollo Productivo, 75, Red de Reestructuracion y Competitividad, Division de Desarrollo Productivo y Empresarial, ECLAC/ UN, Santiago de Chile. Katz, J. (2001), ‘Structural reforms and technological behaviour: the sources and nature of technological change in Latin America in the 1990s’, Research Policy 30(1), pp. 1–19. Katz, J. and Kosacoff, B. (1998), ‘Aprendizaje tecnologico, desarrollo institucional
198
Sectoral systems of innovation and production
y la microeconomia de la sustitucion de importaciones’, Desarrollo Económico Revista de Ciencias Sociales 37(148), pp. 483–502. Kim, L. (1997a), ‘The dynamics of Samsung’s technological learning in semiconductors’, California Management Review 39(3), pp. 86–100. Kim, L. (1997b), Imitation to Innovation: The Dynamics of Korea’s Technological Learning, Harvard Business School Press, Boston, MA. Kim, L. (1998), ‘Crisis construction and organisational learning: capability building in catching-up at Hyundai Motor’, Organisation Science 9(4), pp. 157–172. Kim, L. (2000), The Dynamics of Technological Learning in Industrialization, Discussion Paper Series, INTECH/UNU, Maastricht, http://www.intech.unu. edu (accessed in November 2003). Kishimoto, C. (2004), ‘Clustering and upgrading in global value chains: the Taiwanese personal computer industry’, in Schmitz, H. (ed), Local Enterprises in the Global Economy: Issues in Governance and Upgrading, Cheltenham: Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Knight, P. and Marques, R.A. (2008), ‘Closing the technology gap’, in Harr, Jerry and Price, John (eds), Can Latin America Compete? Confronting the Challenges of Globalization, Palgrave Macmillan, Houndmills, Hampshire. Knorringa, P. (1999), ‘Agra: an old cluster facing the new competition’, World Development 27(9), pp. 1587–1604. Lall, S. (1992), ‘Technological capabilities and industrialisation’, World Development 20(2), pp. 165–186. Lefebvre, E. and Lefebvre, L. (1998), ‘Global strategic benchmarking, critical capabilities and performance of aerospace subcontractors’, Technovation 18(4), pp. 223–234. Lemos, M., Diniz, C., Santos, F., Afonso, M. and Camargo, O. (2000), O Arranjo Produtivo da Rede Fiat de Fornecedores, Report, Instituto de Economia/UFRJ, Rio de Janeiro. Leonard-Barton, D. (1990), ‘Implementing new production technologies: exercises in corporate learning’, in Glinow, M.A. and Mohrman, S.A. (eds), Managing Complexity in High Technology Organizations, Oxford University Press, New York, pp. 160–187. Leonard-Barton, D. (1995), Wellsprings of Knowledge: Building and Sustaining the Sources of Innovation, Harvard Business School Press, Boston, MA. Lundvall, B.A. (1992), National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, Cassell, New York. Lundvall, B., Johnson, B., Andersen, E.S. and Dalum, B. (2001), ‘National systems of production, innovation and competence building’, Paper presented at the Nelson and Winter Conference, DRUIDS, Aalborg. Malerba, F. (1992), ‘Learning by firms and incremental technical change’, Economic Journal 102(413), pp. 845–859. Malerba, F. (2002), ‘Sectoral systems of innovation and production’, Research Policy 31, pp. 247–264. Mani, S. (2001), Government, Innovation and Technology Policies: An Analysis of the Brazilian experience during the 1990s, UNU/INTECH, Maastricht, http:// www.intech.unu.edu (accessed in February 2002). Marengo, L. (1992), ‘Coordination and organisational learning in the firm’, Journal of Evolutionary Economics 2, pp. 313–326. Marques, R. (2004), ‘Evolution of the civil aircraft manufacturing system of innovation: a case study in Brazil’, in Mani, S. and Romijn, H. (eds), Innovation,
The aeronautic sector in Brazil
199
Learning and Technological Dynamism of Developing Countries, United Nations University Press, Tokyo, pp. 77–106. Meyers, P. (1990), ‘Non-linear learning in large technological firms: period four implies chaos’, Research Policy 19, pp. 97–115. Mowery, D. and Rosenberg, N. (1981), ‘Government policy and Innovation in the Commercial Aircraft Industry, 1925–1975’, Brighton: University of Sussex, Science and Technology Research Policy. Nadvi, K. (1999), ‘Collective efficiency and collective failure: the response of the Sialkot surgical instrument cluster to global quality pressures’, World Development 27(9), pp. 1605–1626. Nelson, R. and Rosenberg, N. (1993), ‘Technical innovation and national systems’, in Nelson, R. (ed.), National Innovation Systems: A Comparative Study, Oxford University Press, Oxford. Nelson, R. and Winter, S. (1982), An Evolutionary Theory of Economic Change, Harvard University Press, Cambridge, MA. Nonaka, I. (1994), ‘A dynamic theory of organisational knowledge creation’, Organization Science 5(1), pp. 14–38. Nonaka, I., Takeuchi, H. and Umemoto, K. (1996), ‘A theory of organizational knowledge creation’, IJTM, Special Publication on Unlearning and Learning, 11(7/8), pp. 833–845. Oliveira, L.G. (2005), ‘A cadeia de produção aeronáutica no Brasil: uma análise sobre os fornecedores da Embraer’, Tese de doutorado, DPCT/IG, Unicamp, Campinas. Oliveira, L.G. and Bernardes, R. (2002), O Desenvolvimento do Design em Sistemas Complexos na Indústria Aeronáutica: O Caso de Gestão Integrada de Projetos Aplicada ao Programa ERJ-170/190, ANPAD, Salvador. Padgett, J. (2003), ‘Dogfight: amid a down market in regional jets Brazil’s Embraer and Canada’s Bombardier show their teeth’, Time Europe Magazine, 26 May, 161(21), http://www.time.com/time/europe/magazine/article/0,13005,901030526452785-2,00.html, accessed on 14 July 2003. Patton, M.Q. (1990), Qualitative Evaluation and Research Methods, Newbury Park, CA: Sage Publications, Inc. Patel, P. and Pavitt, K. (1994), ‘National innovation systems: why they are important, and how they might be measured and compared’, Economic Innovation and New Technologies 3, pp. 77–95. Quadros, R., Furtado, A., Bernardes, R., and Franco, E. (2001), ‘Technological innovation in Brazilian industry: an assessment based on the São Paulo Innovation Survey’, Technological Forecasting and Social Change 67, pp. 203–219. Ramamurti, R. (1987), ‘Embraer: combining public power and private initiative’, in Ramamurti, R. State-Owned Enterprises in High Technology Industries: Studies in India and Brazil, Praeger, London, pp. 175–211. Rosenberg, N. (1982), Inside the Black Box, Cambridge University Press, Cambridge. Schmitz, H. (1995), ‘Small shoemakers and Fordist giants: tale of a supercluster’, World Development, 23(1), pp. 9–28. Schmitz, H. (2004), Local Enterprises in the Global Economy: Issues in Governance and Upgrading, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Stiglitz, J. (1987), ‘Learn to learn, localized learning and technological progress’,
200
Sectoral systems of innovation and production
in Dasgupta, P. and Stoneman, P. (eds), Economic Policy and Technological Performance, Cambridge University Press, Cambridge. Tacla, C. L. and Figueiredo, P.N. (2003), ‘Processos de aprendizagem e acumulação de competências tecnológicas: evidências de uma empresa de bens de capital no Brasil’, Revista de Administração Contemporânea 7(3), pp. 101–126. Teece, D.J., Pisano, G. and Schuen, A. (1997), ‘Dynamic capabilities and strategic management’, Strategic Management Journal 18(7), pp. 509–533. Texier, F. (2000), Industrial Diversification and Innovation: An International Study of the Aerospace Industry, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Velho, L. and Saenz, T.W. (2002), R&D in the Public and Private Sector in Brazil: Complements or Substitutes?, Working paper, UNU/INTECH, Maastricht, http://www.intech.unu.edu (accessed in January 2003). Viotti, E. (2002), ‘National learning systems: a new approach on technological change in late industrializing economies and evidences from the cases of Brazil and South Korea’, Technological Forecasting and Social Change 69, pp. 653–680. von Hippel, E. and Tyre, M. (1995), ‘How learning by doing is done: problem identification in novel process equipment’, Research Policy 24, pp. 1–12. Westphal, L.E., Kim, L. and Dahlman, C.J. (1985), ‘Reflexions on the Republic of Korea’s acquisition of technological capability’, in Frischtak, C. and Rosenberg, N. (eds), International Technology Transfer: Concepts, Measures and Comparisons, Praeger, New York, pp. 167–221. Xie, W. and Wu, G. (2003), ‘Differences between learning processes in small tigers and large dragons: learning processes of two color TV (CTV) firms within China’, Research Policy 32, pp. 1463–1479. Yin, R.K. (1944), Case Study Research: design and methods, California (EUA): Sage Publications, Inc.
201 Organizing the already existent process
Implementing quality control step, test step, or another step in the already existent process
Completely new process, for example cell manufacturing, just-in-time, and standards (ISO)
Process
Same machinery and equipment
Organizing the already existent project management process
Completely new software for design, such as adoption of CATIA software, computational fluid dynamics
Completely new project management process, such as implementation of integrated project management and co-design Implementing new steps in the already existent project management process
Completely new machinery, for example, 5 axes CNC machinery and laser machinery, new software for measuring Upgrading the specifications of the already existent machinery and software
Upgrading the already existing software, and/ or implementing new steps in the already existent design procedure Organizing the already existent design procedures
Design procedures
Project management
Organizational processes
Equipment-related
Production
Based on Bell and Pavitt (1993), and on the research.
Same product with improved painting and/or polishing
Low
Source:
New materials; new specification in measures, resistance, durability and/or speed
Completely new product
Product
Technological changes
TECHNOLOGICAL CHANGES: LEVEL AND DOMAIN
Middle
High
Level of impact
APPENDIX 6.1
202
Product-centred
Intermediate
Product engineering activities; in-house design and prototyping activities
Innovative capability Advanced In-house R&D and/ or in partnership with customers/suppliers and/or research institutes/universities substantially changing product design and/or specifications
Capability level
Managing the development of co-design techniques involving the participation of customers/suppliers
Team working to improve design procedures Team working for improving management of multi-firm projects and integration of product components Managing the development of specific machinery and equipment definition for production by an OEM, including monitoring tests
Engineering activities for adapting processes; systematic reverse engineering; continuous process improvement
Design procedures
In-house development of integrated project management techniques, involving the units: product development, production, finance and marketing, among others
Project management
In-house R&D for improving performance of machinery and equipment and for their new components; design and manufacture (machinery and/or equipment); software for attending specific demand
Equipment-related
Organizational processes
In-house process R&D, and/or in partnership with customers/suppliers and/or research institutes/universities
Processes
Production
Technological domains and related activities
APPENDIX 6.2 FRAMEWORK FOR TECHNOLOGICAL CAPABILITY ACCUMULATION
203
Source:
Basic
Routine production coordination across plant; basic quality control; replicating techniques
Managing tests and experiments
Managing tests and experiments for implementing minor adaptations in machinery and equipment and/ or software, adjusting to new raw materials or to improve performance under international certification (ISO 9000); own breakdown maintenance Routine replacement of components in machinery and equipment; routine software upgrading; participation in installation and performance tests
Own elaboration based on interviews, Lall (1992) and Figureiredo (1999).
Replicating specifications; routine quality control; attendance to customer’s requirements
Routine capability PreManaging tests and intermediate experiments in-house to improve product quality
and training operators; preventive maintenance
Basic coordination of project development for accomplishing deadlines; routine management procedures
Team working for improving quality in the internal coordination of projects
Basic control of documentation; routine design procedures (basic CAD)
Managing quality control procedures in design
PART III
Dynamics and evolution of sectoral systems
7.
China’s threat and opportunity for the Thai and Vietnamese motorcycle industries: a sectoral innovation system analysis1 Patarapong Intarakumnerd2 and Mai Fujita3
1.
INTRODUCTION
At present, sector is a key level of analysis for both academics and policy makers. Related to several theoretical traditions such as innovation system approach, evolutionary theory, industry life cycle analysis and others, the concept of sectoral system of innovation and production is better than mainstream industrial economic analysis such as structure, conduct and performance, game theory and transaction cost analysis, as it provides a broader and longer-term view on the evolution and dynamics of sectors. A sectoral system of innovation and production is a set of new and established products for specific uses and the set of agents carrying out market and non-market interactions for the creation, production and sale of those products. Sectoral systems have a knowledge base, technologies, inputs and demand. The agents are individuals and organizations at various levels of aggregation, with specific learning processes, competencies, organizational structure, beliefs, objectives and behaviours. They interact through processes of communication, exchange, cooperation, competition and command, and their interactions are shaped by institutions. A sectoral system undergoes processes of change and transformation through co-evolution of its various elements (Malerba, 2002). The concept may be used to analyse sectors in several aspects, namely for better understanding of the working, dynamics and transformation of sectors, for the identification of the factors affecting performance and competitiveness of firms and countries, and for the development of new public policy proposals. For developing countries, production capability (for overseeing operation of established facilities) is as important as innovation capability (for creating new products and processes). In the earlier phase of technological catching up, production capability may be even more crucial than 207
208
Sectoral systems of innovation and production
innovative capability (Bell and Pavitt, 1995; Amsden, 2001). Therefore, in analysing the situation in developing countries, it is quite possible that sectoral system analysis places more emphasis on production. Also in the context of developing countries, innovation should be seen as something new to the firm, not at the global knowledge frontier. It is likely to be incremental rather than radical. In many cases, innovations happened when imported technologies were first introduced to improve manufacturing processes. Albeit useful, the concept needs further development, especially the verification and elucidation from more empirical research. One of the important fields of research in this area is to study how and why sectoral systems of innovation and production in the same industry but different countries differ from each other, that is their differences in terms of learning processes, technologies, input, demands, types and structures of interaction of players in the sector, underlying institutions and the sector’s evolution. This chapter will shed light on this very important issue. Special attention will be paid to how sectoral systems of innovation and production in the same industry but across countries may evolve differently after facing similar threats and opportunities caused by the same external factor. To elaborate on this, we will use the case of evolution of the automotive sectors in Thailand and Vietnam and their dynamics and transformation when they are facing threats and opportunities from their fierce competitor China.
2.
OVERVIEW OF MOTORCYCLE INDUSTRIES IN THAILAND AND VIETNAM AND THREATS AND OPPORTUNITIES FROM CHINA
Threats and opportunities from China is probably one of the most popular topics for Asian academics and policy makers. Members of the Association of Southeast Asian Nations (ASEAN), in particular, are very much concerned with the rise of China. China is seen as both a threat and an opportunity for ASEAN. Obviously, China’s huge market is very attractive for export and investment from ASEAN. Chinese tourists, increasing rapidly, are one of the main target groups of ASEAN’s tourism promotion authorities. Several Chinese large conglomerates have invested in heavy industries and the energy sector in ASEAN. At the same time, for second-tier newly industrializing countries (Thailand, Malaysia, Indonesia and the Philippines), China is a fierce competitor in their key industries such as automotive, electronics, textile and garments, and so on. For ASEAN’s members which are transitional economies (Vietnam, Laos
The Thai and Vietnamese motorcycle industries
209
and Cambodia), the rise of China can be viewed as even more problematic, as it strongly affects efforts by these countries to industrialize and build up their indigenous technological capabilities. Not only have opportunities for their export-led industrialization strategies recently started been compromised by Chinese goods, but their import-substitution strategies are in jeopardy because of flooding of Chinese consumer and industrial goods. On the global scale, the motorcycle industry is a mid-tech and rather technologically mature industry. The automotive sector of several Asian latecomer economies started with assembling motorcycles and producing their parts by using imported technologies. Within ASEAN, Thailand and Vietnam are large producers, with production capacity of 3.5 million (Thai Automotive Institute, 2006) and 2 million (General Statistics Office, 2006) respectively in 2005. In both countries, markets are basically dominated by Japanese TNCs. Local companies are suppliers, especially second- or third-tier ones. Some of them, especially at the first tier, are joint ventures with foreign component makers. The motorcycle industry is also a typical example of an industry where China’s threat to other Asian countries has been enormous. China is the largest producer as well as the largest market of motorcycles in the world. In 2004 the country produced over 17 million motorcycles, 3.9 million of which were exported overseas (China Automotive Technology Research Center and China Automotive Industry League, 2005). The Chinese motorcycle industry is dominated by local companies that boast a competitive advantage in mass production of low-priced products. Since the end of the 1990s harsh competition at home has driven many Chinese motorcycle assembling companies to search for an export market. The “export drive” of Chinese motorcycle companies has been perceived as a serious threat to motorcycle manufacturers in the region as well as the rest of the world, though its actual impact has varied from country to country.
3.
METHODOLOGY
The studies of the motorcycle industries in Thailand and Vietnam were conducted separately but in the same year of 2004. The study in Thailand was commissioned by the National Science and Technology Development Agency (NSTDA) to the College of Technology and Innovation Management, King Mongkut’s University of Technology Thonburi (KMUTT). Though its main focus is on the Tiger Motorcycle (a purely Thai group of companies) cluster, the study also analyses and synthesizes the motorcycle industry as a whole in Thailand by exploring the roles and
210
Sectoral systems of innovation and production
capabilities of the main agents (both firm and non-firm), their linkages and the learning processes. The analysis of the Thai motorcycle industry was drawn from the data from the questionnaire surveys and in-depth interviews. One hundred and one questionnaires to relevant agents (mainly part suppliers) in the industry were sent, with a return rate of 40 per cent. Together with the survey, ten interviews were conducted to explore at length the characteristics, linkages, activities and interactions of key agents in the industry. Apart from the study, this chapter relies on other secondary sources of information such as industry reports and government plans, and notes from conversations with main senior government officials and company executives. The study on Vietnam was conducted within the framework of the research project “Motorcycle Industry in Asia” by the Institute of Developing Economies (IDE) during 2004–05. IDE commissioned a survey of motorcycle and parts companies in Vietnam from the Vietnam Institute of Economics, Vietnam Academy of Social Science (VIE-VASS), for which VIE-VASS conducted questionnaire surveys of 40 companies, including local and foreign (Taiwanese, Korean and Chinese) motorcycle and parts companies. The author, in collaboration with VIE, also interviewed ten of the companies to obtain qualitative information on the companies’ development process and current operations. In a separate attempt, the author conducted in-depth interviews with Japanese motorcycle and parts manufacturing companies in Vietnam, Thailand and China. In 2005, the author conducted follow-up interviews with some of the local and foreign (Japanese and Taiwanese) companies.
4.
THE MOTORCYCLE SECTORAL INNOVATION SYSTEMS IN THAILAND AND VIETNAM
This section analyses and compares elements of motorcycle sectoral systems of innovation and production of Thailand and Vietnam. These basic elements will be explored: products, agents (types, interactions and learning processes), demand conditions, and institutions (e.g. norms, routines, laws, regulations and standards) that shape agents’ cognition and action and affect the interaction among agents. 4.1.
Products
According to size of engine, motorcycles can be classified into four types: small (50–250 cc), medium (251–750 cc), large (751–1199 cc) and very large (1200 cc or more). Application-wise, there are also four types: standard
The Thai and Vietnamese motorcycle industries
211
motorcycle (small, simple, economical and easy to use), performance motorcycle (normally equipped with an engine of more than 251 and mostly for racing), styling/touring/luxurious motorcycle (normally with medium or large engine) and individual owner customization motorcycle (normally with large engine) (see Kosnik, 1995). Motorcycles produced and used in developing countries including Thailand and Vietnam usually belong to the standard type with economical price and small engine (125 cc or below). This segment has two prominent characteristics: (1) the use of mature technology, and (2) the dominance of Japanese motorcycle companies. In the business motorcycle segment, Honda’s Supercub, equipped with the C100 engine, still remains the dominant model. The C100 was developed back in the 1950s, and its basic technology has not gone through any major innovations for the past 30 to 40 years. Yet, the models developed on the basis of the C100 still keep an overwhelming market share, though individual models are adjusted to meet local conditions such as climate, road conditions, and preferences of consumers (Ohara 2006a, 2006b). Reflecting the proprietary advantage in technology, brand and distribution channels, along with the “dominant model”, Japanese motorcycle companies and especially Honda hold an overwhelming share in the economical business motorcycle segment at the global scale (Ohara, 2006b). Motorcycle units can be classified into two types: completely build unit (CBU) and completely knock-down (CKD) unit (a complete set of all parts ready for assembling). Many developing countries start import substitution of CBUs by simple assembly of imported CKD units, followed by the stage where the government policies force the domestic assembling companies to increase the local content ratio of the motorcycles. The motorcycle industries in both Thailand and Vietnam followed this development path. There are also motorcycle parts. Important parts are body part, exhaust system/filter/fuel system, drive/transmission and steering parts, engine part, suspension/brake/wheel, and instrument and seal. In terms of models produced by Japanese motorcycle companies, the majority of the parts are “custom-made parts” specifically produced according to the specifications of the motorcycle company.4 In contrast, Chinese motorcycle companies have imitated the base models developed by Japanese motorcycle companies, including the C100, often providing some “minor changes” (Ohara, 2006a). This is also the case for Thai motorcycle companies. In China, the base models and engines are shared by companies in the entire industry and, for parts that do not require either originality in appearance design and technology or high accuracy and high quality, ready-made parts available in the market are widely utilized (Ohara, 2006a). Such market-based
212
Sectoral systems of innovation and production
transactions also prevail in the procurement networks of Vietnam’s local motorcycle companies, which started with simple assembly of Chinese parts. 4.2.
Agents: Types, Interactions and Learning Processes
The main agents of the sector in the two countries are both firms (namely own-brand manufacturers and part suppliers) and non-firms (such as education institutes, research institutes, and government sector-specific regulatory and supporting institutes). This section outlines the main agents of the sectors in the two countries prior to China’s export drive. Firms Japanese TNCs which are own-brand manufacturers (OBMs) play significant roles in both countries. Thailand, however, has a longer history of producing motorcycles locally. Japanese TNCs have invested in the countries since the 1960s, while investment in Vietnam started in the mid-1990s. Each TNC (Honda, Suzuki, Yamaha and Kawasaki) has developed its own networks of suppliers. In 2004, Honda had the largest market share of 72 per cent. For suppliers, in total, there were 323 suppliers in Thailand, 122 of which supplied both automobile and motorcycle producers (GMI, 2004). First-tier suppliers are usually medium-size firms and joint ventures between foreign firms (especially Japanese) and Thai entrepreneurs. They used to be original equipment manufacturers (OEMs). However, owing to global sourcing strategies of the TNCs (sourcing parts and components from any supplier worldwide provided that the TNCs’ demanded specifications and prices are met) since the mid-1990s, they were forced to continuously improve efficiency and develop their design capabilities, thereby becoming own-design manufacturers or ODMs5 (Intarakumnerd et al., 2002). The second- and third-tier (OEM) suppliers are usually Thai companies with small factories and limited technological capability and number of employees. They sell motorcycle parts to the first-tier suppliers. Some part makers, the so-called replacement equipment manufacturers (REMs), instead focus on selling parts in replacement markets locally and abroad. Some part makers are both OEMs and REMs. Suppliers in Thailand have longer experience in learning and accumulating technological capabilities than those in Vietnam. Apart from engines and gears, which require the highest technological capabilities, other components can be made locally. In Thailand there are two emerging pure-Thai OBM companies producing motorcycles with their own brand, distributing channels and networks of suppliers. Started in 2000, Tiger Motor became the fourth largest producer of motorcycles in Thailand, with a market share of 3 per
The Thai and Vietnamese motorcycle industries
Table 7.1
Market share of motorcycle industry in Thailand in 2003
Company Thai Honda Manufacturing Thai Suzuki Motor Thai Yamaha Motor Tiger Motor Kawasaki Motors (Thailand) Source:
213
Year of establishment
Market share
1965
72%
1968 1966 2000 1976
13% 10% 3% 2%
GMI (2004).
cent (see Table 7.1). This pure-Thai firm has an aim of not only having its own brand but trying to build and upgrade technological capabilities of a network of pure-Thai suppliers under the slogan “Tiger, Pure-Thai Motorcycle”. However, in practice, Tiger still has to seek technologically sophisticated parts from Japanese joint ventures in Thailand. Another is a group of Thai suppliers, SME 007 Plus. They began with supplying motorcycle parts to Japanese TNCS and selling their products in replacement markets both in Thailand and abroad. With five core companies, the network of almost 100 SMEs has been built. They have jointly developed a few motorcycle parts (e.g. chain and choke) and sell them under the cobrand SME 007 Plus. They expanded their distribution network to cover repair shops all over Thailand. Recently, they teamed up with a Thai electrical appliance manufacturer, DiStar, which has financial strength and an extensive distribution network, to produce a prototype of the whole motorcycle, initially expected to be sold in 2006. There are also companies which are capital goods (mostly engine) suppliers to motorcycle assemblers and part makers. The local capital goods industry in Thailand is relatively weak. Most types of engines are, therefore, imported from abroad. The linkages between producers in the motorcycle industry and their capital goods suppliers are limited only to market transactions and advice on how to ‘operate’ the machines. There is virtually no knowledge exchanges leading to technological upgrading (GMI, 2004). The motorcycle industry in Vietnam started to develop only in the early to mid-1990s, when the Vietnamese government launched an import substitution policy for motorcycles by erecting trade barriers and providing incentives for foreign direct investment. By the late 1990s, the major motorcycle companies in Vietnam included one Taiwanese TNC (VMEP, a subsidiary of Taiwan’s Sanyang Motors6) and three Japanese TNCs
214
Sectoral systems of innovation and production
Table 7.2
Major foreign motorcycle companies in Vietnam
Company
Year of licence
Ownership structure
Vietnam Manufacture and Export Processing Co., Ltd. (VMEP) GMN Automobile and Motorcycle Parts Manufacture JV Co., Ltd. (GMN) Vietnam Suzuki Corp. Honda Vietnam Co., Ltd.
1992
Taiwan (100%)
1995
Yamaha Vietnam Co., Ltd.
1998
Lifan Motorcycle Manufacturing JV Co.
2002
JV (Thailand, Laos, Vietnam) JV (Japan, Vietnam) JV (Japan, Thailand, Vietnam) JV (Japan, Malaysia, Vietnam) JV (China, Vietnam)
Note:
1995 1996
JV stands for ‘joint venture’.
Source: Survey by the Vietnam Institute of Economics, Vietnam Academy of Social Science in 2004 and author’s interviews.
(Suzuki, Honda and Yamaha) (Table 7.2). Some Taiwanese and Japanese parts manufacturers followed the motorcycle companies to invest in Vietnam, producing such parts as tyres, batteries, electric parts, brakes and plastic parts. The number of Taiwanese suppliers that initially followed Sanyang amounted to 13 (Chen and Jou, 2002). The number of Japanese suppliers was smaller, reflecting Vietnam’s small market size and unstable investment environment in the mid-1990s. The number of local parts suppliers was even more limited. Although there were numerous local companies producing replacement parts, they were outside the procurement networks of foreign motorcycle companies for two reasons. First, the foreign motorcycle companies were not compelled to increase the local content ratio at this stage. Second, the parts produced by local companies did not meet the quality standards of TNCs. Apart from a limited number of state-owned machinery companies that started to produce parts for Honda Vietnam in the late 1990s, virtually no local parts suppliers participated in the procurement network of the TNCs. As of the late 1990s, Japanese motorcycle companies depended largely on parts produced in-house and parts imported from abroad, especially Thailand. Government Apart from firms, differences in roles played by non-firm agents contribute to differences in the evolution of the two countries’ sectoral systems of innovation and production. The Thai government, for the first time in
The Thai and Vietnamese motorcycle industries
215
the history of the country’s industrial policy, has selective policies under the Thaksin administration, which started in 2001 (see Intarakumnerd, 2006). The government has an aspiration to make Thailand the Detroit of Asia, that is to be a global centre for producing cars, motorcycles and automotive parts. It sets a clear target for the country to produce 10 million cars by 2010. To realize this mission, the ten-year National Science and Technology Strategic Plan (2004–13) aims to produce 1000 researchers and 4000 specialized engineers in the automotive industry (National Science and Technology Policy Committee, 2004). Implementation of these government plans by concerned agencies is still far from being coherent and synergistic (Intellectual Property Institute, 2006, p. 29). However, there is a specific organization responsible for development of the country’s automotive industry. The Thai Automotive Institute was set up as an independent public organization under the Ministry of Industry in 1988. It aims to operate like a private sector organization with a high level of flexibility and efficiency. The main objective is to be the centre for supporting development of the Thai automotive sector to be a main exporting base in the world. The institute has broad functions. It plays an important role in formulating policies for the industry, coordinating to implement those policies, setting and enforcing industrial standards, providing technical services such as testing, calibrating and quality assurance, and market information, applying results from research to upgrade technological capabilities and quality control systems to the global standard, and finally developing high-calibre human resources for the industry. The institute is trying to act as an intermediary bringing in external technology and knowledge to upgrade the technological capabilities of local suppliers. One of the important programmes of the institute was the Automotive Experts Dispatching Programme (2003–05) to bring Japanese experts to transfer key production technologies and skills, including those for mould and die design, and plant management to engineers and technicians of 200 Thai OEMs and REMs which joined the programme. The programme was a collaborative programme between the Thai and Japanese governments. Japanese experts were dispatched from Japan, including those who used to work for Japanese TNCs before (Thai Automotive Institute, 2006). Vietnam has regarded the motorcycle industry as a “key industry” since the mid-1990s. However, a comprehensive strategy for development of the motorcycle industry was not promulgated until 2006.7 From the mid-1990s, individual policy instruments, such as import protection, incentives for foreign direct investment, and product quality and safety standards, have been formulated in an ad hoc and often inconsistent manner. The involvement of different ministries such as the Ministries
216
Sectoral systems of innovation and production
of Industry, Trade, and Science, Technology and Environment and the lack of coordination between them were partly responsible for these problems. Frequent policy changes and weak enforcement have been noted by foreign companies as serious problems. To illustrate, despite the fact that imports of CBUs were controlled by import quotas up to 1998 and were prohibited from 1998, new and secondhand CBUs made in Thailand and Japan continued to be smuggled into Vietnam in large numbers. Recurrent complaints by the foreign companies about the Vietnamese government policies8 clearly demonstrate the general absence of trust between the government and foreign companies. Public–private collaboration in the motorcycle industry as seen in the case of Thailand did not emerge. Supporting knowledge-producing agents Given the size and significance of the automotive industry in Thailand, special education programmes for the industry are very much needed. There are a few education programmes for the automotive industry. Chulalongkorn University’s Faculty of Engineering has the most recognized and specific one, as it offers a B.E. in Automotive Engineering. A Thai-language programme started in 1995. It produced 15 high-quality graduates annually. Toyota has been providing significant support in terms of equipment and instructors. After graduation, many students were recruited by Toyota. Since 2005, the faculty has offered an Englishlanguage programme aiming at producing 100 graduates per year. The instructors came from both Thai and foreign academics as well as guest lecturers from the automotive industry (Boonchukosol, 2006). Besides education, the linkages and knowledge flow between firms in the motorcycle industry and education institutes in Thailand are rather weak and fragmented. Firms have linkages with individual faculty members rather than organizational linkages. The organizational linkages are mostly limited to sending students to internships with companies, testing and consultancy for solving basic production problems. In Thai universities, there are some research projects on automotive technologies, but there is none specifically on ‘motorcycle technologies’, needless to say about the collaborative research between firms and universities (GMI, 2004). The linkages with research and technological supporting organizations are insufficient but better than those with education institutes. For example, the Thai German Institute, the largest and most advanced training centre for industrial technologies in Thailand, provides technical services and course-based or tailor-made training to entrepreneurs in rather advanced technology-related design and production systems, especially in mould and die technology and computer-aided design and computer-aided manufacturing, automated production and precision machining (Thai German
The Thai and Vietnamese motorcycle industries
217
Institute, 2006). The National Metal and Materials Centre (MTEC) under the National Science and Technology Development Agency (NSTDA) has been conducting research quite relevant to the automotive industry, such as finite elements, which can be used for designing many parts of motorcycles and automobiles, computer-aided design, engineering and manufacturing, failure analysis and material degradation. It provides testing and training services to the private sector in the aforementioned areas as well (National Metal and Materials Centre, 2006). As mentioned above, because of higher-degree competition through the global sourcing strategies of TNCs and China’s threats, the Thai companies are trying to forge more and deeper collaboration with local universities and research institutes. In Vietnam, while formal education and training institutions targeting the demands of the industry remain insufficient, in-house training within foreign companies seems to have worked relatively well.9 Hanoi and Ho Chi Minh City Universities of Technology, with programmes on mechanical engineering, were cited as important sources of engineers for Japanese motorcycle manufacturers. The newly recruited engineers generally go through substantial in-house training for more practical skills such as specific methods of management and requirements of the Japanese companies. Most Japanese companies interviewed by the author emphasized the high quality of the Vietnamese engineers and workers, a key to improving the quality and production efficiency within the factory. For Japanese motorcycle and parts companies in Vietnam, knowledge flows mainly originate abroad, especially their headquarters (research institutes) in Japan. For instance, Honda has a regional R&D centre in Thailand, which cooperates with headquarters in Japan closely in product development and design for Honda’s subsidiaries in South-East Asia. Honda’s factory in Vietnam receives substantial technical support from the headquarters in Japan and R&D centre in Thailand, in launching new models, finding potential local suppliers and providing technical assistance to them, and so on. Vietnam does not have a research institute specifically targeting the automotive industry, but has a few public research institutes engaged in the wider field of machinery industry. Among them, the Research Institute of Technology for Machinery (RITM) under the state-owned Vietnam Engine and Agricultural Machinery Corporation and Industrial Machinery and Instruments (IMI) Holding under the direct management of the Ministry of Industry are engaged in research and training in the fields closely related to the production of motorcycle parts, such as die casting, forging, metal stamping, testing, and production of moulds and dies. In particular, IMI Holding has been successful in upgrading the
218
Sectoral systems of innovation and production
level of equipment and technology through assistance from international organizations and companies in developed countries.10 However, even IMI Holding is fraught with weaknesses, such as limited linkages between its R&D and training activities and the actual demands in the industry. Virtually none of the local or foreign-invested companies surveyed by the author, including state-owned ones, had substantial linkages with the public research institutes. 4.3.
Demand Conditions
In general, Thai consumers have more purchasing power than Vietnamese. The country GNI per capita was US$2750 compared to US$620 for Vietnam in 2005 (World Bank, 2006). Demand in Thailand for the motorcycle is also more sophisticated. The country embarked on industrialization in the early 1960s, with a relatively high economic growth of 6–7 per cent during the last 40 years. The amount of middle-class people who require sophisticated products is rather large. Moreover, the motorcycle was adopted as a major means of transportation much sooner in Thailand. Some sections of consumers, especially the middle class, have more time than their Vietnamese counterparts to develop special tastes, for example demanding motorcycles with a higher capacity and better appearance. The more sophisticated demands in Thailand also pressure local companies and Japanese subsidiaries to deepen their technological and innovative capabilities in order to respond to such demands. In Thailand, motorcycles and their parts produced in China are perceived as products inferior to those produced locally. Japanese motorcycles have very established brand names. Vietnam embarked on market-oriented economic reform in the late 1980s, and rapid economic growth started only in the early 1990s from a low starting point. The shift from bicycles to motorcycles started in the 1990s within the newly emerging “salaried urban middle class” (Fforde, 2003, p. 50), but this remained a minimal portion of the nation, where 80 per cent of the population still lived in the rural area as of 1995 (General Statistics Office, 2006). The prices of locally assembled and imported motorcycles were still far above the reach of the vast majority of the population, exceeding US$2000 in the late 1990s. TNCs that entered Vietnam in the late 1990s failed to recognize the vast “potential” demand for low-priced motorcycles, and such missed opportunities were eventually grabbed by the Chinese firms. Until very recently, the Honda brand dominated the Vietnamese market, because the secondhand Honda C100 motorcycles had been in the country for decades, with a reputation of being durable and economical.11
The Thai and Vietnamese motorcycle industries
219
All the above differences have implications when Chinese motorcycles started to penetrate the Thai and Vietnam markets. 4.4.
Institutions
Institutions like laws, norms, routines, standards and so forth play an important role in shaping the behaviour of agents and their interaction among each other. One example is in product standards. In Thailand, industrial standards concerning quality, efficiency, consumer safety and the environment were introduced and enforced more strictly than in Vietnam. These, in effect, block inferior Chinese imports. Therefore, Chinese motorcycles and parts were not as successful as in Vietnam in penetrating into the Thai market. On the other hand, in Vietnam, quality and environmental standards, safety regulations and other policies were newly emerging, and enforcement was seriously weak.
5.
5.1.
EVOLUTION OF THE SECTORAL INNOVATION SYSTEM UNDER CHINA’S THREATS AND OPPORTUNITIES The Extent of China’s Impact on the Motorcycle Industries in Thailand and Vietnam
As noted in section 2, the “export drive” of Chinese motorcycle companies has accelerated since the late 1990s. Though the Chinese products penetrating both markets are the same (cheap and standardized), their impact differed significantly between Thailand and Vietnam. In Thailand, imports of motorcycles and parts from China remained insignificant throughout the period, though a gradual increase in the imports of parts was observed after 2003 (Table 7.3). The market structure also remained relatively stable, major Japanese TNCs keeping the lion’s share (Table 7.1). In contrast, in Vietnam, imports of motorcycles surged dramatically from 1999 to 2001, followed by a sudden drop in 2002 and a steady increase in imports of parts from 2002 onwards. The import of motorcycles reached over 1.8 million in 2001, more than triple the annual sales in the mid-1990s. In 2000, so-called “Chinese’ motorcycles, motorcycles assembled by Vietnamese companies using the knocked-down kits imported from China, accounted for nearly 80 per cent of the Vietnamese motorcycle market (see Figure 7.1), significantly reducing the share of Japanese motorcycle companies. The low-priced “Chinese’ motorcycles
220
1996
0 2
2 0 0
0
0 1
1 0 0
0
6 2 417
1995
378 3 801
1996
0
2 0 0
0 2
1997
84 861
1997
1
2 0 1
0 2
1998
1 548
1998
2
2 0 2
0 2
1999
361 89 778
1999
24
4 5 19
0 4
2000
2001
85
5 50 35
0 5
2001
1 041 1 833 073
Unit: million
384 1 229 195
2000
132
11 80 52
3 8
2002
13 054 284 194
2002
59
20 23 36
9 12
2003
14 706 31 193
2003
92
16 46 46
6 10
2004
15 906 25 466
2004
Source:
World Trade Atlas.
Note: HS codes corresponding to each of the categories are as follows: motorcycles (8711), engines (840732) and parts other than engines (871419).
Thailand: Engines Other parts Total Vietnam: Engines Other parts Total
(2) Parts
Thailand Vietnam
1995
Unit: number of motorcycles
China’s export of motorcycles and parts to Thailand and Vietnam
(1) Motorcycles
Table 7.3
62
25 31 32
8 17
2005
39 158 68 203
2005
The Thai and Vietnamese motorcycle industries
221
100% 80% 60%
Others Honda Vietnam Co., Ltd. ‘Chinese’ motorcycles
40% 20% 0% 1998
1999
2000
2001
2002
2003
Notes: 1. ‘Chinese’ motorcycles include motorcycles assembled by Vietnamese companies mainly using parts imported from China. 2. ‘Others’ include: (a) motorcycles produced by foreign motorcycle assembling companies based in Vietnam other than Honda Vietnam (e.g. Vietnam Export Manufacturing Processing Co., Ltd. (VMEP, a subsidiary of Taiwan’s Sanyang Motors), Vietnam Suzuki, and Yamaha Vietnam); and (b) imported motorcycles including those made by Honda’s subsidiaries abroad (e.g. Thailand). Source:
Authors’ interview at Honda Vietnam in September 2004.
Figure 7.1
Market share of the motorcycle industry in Vietnam
significantly enlarged Vietnam’s motorcycle market to the middle- to lowincome group in the urban and rural area. However, the years after 2002 witnessed a sudden drop of motorcycle imports and the quick recovery of market share by Japanese TNCs, including Honda Vietnam, owing to reasons which will be elaborated later. The difference can be explained, first and foremost, by the difference in the effectiveness of the import controls by the respective governments. While Vietnam prohibited import of CBUs and imposed tariffs on parts according to the local content ratio, Chinese motorcycles were imported into the country by local traders who claimed a false local content ratio to evade tax. In Thailand, import controls were effectively enforced. Second, price differentials between the locally produced motorcycles and Chinese motorcycles were much larger in Vietnam than in Thailand. In Vietnam, the Chinese motorcycles were priced at one-third or even onequarter (around US$500 to US$800, compared to US$2000 or over for Japanese brand motorcycles) of the price of motorcycles produced locally by Japanese motorcycle companies. Third, Japanese TNCs took a much
222
Sectoral systems of innovation and production
stronger hold of the market, in terms of brand diffusion and nationwide distribution and service networks, in Thailand than in Vietnam, reflecting the history of operation in the respective countries. Last but not least, the aforementioned differences in demand conditions in the two countries contributed to different results. Thai consumers, especially the sizeable middle class, had more sophisticated demand than their Vietnamese counterparts. Chinese imports were considered as inferior in terms of technology and appearance to locally produced motorcycles. 5.2.
How Thailand and Vietnam’s Sectoral Innovation Systems Were Transformed
Whereas data on imports, sales and market share demonstrate the difference in terms of the extent to which the motorcycle industries in the two countries were affected by China’s export drive, examination of the dynamics of the sectoral innovation system in the two countries provides much deeper insights into similarities and differences in the way different agents and institutions reacted to the new challenges. Generally, the longer presence of Japanese TNCs and the higher indigenous technological and marketing capabilities of Thai assemblers and part suppliers made Thailand less vulnerable to China’s threat. Different types of firms behave differently when facing a common external factor. To Japanese companies in Vietnam and Thailand, China is undoubtedly a competitor. Here we can observe similarities in strategies within the same company across countries (i.e. Thailand and Vietnam), yet differences across companies which reflect the regional strategies of the TNCs. In the case of Honda, the company pursued a common strategy of launching low-priced models in both countries. In Vietnam, where the company’s market share diminished sharply, the Wave a was launched in January 2002 with a price nearly one-third that of the company’s previous models. In Thailand, where the company perceived a “potential threat” from China, Wave 100 was launched in June 2002.12 Wave a was developed mainly in the Thai factory of Honda (Thai Honda Manufacturing Co., Ltd.), which was Honda Vietnam’s mother factory, in close collaboration with Honda R&D Southeast Asia in Thailand (Ohara et al., 2003). The model was developed using the engine and body of Honda’s existing models, as well as low-priced Chinese parts to bring down production costs. The new low-priced models quickly gained popularity in both countries. Especially in Vietnam, Honda quickly recovered market share after 2002 (Figure 7.1). In contrast, Yamaha has pursued ‘higher value added’, with more emphasis on brand, design (for a new model) and quality in both countries, a strategy fundamentally different from that of Honda.
The Thai and Vietnamese motorcycle industries
223
In terms of the local companies, we can observe clear differences between the two countries. For a more established Thai own-brand manufacturer like Tiger, China is viewed both as a competitor and as an opportunity provider. To differentiate from Chinese motorcycles, Tiger is trying to produce higher-quality products for the ‘upper’ market. It attempted to increase its own technological capabilities by setting up a design and development department and starting collaboration with local universities and public research institutes. At the same time, some Chinese imported parts, not locally produced, were accepted. This gives Tiger flexibility in choosing the best parts for its motorcycles. For SME 007 Plus, the impacts of China are quite substantial. It is a new group of part producers with a rather low level of technological capabilities (relative to Tiger), and it cannot produce all its parts through its members. Therefore, it decided to change its strategy from assembling the whole motorcycle with its own brand name to importing designed frame and major components which cannot be produced by its group’s members, such as the engine and crush from China. The plan to produce the branded motorcycle has been postponed. Instead, the group moved downstream in the value chain to recruit repair shops as members and try to modernize these shops by introducing new shop layouts and modern management. By doing that, the group aims to stimulate demand from downstream (repair shops) members for components produced by the group’s upstream (manufacturing) members (Katikarn, 2006). The leaders of the group are very entrepreneurial and very capable in developing relationships with government agencies. The new movement to upgrade repair shops has got financial support from the National Innovation Agency and Department of Industrial Promotion (under the Ministry of Industry) and technical support from the Faculty of Engineering, Kasetsart University. It also received financial support to develop a software program for managing its information system and supply chain, especially franchising, management from the Software Industry Promotion Agency and the Department of Business Development, Ministry of Commerce (see Intellectual Property Institute, 2006). For part suppliers, there are differences between Japanese (pureJapanese or joint ventures) suppliers and local (pure-Thai) suppliers. The local pure-Thai suppliers have lower technological capabilities than their Japanese counterparts, which have technological transfer from their mother companies in Japan. Also first-tier Japanese part makers in Thailand usually have long-term relationships with their customers (Japanese OBMs). In some cases, engineers from Japanese OBMs were sent to co-develop parts or components manufactured by part makers or give advice on how to upgrade the production system, but the cooperation
224
Sectoral systems of innovation and production
has not reached the level of joint R&D, which is mostly done in Japan. Undoubtedly, the threat from China is more intimidating for the pureThai suppliers, especially the second- and third-tier ones. In Vietnam, the surge of imports from China provided new opportunities for local businesses. Since Vietnam prohibited imports of CBUs from 1998 to 2003, Chinese motorcycles had to be imported as knockeddown parts, and had to be reassembled in Vietnam.13 As a result, over 50 local companies assembling imported Chinese motorcycle parts emerged. However, they faced difficulties from 2002 onwards, when the Vietnamese government strengthened both the substance and the enforcement of regulations on the motorcycle industry by the following measures: (1) stronger enforcement of import tariffs according to the local content ratio,14 (2) “standards for motorcycle manufacturing companies”, requiring, among other things, in-house production of key parts,15 and (3) renewed quality and environmental standards.16 While many local companies left the industry, some local companies remained, investing in in-house production of parts. A few even started to develop their own motorcycle brands and distribution networks. These local companies continue to keep a certain market share owing to their price advantage vis-à-vis Japanese products (see Figure 7.1). Yet accumulation of technological capabilities within these local companies remained limited. First, unlike in Thailand, they have not succeeded in developing their own motorcycle models instead of ‘copies’ of Japanese models. Second, many are largely dependent on foreign partners for in-house production of parts, or largely use imported parts or parts produced by Chinese companies based in Vietnam. Among the companies interviewed by the author, Company A, one of the largest local motorcycle companies, whose production reached 200 000 units in 2004, was heavily reliant on its Chinese partner. The company’s basic strategy was: “The Vietnamese would look after the sales and management, while the Chinese engineers would take care of the production.” The company had a joint venture with a Chinese motorcycle company to produce motorcycle parts, where 50 Chinese engineers were stationed to assist Vietnamese workers in 2004, two years after the factory started operations. Company A also cooperated closely with Chinese parts companies located in the neighbourhood. The close cooperation with Chinese partners enabled Company A to rapidly increase production and bring down the production cost. In contrast, Company B pursued a more self-reliant strategy to avoid over-dependence on foreign partners. It has become clear, however, that upgrading its own technological level through trial and error is too time-consuming when it has to compete in a highly competitive market consisting of the powerful Japanese TNCs and numerous local companies.
The Thai and Vietnamese motorcycle industries
225
Against this background, Company B had to increase the use of imported parts from China in recent years. The increased use of Chinese parts was widely observed among many local motorcycle companies. Table 7.3 also confirms that imports of engines and parts from China has increased rapidly since 2003. More positive developments, however, can be observed in motorcycle part production. As pointed out in section 4.2, companies producing motorcycle parts in Vietnam were extremely limited prior to China’s export drive. However, the emergence of numerous local motorcycle companies created substantial demand for low-priced motorcycle parts, giving impetus to numerous local companies, previously producing motorcycle replacement parts, bicycle parts or machinery parts, to enter into motorcycle parts production. While many of the new entrants are suffering from the falling demand owing to the demise of the local motorcycle companies, some of these new entrants have managed to participate in the procurement networks of Japanese and Taiwanese TNCs mainly as second-tier suppliers.17 Two of the three second-tier suppliers interviewed by the author received regular assistance in production technology and quality controls either from the first-tier supplier or directly from the Japanese motorcycle company. Apart from the emergence of new firms and changes in existing firms’ strategies and capabilities, the China factor also induces ‘systemic’ changes as follows: ●
●
●
Inter-firm linkages have been strengthened. In order to survive, Tiger and SME 007 were pressured to interact more extensively, for instance frequent meetings and joint activities like developing closer and stronger supply chains, groups’ identity and brand names. In Vietnam, as mentioned before, new inter-firm linkages between local firms and Chinese competitors and part suppliers in the form of joint ventures and/or producer–supplier relationships have been initiated. Some Vietnamese firms started to become OEM suppliers of Japanese TNCs, a relationship which had not happened before. The relationships between firms and non-firm agents such as government agencies, research institutes and universities became stronger. This is more obvious in the case of Thailand. As pointed out before, Thai own-brand manufacturers tried to move up the value chain and develop products superior to the Chinese imports by starting collaborative research with local universities and public research institutes. Institutional changes happened as well. The Vietnamese government, in response to the problems caused by the sudden influx
226
Sectoral systems of innovation and production
of Chinese motorcycles, such as traffic accidents and congestion, improved the enforcement of import controls and introduced new regulations on industrial and environmental standards. Both the substance and the enforcement of policies continued to improve gradually over time. The analysis above suggests that the external threats from China induced changes in firms’ strategies and market environment, entry of new firms, and creation of new relationships among firms and between firms and non-firm agents. These, to a certain degree, systemically transformed the motorcycle sectoral systems of innovation and production of Vietnam and Thailand to be stronger and less fragmented.
6.
CONCLUSION
The above findings shed some light on the evolution of the sectoral system of innovation and production. Theoretical implication to the concept of the sectoral system of innovation and production can be highlighted. The same external factors can be both threats and opportunities that can influence the transformation of a sectoral system of innovation and production, as shown in the cases of the motorcycle sectoral systems of innovation and production in Thailand and Vietnam. Under such transformation ‘new’ types of agent, new inter-firm relationships and new types of collaboration between firms and non-firm agents can emerge, while ‘existing’ ones can be strengthened, weaken or disappear. Different sectoral systems of innovation and production can evolve differently when they are facing similar threats and opportunities. The direction and the pace of evolution depend very much on the existing capabilities of agents, the strength of their linkages and their processes of collective learning to withstand the threats and exploit the opportunities. To illustrate, Thailand can withstand the threats and exploit the opportunities better than Vietnam (i.e. being more equal partners with Chinese part makers) because, despite still being rather weak and fragmented, its motorcycle sectoral system of innovation and production has relatively more capable agents (i.e. longer-present and more technologically sophisticated TNCs, local champions who are own-brand manufacturers, a government with more vivid and targeted government strategies for the automotive sector, and more sectoral-specific and active government supporting agencies, universities and research institutes), more sophisticated demand conditions, and relatively more interaction, especially knowledge transfer, among agents.
The Thai and Vietnamese motorcycle industries
227
An external factor, in this case threats and opportunities from China, induces changes in the strategies of agents, a higher degree of interactions among agents, and more variation in the sectoral system of innovation and production. For example, a Thai OBM producer, Tiger, and some locally owned suppliers were forced to make greater in-house efforts to deepen their own technological capabilities and develop more outwardlooking strategies such as forging collaboration with local universities and research institutes to improve existing products and production processes in order to outperform Chinese competitors. Local entrepreneurs in Vietnam seized the opportunity by collaborating with Chinese firms to start new businesses as motorcycle part suppliers. Given external threats and opportunities, firms in both countries do play more important roles than non-firm agents in transforming the sectoral system of innovation and production. In the case of Vietnam, however, relatively weaker dynamic linkages between firms and non-firm agents such as universities and research institutes allow firms to play an even more decisive role in the transformation of the sectoral system. While the direction and the capacities of the local companies to respond to the external threat were largely shaped by the capabilities of existing agents, local demand conditions, and institutions in the country, explaining the strategies of TNCs calls for a regional perspective. Honda, for instance, perceived what happened in Vietnam in 2000 to 2001 as a fundamental threat to the company’s operations in Asia as a whole, which had to be countered by close collaboration between its subsidiaries in the region. The new low-priced model was developed mainly in the company’s production and R&D bases in Thailand, and the new model was launched not only in Vietnam but also in Thailand, where China’s impact was still only a threat. In contrast, Yamaha has tried to increase the value added of its products through brand, design and quality, a strategy common to both Thailand and Vietnam. The different strategies of the two Japanese TNCs also illustrate that, even within the same sectoral system of innovation and production, different firms adopt different strategies and change differently when facing similar external factors. Within the same sectoral system of innovation and production, while there might be an opportunity and a threat created from an external factor for many firms, the ‘realization’ of that opportunity and preventing that threat depend very much on firms’ abilities to seize such opportunities and fend off threats. For instance, while SME 007 Plus needed to postpone its plan to build the whole motorcycle under its own brand name, Tiger managed to survive and turn threats into opportunities to be able to source some motorcycle parts from China. This could be achieved because its level of technological capabilities was higher than that of
228
Sectoral systems of innovation and production
SME 007 (and relatively high compared with Chinese competitors). For part suppliers, the ability to integrate into global production networks of TNCs is also an important factor in fending off the threat. In Vietnam, more accumulation and learning happen in local parts companies that are incorporated into the procurement networks of TNCs as first- or second-tier suppliers. Finally, what general lessons can be drawn from the case study for researchers and policy makers concerned with enhancement of technological capabilities in developing countries? The most obvious lesson is the key roles played by foreign affiliates in facilitating the process capability building by providing training, developing linkages with suppliers and local institutions, and so on. This finding basically concurs with that of the existing literature on global production networks and technology transfer via FDI (Lall and Mortimore, 2000; UNCTAD, 2001; Ernst and Kim, 2002). However, our case study also showed that there were limits to the contributions of foreign affiliates in enhancing the technological capabilities of local firms. In both Thailand and Vietnam, first-tier suppliers of Japanese motorcycle manufacturers were largely foreign affiliates (Japanese–Thai joint ventures in the case of Thailand, and Japanese and Taiwanese in the case of Vietnam), and local suppliers were largely in vulnerable positions as second- and third-tier suppliers. The fact that joint R&D activities between motorcycle manufacturers and first-tier suppliers are likely to remain in Japan suggests limits to the roles of Japanese affiliates as both countries eventually shift their focus from production to innovation capabilities in the future. In view of such limits, it may be worth pointing out the emerging local firms’ initiatives that are independent of the production networks of Japanese motorcycle manufacturers, such as SME 007 in Thailand and local assemblers and their suppliers in Vietnam. Those firms were shown to be flexible to adapt to the new opportunities by taking advantage of their strengths in grasping the needs of the local market and having access to local distribution channels. Further research is needed to explore their longer-term prospects for growth and capability building.
NOTES 1. 2. 3.
The authors would like to give special thanks to Cristina Chaminade for her valuable comments as a discussant at the GLOBELICs India 2006 Conference in Trivandrum, 4–7 October 2006. Project Manager of College of Innovation, Thammasat University, Thailand. Researcher, Southeast Asian Studies II Department, Area Studies Center, Institute of Developing Economies, Japan External Trade Organization.
The Thai and Vietnamese motorcycle industries 4. 5.
6. 7.
8.
9. 10. 11.
12. 13.
14.
15. 16.
17.
229
Otahara (2006) notes that 90 per cent of the motorcycle parts that Japanese motorcycle companies procure from parts manufacturers are specifically designed for each model. OEM and ODM are specific forms of subcontracting. Under original equipment manufacture (OEM), a latecomer firm produces a finished product to the precise specification of a foreign transnational corporation (TNC), which will market under its brand name via its own distribution channels. Under own-design manufacture (ODM), a latecomer firm also carries out some or all of the product design (Hobday, 1995, p. 37). Chinfon Group, the parent company of Sanyang Motors, holds a 100 per cent stake in VMEP. After a long period of discussion and preparation, “Decision of the Ministry of Industry No. 33/2006/QD-BCN approving the development strategy for the motorcycle industry till 2015 and orientation towards 2020” was promulgated on 13 September 2006. It aims to develop Vietnam into a major producer and exporter of motorcycles, parts and components in the region. There were numerous occasions when Vietnam’s policies towards the motorcycle industry received criticisms from companies, both foreign and local. The most prominent example was observed in September 2002, when the Vietnamese government suddenly announced the import quota for motorcycle companies for the whole year (“Motorcycle makers to send reps to Vietnam”, Japan Times, 10 October 2002; “Japanese Automobile Association to visit Hanoi”, Vietnam Economic Times, 15 October 2002). Owing to the sudden imposition of the quota, a number of foreign motorcycle companies that had used up the quota, including Honda Vietnam and Yamaha Vietnam, had to temporarily stop their operations because they could not import parts. Information provided in the following two paragraphs is based on the author’s interview with Honda Vietnam, Yamaha Vietnam, and several Japanese parts manufacturers in Vietnam. This remark was made by two Japanese engineering specialists who visited the company in 2001 under a project by the Japan External Trade Organization (JETRO) aiming at strengthening supporting industry in Vietnam. The first Honda boom in Vietnam took place in the 1960s, in the midst of the Vietnam War. Tens of thousands of Honda Supercubs were imported into the country and used mainly by American military officers stationed in South Vietnam. After the reunification of the country in 1975, secondhand Honda Supercubs continued to be traded within the country owing to their remarkable durability, high fuel efficiency, and ease of maintenance and repair. Interestingly, Honda did not launch low-priced models in Indonesia. Since assembled motorcycles could not be imported, Chinese motorcycles went through the Vietnamese customs as knocked-down parts. The parts were reassembled by Vietnamese companies. But the Chinese customs statistics (Table 7.3) show that China’s exports in the years 1999–2001 were in the form of assembled motorcycles, suggesting that the motorcycles were knocked down mainly for the purpose of getting through Vietnam’s customs. The inspection carried out by the inter-ministerial fact-finding team in 2002 revealed that all of the 52 local motorcycle companies claimed false local content ratio to get access to preferential import tariffs. The Vietnamese government prohibited these companies from importing motorcycle parts until they repaid the unpaid tax amount (Viet Nam News, 28 July 2002). Decision No. 24/2002/QD-BCN dated 10 June 2002 of the Ministry of Industry issuing criteria for motorcycle manufacturing and motorcycle assembling enterprises. Decision No. 2557/2002/QD-BGTVT dated 16 August 2002 of the Ministry of Transport promulgating the regulation on the inspection of quality, technical safety and environmental protection in the manufacture and assembly of motorcycles and mopeds of all kinds. Six foreign parts suppliers based in Vietnam interviewed by the author (Japanese, Taiwanese and Korean first-tier suppliers) used an average of 27 second-tier suppliers,
230
Sectoral systems of innovation and production 22 of which were local companies. The majority of the second-tier local suppliers had newly entered into the production of motorcycles to supply parts to local motorcycle companies, and were located near the first-tier suppliers.
REFERENCES Amsden, A. (2001), The Rise of the Rest: Challenges to the West from LateIndustrialising Economies, Oxford University Press, New York. Bell, M. and Pavitt, K. (1995), ‘The development of technological capabilities,’ in Haque, I. (ed.), Trade, Technology and International Competitiveness, World Bank, Washington, DC. Boonchukosol, K. (2006), Personal communication, 22 August. Chen, Dung-Shen and Jou, Sue-Ching (2002), “Weakening transplanted production networks: a case study of Taiwan’s motorcycle production network in Vietnam,” Paper presented at Conference on Taiwanese Businesses in Vietnam, organized by Academia Sinica and National Taiwan University, 11 October. China Automotive Technology Research Center and China Automotive Industry League (2005), China Automotive Industry Yearbook 2004–2005, Tianjin, China, Editorial Office of China Automotive Industry Yearbook. Ernst, D. and Kim, L. (2002), ‘Global production networks, knowledge diffusion, and local capability formation’, Research Policy 31, pp. 1417–1429. Fforde, Adam (2003), “Vietnam: dyed-in-the-wool tigers?”, in Drummond, Lisa B.W. and Thomas, Mandy (eds), Consuming Urban Culture in Contemporary Vietnam, Routledge Curzon, London and New York. General Statistics Office (2006), Statistical Yearbook of Vietnam 2005, Statistical Publishing House, Hanoi. Graduate School of Management and Innovation (GMI), King Mongkut’s University of Technology Thonburi (2004), Final Report of the Project of Enhancing Technological Capabilities of Thai Motorcycle Cluster, submitted to National Science and Technology Development Agency, October. Hobday, M. (1995), Innovation in East Asia: the Challenge to Japan, Edward Elgar Publishing, Aldershot, UK and Brookfield, VT, USA. Intarakumnerd, P. (2006), ‘Thailand’s National Innovation System in Transition’, in, Lundvall, B., Intarakumnerd, P. and Vang, J. (eds), Asia’s Innovation Systems in Transition, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Intarakumnerd, P., Chairatana, P. and Tangjitpiboon, T. (2002), ‘National innovation system in less successful developing countries: the case of Thailand’, Research Policy 31(8–9), pp. 1445–1457. Intellectual Property Institute (2006), ‘Evaluation of technological and innovative capabilities of motorcycles and parts cluster’, a Progress report submitted to National Science and Technology Agency,12 September. Katikarn, C. (2006), Personal communication, 21 August. Lall, S. and Mortimore, M. (2000), ‘Competitiveness, restructuring and FDI: an analytical framework’, in The Competitiveness Challenge: Transnational Corporations and Industrial Restructuring in Developing Countries, United Nations Conference on Trade and Development, Geneva.
The Thai and Vietnamese motorcycle industries
231
Malerba, F. (2002), ‘Sectoral systems of innovation and production’, Research Policy 31(2), pp. 247–264. National Metal and Materials Centre (2006), retrieved from http://www.mtec. or.th. National Science and Technology Policy Committee (2004), Science and Technology Strategic Plan 2004–2013, National Science and Technology Development Agency (NSTDA), Bangkok. Ohara, Moriki (2006a), Interfirm Relations under Late Industrialization in China: The Supplier System in the Motorcycle Industry, IDE Occasional Paper Series No. 40, Institute of Developing Economies, Chiba. Ohara, Moriki (2006b), “Nihon no nirin kanseisha kigyo: attoteki yuui no keisei to kaigai shinshutu” [Japanese motorcycle companies: formation of overwhelming dominance and advancement abroad], in Sato, Yuri and Ohara, Moriki (eds) Ajia no nirinsha sangyo [Motorcycle industry in Asia], Institute of Developing Economies, Chiba (in Japanese). Ohara, Moriki, Tian, Fenglun and Lin, Hong (2003), “The overseas moves of Chinese local manufacturers: the export and FDI of Chongqing’s motorcycle makers in Vietnam”, in Ohara, Moriki (ed.), The Growing Importance of China and the Machinery-Related Industries, Institute of Developing Economies, Chiba. Otahara, Jun (2006), “Nihon no nirinsha buhin sapuraiya: bungyo kozo to torihiki kankei’ [Japanese motorcycle parts suppliers: structure of the division of labour and transaction relationships], in Sato, Yuri and Ohara, Moriki (eds) Ajia no nirinsha sangyo [Motorcycle industry in Asia], Institute of Developing Economies, Chiba (in Japanese). Thai Automotive Institute (2006), retrieved 4 July 2006 from http://www.thaiauto. or.th. Thai German Institute (2006), retrieved from http://www.tgi.or.th/index.php. UNCTAD (2001), World Investment Report: Promoting Linkages, United Nations Conference on Trade and Development, New York and Geneva. World Bank (2006), retrieved from http://devdata.worldbank.org/external/ CPProfile.asp?PTYPE=CP&CCODE=THA.
8.
‘Low-tech’ industry: a new path for development? The case of the salmon farming industry in Chile Michiko Iizuka
1.
INTRODUCTION
Conventional distinctions between ‘low’- and ‘high’-tech sectors have been closely linked to a sectoral ‘ordering’ in accordance with technological intensity instead of innovation intensity.1 The emphasis placed on ‘technological intensity’ is deeply rooted in theoretical considerations that go back to the early economic growth theories of the post-Second World War era (such as Singer, 1950; Prebisch, 1962). Several changes have taken place at the macro-level as far as the discussion of ‘high’- and ‘low’-tech sectors is concerned. First, the boundary between ‘low’- and ‘high’-tech sectors became more blurred owing to the advent of newly emerging technologies – such as biotechnology and ICT – which had impacts on both high- and low-tech sectors. Second, sectoral dynamism can be observed in ‘low’-tech sectors as well as ‘high’-tech sectors, for example with evidence of growing backward and forward linkages (de Ferranti et al., 2002; Kuwayama and Duran, 2003; Pietrobelli and Rabellotti, 2004). Third, the knowledge base required in ‘low-tech’ sectors – although different from that of ‘high-tech’ – equally demands high levels of capability, particularly collective capabilities, to employ the existing heterogeneous knowledge domains by combining them flexibly (HirschKreinsen, 2004; Mendonça and von Tunzelmann, 2004; von Tunzelmann and Acha, 2005) to stay abreast of global competition. Furthermore, the existence of collective capability at the sectoral level seems to have particular relevance to developing countries in relation to governance – such as negotiating institutional frameworks (rules and standards) – in global markets (Raynolds, 2004).
232
The case of the salmon farming industry in Chile
2.
2.1.
233
THEORETICAL FRAMEWORK: SECTORAL SYSTEMS OF INNOVATION FOR ‘LOW-TECH’ SECTORS WITH REFERENCE TO DEVELOPING COUNTRIES Sectoral Systems of Innovation
Malerba (2002, 2004) defines a sectoral system of innovation (SSI) as “a set of new and established products for specific uses and the set of agents carrying out market and non-market interactions for the creation, production and sale of those products” (Malerba, 2002, p. 250). The SSI locates the system in relation to the product instead of the technology employed, hence incorporating market demand centrally via the product structure. The building blocks of a sectoral system of innovation are: (1) its knowledge and technology base; (2) actors and agents; and (3) institutions (Malerba, 2004). The SSI lens allows the dynamics of sectors to be analysed by underscoring the roles of (1) structure and boundaries of sectors, (2) relevant agents and their interactions, (3) learning and innovation processes, and (4) types of sectoral transformation, with these blocks being combined in different ways from one sector to another. The SSI emerges from the evolutionary theory of innovation that emphasizes the differences in opportunities and conditions in which agents operate, since “the learning, behaviour and capabilities of agents are constrained and “bounded” by the technology, knowledge base and institutional context in which firms act” (Malerba, 2004, p. 15). However, the sectoral focus through activities that emerges from products may be able to extend the technology and knowledge base beyond its given ‘boundary’ through networks. The understanding of such inner mechanisms through a sectoral lens should facilitate the formulation and implementation of effective policy measures (Malerba, 2004). 2.2.
Innovation in ‘Low-Tech’ Sectors
The conventional idea of ‘low-tech’ sectors as non-innovative is currently undergoing major revision. Evidence from a range of developed countries (e.g. Maskell, 1998; Hirsch-Kreinsen, 2004; Mendonça and von Tunzelmann, 2004; Hirsch-Kreinsen et al., 2005; von Tunzelmann and Acha, 2005) has demonstrated widespread innovation ability in many if not most of the low-tech sectors. These findings suggest that innovation in ‘low-tech’ sectors is possible, contrary to conventional views (Mendonça and von Tunzelmann, 2004). The discussion of innovation in ‘low-tech’ industry emphasizes the importance of networks in sourcing external
234
Sectoral systems of innovation and production
knowledge bases through encompassing actors who are independent of each other and who often come from different branches of trade and technology fields (Hirsch-Kreinsen, 2004; Mendonça and von Tunzelmann, 2004; von Tunzelmann and Acha, 2005). In other words, innovation in ‘low-tech’ sectors combines the technology, which is often created outside the product’s technological domain, in ways that best suit their purpose to penetrate the market. More specifically, the improvement in the sector takes place with the process part of production through inputs, supplies and services. The overall upgrading requires actively extending the knowledge base through interactions and involving new agents, which may subsequently redefine the sectoral boundaries. The innovation process in ‘low-tech’ sectors is also characterized differently from that in ‘high-tech’. Hirsch-Kreinsen (2004) differentiated innovations by type of product: (1) simple products (such as simple electric elements for industrial applications with no significant technological advance); (2) niche products (e.g. fashion-oriented design products); and (3) simple products based on high-tech processing (e.g. paper, food and drink, and agricultural products). Hirsch-Kreinsen argues that the type of innovation that is employed by ‘low-tech’ sectors is characterized as “incremental” and particularly “architectural”, but less as “radical”, innovation, following the definitions of Henderson and Clark (1990). This underlines the importance of ‘low-tech’ sectors (such as agriculture) linking with ‘high-tech’ sectors (such as biotechnology) to innovate by combining the existing knowledge and technology best suited to a set of given conditions. Such innovation may feed back into creating new products and processes, conceptually resembling the ‘reverse product cycle’ claimed for the service sector (Barras, 1986) as well as the OEM–ODM–OBM model in the manufacturing sector (Hobday, 1995). 2.3.
‘Low-Tech’ Sector Development Strategy for Developing Countries
The new focus on ‘low-tech’ sectors has major policy relevance to developing countries, particularly in the natural-resource-rich countries, despite the above arguments being made based on evidence presented in developed countries. The argument of an alternative development path through ‘low-tech’ sectors is especially pertinent to the present-day context where the conventional sectoral ordering by technological intensity is blurred. Perez (2007) makes an interesting proposal for the developing countries to enhance their competitiveness through “market hyper-segmentation” – creating market niches for the developing countries to compete in the global market. She categorizes the products in two dimensions to create various gradations of market niches, based on a theory-driven comparison
The case of the salmon farming industry in Chile
235
between the Fordist mass production paradigm and the flexible specialization paradigm (Piore and Sabel, 1984). The horizontal dimension is distinguished by the targeted market: a price competitive market for mass-produced products as against a niche market for flexibly adaptable products. The vertical dimension is distinguished by the type of products: standardized as against special. Here, Perez emphasizes ‘special quality’, which I group as innovation intensity that includes technological intensity together with brands, standards and patents. The uniqueness in her approach is the emphasis placed on overall strategy in finding market niches rather than just on increasing technological intensity, which often is a challenge for developing countries to achieve in the short run. The ultimate aim of her strategy is to increase the possibility of penetrating and persevering in the market through identifying one’s ‘core competence’ more holistically, going beyond conventional ‘technological intensity’ (Prahalad and Hamel, 1990). Drawing attention away from focusing only on increasing ‘technology intensity’ to a strategy of niche finding opens the door to opportunities for many developing countries with rich natural resources but less technological capability. However, how developing countries can ‘interact’ with global markets to find their ‘core competence’ in the market and develop such competence – enhancing capabilities – still remains unclear. Figure 8.1 combines the market segmentation approach of Perez (2007) for developing countries, the ‘low-tech’ product typology of HirschKreinsen (2004) and the type of innovation according to Henderson and Clark (1990), with some modifications. The quadrants (1), (2) and (3) are the domains that mainly concern ‘low-tech’ sectors in developing countries. It is considered that, once a sufficient number of agents interact within the framework to define a sector, then the collective capability –through either private or public effort – would be equipped to search for the market niche. The process of selection does not necessarily involve the enhancing technological intensity but diversity in combination – ‘architectural innovation’. However, to enable such activities, the capabilities at sectoral level need to be enhanced. 2.4.
Governance in the Global Market and its Implication for Developing Countries
The difficulties in obtaining capabilities – particularly technological – in developing countries have been discussed in the past (Lall, 1992; Bell and Pavitt, 1993; Kim, 1998). Recent studies of globalization and the global division of knowledge creation (Lundvall and Johnson, 1994; Ernst, 2001; Cantwell and Iammarino, 2003) add yet another dimension, emphasizing
236
Sectoral systems of innovation and production Degree of innovation intensity in process
HIGH
LOW
High profitability due to SPECIAL QUALITIES Markets protected by INNOVATION, TECHNOLOGY, PATENTS and BRANDS Profitability attained through VOLUME Markets protected through LOW-COST AND STANDARDIZED QUALITIES
(3) Simple product and highly innovative process
(4) Complex product and highly innovative process
ARCHITECTURAL AND SOME RADICAL INNOVATION
RADICAL AND ARCHITECTURAL INNOVATION
(1) Simple product
(2) Niche, design and flexible product
INCREMENTAL INNOVATION
INCREMENTAL AND ARCHITECTURAL INNOVATION
PRICE COMPETITION Advantages in COST
LESS/LOW
COMPETITION IN ADAPTABILITY Advantages in client access, flexibility and rapid response
MORE /HIGH
Degree of product differentiation/Market niche
Examples: (1) paper; (2) customized cut and coloured paper; (3) non-chemical bleached paper, paper with sustainable forest certificate such as FSC; (4) biotechnologically engineered materialbased paper for special use. Source: Based on Henderson and Clark (1990), Hirsch-Kreinsen (2004) and Perez (2007), with some modification by the author.
Figure 8.1
Low-tech sector innovation typology
the differences in the way knowledge is created. Recent studies allocate a greater role to local knowledge and require different competences in developing countries, so that knowledge flows are both ‘bottom-up’ and ‘topdown’ (Iammarino, 2005). In other words, as Perez (2007) rightly pointed out, the attainment of competence and learning emerging from a strategy to identify ‘core’ competence is increasingly gaining relevance. In governance – the collective decision-making process (Jessop, 1996; Rhodes, 1996; Stoker, 1998; von Tunzelmann, 2003) – it is important to understand who is involved and how such decisions are taken. The participation in such a decision-making process is influenced by the existing knowledge/ technological base, actors and agents, and institutions which constitute a sector. This brings out the importance of governance in determining sectoral boundaries; furthermore, it demonstrates the particular issue in dealing with developing countries, since many developing countries are
The case of the salmon farming industry in Chile
237
often being ‘governed’ as the disadvantaged partner in decision-making processes at the global level. For instance, developing countries can upgrade in technological content of what is produced but they can still be ‘governed’ by developed partners in how products should be produced. The importance for developing countries to establish capacity to negotiate – collective capability – emerges in this context. This is particularly true in determining the institutional frameworks at the global level. An obvious way to observe how governance is exercised in specific institutional frameworks would be to look at the case of private international standards. Various authors have mentioned that globalization has brought rule setting under the control of the private sector (Clapp, 1998; Cutler et al., 1999; Nadvi and Waltring, 2004; Vandergeest, 2007), which has made the local institutional capacity more important than ever in terms of both product differentiation and branding through incorporating local conditions into the standard-setting processes (Ponte, 2002; Vandergeest, 2007). The case of the salmon industry is considered in the following section. The emergence of the salmon farming industry in Chile demonstrates the creation of a sectoral system and, in particular, its transformation through its attempts to penetrate and persevere in the global market. The case is also of interest in demonstrating how ‘low-tech’ products – food and so on – can actually be innovative through connecting with varieties of agents. Furthermore, this case, through looking at standards setting, would demonstrate whether developing countries like Chile could take an active role in governance structure.
3. 3.1.
EMPIRICAL CASE STUDY OF THE CHILEAN SALMON INDUSTRY Emergence of the Salmon Farming Industry and Current Realities at the Macro-Level
The development of salmon farming techniques in the 1970s totally changed the key producers of salmon at the global level (Table 8.1). In the 1980s, extractive fishery produced more than 99 per cent of the salmon consumed worldwide (Eagle et al., 2004). Today, extractive fisheries contribute only about 40 per cent of the world’s salmon, with the rest originating from farms situated along the coast of Norway, Chile, Scotland, Canada and other countries. During the transition to fish farming, Chile advanced from non-exporter of salmon to one of the top exporters in about two decades.
238
Sectoral systems of innovation and production
Table 8.1
1 2 3 4 5 Note:
Transition of ranking in salmon production in volume by country, 1981–2006 1981
1991
2001
2006
USA Japan USSR Canada Norway
USA Japan Norway Canada UK
Norway USA Chile Japan Canada
Norway Chile USA Japan Canada
USA includes Alaska.
Source:
Based on data from Bjorndal (2002) and SalmonChile (2007). 700 600
thousand ton
500
Norway Chile
400
UK 300
Canada Faroe Islands
200 100
Source:
06 20
04 20
00
02 20
20
98 19
96 19
92
94 19
SalmonChile (2007).
Figure 8.2
3.2.
19
19
90
0
Export volume of farmed salmon in 1990–2002 by major countries
The Global Position of the Chilean Salmon Farming Industry
The Chilean salmon farming industry has demonstrated strong export growth since its commercial establishment in the mid-1980s. In 2006, this industry exported approximately 628 000 tons and earned US$2 billion, making it the top exporter of farmed salmon in the world after Norway (SalmonChile, 2007). The Chilean contribution to the world supply of salmon has increased tremendously in the past ten years (Figure 8.2). As compared to the 1980s, farmed salmon currently has 70 per cent of total
Source:
239
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
9 8 7 6 5 4 3 2 1 0 1980s
US$Fob/kg
The case of the salmon farming industry in Chile
Elaborated from data from Technopress in various years.
Figure 8.3
Transition in the price of Chilean salmon from the 1980s to 2005
production in the market. It is worth mentioning that half of that, 35 per cent, is produced in Chile. Despite the high growth of production, the increase in supply resulted in a price fall. The average price per kilogram of salmon produced in Chile has dropped since the 1980s (Figure 8.3). This seems to suggest that this industry also follows the typical tendency of a price decline for standardized food commodities. The increasing price competition speeded up the merger/acquisition process to increase effectiveness and raised the concentration of salmon farming firms at the global level. The firms in Chile were no exception to this trend in concentration. 3.3.
Structural Changes in the Salmon Farming Industry in Chile: Catching Up
Several changes took place in the Chilean salmon farming industry owing to intensification of global competition. These changes are in line with the sectoral system of innovation in redefining the sectoral boundaries. These changes concern the agents and actors, knowledge base and institutions which brought the sector’s redefinition. It is noteworthy that the impact of change emerges from the interaction with others that goes beyond the national boundaries through the market, addressing the relevance of the sectoral innovation system.
240
Sectoral systems of innovation and production
Here the changes in the sector are explained to understand how new actors and agents are involved as this sector interacted with the market and transformed, leading to a redefinition of the sectoral boundary and more importantly to an increase in the capability at sectoral level. Many factors are interlinked; however, these are explained in the following order: (1) a concentration of industries, creating a margin for innovation processes; (2) an increase in value-added products, inducing development of forward linkages; (3) an increase in input suppliers, inducing development of forward linkages; (4) an increase in the variety of products, raising the scope for technical progress; and (5) enhancing collective capability in the industry to establish and comply with standards, allowing imperfect competition and making the product more elastic to increases in income and less elastic to declines in price.
70
6 000
60
5 000
50
4 000
40 3 000 30 2 000
20
net tonnes
Number of firms
Concentration of industries: creating a margin for innovation processes At the start of this industry in the 1980s, number of salmon farming firms had increased drastically. However, as the industry is exposed to international competition, these firms are required to make adjustment in order to meet the price competition. To increase cost efficiency, the concentration of firms in the salmon industry (farming phase) has grown since 1992. The number of firms in the farming business decreased from 63 in 1992 to 40 in 1999, while the average production volume per firm increased from 790 tonnes to almost 5500 tonnes in 1999 (Figure 8.4). Despite the fact that foreign direct investment is one of the driving forces
Number of firms Average production per firm in volumes
1 000
10 0
19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99
0
Source:
Based on data from Technopress, Revista Aquanoticias 50.
Figure 8.4
Changes in the number of firms in production and average production per firm in volume in 1992–1999
The case of the salmon farming industry in Chile
Table 8.2
241
Share of exports in value by the capital ownership of firms in Chile, 2002 Percentage of total exports
Canada Japan Spain Holland Norway Chile Total Source:
1 5 4 12 16 62 100 SalmonChile, (2003).
in the concentration of firms in salmon farming, Chilean capital still plays a key role, as indicated by the share in terms of export value (Table 8.2). As far as current entrance and exit of firms in this sector are concerned, there are some evolutional transitions depending on the types and maturity of firms in the sectors. For instance, since 2000, merger and acquisition have taken place among already well-established domestic firms and foreign firms, while the entrance of new domestic firms is observed in emerging suppliers of inputs and services. The entrance to the current ‘established segment’ of the sector – fish farming – by both domestic and foreign firms was at its peak during the 1990s. Increase in value added in products: development of forward linkages At the initial phase of its production, Chilean salmon exports were dominated by non-processed products called HG, salmon with head and internal organs taken out. However, since the 1990s, more value-added products such as fillet, salted, conserved and smoked salmon are being produced. Currently, the processed product (non-HG) constitutes 60 per cent of total production (see Figure 8.5). In other words, individual firms intend to move toward diversification of products through applying minor changes in technology to move away from the simple product for price competition. The increase in production, which created economies of scale, also allowed further input suppliers to emerge. There are no official statistics to gauge the increase in suppliers for salmon farming; however, an annually produced directory for aquaculture lists firms providing inputs and services. The number of supplier firms listed in this directory increased dramatically from 75 in 1993 to 461 in 2003. The proportion of those with legal residence in Region X2 increased from around 20 per cent in 1993
242
Sectoral systems of innovation and production
millons of dollars FOB
1200 Other products
1000
Conserved 800
Smoked Dried and salted
600
Frozen fillet
400
Fresh fillet 200
Non processed
0 1990 Source:
1993
1996
1999
2002
Elaborated based on SalmonChile (2003).
Figure 8.5 Transition of production in value by type of value-added product Table 8.3
Increase in the supplier firms for aquaculture
Years
National
Of which foreign
Of which Region X
Percentage foreign
Percentage Region X
1993 1998 2003
75 240 461
20 35 96
14 119 228
26.7 14.6 20.8
18.7 49.6 49.5
Note: Source:
Region X denotes those firms with legal residence in Region X Technopress (1993, 1998, 2003).
to almost 50 per cent from 1998 onwards (Table 8.3). The sheer number of foreign firms for supply increased, though actually decreasing in total share. Region X is currently diversifying its activities to other products such as shellfish and algae. It seems possible to say that having such competitive suppliers within the region helped related aquaculture to develop and enhance its competitiveness. Compared with the early period of its establishment, when the industry was almost self-sufficient in many of the inputs, specialization and progress of technology brought the region a cluster of industries dependent on exporting salmon firms. There were also attempts made to increase and enhance the domestic production of inputs such as salmon eggs. Table 8.4 shows the varieties of inputs produced. This also demonstrates the diverse varieties of technologies – such as metallurgy, chemistry (plastics),
243
Goods
Source:
Montero (2004).
Imported Automatic feeders, Genetic services. computers, oxygen system, machine to count the eggs and alevins.
Maritime transport (Trucks, tractors, ship and well boat), maintenance of cages, nets, harvesting services, veterinary services (vaccine), pathology. Automatic feeders, Laboratory computers, sensors, services. underwater cameras, nets (for some), pigments and medicine (vaccines). Fish feed, cages and buoys, nets, medicines (vaccine, antibiotics, immune depressors), pigments, smolts, ultrasonic, iodine.
Maritime and land transport (trucks, tractors and ships), maintenance of cages and nets, veterinary services.
Services
Goods
Cultivation phase
Services
Fresh water phase
Principal supplies to the salmon farming industry by phase
Domestic Fish feed, tanks, nets, buoys, cages, eggs, iodine, some simple machinery.
Table 8.4
Transport (trucks, ships, air), traders, clearing services. Salmon and trout, packaging materials (plastic bags, aluminium-coated trays, polythylene trays etc.), salt, sugar, detergents, iodized soaps, charcoal. Cutting machineries, skinning machines, smoking machine replacement knives, parts, detergents, injectors.
Transport, trading, marketing, retailing.
Services
Goods
Processing stage
244
Sectoral systems of innovation and production
basic machinery and biotechnology – that need to be developed to enhance the competitiveness of this industry. The increase in processing of the product through more value added was accompanied by salmon farming firms investing in and integrating the independent processing plants. This, therefore, led to further structural transformation of the industry. The value chain of products and agents and actors involved was extended. New agents and actors to the industry – the operational allies, suppliers – within the product value chain added different dynamics to the type of interaction from before, which had been dominated mainly by competitors. This demonstrates that ‘architectural innovation’ is taking place with ‘incremental innovation’. Skill development The overall figures on the educational level of the labour force in the Chilean salmon industry demonstrate that this sector is still a lowskilled, labour-intensive industry, as 52 per cent of its labour are without completion of basic education (SalmonChile, 2007). However, the skills required by the salmon industry have changed drastically owing to the structural changes that have taken place in the past few years. This is reflected in the emergence of various local technical schools and universities with specific courses on salmon farming and related areas of study such as marketing and sanitation. Furthermore, a fieldwork interview revealed a labour-intensive processing plant owner stating that “We need a labour force that has finished secondary education to operate the machines to meet the standards at a global level, (Interview, 2004). However, such needs created by the industry are not easily met by the labour market owing to the plants being located in remote regions of the country where demands for such educated labour have been low. To solve this problem, some plants are introducing night and weekend schools for their labour so that they can at least obtain secondary school certificates. Conclusion The salmon farming firms’ specialization in the core activities of their business – such as fish rearing, processing and exporting – accompanies the increased use of suppliers. The intensification of core activities with externalization of auxiliary activities to third parties created different up- and downstream agents into the cluster. Furthermore, the skills required for the industry also increased. This has brought a big change in the culture of the firms in the short term, as many preferred to be independent and self-reliant in many of their supplies of inputs, and opened a new dimension for the potential technological and knowledge base to be developed
The case of the salmon farming industry in Chile
245
in the future. Furthermore, now the firms became able to ‘tap’ varieties of technology and the knowledge base that exist either upstream (vaccines and pharmaceutical products for feeds) or downstream (packages, processing machinery and materials for nets) of their activities. This is well in line with the conceptual model presented in Figure 8.1. It must be reiterated that structural transformation of the industry influenced the sectoral boundary through design, variety and knowledge content of agents included in the ‘sector’.
4.
4.1.
APPLICATION OF INSTITUTION: STANDARDS COMPLIANCE IN THE SALMON FARMING INDUSTRY AND ITS IMPACT ON THE SECTORAL BOUNDARY Standards Compliance for Developing Countries in the Global Context
Standards compliance for food exporting industries is important in gaining access to high-end markets as well as differentiating their product to create value by reflecting consumer demand. In addition to that, globalization transformed the way rules and standards are determined (e.g. Cutler et al., 1999). Keeping abreast of new global standardization gives the product a quality edge and can prevent producers falling into the trap of declining commodity prices, and increases productivity in value terms. The compliance with the standards, therefore, is conventionally considered to be stimulated through firms having exposure to the market through exports or through foreign direct investment. Indeed, the pattern of certification for ISO 9000 and ISO 14000 – the most representative and popular international management standards – in South American countries demonstrated a higher compliance rate in the export-oriented sectors, such as natural resource-based sectors. Moreover, further exploration of compliance level in relation to other economic factors applying econometric analysis suggested that internationalization and the economic development level of the country are the possible factors affecting the diffusion of the international management standards ISO 9000 and ISO 14000 (Freitas and Iizuka, 2007). This section intends to demonstrate the inner mechanism of standard compliance to illustrate how institution’s standards will change the boundaries of the sector and type of interaction among the actors and agents, taking the case of the Chilean salmon farming industry to describe how institution can also shape the sector by involving different actors
246
Sectoral systems of innovation and production
and agents and creating different interactions at the sectoral level, which enable a collective decision of the future trajectories through determining the technological and knowledge domains to be included. 4.2. Standards in the Chilean Salmon Industry: A Historical Perspective In the Chilean salmon farming industry, there have been several initiatives to control the quality of the product to enhance international competitiveness since the mid-1980s. The first attempt to develop a local quality standard for Chilean salmon was undertaken by the private sector. In 1987, the Association,3 with the technical cooperation of the Fundacion Chile, the privately run institution with the public purpose of promoting technological transfer, created the local private standards called ‘quality seal’ (sello de calidad) and ‘code of good practice’ (codigo de buenas practicas). The ‘quality seal’ outlined the sanitary procedures for the fish processing plant for exports. The Association monitored and controlled this certification, enforcing the compliance of all exporting members. In this way, the Association controlled and differentiated the quality of the products of member firms from that of others. This effort of the Association is considered to have contributed to the success in enhancing competitiveness (Perez-Aleman, 2005). The public sector followed this private initiative. In particular, in 1985, the National Fishery Service (Servicio Nacional de Pesca: SERNAP, later SERNAPESCA) started developing the guideline POS (Procedimiento Operacion de Saneamiento: Sanitary Operation Procedure), based on the international standard HACCP (Hazard Analysis and Critical Control Point), for fish processing plants. From the mid-1990s, SERNAP required all the exporting processing plants to certify to the HACCP standard. SERNAPESCA also started to monitor and regulate this standard (PAC: Programa asegrameinto de Calidad, hereafter referred to as HACCP-PP) (interview SERNAPESCA, 2004). The introduction of HACCP-PP by SERNAP in the mid-1990s replaced the privately initiated “quality seal” (Alvial, 2005). In the early 2000s, HACCP was extended to the fish rearing farms (hereafter referred to as HACCP-CC) (interview SERNAPESCA, 2004). The involvement of the public sector in the standardization of quality standards control and in the certification ensured transparency in the certification system and consequently the wider diffusion of standards. These efforts permitted firms to reduce transaction costs in searching for the information, ensured technical assistance and gave firms international credibility, over and above supporting them technically in de-codifying the
The case of the salmon farming industry in Chile
247
standards to the specificity of the sector and local context and re-codifying to facilitate further the compliance. Moreover, owing to the interdependency of activities – upstream and downstream – the diffusion in usage of standards was crucial in achieving cost efficiency. In the 2000s, international environmental concerns increased, and some Chilean salmon producers started obtaining ISO 14001 certification to comply with global demands. Aiming at responding to such market demand as well as local needs in meeting environmental sustainability, various local standards and regulations emerged. For instance, in 2002, the Association and the public sector4 developed the protocol Cleaner Production Agreement (Acuerdo de Produccion Limpia: APL) to ensure that Association members would meet the target agreed on environmental issues. From 2004, firms which participated and complied with the target set by the APL were given the APL certificate. This collaboration between industry and the public sector towards the setting of national environmental standards ensured transparency in monitoring and regulation (interview with SalmonChile, 2004). Here, the link between public sector and Association was already made owing to the previous interactions on the quality seal. In addition to APL, the Association was involved in the process of establishing sectoral regulations on the environment (RAMA in 2001) and on sanitation (RESA in 2002). In 2003, the Association itself created SIGes (Sistema Integrada de Gestions: Integrated Management System) as a voluntary scheme that incorporates several standards. This is designed to facilitate compliance with all the important standards used within the salmon industry.5 Consequently, SIGes is expected to demonstrate a signal of the firms’ engagement on compliance with international standards, such as ISO 9000, ISO 14001 and OHSAS 18001.6 This example demonstrates that the Association has gone a step further from the public sector and located itself closer to market needs, to meet the global competition. 4.3.
Factors that Make Firms Comply with Standards: Statistical Analysis Based on the Survey
The statistical analysis based on the fieldwork survey of the Chilean salmon industry7 illustrates an interesting picture in terms of what factors enabled these firms to comply with the standards. The analysis used factors that are considered to contribute to compliance with international mangement standards such as ISO 9000 and ISO 14000. These factors are: an openness factor (variables such as share of export and investment), an absorptive capacity factor (variables such as experience, size of sales, past experience and share of investment in R&D for compliance) and a
248
Sectoral systems of innovation and production
collaborative factor (variables such as membership of the Association, collaboration with suppliers and collaboration with clients). Ordinal and binary logit models are computed for the ISO standards8 (ISO 9000, ISO 14000) variable categorical level of compliance9 and for the dummy variable certification with international standards. Results suggest that compliance with ISO 9000 is more likely for firms that are members of the Association. Looking instead at the level of compliance, it was found that the level of compliance with ISO 9000 is expected to be higher for firms with larger sales, in other words larger firms. The level of compliance with ISO 14000 is also expected to be higher for firms that are members of the Association as well as for firms that collaborate with suppliers in obtaining certification. The above empirical evidence demonstrates that the firms’ level of compliance with international standards might be a consequence not only of the interaction with global buyers but also of the sectoral efforts to improve the quality, safety and environmental standards of salmon production to differentiate its products. 4.4.
Collective Capabilities and the Role of the Association of the Salmon Industry in Chile
Both historical and statistical observation seems to suggest that the Association of the Salmon Industry in Chile plays an important role as a coordinating agency for institution building, negotiation and eventually shaping the sectoral strategy through alignment of interests among actors and agents. The qualitative evidence also demonstrates that the Association played an active role in the establishment of regulations specific to the aquaculture sector, collaborating closely with the government. As mentioned earlier, the Association was involved in rule setting on the environment and sanitation for aquaculture (RAMA and RESA). The Cleaner Production Agreement (Acuerdo de Produccion Limpia: APL) is another example of such partnership. This demonstrated that the Association was capable not only of bringing firms together to engage in voluntary setting of their own standards but also of monitoring those who subscribed to this agreement. The Association also participated in the National Commission for Aquaculture (Comision Nacional de Acuicultura) and the elaboration process of the National Aquaculture Policy (Politica Nacional de Acuicultura en Chile: PNAC) in 2004 with the Under-secretary of Fisheries (Subsecretaria de Pesca, 2003). Again, the presence of the Association in formulating future directions for this industry was considered crucial. Apart from these interactions with the public sector, each exporting member firm of the Association is complying with the firm-specific standards as it does business with global buyers
The case of the salmon farming industry in Chile
249
from Japan, the USA and the EU. In this context, it is possible to consider that the Association had a wide range of information in terms of what kind of platform standards are needed for salmon products. 4.5.
New Development: Collective Capability
The local standards set up by the Association, SIGes, are considered as successful. This is a local set of standards that tries to encompass all the relevant standards for this industry. The attempt is not new to the industry, as it has already experienced such work with the quality seal. Nevertheless, the crucial difference lies in the way it started to influence external standard-setting procedures. In 2004, standards based on SIGes were adopted as industry-wide standards among Chilean, Canadian and American salmon farming firms associated with SOTA (Salmon of the Americas), formally qualified as Safe Quality Food (SQF-SOTA). In other words, the Chilean standards are potentially having an important influence on standard setting at the level of the American continent. Furthermore, SIGes is currently adopted by Wal-Mart as a standard for procurement for salmon (Aquanoticia, 2006). This suggests that the Association, through collaborating with various stakeholders in bringing standards compliance, became the active, path-finding institution capable of accessing various different sources of knowledge and coordinating, sometimes even negotiating with, different stakeholders to maintain a common platform of standards for agents involved. Figure 8.6 provides a conceptual map of how the Association is actually linked with the many different actors together with the collaborative rule-setting process. Here, SalmonChile is involved with various different stakeholders. It is involved with sector-specific government institutions such as SERNAPESCA for setting up regulations on the environment (RAMA) and sanitation (RESA). It is involved with private NGOs such as Fundacion Chile to established codes of good practice (codigo). It is involved with the public sector in regulating environmental issues related to aquaculture to implement the Cleaner Production Agreement (APL). Also, it needs to be mentioned that member firms of the Association interact with the private sector – global buyers – in setting up their own product specifications. The SIGes, created by SalmonChile, incorporates much of what the industry needs to comply, and owing to this platform nature it was possible to influence the standards that apply to American and Canadian producers. In other words, the involvement and linking with different stakeholders actually enhanced the negotiation as the Association gained collective capability to mobilize knowledge necessary for its goal. This example shows that an attempt to comply with standards in an
250
Sectoral systems of innovation and production Haccp-cc, Haccp-pp
Association of Salmon Farmers in USA, Canada
Under-secretary of Fisheries, National Fishery Service Related fishery institution and private sector related to fishery Commission for Aquaculture RESA, RAMA
SalmonChile
Firm- SIGes Global specific buyers from standards Member firms Japan, USA, EU
Salmon of Americas SQF-SOTA
Fundacion Chile CODIGO
Cleaner Production Agreement APL Group of the public sector CONAMA RCA: related to the environmental environment qualification
Note: Names of projects are in italics, participants are in ordinary font; underlined italics are the names of standards.
Figure 8.6
Conceptual map of the Association (SalmonChile) as the interface of different stakeholders through standards: example of establishing regional standards, SQF-SOTA
interdependent sectoral framework enhanced the collective capability to negotiate on institutions with other stakeholders at the global level. At the same time, the example of standards demonstrated the importance of networks and the involvement of new actors in standards setting, which would eventually influence the definition of sectoral boundaries.
5. 5.1.
CONCLUSION Conclusion Drawn from the Case of the Salmon Farming Industry
The example of the salmon industry in Chile has demonstrated that ‘lowtech’ sectors can be innovative and that there is a potential alternative path for development through building up the collective capability at the sectoral level. The collective capability would be the key in negotiating and influencing the decision-making process for governance. This collective capability, in expressing a more proactive nature, distinguishes itself from collective efficiency (Schmitz, 1999). It would be the capability to negotiate through searching for new technology and a knowledge base for problem solving. The example of standards compliance demonstrated that
The case of the salmon farming industry in Chile
251
the Association was able to influence Canadian and American producers because it placed itself in the integrating node position of the numerous agents and actors that constitute the sector. Moreover, the collective capability – in its capacity to expand the network – is strongly associated with determining sectoral boundaries. In other words, with the establishment of sufficient collective capability, redefining the sectoral boundaries can be more strategic and dynamic in favour of developing countries. The start of the Chilean salmon farming industry benefited from technological progress at the global level, which was transferred through various means to local producers, and the good environmental conditions. However, these two conditions were not sufficient factors for its success in the export markets. As the descriptive data indicated, many changes had taken place: (1) improvements in value added with the development of forward linkages; (2) increasing variety of products from input suppliers with the development of backward linkages; (3) an increase in the diversity of products produced at the local level with the improvement in variety and the depth of technology and the knowledge base; (4) development of collective capability at the sector level to sustain competitiveness through establishing common institutions for learning, directing future investments and negotiating with external stakeholders. The Chilean salmon farming industry still has ample areas for future improvement, especially in terms of R&D, as Katz (2004) indicated, to sustain long-term competitiveness as well as local environmental sustainability. However, with the development of collective capability, it may be able to sustain its development in the future. Figure 8.7 is based on the case of the salmon farming industry. This demonstrates an example of steps that low-tech sectors in developing countries may be able to take. As demonstrated earlier, low-tech simple products have two strategies to improve their position in the market: process improvement and/or product differentiation. Both require new sets of partners, agents and actors, and it is considered that these new interactions will subsequently change the boundaries of the sector. In the case of the Chilean salmon industry, the majority of firms engaged in cumulative and architectural innovation owing to the structural changes that took place, which made them interdependent on each other – forcing them to perform as a cluster. At the same time, the development of collective capability in the salmon farming industry, as observed in the case of standards setting, enabled the priority areas of technological development to be identified, giving some clearer indication of possible future directions for innovation in the industry. For instance, currently some collaborative approaches are taking place in the development of innovation in vaccines, fish feeds and the system of transportation to enhance the overall
252
Sectoral systems of innovation and production Degree of innovation intensity in process
HIGH
LOW
High profitability due to SPECIAL QUALITIES Markets protected by INNOVATION, TECHNOLOGY, PATENTS and BRANDS Profitability attained through VOLUME Markets protected through LOW-COST AND STANDARDIZED QUALITIES
(3) Simple product and highly innovative process
(4) Complex product and highly innovative process
ARCHITECTURAL AND SOME RADICAL INNOVATION
RADICAL AND ARCHITECTURAL INNOVATION
(1) Simple product
(2) Niche, design and flexible product
INCREMENTAL INNOVATION
INCREMENTAL AND ARCHITECTURAL INNOVATION
PRICE COMPETITION Advantages in COST
LESS/LOW
COMPETITION IN ADAPTABILITY Advantages in client access, flexibility and rapid response
MORE /HIGH
Degree of product differentiation/ Market niche
Source:
Author.
Figure 8.7
Progression of strategy for the Chilean salmon industry
competitiveness of the industry. The radical innovation is more likely to happen in these newly incorporated areas of sector than the conventional areas when R&D is applied. The market-centred nature of SSI (Malerba 2002, 2004) would enable the capture of the sectoral dynamics and the reality of how a ‘low-tech’ sector can enhance its capabilities. The new technology and knowledge base which were incorporated by the sector through extending its network may influence the type of product the sector can produce. In other words, this casts a doubt on the point raised earlier, that of the sector being ‘bounded’ by existing technology and the existing knowledge base: ‘the learning, behaviour and capabilities of agents are constrained and “bounded” by the technology, knowledge base and institutional context in which firms act” (Malerba, 2004, p. 15). The SSI can offer much wider and more extensive analysis on the evolution of the ‘sector’ through looking at its structure, its knowledge base and the agents involved in constituting a sector. Moreover, as demonstrated throughout the discussion, the collective capability at the sectoral level plays an important role in fostering the power to negotiate, coordinate and determine the technological trajectories. The importance of this capability may be more pronounced for a
The case of the salmon farming industry in Chile
253
‘low-tech’ natural resource-based industry than for a high-tech industry, owing to the very wide areas of knowledge entailed in the industry.10 5.2.
Policy Implications
Catching up in the globalizing economy is not an easy task for many developing countries. The role of state and policy in nurturing the industrial and knowledge base of domestic firms is increasingly limited in the current context. For this reason, the present-day catching-up process needs to be viewed from a holistic perspective through sectoral boundaries going beyond the national ones, and the potentiality of the sector needs to be evaluated through the interaction with wider ranges of stakeholders. Many developing countries are endowed with natural resources – the sources of the ‘low-tech’ sector – yet some countries stay as commodity exporters while others intend to move to producing manufactured goods, such as simple computer chips. However, if a country does not have a market niche in what it is producing, the characteristics of commodity goods apply to any manufactured goods. In this context, instead of changing ‘what’ is produced – going up the technological intensity ladder for products – it is important to consider ‘how’ the ‘low-tech’ sector can also be an alternative path for development through enhancing the degree of innovation in the production process and identifying market niches.
ACKNOWLEDGEMENT I appreciate the stimulating exchange on ‘low-tech’ industry and comments on the draft from Professor Nick von Tunzelmann in drafting this document. I also appreciate comments and suggestions made by Professor Malerba in finalizing the draft of this chapter.
NOTES 1.
2.
For instance, the OECD uses technological classification in two different approaches: sector and product. The sector approach classifies industries according to their technological intensity, while the product approach classifies it according to finished products (ISIC). In both ways, the OECD classifies sectors and products rather similarly. Industries such as aerospace, pharmaceuticals, computers, photography and photocopying are considered as ‘high’-tech, while wood and paper, food, drink and tobacco are ‘low’-tech, on the grounds of conducting low levels of their own R&D and making low use of high-tech inputs. The 10th region of Chile currently produces 80 per cent of total farmed salmon in Chile.
254
3. 4. 5.
6. 7. 8. 9. 10.
Sectoral systems of innovation and production The offices of most of the salmon firms are also present in Puerto Montt, the capital of the region. The field survey was conducted in this region. The Association of Salmon and Trout Producers of Chile (APST) was established in 1986 by producers. In 2001, the membership was extended to suppliers and its name changed to the Association of the Salmon Industry in Chile (SalmonChile). The agreement was made between a national and industrial group of regulatory bodies for environmental issues on fishery and a group of firms. The SIGes includes the elements of: APL, RAMA, RESA, Code of Good Practice for the Environment, ISO 14000, ISO 9000, OHSAS 18000, Safe Quality Food (SQF), HACCP-PP, HACCP-CC and RCA (environmental qualification resolution). It is also currently used by Wal-Mart in its procurement of salmon. For the detail of the standards, please refer to the Appendix. The firms covered by the survey constitute 80 per cent of the total Salmon exports of Chile at that time. Only ISO standards are compared in this context, owing to their generic nature, which offered a larger number of samples. However, another analysis was done as well. See Freitas and Iizuka (2008). The levels of compliance are categorized as follows: (1) no compliance; (2) planning to comply; (3) in the process of compliance; (4) complied. These are introduced to clarify the differences among firms which have not complied with standards yet. Mendonça and von Tunzelmann (2004).
REFERENCES Aquanoticia (2006), ‘Wal-mart solicitara certificaciones Siges-SalmonChile y Siges SQF a todos sus proveedores chilenos de salmón’, Noticias, March 31, available at http://www.aqua.cl Alvial, Adolfo (2005), Voluntary Initiatives in the Salmon Industry in Chile: Role and Prospects, INTESAL, Santiago. Barras, R. (1986), “Towards a theory of innovation in services”, Research Policy 4, pp. 161–173. Bell, M. and Pavitt, K. (1993), “Technological accumulation and industrial growth: contrast between developed and developing countries”, Industrial and Corporate Change 2(2), pp. 157–210. Bjorndal, Trond (2002), “The competitiveness of the Chilean salmon aquaculture industry”, Aquaculture Economics and Management 6(1–2), pp. 97–116. Cantwell, John and Iammarino, Simona (2003), “Regional system of innovation and Europe and the globalisation of technology”, in Cantwell, John and Iammarino, Simona (eds), Multinational Corporations and European Regional System of Innovation, Routledge, London. Clapp, Jennifer (1998), “The privatization of global environmental governance: ISO 14000 and the developing world”, Global Governance 4, pp. 295–316. Cutler, C., Haufler, Virginia and Porter, Tony (eds) (1999), Private Authority and International affairs, University of New York Press, Albany. De Ferranti, David M (eds) (2002), From Natural Resources to the Knowledge Economy: Trade and Job Quality, World Bank, Washington, DC. Eagle, Josh, Naylor, Rosamond and Smith, Whitney (2004), “Why farm salmon out compete fishery salmon”, Marine Policy, 28, 259–70. Ernst, Dietor (2001), Global Production Networks and Industrial Upgrading: A Knowledge-Centred Approach, East–West Center, Honolulu, HI, p. 26.
The case of the salmon farming industry in Chile
255
Freitas, Isabel and Iizuka, Michiko (2007), “Can compliance to the international management standards enhance the competitiveness of firms in Latin America?”, Paper presented at GLOBELICS Russia, September. Freitas, Isabel and Iizuka, Michiko (2008), “Standards compliance as an alternative learning opportunity under globalization in Latin America”, SPRU Electronic Working Paper Series 172, University of Sussex, SPRU, Science and Technology Policy Research. Henderson, R. and Clark, K. (1990), “Architectural innovation: the reconfiguration of existing product technologies and the failure of established firms”, Administrative Science Quarterly 35(1), pp. 9–30. Hirsch-Kreinsen, H. (2004), “Low-technology: a forgotten sector in innovation policy”, Paper presented at International ProACT, Tampere, Finland. Hirsch-Kreinsen, H., Jacobson, David, Laestadius, Staffan and Smith, Keith (2005), “Low and medium technology industries in the knowledge economy: the analytical issues”, in Hirsch-Kreinsen, H., Jacobson, David, Laestadius, Staffan and Smith, Keith (eds), Low-Tech Innovation in the Knowledge Economy, Peter Lang, Frankfurt am Main, pp. 11–30. Hobday, Michael (1995), Innovation in East Asia: the challenge to Japan, Edward Elgar Publishing, Aldershot, UK and Brookfield, VT, USA. Iammarino, Simona (2005), “An evolutionary integrated view of regional system of innovation: concepts, measures and historical perspectives”, European Planning Studies 13(4), pp. 497–519. Jessop, B. (1996), “Partnership in Greater Manchester and the Thames Gateway”, paper for DoE Seminar. Katz, Jorge (2004), Economic, Institutional and Technological Forces Inducing the Successful Inception of Salmon Farming in Chile, DfID/World Bank, Washington, DC. Kim, Linsu (1998), “Crisis construction and organizational learning: capability building in catching up at Hyundai Motor”, Organization Science 9(4), pp. 506–521. Kuwayama, Mikio and Duran, Jose (2003), La Calidad de la Insercion Internacional de America Latina y el Caribe en el Comercio Mundial, Serie de Comercio Internacional, ECLAC, Santiago. Lall, S. (1992), “Technological Capabilities and Industrialization”, World Development, 20(2), 165–86. Lundvall, Bengt-Ake and Johnson, Bjorn (1994), “The learning economy”, Journal of Industry Studies 2(1), pp. 23–42. Malerba, Franco (2002), “Sectoral systems of innovation and production”, Research Policy 31, pp. 247–264. Malerba, Franco (2004), Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe, Cambridge University Press, Cambridge. Maskell, Peter (1998), ‘Learning in the village economy of Denmark: the role of institutions and policy in sustaining competitiveness’, in Braczyk, H.-J., Cooke, P. and Heidenreich, M. (eds), Regional Innovation Systems, London: Routledge, pp. 190–213. Mendonça, Sandro and von Tunzelmann, Nick (2004), ‘Brave old world: accounting for “high tech” knowledge in “low-tech” industries’, Paper presented at Industrial Dynamics, Innovation and Development, Denmark. Montero, Cecilia (2004), Formacion y Desarrollo de un Cluster Globalizado: El
256
Sectoral systems of innovation and production
Caso de la Industria del Salmon en Chile, ECLAC, United Nations, Santiago de Chile. Nadvi, Khalid and Waltring, Frank (2004) “Making sense of global standards”, in Schmitz, Hubert (ed.), Local Enterprises in the Global Economy: Issues of Governance and Upgrading, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA. Perez, Carlota (2007 forthcoming), ‘Respecialization and the deployment of the ICT paradigm: an essay on the present challenges of globalisation’, in Compano, R., Pascu, C., Bianchi, A., Burgalman, J.-C., Barrios, S., Uttrich, M. and Maghiros, I., (eds), The Future of the Information Society in Europe: Contribution to the Debate, Technical Report Series, Institute for Prospective Technological Studies, Seville. Perez-Aleman, Paola (2005), “Cluster formation, institutions and learning: the emergence of clusters and development in Chile”, Industrial and Corporate Change 14, pp. 651–677. Pietrobelli, Carlo and Rabellotti, Roberta (2004), Upgrading in Clusters and Value Chains in Latin America: Role of Policies, Inter-American Development Bank, Washington, DC. Piore, Michael and Sabel, Charles (1984), The Second Industrial Divide, Basic Books, New York. Ponte, S. (2002), “The “Latte Revolution”? Regulation, markets and consumption in the global coffee chain”, World Development 30(7), pp. 1099–1122. Prahalad, C.K. and Hamel, G. (1990), ‘The core competence of corporations’, Harvard Business Review, May–June 68(3), 79–87. Prebisch, Raul (1962), ‘The economic development of Latin America and its principal problems’, Economic Bulletin for Latin America 7, 1–22. Raynolds, L. (2004), “The globalization of organic agro-food networks”, World Development 32(5), pp. 725–743. Rhodes, R. (1996), “The new governance: governing without government”, Political Studies 44, pp. 652–667. SalmonChile (2003), ‘Proyecciones y oportunidades de desarrollo en la industria del salmon hacia el ano 2010’, Presentation material by president of SalmonChile, presented by Rodrigo Infante in Puerto Montt, April. SalmonChile (2007), www.salmonchile.cl. Schmitz, Hubert (1999), “Collective efficiency and increasing returns”, IDS Working Paper 50, IDS at University of Sussex. Singer, Hans W. (1950), ‘US foreign investment in underdeveloped areas: the distribution of gains between investing and borrowing countries’, American Economic Review 15, pp. 473–485. Stoker, G. (1998), “Governance as theory: five propositions”, International Social Science Journal 155, pp. 17–28. Subsecretaria de Pesca (2003), “Politica Nacional Acuacultura de Chile”, Valparaiso, Chile. Technopress, various years, Revista Aquanoticias. Technopress (1993), Compendium of Chilean Aquaculture and Fisheries; Directory of Chilean Aquaculture and Fisheries, Technopress, Santiago Chile. Technopress (1998), Compendium of Chilean Aquaculture and Fisheries; Directory of Chilean Aquaculture and Fisheries, Technopress, Santiago Chile. Technopress (2003), Compendium of Chilean Aquaculture and Fisheries; Directory of Chilean Aquaculture and Fisheries, Technopress, Santiago Chile.
The case of the salmon farming industry in Chile
257
Vandergeest, P. (2007), “Certification and communities: alternatives for regulating the environmental and social impacts of shrimp farming”, World Development 35(7), pp. 1152–1171. von Tunzelmann, Nick (2003), “Historical co-evolution of governance and technology in the industrial countries”, Structural Change and Economic Dynamics 14(4), pp. 365–384. von Tunzelmann, Nick and Acha, Virginia (2005), ‘Innovation in “low-tech” industries’, in Fagerberg, Jan, Mowery, David and Nelson, Richard (eds), The Oxford Handbook of Innovation, Oxford University Press, New York.
258
Sectoral systems of innovation and production
APPENDIX APL: Acuerdo de Produccion Limpia (Agreement for Cleaner Production). A local certificate that emerges from a voluntary scheme to meet cleaner production guidelines agreed between the industry and the public sector (local and national). Codigo: Codigo de buenas practicas (code of good practice). Local firmlevel standards. This is a code of good practice in written form for internal use in the firm. It could vary from firm to firm depending on the activity. The importance is in its written format. HACCP-CC: Hazard Analysis and Critical Control Point, a food safety methodology for fish cultivation centres. This was originally an international standard; however, the Chilean government adapted this standard to the national level, and the National Fishery Service (SERNAPESCA) diffuses the idea and monitors and regulates all of the farmed fish exported abroad. HACCP-PP: Same as HACCP-CC but for the fish-meat processing plant. ISO 9000: A global standard for quality management. ISO 14000: A global standard for environmental management. SIGes: Integrated Management System (Sistema Integrado de Gestion: SIGes). A local standard created by the Association of the Salmon Industry in Chile that tries to integrate the necessary standards both international and national, adapting to local conditions, with an intent to differentiate those firms that are in the process of compliance from the others. This includes a code of good conduct for sustainable aquaculture by Fundacion Chile. Currently this standard conforms to SQF (Safe Quality Food) standards with the Association of Salmon Farming in Canada and the USA. This is also currently used by Wal-Mart in its procurement of salmon in Chile.
9.
Making a technological catch-up in the capital goods industry: barriers and opportunities in the Korean case* Yoon-Zi Kim and Keun Lee
1.
INTRODUCTION
While a successful catch-up has the tendency to be an exception rather than a rule, an increasing recognition is that one of the most serious barriers and thus important paths to catch-up is to grow internationally competitive enterprises. However, this turns out to be even more difficult, as globalization tends to expose local firms to a fierce global competition even within their home economies. In light of this, the current chapter deals with the important question of how to grow local corporate capability, using the case studies of capital goods firms in Korea. While creating world-class giants from developing countries is almost impossible, Korea has generated several, such as Samsung, LG and Hyundai. However, these are all companies involved in the consumer goods industry. In capital goods industries, this is not the case. The weakness of the intermediate capital goods industry originates from the beginning of the Korean economic development in the 1960s. Korean economic growth has focused on the industry of final products while relying upon the imports of core parts, intermediate materials and supplies. The main source of these intermediate goods is Japan. However, Korea still experiences persistent trade deficits with Japan, given the structure of more exports of final products requiring more imports of intermediate goods from Japan. This chapter then focuses on the question of why making a catch-up is even more difficult in capital goods industries which are usually led by small or middle-sized companies. It relies upon the sectoral systems of innovation (Malerba, 2004) as a theoretical framework for analysis. The patterns of catch-up in the capital goods industry are quite different from those in the consumer goods industry because the firms in the former industry now deal with other client firms rather than consumers only. 259
260
Sectoral systems of innovation and production
Small firms in the capital goods industry are usually specialized suppliers to big final goods assembly firms in the consumer goods industry or other industries, and thus the tacit knowledge accumulated from the interface between the producer and the customer firms is very important. In this industry, important knowledge on production cannot simply be embodied in production equipment, and technical licensing alone cannot solve the problem of poor design capability in the product development stage (Lee and Lim, 2001). This chapter sheds light on the following three sources of difficulties of catch-up in the capital goods industry. First, in terms of the accumulation of tacit knowledge and marketing, a serious difficulty lies in the fact that local client firms are reluctant to use locally made capital goods, owing to their poor quality and low precision level. In this matter, even government policies that encourage the use of domestic products are not and cannot be effective. Second, a difficulty arises because incumbent firms often react by charging predatory prices upon news of the local development of capital goods by latecomer firms. Third, if the catch-up firms overcome this barrier, then the next strategy used by incumbent firms is often to charge latecomers with legal actions for patent violations. This chapter is organized as follows. Section 2 provides an overview of the capital goods industry in Korea, particularly the machine tools industry. Section 3 discusses the sectoral systems of innovation of machine tools and derives theoretically the intrinsic difficulty of catch-up in capital goods industries. Section 4 provides an in-depth elaboration of the three barriers to catch-up in machine tools. Then section 5 discusses available opportunities for slow but gradual catch-up in terms of the role of the government, market conditions, and technological changes or paradigm shifts. Section 6 concludes the paper with a summary and some remarks.
2.
IMPORTANCE OF THE CAPITAL GOODS INDUSTRY AND THE KOREAN EXPERIENCE
Adam Smith stated that the division of labour, especially the development of the intermediate goods industry, has aggrandized the wealth of a nation. The importance of intermediate or capital goods has continuously been emphasized by various researchers such as Young (1928), Stigler (1951), Romer (1990), and Rodríquez-Clare (1996). Particularly, Porter (1992) emphasized that only after the supplier of intermediate goods becomes domestic is it possible to utilize factors more effectively, more quickly and most efficiently. Meanwhile, the recent growth theory also pays more attention to the role of the intermediate goods industry in
Barriers and opportunities in the Korean case
261
understanding economic growth and technological development (Rodrik, 1996; Rodríquez-Clare, 1996). According to these theories, strong mutual dependence and causality exist between the final and intermediate goods industries, and thus the different degrees of the development of the intermediate goods industry may give birth to multiple equilibria in the growth path, causing either a vicious or a benign cycle (Rodrik, 1996; RodríquezClare, 1996). The idea is that the higher the expertise and diversity that the intermediate goods industry’s development results in, the more productive the final goods industry will be. In turn, this yields higher returns and eventually brings an increase in the demand of intermediate goods and improves the expertise and diversity of the intermediate goods industry. A repetition of this results in continual growth and high-technology equilibrium. In contrast, if the initial condition is not satisfied, or the establishment of a structure in industry linkage over some critical point is not reached, low-technology equilibrium or the underdevelopment may result. Rodrik (1996) pointed out that, although middle-income countries with a high quality of human resource have the potential to transit from a low-technology equilibrium to a high-technology one, if they fail to harmonize different decisions of enterprises with expertise and to enter into the successful development of the intermediate goods industry, they might be stuck in the low-technology equilibrium. Since the goods from hightechnology sectors are often highly priced, high-technology equilibrium is not only Pareto-superior to low-technology equilibrium (RodríquezClare, 1996) but also more desirable in terms of social welfare, as it provides higher wages than the low-technology states (Rodrik, 1996). This implies that some effort should be exerted into moving from the low-technology equilibrium into the high-technology one. However, it is true that the level of concentration is much higher in the capital goods industry than in the consumer goods industry because the former is more monopolized by advanced countries. The growth of the capital goods industry can thus serve as a real barometer to determine whether a country is indeed in the ranks of advanced countries or not. From this perspective, let us consider the case of Korea in the following discussion. While rapid economic growth has been the symbol of the Korean economy, it is noteworthy that the growth has been mainly recorded in the production and export of final consumer goods led by large, diversified and family-controlled conglomerates, the so-called chaebols. It is widely known that, to produce final goods for export, conglomerates rely upon imported facilities and machinery as well as key intermediate materials and parts. The case of machine tools is a typical example. Since the 1960s, Korea has imported almost all its main machine tools and has
262
Sectoral systems of innovation and production
100 000 The Amount of Export The Amount of Import
90 000 80 000 70 000 60 000 50 000 40 000 30 000 20 000 10 000
0 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Year
Source:
Korea International Trade Association (www.kita.net).
Figure 9.1
Trends in the Korean export/import of machines
thus suffered from large trade deficits (Figure 9.1). It is only in very recent periods that trade in the machine tools industry has either struck a balance or indeed had a surplus. Japan has consistently been the biggest supplier country of machines, but there is no sign of Korea striking a balance of machine trade with Japan. Table 9.1 shows the persistent trade deficit of Korea in terms of its trade with Japan in machinery. The huge trade deficit with Japan is due to a structural reason. At present, the main Korean export products are final consumer products and general-purpose electronics parts. However, the precision electronics parts and machinery used rely on Japanese firms. In short, the Korean industry is confined to assemblies with the use of machinery imported from Japan. The same is the case for semiconductor products, one of the representative export items of Korea. Only 20 per cent of the production equipment in the semiconductor industry is made within Korea. About 80 per cent of the front-end process equipment is imported primarily from Japan and the United States. Confronted with the foreign exchange crisis and the financial crisis in 1997, the Korean economy undertook a severe structuring, and the imposition of harsh policies turned the tide toward trade surplus in machine tools after 1997. Since then, export has expanded at the rate of over an average of 10–30 per cent, with the balance turning to surplus. This surplus is earned by exporting to lower-tier countries like China because
263
425 031 530 507 460 766 532 248 524 733 465 306 590 730 830 557 818 145 893 517 953 156 839 145 1 241 223 1 164 029 1 170 756 1 459 350 2 087 793 2 612 463 2 804 695
To Japan $’000
4 4 4 4 5 4 4 5 5 6 8 5 6 7 8 8 10 11 11
%
Exports
12 004 069 13 456 797 12 637 879 12 355 839 11 599 454 11 564 418 13 522 860 17 048 871 15 766 827 14 771 155 12 237 587 15 862 448 20 466 016 16 505 766 15 143 183 17 276 137 21 701 337 24 027 438 26 534 015
Total* $’000
Source:
Korea International Trade Association (www.kita.net)
4 567 981 5 317 878 5 758 110 7 190 513 6 408 835 6 509 939 8 733 633 10 958 586 11 122 570 8 474 827 3 664 238 5 536 274 8 339 553 6 073 020 6 370 180 8 355 989 10 965 445 11 590 870 12 614 682
From Japan $’000
Trade deficits with Japan in machinery
Notes: * Total machinery export to Japan. ** Total machinery import from Japan. *** Total machinery trade deficit with Japan.
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Year
Table 9.1
29 30 31 34 33 33 34 34 35 30 22 23 26 23 21 23 24 24 24
%
Imports
15 928 766 17 448 627 18 573 851 21 120 216 19 457 650 20 015 519 25 389 988 32 606 368 31 448 636 27 907 108 16 840 409 24 141 990 31 827 943 26 633 372 29 856 228 36 313 091 46 144 463 48 403 183 51 926 292
Total** $’000 4 142 950 4 787 371 5 297 344 6 658 265 5 884 102 6 044 633 8 142 903 10 128 029 10 304 425 7 581 310 2 711 082 4 697 129 7 098 330 4 908 991 5 199 424 6 896 639 8 877 652 8 978 407 9 809 987
With Japan $’000 106 120 89 76 75 72 69 65 66 58 59 57 62 48 35 36 36 37 39
%
Trade deficits
3 924 697 3 991 830 5 935 972 8 764 377 7 858 196 8 451 101 11 867 128 15 557 497 15 681 809 13 135 953 4 602 822 8 279 542 11 361 927 10 127 606 14 713 045 19 036 954 24 443 126 24 375 745 25 392 277
Total*** $’000
264
Sectoral systems of innovation and production
Korean firms have succeeded in entering the emerging market with price competitiveness in middle-technology products. However, Korea continued to have deficits with Japan and other advanced countries. While the trade surplus in both total trade balance and machine tools items certainly exists, trade imbalance with Japan has continued to worsen. The import volume from Japan is more than twice the export volume to it, and the growth rate of import from Japan outnumbers that of the export to it, allowing the trade imbalance to expand, especially to electronic, chemical and machine products. The trade deficit of machine products, in particular, outnumbers the total trade deficit in the 1980s, and the proportion of the import volume to Japan has not declined significantly. In the sections that follow, we will focus on the issue of why localization and the growth of machine tools industries take more time and are thus more difficult than those of other sectors. For this, we relied upon the conceptual framework of the sectoral systems of innovation (Malerba, 2004, and others).
3.
SECTORAL SYSTEM OF INNOVATION AND THE MACHINE TOOLS INDUSTRY
Drawing from several intellectual precedents such as the innovation system literature (Edquist, 1997), the national systems of innovation literature (Freeman, 1987; Lundvall, 1992; Nelson, 1993) and the technological systems concept (Hughes, 1984; Callon, 1992; Carlsson and Stankiewitz, 1995), Malerba (2002, 2004) defined a sector as a set of activities that are unified by some linked product groups for a given or emerging demand and that share some common knowledge. Firms in a sector have some commonalities and at the same time are heterogeneous in terms of learning processes and capabilities. Thus, the building blocks of Malerba’s SSI consist of (a) the regimes of knowledge and technologies, (b) demand conditions (or market regimes), (c) actors and networks and the coordination among them, and d) the surrounding institutions, including IPRs, laws, culture and the like. These elements are supposed to interact to generate variety subject to selection and coevolution. This chapter, together with the other chapters in this volume, extends the original SSI framework to the context of catch-up in developing or latecomer countries. Thus, while we will apply the same framework, we can expect that some modification or adaptations are necessary to make it suitable to the context of developing countries. Similar adaptations have been made by Lee and Lim (2001), Lim et al. (2005), Mani (2005) and Mu and Lee (2005) when they analysed the industry cases from China, Korea
Barriers and opportunities in the Korean case
265
and India with theoretical concepts such as the sectoral innovation system or technological regimes as its sub-components. Regimes of Knowledge and Technologies The machine tools industry is referred to as the “machine that makes machines”, that is it is the “mother machine”. In other words, the machine tools industry is the foundation of the manufacturing industry, playing a pivotal role as the determinant of productivity and quality of products. Thus, the machine tools industry influences the development of other industries, and furthermore it serves as the criterion of the nationwide industry level. According to Pavitt (1984)’s classification, the machine tools industry is a typical specialized supplier industry in which the tacit knowledge accumulated from the interface between the producer and customer firms is critical. In this industry, crucial know-how is not easily embodied into the production equipment because the equipment used in the production process is usually a general-purpose machine. It follows that the skills accumulated by workers are much more important. In addition, licensing is confined within a few specific models and therefore does not give much help in acquiring or improving design capability. Furthermore, a producer needs the ability to skilfully revise the design of a machine for various products in order to cater to the diverse requests of customers, but this ability is not acquired only with technology licensing. This fact partly explains why catch-up is difficult in the machine tools industry, which has low innovation speed. The implicit knowledge feature of the machine tools industry makes the path for technological catch-up different from that in the electronics industry. As mentioned above, in the machine tools industry, experience over the long term carries more weight, while intensive research and development (R&D) for a short period hardly leads to an effective catch-up. The above discussion explains why catch-up is not easy despite the fact that low frequency of innovation and low volatility have traditionally been the feature of the machine tools industry. Recent changes of the technological regime, however, have enlarged the room for subverting traditional competitive advantage conditions. This is due to the increasing adoption of computer technology in this industry, which has been changing such old industry features. An increase in the frequency of innovation as well as fluidity of technological trajectory can now be seen. We can find examples of this in the growing mechatronics items which are rooted in the IT industry, such as the numerical control (NC) machine, CAD/CAM, PLC and the like. The introduction of this line of products partially improves
266
Sectoral systems of innovation and production
the possibility of catch-up. Since the production of these products involves electronics and software applying new technology to machines, accumulativeness is relatively low. Moreover, Korea is highly competitive in its IT industries; as such, it is highly likely to benefit because it has an edge in terms of technological advantage. Technology latecomers are usually weaker in terms of designing machines, but, with the help of process tools like CAD, a simulation package, the tacit knowledge in machine development that is accumulated through “learning by doing” is converted into explicit knowledge. In other words, software works as a bridge of tacit and explicit knowledge, which is acquired only through learning by doing and has now been adopted by formal learning (Antonelli, 1997). Therefore, the importance of scientific knowledge is now emphasized, and the interaction with public research institutions and universities becomes increasingly important for technological catch-up in the industry. Universities become important not only as providers of innovative scientific knowledge but also as a pool of human resources that have the ability to bring the worker closer to the field. Demand Conditions or Market Regimes While the above discussion suggests a somewhat mixed picture of the possibility of catch-up in the machine tools industry, it is our view that it is the demand conditions that give rise to more difficulties for catch-up. Stable or long-term market demand is critical because real R&D capability in machine tools is acquired from the tacit knowledge accumulated in the process of developing and producing the products over longer-term interactions with the user firms. However, in most developing countries that tend to specialize in producing final consumer goods, the user firms are seriously reluctant to use locally made machine tools, owing to their poor quality and low level of precision, which could hurt the competitiveness of their final goods in an unpredictable manner. Since the quality of a machine directly determines the quality of the final consumer goods made by the machine, the final goods industry, sensitive to the quality of its own products, shuns the use of locally made, poor-quality machine tools; this has also been the case in Korea. From the user firm’s perspective, the risk of adopting locally made capital goods is simply too high. Outside of the export market, the domestic market is in itself weak, making the accumulation of tacit knowledge difficult by expanding production and interacting with various user firms. Latecomers, therefore, cannot expect any comparative advantage in either cost or quality (Lee and Lim, 2001). Furthermore, latecomer firms have few incentives to research and develop; this is in line with the fact that they
Barriers and opportunities in the Korean case
267
perceive a low possibility of success in the market, since they cannot expect any of the benefits such as cost edge, quality differentiation or first-move advantages (Lim, 1997). The next section will discuss this issue further, namely the difficulty posed by the uncertain demand from user firms. Role of the Actors: Governments and the Incumbent Firms While it has been typical for the government to intervene through helping latecomer firms in overcoming the barriers to catch-up in various forms, we find that, in the capital goods industry, government activism tends to have limited effectiveness for several reasons. Lee and Lim (2001) aptly pointed out that even government policies which encourage the use of domestic products were not and cannot be effective. Since the quality of the machine tools employed directly determines the quality of the output, customer firms, sensitive to the quality of their own products, cannot afford to use domestically produced machine tools following the “order” of the government. Knowing the risk involved, even the government realized that it could not help in this situation. Therefore, the only solution is to produce local capital goods which are high in quality and are cheaper, which is in turn very difficult. An additional difficulty arises from the reactions of incumbent firms when a latecomer firm successfully develops its own machine tools and starts to sell them in markets or to local user firms. The strategies of incumbent firms, often foreign, include charging predatory or dumping prices so that the entrant firms will not be able to attract users, and filing lawsuits against the entrant firms for violation of the IPR rights of the incumbent firms. These roles of the actors will be discussed in detail in the next section.
4. 4.1.
THE THREE BARRIERS TO CATCH-UP IN THE CAPITAL GOODS INDUSTRY Weak Demand and Weak R&D
The first difficulty encountered by local machine tools firms is how to create and maintain market demand from user-client firms, which are largely big conglomerates producing final consumer goods. As mentioned above, since capital goods, machine tools in this case, directly determine the quality of the final products, the user firms or large enterprises that produce final goods are likely to be very fastidious in choosing which
268
Sectoral systems of innovation and production Low quality or technology
42.9
Uncertainty in deliverer’s request for large margin
25.2
High logistic costs
8.8
Request for enough margin
8.2
Government regulation Etc.
No answer
3.4 6.1 5.4
The Survey Answers of the User Firms
Note: Surveyed: 147 firms in the manufacture, construction, distribution, etc. sectors; Period: 20 February, 2004 to 5 March, 2004. Unit: percentage. Source:
Federation of Korean Industries (2004).
Figure 9.2
Difficulties in purchasing the products made by SMEs
machine tools they should use for their production processes (Lee and Lim, 2001). User firms cannot be blamed for this, because local machine tools firms used to either produce low-end or low-technology machines for general purposes or show poor performance in making high-end or hightechnology machines. In the meanwhile, the user firms have been able to import from Japan more reliable machine tools, often at affordable prices. Given the existence of a big gap in the quality of the products made in Japan and Korea, it is not a surprise that the user firms preferred to use Japanese products, and thus no user–producer interaction was possible in Korea. This is different from the case of Europe, where the difference in quality among the producers is small, and thus geographical proximity is an important factor in choosing the products, thereby allowing more chances for local user–producer interaction. In Figure 9.2, a survey on how to promote co-prosperity partnerships between SMEs and large firms conducted by the FKI (the Federation of Korean Industries) in 2004 reveals that one of the most frequent answers for the difficulties in the partnership was “lack of quality or low level of technology of the products by the SMEs” (42.9 per cent). The low level of trust in the quality of the products by the SMEs in machine tools leads big user firms to adopt a “wait and see” attitude, avoid being the first user, or take advantage of the fact that there are few buyers of the products by the SMEs. In other words, the typical buyers’ market conditions in the capital goods industry give an upper hand to big buyers,
Barriers and opportunities in the Korean case
269
firms in their dealings with the seller SMEs. The buyer firms often ask for extraordinary discounts in transactions. An annual survey on the difficulties confronted by SMEs as conducted by the KFSB (Korean Federation of Small and Medium Business) points out the problem of subcontracting which is associated with the requests for discount of buyer firms (Table 9.2). This still holds for the development process; the cost for developing new items is not fairly compensated, owing to the conglomerate’s demand for a lower price, thereby taking away incentives or opportunities for development or incurring expenditure for the next rounds of R&D. The uncertain or unfair demand from user firms makes machine tool firms hesitant about investing big chunks of money to develop capital goods, whereas the size of R&D expenditure to develop machine tools is usually quite big, especially compared to the size of the firms in charge of this business. They are usually small or middle- sized specialized firms and as such often do not have enough financial capacity for big R&D projects. The results of a survey on the difficulties in the machine parts industry by FKI in 2005 confirm this (Figures 9.3 and 9.4). A machine parts producer considers “insufficient R&D funds” (32.7 per cent) as the most serious difficulty in the development process. This insufficiency of R&D funds does not simply mean that the funds are not enough but that there are structural barriers for the SMEs to put more money into R&D. The barriers are the uncertain demand from local users, the relatively bigger size of the required R&D budgets, and the involved risks as compared to the size of the firms. The same survey also points out the “long turnover period” (24.6 per cent) as an additional source for the difficulty in conducting R&D. Moreover, the SMEs are not paid for the delivery of goods as scheduled in the contracts with the user firms. Instead, the user conglomerates have tended to delay payments, taking advantage of their bargaining power. This suggests that the problem of insufficient funds in SMEs is connected with the power imbalances between the buyer and the seller firm, worsening the condition of R&D for SMEs. As a CEO of a Korean semiconductor machine firm (Jusung Engineering Co.) said in an interview, “only with a quality 30 per cent higher for a 30 per cent lower price can locally produced machine tools be chosen by user conglomerates”. While the local development itself of products with an enhanced quality level is already quite difficult, the even lower prices requested by user firms add further to the problem. This situation suggests the difficulty associated with the local development of capital goods. It might thus be better for machine tools firms to form a long-term alliance or network with large user firms from the beginning of the business in order to develop certain items. In some cases, big conglomerates themselves have started to develop machine tools or parts.
270
Sectoral systems of innovation and production
Table 9.2
Difficulties facing SME suppliers in transactions with large user firms 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 % % % % % % % % % % %
Difficult to meet the demand for high quality Delayed delivery/ receipt by user firms Request for price discount by user firms Long settlement date (delayed payment) by user firms Unpredictable order Short delivery deadlines Uncertainty about long-term contracts Forced settlement in the form of products by the buyer firms Absence of local L/C (letter of credit) in foreign trading Irregular practices with payments using CP (commercial paper) Low Profits due to competitive bidding Conflict with client firms over end costs (prime cost)
25.7 28.9 30.8 32
3.6
3.1
6.7
7.0
33.1 31.6 35.6 38.1 38.5 37.7 35.4
6.9
4.9
2.5
3.1
4.2
4.1
3.0
74.6 73.1 75.4 68.4 61.3 74.6 68.8 72.1 80.2 71.5 70.3
41.6 47.8 40.7 44.9 53.3 45.9 37.9 37.4 24.6 31.2 32.1
35.1 40.0 44.5 48.3 41.5 48.1 52.2 51.3 53.9 53.6 46.3 32.9 34.4 35.0 34.2 29.6 43.0 45.0 45.8 47.1 39.7 40.8 20.5 27.5 23.1 26.2 22.7 12.5 17.8 11.5 12.5 16.5 17.1 –
–
–
–
–
3.4
1.8
4.0
–
–
–
–
–
–
–
–
1.3
1.8
3.0
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
17.3 13.5 12.8
–
–
–
–
–
–
–
–
18.0 13.4 14.9
23.9 22.3 18.8 18.3 13.7 13.8
Notes: Multiple answers allowed. Period: 1993–2003. Source:
4.2.
Korea Federation of Small and Medium Business (2004).
Dumping Pricing by the Incumbent Firms
The second barrier facing the local producers of capital goods is the counter-attacks of the incumbent firms such as charging dumping prices to crowd out new and late-entrant firms. The case of the industrial robot
Barriers and opportunities in the Korean case
271
Insufficient R&D funds
32.7
Shortage in capable R&D staffs
25.9
Weak in-house technology and infrastructure
15.7
Insufficient information about technology trends
10.4
Weak knowledgment of technology trends
7.4
Weak tax subsidies for the R&D or technology transfer
5.6
No promising opportunity for R&D Etc.
2.2 0.2 0.0
Note:
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Period: April 2005; 193 enterprises with over 8 billion Won Sales; Unit: percentage.
Source:
Federation of Korean Industries (2005).
Figure 9.3
Difficulties in the development of technology
Long turn-over period
24.6
Uncertain sales
20.5
Difficulties in bank borrowing
19.3
Restriction on maximum amount of borrowing
17.1
Too high interest on borrowing Bankruptcy of the user firms Etc. Difficulties in re-financing with the CP discounting 0.0
Note: Source:
12.4 2.5 1.9 1.5 10.0
20.0
30.0
Period: April 2005; 193 enterprises with over 8 billion Won sales; Unit: percentage. The Federation of Korean Industries (2005).
Figure 9.4
Difficulties in financing for R&D
272
Sectoral systems of innovation and production
Table 9.3
The anti-dumping and safeguard cases investigated by the authorities 1987–2003
Consumption goods Intermediate goods Total Note: Source:
2004
Total
AD
SG
Total
AD
SG
Total
AD
SG Total
27 52 79
20 13 33
47 65 112
1 4 5
– – 0
1 4 5
28 56 84
20 13 33
48 69 117
Unit: case; AD: anti-dumping, SG: safeguard. Korean Trade Commission (2004).
with a six-axes vertical multi-articulation structure that was judged to be anti-dumping by the Korean Trade Commission in April 2005 is a typical example. Up to 2000, Korean firms including Hyundai Heavy Industry have occupied over half of the domestic market share, whereas world-class robot manufacturers such as Najji, Kawasaki, Yaskawa and Hwanak accounted for 53.3 per cent of the Korean market share in 2004. These incumbent firms were selling their products to Hyundai Motors, Kia Motors, and GM Daewoo and then staged a price war to kill Hyundai Heavy Industries by supplying their products at dumping prices from 2003. As a result, the market share of domestic robots dropped sharply to as low as 30 per cent. There have been numerous cases like this, namely dumping pricing by the incumbent firms upon news of the entry of local producers into markets that they had formerly monopolized. The cases filed with the KTC (Korean Trade Commission) by Korean enterprises since 1988 for the investigation of dumping charges reveal this tendency of dumping by foreign firms (see Table 9.3, Appendix Table A9.1). In Appendix Table A9.1, firms seeking for relief from dumping attacks usually belong to a sector of intermediate materials or parts industry, whose share is over 70 per cent compared with that of final goods or consumer goods firms. We also noticed a certain change in the main items filed over time. Until the mid-1990s, the items belonging to sectors on basic intermediate materials and parts were at the top of the list, whereas after 2000 the items belonging to high-technology goods such as pre-sensitized printing plates,1 lithium batteries,2 PVC plates,3 CD-Rs and industrial robots had increased in number. Given the fact that it takes over a year to file anti-dumping cases and many SMEs have given up filing owing to the complicated process, the real number of cases is deemed to be much larger than that reported.
Barriers and opportunities in the Korean case
4.3.
273
Filing IPR Lawsuits against the Catching-up Firms
When latecomer firms achieve initial success in markets despite all the difficulties, a final difficulty that might await them is the IPR-related lawsuits from incumbent firms. The active litigation by the incumbent firms aims not only at collecting returns in the form of royalties or fees but, more importantly, at stopping the activities of latecomers in the market. In particular, the likelihood that the latecomers will face lawsuits increases when they enter international markets. Because the latecomer SMEs often do not have a capable IPR department or personnel, the risks they face are very high. An interesting case in this regard is a Korean company called Sunstar, which lately entered the market by indigenously developing a computercontrolled automatic embroidery machine. A Japanese firm, Tokai, has been the unchallenged leader in the embroidery machine market since the 1990s. However, in 1997, the new embroidery machine developed by Sunstar Co. was able to enter the market after five years of relentless effort. Tokai got alarmed and filed a lawsuit against Sunstar in March 1998, arguing that the communicational input tool used by Sunstar violated its patent. Sunstar managed to have the charges dropped by presenting critical evidence that the technology had already been in use in the industry even before Tokai adopted it for its products. It was an obviously critical moment for the existence of Sunstar. Since then, Sunstar has successfully caught up with Tokai, occupying as much as 35 per cent of the market share, and is now considered as one of the top firms in the world in its field. Another example would be the Joosung Engineering case. Established in 1996, Joosung Engineering has manufactured equipment for semiconductors and LCDs (liquid crystal displays) such as those used in CVD (chemical vapour deposition) machines. CVD is the process of plating chemicals on a wafer substrate, and it used to be imported to Korean user firms. Chul-Ju Hwang, the founder and CEO of Joosung Engineering, used to be an after-sales service engineer and local agent for ASM, a semiconductor machine company in the Netherlands that used to do business with Samsung Electronics. After leaving ASM, he established his own company and developed the equipment for the front-end process. Since 2002, Joosung has expanded its production scope to PECVD (plasma enhanced chemical vapour deposition) for LCDs, and branched its sales out to the United States, Japan, Taiwan and other countries. Applied Material Inc. (AMAT) filed a lawsuit against Joosung for patent violations in Korea and filed in a Taiwan court for provisional disposition, which meant a sudden halt in all of Joosung’s commercial activities in the emerging LCD machine market in Taiwan. Litigations for a period of
274
Sectoral systems of innovation and production IPR-related legal dispute
64.3
Shortage in capable R&D staffs
50.0
Insufficient R&D funds Uncertain demand from the local user firms
42.9 50.0
Weak technology capacity
35.7
Increasing market competitions
14.3
Dumping by the foreign rival firms
14.3
Etc.
7.1
Labour disputes 0.0 0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
%
Note: Period: March–May 2004. Answerer: 114 semiconductor machine equipment firms (multiple answers allowed). Source:
Center for Corporate Competitiveness, Seoul National University.
Figure 9.5
Obstacle for localization: semiconductor machine equipment firms
over a year took place, ending in Joosung being cleared of all the charges. Despite this, it created huge damage to Joosung because of its negative image as a “patent robber” in the Taiwanese market. If the SMEs are entangled in lawsuits on industrial property rights, the litigation usually hurts the firms in many ways and not only in terms of sales. Prohibitive patent licence fees as well as marketing channels can be lost during the extended lawsuit period. Seeing these kinds of difficulties, most SMEs feel very much concerned about patent lawsuits, especially during the stage when they are starting to develop a technology. The survey by the Center for Corporate Competitiveness of Seoul National University in 2004 shows this (Figure 9.5). In this survey, semiconductor equipment firms said that the localization of intermediate materials and goods is not really difficult, estimating its feasibility as “very high’ (40.9 per cent) and “high’ (59.1 per cent). However, they regarded the biggest obstacle to localization as “IPR-related legal dispute’ (64.3 per cent). Since the 1980s, patent lawsuits by American firms have caused enormous losses or fees against target Japanese firms. Through these experiences, the CEOs of Japanese firms have learned the importance of IPR rights. Thus, in the same fashion, the filing of litigations by Japanese firms whose export products coincide with those of Korean firms has
Barriers and opportunities in the Korean case
275
been dramatically increasing. For instance, in 2004 alone, Japanese firms filed lawsuits against: Samsung SDI (April); Samsung Electronics, LG Electronics and Kiryung Electronics (May); Daewoo Electronics (September); LG Electronics of Korea (November); Nanya Technology (February); AUO (July); and E&E of Taiwan (July). Particularly, Japan has strengthened intellectual property rights protection when it revised its Customs Tariff Law in 2003, which enables the government to request from the Customs Office the suspension of the importation of products when they have infringed patent rights or the design rights of Japanese firms. Once requested from the Industrial Property Office, Japanese authorities investigate the imported product, and if it is found to be infringing patent rights it is banned immediately. This is definitely a strong measure, and only upon the domestic firms’ request can there be a retention of customs clearance. In general, technologically advanced countries have strengthened their IPR protection policies and rules, and this has emerged as one of the important challenges that needs to be addressed by catching-up firms.
5. 5.1.
OPPORTUNITIES FOR CATCH-UP The Role of the Government
As noted in the preceding section, it is difficult for the capital goods industry to grow in latecomer economies. Since no firms are willing to enter this industry, latecomer economies are more likely to end up being underdeveloped. Given such possibility of low-technology equilibrium, Rodrik (1996) emphasized the need for government intervention that will move firms out of a “bad” equilibrium to a “good” one. Rodrik argued that, as long as the high-technology industry is much more capital-intensive, policy support such as subsidies and wage promotion policy is effective in the transition to good equilibrium. Consistent with Rodick’s view, the Korean government has also been involved heavily in this sector. In the ex post sense, the Koreans saw the slow but steady success in localizing the production of formerly imported machinery over the last three decades. However, as emphasized above, the limited success was inevitable. In what follows, a brief review of related government policies is provided. There have long been government efforts for the home production of machines, starting from the project on the promotion of the machine industry as early as the late 1960s. In the 1970s, localization policies on machinery were in the form of prohibition of imports. With a specific target of localization, a specific enterprise is selected and
276
Sectoral systems of innovation and production
financed for the introduction of related technology or equipment. There was also the establishment of a joint company with a foreign partner. For example, in the automobile industry, although assembling firms are permitted to import engines, they were compelled to procure other parts from local firms, and so part-supplying firms were intensively supported. In the 1980s and thereafter (1977–1999), localization policies were pursued in the name of the import-source diversification policy, which aimed at curbing the sharp increase in the importation of Japanese products.4 The number of items under import source control increased from 261 to 924, and from 1993 it gradually decreased to 15 until the abolition of import control in 1999. Policies in the 1980s were in essence an extension of the import substitution policies, as they also aimed for the localization of low-technology, general-purpose parts and materials of 4202 items. On the other hand, this policy is perceived as too protectionist in that it limits market competition by often protecting items that failed in such competition. Thus, it was abolished in 1999 and some perception was that it was only after this action that the competitiveness of machine tools industries started to increase. Policies after 2000 were focused on countering the negative effects of previous localization policies, which were too protective. Thus, the main goals were now the development of technologies and R&D efforts to meet the demands of markets. According to the Act for Encouraging the Machine Parts Industry that was enacted in 2001, the so-called “R&D which bridges demand and supply” was implemented, and some firms were issued with a “certification of credibility by the government” regarding the quality of the products. During this period, some firms achieved a certain level of capabilities that allowed them to supply their products globally. Nevertheless, it was recognized that the marketing of locally produced capital goods was quite difficult, and big user firms were not patronizing local products. In a nutshell, up to now, localization policies pursue quantitative expansion in a short span and succeed in substituting some of the formerly imported products using low- or mid-level technology that is easy to catch up within a short period. On the other hand, they have also shown their limitation in promoting core or high-technology capabilities, as shown by the persistent trade deficits with Japan. 5.2.
Opportunities from Firms’ Perspectives
Sections 3 and 4 emphasize the intrinsic difficulty of catch-up in the machine tools sector as implied by its knowledge regime, its demand conditions, and the reactions of incumbent firms. Despite this, Korean
Barriers and opportunities in the Korean case
277
firms have made slow but steady catch-up. This was made possible owing to several factors, including the strenuous effort of the government in response to this issue. First, while the accumulation of knowledge in this industry requires close interaction with the user firms, general-purpose machine tools were less subject to this constraint. The growth of the machine tools industry in Korea has been fast in the general-purpose machines sector as compared to that in the specific tools sector (Lim, 1997). One of the reasons for the growth of the general-purpose machine tools sector is the fact that it neither requires much interaction with the user firms nor entails high-level technology. These machine tools can be standardized and sold in mass markets to any user. For this reason, the producers of these tools can be freer from the whims of specific user firms. Second, recent changes in the knowledge regime owing to the IT revolution have opened up the new possibility of catch-up along stage-skipping or leapfrogging strategies. The share of mechatronics like NC processors, CAD/CAM and PLC, the base of which is the IT industry, has recently increased, and therefore stage-skipping catch-up becomes more feasible. Since this area combines electronics and software, there is a lower accumulativeness than before. Moreover, Korea is likely to gain an edge in terms of technology with a strong base of the IT industry. This is evident in Figure 9.1, which shows the sudden surge of machine tools exports since the late 1990s, which coincides with the emergence of new digital paradigms. Third, the strong increase of demand from late-emerging economies, headed by China and the BRIC countries, has served as markets for Korean-made machine tools. The recent improvements in machine trade sales in Korea are actually due to the expansion of these new markets. While the slowdown of the machine tools market in the world is evident in the negligible market growth rate of 0.1 per cent from 1999 to 2003, the Chinese and Asian market growth rate is as high as over 14 per cent. Korean firms penetrated these markets successfully with competitive quality and reasonable prices. Table 9.4 shows the increasing importance of China markets for Korea. While the major export partners in machine parts have been the United States and Japan up to 2001, the Chinese share in exports has become larger and larger, with shares increasing from 19.1 per cent (2002) to 27.1 per cent (early 2004).
6.
SUMMARY AND CONCLUDING REMARKS
This chapter has dealt with the question of why making a catch-up is even more difficult in capital goods industries that are usually led by small or
278
Source:
Note:
574 770 587 83 110 2124 4044
Value
1.9
19.0 10.3 9.4 11.4 7.6
Growth rate %
2001
611 762 820 92 86 2371 4298
Value 6.4 –1.0 39.7 10.8 21.8 11.6 6.3
Growth rate % 14.2 17.7 19.1 2.1 2.0 55.2 100
Share % 711 820 1252 94 102 2979 5423
Value 16.4 7.6 52.7 2.2 18.6 25.6 26.2
Growth rate %
2003
13.1 15.1 23.1 1.7 1.9 54.9 100
Share %
488 425 923 49 69 1954 3405
Value
34.3
52.4 –1.0 58.3 4.4 62.7
Growth rate %
14.3 12.5 27.1 1.4 2.0 57.4 100
Share %
First half of 2004
Ministry of Commerce, Industry, and Energy, Reorganized from “Parts and Material Industry Statistics DB” (www.pmsd.or.kr).
14.2 19.0 14.5 2.1 2.7 52.5 100
Share %
2002
Export markets of general-purpose machines of Korea
Unit: million dollars.
Japan US China UK Germany Total Sum of exports
Nation
Table 9.4
Barriers and opportunities in the Korean case
279
middle-sized companies. It relies upon the sectoral systems of innovation (Malerba, 2004) as a theoretical framework for analysis. From the findings, the chapter has identified three sources of difficulties in the catch-up of the capital goods industry, particularly in the machine tools sector. First, while small firms in the capital goods industry are usually specialized suppliers to big final goods assembly firms in the consumer goods industry or other industries, and thus the tacit knowledge accumulated from the interface between the producer and the customer firms is very important, a serious difficulty lies in the fact that local client firms are reluctant to use locally made capital goods owing to their poor quality and low precision level. In this matter, even government policies which encourage the use of domestic products are not and cannot be effective. Second, while a successful catch-up first requires the ability to produce goods of better quality and lower prices than those produced by the incumbent firms from advanced countries, a typical difficulty arises because incumbent firms often react by charging predatory prices upon news of the local development of capital goods by latecomer firms. Third, if the catch-up firms overcome such barriers, then the next strategy used by incumbent firms is to charge latecomers with legal actions for patent violations. Despite these intrinsic difficulties, the Korean economy has achieved a very slow but gradual catch-up in the capital goods industry, with several successful companies having emerged. The chapter has attributed such achievement to several factors, such as the strenuous effort of the government, niche markets in general-purpose machine tools and emerging economies, the so-called BRICs, and finally the increase in the introduction and adoption of IT or digital technologies in machine tools. The three sources of barriers to catch-up also imply that any latecomer firms that wish to record a successful catch-up should have these barriers in mind from the beginning of the road toward catch-up. We observe that a successful catch-up requires the ability to produce goods of better quality and lower prices than those produced by incumbent firms from advanced countries. Then, after the initial success, it is imperative for these firms to be well prepared against eventual or possible attacks by the incumbent firms in the forms of predatory pricing and IPR charges.
NOTES *
This paper was previously published in Global Economic Review, June 2008, Taylor and Francis. 1. This is used in press papers.
280
Sectoral systems of innovation and production
2. This is used in the batteries of cameras, electronic lockers and communication equipment. 3. This is the main material of semiconductors and is the LCD·PDP frame of production equipment. 4. The import-source diversification policy is meant to control imports from countries with which Korea has too much trade deficit. The importing firms are then guided to import from other countries. This policy came into effect in 1977 and was abolished in July 1999 following the recommendation of the WTO.
BIBLIOGRAPHY Antonelli, C. (1997), Localized Technological Change, New Information Technology and the Knowledge-Based Economy: The European Evidence, Laboratorio di Economia dell’ Innovazion, Universita di Torino, Torino. Callon M. (1992), ‘The dynamics of techno-economic networks’, in Coombs, R., Saviotti, P. and Walsh, V. (eds), Technical Change and Company Strategies, Academy Press, London. Carlsson, B. and Stankiewitz, R. (1995), ‘On the nature, function and composition of technological systems’, in Carlsson, B. (ed), Technological Systems and Economic Performance, Kluwer, Dordrecht. Edquist, C. (ed.) (1997), Systems of Innovation: Technologies, Institutions, and Organizations, Pinter, London. Federation of Korean Industries (2004), The Survey to Improve the Cooperation of Conglomerates and SMEs (in Korean), www.fki.or.kr. Federation of Korean Industries (2005), The Survey on the Difficulties in the Machine Parts Industry by FKI in 2005 (in Korean), www.fki.or.kr. Freeman, C. (1987), Technology and Economic Performance: Lessons from Japan, Pinter, London. Hughes, T.P. (1984), ‘The evolution of large technological systems’, in Bijker, W., Hughes, T., Pinch, T. (eds), The Social Construction of Technological Systems, MIT Press, Cambridge, MA. Korea Federation of Small and Medium Business (2004), The Annual Survey on the Difficulties of SMEs (in Korean), www.kbiz.or.kr. Korean Trade Commission (2004), Handbook of the Measures for the Trade Dispute Settlement (in Korean), www.ktc.go.kr. Lee, K. and Lim, C. (2001), “Technological regimes, catching-up and leapfrogging: findings from the Korean industries”, Research Policy, 30(3), 459–83. Lee, Keun, Lim, C. and Song, W. (2005), “Emerging digital technology as a window of opportunity and technological leapfrogging: catch-up in digital TV by the Korean firms”, International Journal of Technology Management, 29(1/2), 40–63. Lim, C. (1997), Sectoral Systems of Innovation in the Period of Cluster Forming: The Case of the Korean Machine Tool Industry, Science Policy Research Unit, University of Sussex, Brighton. Lundvall, B.A. (ed.) (1992), National Innovation Systems: Towards a Theory of Innovation and Interactive Learning, Pinter, London. Malerba, F. (2002), “Sectoral systems of innovation and production”, Research Policy, 31(2), 247–64.
Barriers and opportunities in the Korean case
281
Malerba, F. (2004), Sectoral Innovation System, Cambridge: Cambridge University Press. Malerba, F. and Breschi, S. (1997), “Sectoral innovation systems: technological regimes, Schumpeterian dynamics, and spatial boundaries”, in Edquist, C. (ed.), Systems of Innovation, Pinter, London. Malerba, F. and Orsenigo, L. (1990), ‘Technological regimes and patterns of innovation: a theoretical and empirical investigation of the Italian case’, in, A. Heerjte and M. Perlman (eds), Technologies and Market Structure, Michigan University Press, Ann Arbor, pp. 283–306. Mani, Sunil (2005), “The dragon vs. the elephant: comparative analysis of innovation capability in the telecom industry of China and India”, Economic and Political Weekly 40(39), pp. 4271–4283. Mu, Qing and Lee, Keun (2005), “Knowledge diffusion, market segmentation and technological catch-up: the case of the telecommunication industry in China”, Research Policy 34(6), August, pp. 759–783. Nelson, R. (ed.) (1993), National Innovation Systems: A Comparative Analysis, Oxford University Press, New York and Oxford. Pavitt, K. (1984), “Sectoral pattern of technical change: towards a taxonomy and a theory”, Research Policy 13, pp. 343–373. Porter, M. (1992), The Competitive Advantage of Nations, Free Press, New York. Rodrik, D. (1996), “Coordination failures and government policy: a model with applications to East Asia and Eastern Europe”, Journal of International Economics, 40(1–2), 1–22. Rodríquez-Clare, A. (1996), “The division of labor and economic development”, Journal of Development Economics, 49(1), 3–22. Romer, P. (1990), “Endogenous technological change”, Journal of Political Economy 98, pp. 71–S102. Stigler, G. (1951), “The division of labor is limited by the extent of the market”, Journal of Political Economy, 61, pp. 185–193. Wengel, J. and Shapira, P. (2004), “Machine tools: the remaking of a traditional sectoral innovation system”, in F. Malerba (ed.), Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe, Cambridge University Press, Cambridge. Young, A. (1928), “Increasing returns and economic progress”, Economic Journal 38, pp. 527–542.
282
Sectoral systems of innovation and production
APPENDIX Table A9.1 No
Cases of anti-dumping appeals by firms
Date of application
Items
Applicant
1
88.6.3
2
88.8.29
3
89.10.6
Dicumyl peroxide Daehwa Fine Chemical Alumina cement Union Corporation Polyacrylamide Eyang Chemical
4
90.5.8
Polyacetal
5
92.6.29
6
92.7.3
7
92.7.6
8
93.1.29
9
93.3.12
10
93.8.9
11
93.10.14
12
94.4.6
13
94.11.14
14 15
95.6.8 95.8.8
16 17 18
Korea Engineering Plastics H-acid Pungkuk Oil Refinery Refined Korea Specially phosphoric acid Chemical Industry Association Ball bearing HanKook Precision Sodium carbonate Dong Yang Chemical Pre-sensitized Hoechst Industry printing plate E-glass fibre Vetrotex Korea
Machine Object part country ❍
Japan, Taiwan
❍
France
❍
UK, Germany, France US, Japan
China, India, Japan China
❍
Thailand
❍
China
❍
Japan
❍
US, Japan, Taiwan Russia
Disintegrated Daljay Chemistry calcium phosphates Sodium hydroxide Korea Soda Products Industry Association Zinc ingot Korea Zinc
❍
❍
95.9.18
Copper foil Duk San Metals Refined phosphoric Dong Bu Preyon acid E-glass fibre Korea Oribest
96.1.12 96.2.7
Ethanolamine Lithium battery
Korea Polyol Teckraf
❍
❍
❍
US, China, Belgium, France China, Russia, Kazakhstan, Uzbekistan Japan China US, Japan, Taiwan US Japan, US
Barriers and opportunities in the Korean case
Table A9.1
283
(continued)
No
Date of application
Items
19 20 21
96.2.13 96.3.4 96.4.3
22
96.6.14
23
96.6.28
24
96.7.22
25
96.8.7
26
96.10.26
27
97.1.3
28
97.2.4
29
97.2.10
EVA
30
97.4.4
Furfuryl alcohol
31
97.4.4
32
97.5.16
33
97.5.16
34
97.9.25
35
97.10.14
36
97.10.16
Medium-density fibreboard Carbonless selfcopy paper Hydroxypropylmethylcellulose Electric smoothing irons Silicon manganese Polyvinyl alcohol
37
97.12.8
Soybean oil
38
98.4.25
Lithium battery
Applicant
Machine Object part country
E-glass fibre Choline chloride Pre-sensitized printing plate Electric shavers
Hankok Fiber Kolin Chemical U-IL Corporation Kaiser
❍
Disodium carbonate Industrial air handling unit Disodium carbonate H-beam
Dong Yang Chemical Seobun Engineering Dong Yang Chemical Kangwon Industry Seobun Engineering Korea Lighter Industries Cooperative Teayoung Chemical Samsung Fine Chemicals Hansol Forem
❍
Netherlands, Germany, Japan, China China
❍
Switzerland
❍
Bulgaria
❍
Russia
❍
Liechtenstein
❍
China
❍
Taiwan
❍
China
Hansol Patech
❍
Germany, UK
Samsung Fine Chemicals Cobalt Electronic
❍
US, Germany
❍
Singapore, China, France China
❍
Japan
Industrial air handling unit Disposable lighters
Dongbu Corporation Dong Yang Chemical Korea Soybean Industries Association Sanyo Electric
❍ ❍
Japan, US US, China Japan
US, Malaysia
Argentina, US, Brazil ❍
Japan
284
Sectoral systems of innovation and production
Table A9.1
(continued)
No
Date of application
39
98.5.7
40
Items
Applicant
98.9.30
Pre-sensitized printing plate Particle board
Agfa Korea Hansol Forem
41
98.10.10
E-glass fibre
Vetrotex Korea
42
99.2.12
43
99.3.20
Compound sizing agent Disposable lighters
44
99.3.29
Kong Young Chemical Korea Lighter Industries Cooperative Agfa Korea
45
99.8.31
46
99.9.13
47
99.11.15
48
00.5.26
49
00.6.2
50
00.6.19
51
00.7.14
52
00.9.8
53
00.11.1
54
00.12.28
55
00.12.29
Machine Object part country ❍
❍
Japan Indonesia, Malaysia, Thailand US, Japan, Taiwan Japan
❍
China
Pre-sensitized printing plate Aluminium cans Deahan Steel and easy-open end Alkali manganese STC Corporation batteries
❍
Netherlands
Disodium carbonate Pre-sensitized printing plate Bicycle and components Medium-density fibreboard Medium-density fibreboard Carbonless selfcopy paper Electric smoothing irons Cotton yarn
Dong Yang Chemical AS Ink & Chemical Korea Bicycle Association Korea Wood Panel Association Korea Wood Panel Association Hansol Patech
❍
Singapore, China, Japan, US China
❍
Japan
❍
China
❍
Germany
Bubang Techron
❍
Spinners and Weavers Association Korea Hansol Patech
France, China, Singapore India, Indonesia, Pakistan
❍
Japan, China, Indonesia, US, Thailand
Carbonless selfcopy paper
Taiwan ❍
Thailand, Indonesia Malaysia
Barriers and opportunities in the Korean case
Table A9.1
285
(continued)
No
Date of application
56
01.5.31
57
01.6.16
58
01.7.3
59
01.7.4
60 61
01.10.12 01.12.6
62
02.1.8
63
02.2.4
64
02.4.26
65
02.5.2
66
02.6.27
67
02.6.29
68
02.8.27
69
02.9.30
70
02.10.8
71 72
02.10.8 02.10.12
Items CD-R (compact disc recordable) Pre-sensitized printing plate Pre-sensitized printing plate Disposable lighters
Applicant
Machine Object part country
SKC
❍
Taiwan
Europe Graphic
❍
Japan
Agfa Korea
❍
Japan
❍
China
❍
Russia China
❍
China
❍
Taiwan
❍
China
❍
China
❍
Japan
❍
US, Germany, France, Poland
❍
Vietnam, Indonesia
❍
Indonesia, China China
Korea Lighter Industries Cooperative H-beam INI Steel White Portland Union cement Corporation Disodium G-Won carbonate International Ethylene-vinyl Air Product acetate Korea Alkali manganese STC batteries Corporation, Rocket Electric Disposable Korea Lighter lighters Industries Cooperative Aluminium Hankuk hydroxide Composite Chemical 2-ethylhexyl LG Chem, Korea alcohol Petrochemical Industry Association Disposable Korea Lighter lighters Industries Cooperative Uncoated wood- Shinho Paper, free paper Hansol Paper Ferro-silico Hannong Co manganese Polyvinyl alcohol DC Chemical Alkali manganese STC batteries Corporation, Rocket Electric
❍ ❍ ❍
Japan China, Japan, Singapore
286
Sectoral systems of innovation and production
Table A9.1
(continued)
No
Date of application
73
03.4.30
74
03.5.12
75
03.5.14
76
03.6.9
77
03.7.7
78 79
03.7.16 03.10.31
80 81 82
04.3.9 04.4.6 04.7.15
83
04.8.19
84 04.12.20 Total Source:
Items
Applicant
Korea Wood Panel Association Stainless steel bar Kumkang Industrial Sodium silicate Rhodia Silica Korea Sodium dithionite Hansol Chemical, PooHung Photo Chemical Hyundai Heavy Industrial robot Industries with a six-axes vertical multiarticulation structure Lithium battery Vitzrocell Choline chloride Kofavet Special
Machine Object part country
Particle board
Titanium oxides PVC Plate Pre-sensitized printing plate Industrial robot with a six-axes vertical multiarticulation structure H-beam
❍
Belgium, Spain, Italy Japan, India, Spain China China
❍
Japan
❍
Cosmo Chemical Crown Agfa Korea
❍ ❍
Japan, US US, India, China, Canada China Japan Japan
Hyundai Heavy Industries
❍
Japan
❍
Russia
INI Steel
Korean Trade Commission (www.ktc.go.kr).
❍
75 %
10.
From ‘nuts and bolts’ to ‘bits and bytes’: the evolution of Taiwan ICT in a global knowledge-based economy Ting-Lin Lee
1. 1.1.
INTRODUCTION Motivation and Objectives
As we have entered the information age and developed a digital economy, with the rapid development of satellite communications, the universal penetration of the internet, and the emergence of intelligent industries, ‘speed’ and ‘innovation’ have become the main factors spurring industrial development. Taiwan, with its universal education, highly educated citizenry, high-quality workforce, and facility in cultivating technical personnel, is very well positioned for the development of knowledge-intensive hightechnology industries. This is where Taiwan’s comparative advantage lies. Industry output by value of the ICT hardware sector and IT software sector respectively was US$684.1 hundred million and US$49.4 hundred million in 2004. The proportion of ICT expenditure to GDP was then 1.7 per cent. And, according to a survey conducted by FIND, the number of internet subscribers (internet access accounts) in Taiwan reached 9.98 million as of December 2004. In 2004, 61 per cent of households in Taiwan were connected to the internet, and 81 per cent of Taiwan enterprises had internet access. The bandwidth used for international internet connection in Taiwan exceeded 70 Gbps; there were over 5 million mobile internet subscribers in Taiwan; and the Taiwan government offered 847 government services online as of the end of 2004 (FIND, 2005). Taiwan’s ICT policy developments have had fruitful outcomes. In June 2002, the Taiwan government proposed the Two Trillion Twin Stars programme to establish digital content as one of the industries with an annual production value of over NT$1 trillion. Facing the changes of the digital 287
288
Sectoral systems of innovation and production
world, the Taiwan government has actively worked to promote digitization through a number of initiatives in recent years to improve the nation’s IT proficiency and the competitiveness of domestic IT industries. In May 2002, NICI (National Information and Communications Initiative Committee) and other government agencies worked together to launch the e-Taiwan Program as a part of the Challenge 2008 Program. With the need for a sound e-business framework and application standards, DoIT of MOEA commissioned ACI of III1 to undertake long-term research and promotion work with regard to the E-Business Standard Research Plan. Besides, in order to continually strengthen the enterprises’ digital capacities, Projects A, B, C, D and E aim to create an e-business supply chain system and lay the foundations for a new business model where “orders are received in Taiwan; production can take place anywhere in the world”. In addition, the Taiwan government proposed the M-Taiwan Program to promote a ubiquitous network and e-services in Taiwan with a budget of NT$37 billion over five years. Fuller explanations of initiative policies for ICTs will be presented in the following section. Meanwhile, in contrast to the situation of other countries, the most special aspect of Taiwan lies in its deep historical origins and geo-relations with Mainland China, which cause both sides to have an inseparable relation in economic development, even though they have opposing stances in politics at the present stage. All these complex factors help shape the intricate appearance of the development process in the formidable Taiwan ICT industry. Later I shall try to give a more precise account of the patterns of Taiwanese ICT investment in Mainland China. 1.2.
Methodology and Analytical Framework for the Taiwan Study
This case study describes the development and transition of the whole knowledge-based economic society in Taiwan, based on an analysis of the structure of the National Innovation System (NIS) from the viewpoint of coevolution. In this study we shall focus on a particular industry (especially ICTs), in a particular country, Taiwan, and thus a hybrid of national and sectoral systems. In the Taiwan case, the sectoral industry (ICTs) is embedded in the “national” innovation system, and is nurtured and restricted by it. Moreover, the base and development of ICTs industry in Taiwan originated from the IT industry’s prosperity (as with the computer hardware industry). The definition of the ICT industry under this study mainly includes two parts: information and electronic (computer, motherboard, etc.), telecommunication (telephone, modem, internet, etc.) and their peripheral equipment and software. This term is sometimes extended to cover the notion of the OECD’s definition. From the historical viewpoint,
The evolution of Taiwan ICT in a global economy
289
the rise and development of the telecommunication industry is based on the infrastructure, the technology capabilities and its knowledge accumulated by the IT industry. It is hard to separate from the historical aspect. In general, we can use IT to explain the concept of ICT partially. Although they are not the same in nature, they have common components. Section 2 starts by sketching Taiwan’s economic growth. Section 3 analyses the changes in the Taiwan NIS. Specific attention is devoted to the range of indicators, and to assessing the related actors, agencies and organizational and institutional changes, as well as the conditions for knowledge-based techno-economic growth observed in the period 1990–2005 (approximately). Section 4 is devoted to the major industrial policies and governance in Taiwan. This part of the study explores the related Taiwanese ICT programmes and other related initiatives: campaigns at the national level to increase the level of ICT usage, FDI policies, and financial and non-financial incentives to attract higher participation among local and foreign companies to invest in the ICT industry. Section 5 addresses the role played by ICT. This section covers the ICT sector’s evolution, availability of ICTs that allow access to knowledge and communication, in particular the internet (fixed, mobile, wireless, etc.), and access to and usage of ICTs (phone lines, number of personal computers, internet users, internet hosts, etc.). Section 6 aims to describe possible strategies that Taiwan’s ICT industry may choose in the face of the tendency to internationalization and globalization, and explain how the development of Taiwan’s ICT industry should be adjusted. The chapter finishes with the offshore development of Taiwan’s ICT/IT in Mainland China in section 7. This section aims to assess how Taiwan IT manufacturers launched investment patterns, the forms of division of labour, and the arrangements of production, R&D and marketing, from the viewpoint of evolution.
2. 2.1.
GROWTH DRIVERS SINCE 1990 Industrial Structure Change
To understand the development of the national industrial system, it will be most insightful to break down Taiwan’s economic development into three stages, as in Table 10.1 (Yu, 1999, pp. 6–22). For many years the government has encouraged industry to engage in technology development. For instance, every year the Ministry of Economic Affairs (MOEA) provides funding to non-profit research institutions for the development and transfer to the private sector of “critical, common and forward-looking” technologies, and allows public and private enterprises
290
Sectoral systems of innovation and production
Table 10.1 Period
Economic growth and structural change in Taiwan Economic growth rate (%)
1952–80 (68SNA) 1982–95 (68SN) 1995–2004 (93SNA) Source:
Agriculture (%)
Industry (%)
Services (%)
GDP Sectoral GDP Sectoral GDP Sectoral percentage growth percentage growth percentage growth share rate share rate share rate
9.21
19.68
4.24
31.34
12.39
49.00
9.08
7.86
5.01
1.33
40.17
6.18
54.82
9.33
4.5
2.13
–0.67
29.09
3.77
68.77
5.01
DGBAS (2006).
to participate in technological research projects. In order to encourage and support industrial R&D, the NSC has promoted the Research and Development of Key Parts, Components and Products Program within the Hsinchu Science-based Industrial Park, and also regularly provides market information and technical assistance in order to reduce market risk and stimulate industry’s willingness to engage in R&D work. 2.2.
Trends in Productivity
The inferior performance of Taiwan’s labour productivity is associated with the industrial structure and business operation mode. Taiwan’s industrial structure is largely concentrated on so-called ICT domain in the “high-tech industry”, although in the past, owing to the fast growth of the ICT industry, Taiwan’s overall economic growth was supported. However, in recent years the IT hardware industry has been facing the threat of the latecoming countries, the quick drop of product prices and the constraint of large international companies; thus the overall ICT industrial added value and gross profit have been gradually reduced. Additionally, the low growth of service sectors and the low growth rate of labour productivity in recent years, which is even lower than the growth rate of the manufacturing labour productivity (see Figure 10.1), have caused the insufficient growth of Taiwan’s overall labour productivity (Lin and Lin, 2005). 2.3.
Trends of Venture Capital and FDI
At the initial stages, as Taiwan started to develop in the high-tech field, there was no venture capital business. Thus, the government established
The evolution of Taiwan ICT in a global economy real growth rate % 14
10
GDP Growth Rate Service Industry Manufacturing
12.10 11.30
12
9.60
9.2 9.20
6
7.00 4.90
3.70
4
9.36 8.25
7.2
8
291
5.8
5.50 4.40
3.61
4.77
5.71 4.83
3.31 3.30
2 0 1971–1987
Source:
1988–1994
1995–1999
2000–2004
2004
DGBAS (2006).
Figure 10.1
Average growth rate by sector in Taiwan, 1971–2004
the Development Fund of Executive Yuan in 1973 to invest in venture capital and coordinated Chiang Tung Bank to provide refinance for venture capital. However, the preliminarily introduced venture capital created only a few successful cases that led the industrial development. The first venture capital company was established in 1984 after reinvestment by Acer. The number of venture capital companies increased slightly to the early 1990s (Wu et al., 2002). The number of venture capital firms in Taiwan then grew to 259 in 2004. Their accumulated capital increased from NT$200 million in 1984 to NT$ 184.5 billion in 2004, growing over 922 times within two decades. The investment in electronic and electrical appliances by overseas Chinese and foreign companies in Taiwan has had a tendency to rise gradually in recent decades. The year 2000 was a watershed, reflecting the first rotation of political party in Taiwan, when FDI attained high levels. However, the political and economic situation after the political change (such as political infighting among the parties, the economic emergence of Mainland China and India, etc.) was not as good as anticipated, and the cases and amounts of overseas Chinese investment and FDI have reduced year by year since 2001. In 2004 the total of approved cases dropped to the minimum since 2000 but increased considerably in amounts. This phenomenon was mainly because the government opened up an increasing range of sectors to foreign participation. This was particularly so in ICT, where there were changes to foreign investment regulations, particularly in foreign ownership levels in the telecommunications sector.
292
3.
Sectoral systems of innovation and production
THE TRANSFORMATION OF TAIWAN’S INNOVATION SYSTEM
The government played a central role in the transformation of the economy and society. It has implemented strong and coherent planning mechanisms for the economy, science and government, and for close collaboration with the private sector, and has made a heavy investment in education, research and infrastructure. The Ministry of Economic Affairs (MOEA) is chiefly responsible for industrial technology applications research, and it transfers the results of research to the corporate sector for product development and commercialization via technical assistance, information sharing and manpower training. To accomplish these goals, the MOEA relies on its own subordinate research organizations, the research departments of state-owned enterprises, and research organizations hired on a case-by-case basis. Industrial technology development work is conducted primarily via in-house R&D and secondarily via technology acquisition. The MOEA is working to strengthen interaction between industry, government, universities and research institutions as a means of promoting technological upgrading throughout the industrial sector. Besides, unique institutional arrangements have been made, such as the quasi-governmental Institute for Information Industry (III), which serves as a think-tank and research centre for both government and business. Furthermore, government-constructed science parks support innovating and the incubation of new ideas, build synergies among growing businesses, and make efficient use of the best available human resources, facilitating growth and wealth generation (see Figure 10.2). 3.1.
Trends in R&D Input
According to the annual Survey of National Science and Technology Activity, total public and private R&D expenditures amounted to NT$197.6 billion, or 1.94 per cent of GNP, in 2000, and NT$260.9 billion, or 2.34 per cent of GNP, in 2004. Furthermore, the nation’s total R&D spending as a proportion of GDP remained at an average of around 2.2 per cent during these five years. R&D expenditure as a percentage of GNP showed a gradual increase during the years from 1996 to 2004, although the growth rate of R&D expenditures declined sharply in 2000. However, the annual average growth rates of these expenditures are only 6.5 per cent in the five most recent years. Of overall national R&D spending of NT$260.85 billion in 2004, government agencies contributed NT$88.47 billion (33.9 per cent), business enterprise NT$168.1 billion (64.4 per cent), higher education NT$3.1
The evolution of Taiwan ICT in a global economy Government
In-house Research & Development
Academic Research
National Science Council Tax Credit
University
Ministy of Economic Affairs Subsidiary or Matching
Industry Technology Consultation Personnel Training
Source:
Technology Development Projects Research Institutes/ University
Technology Transfer
Technology Transfer
Technology Acquisition
293
Overseas Enterprises or Research Institues
Technology Consultation Information Supply Personnel Training Technology Acquisition/ Joint Research/ Foreign Technological Cooperation and Investment/ Strategic Alliance
National Science Council (2002).
Figure 10.2
The interactions model among actors for promoting industrial technology upgrading
billion (1.2 per cent), private non-profit organizations NT$1.1 billion (0.4 per cent) and foreign institutions just NT$60 million (0.0 per cent). As described, 64.4 per cent of R&D funds came from enterprise in 2004. The proportion of enterprise R&D funds has tended to fall, which differs from the increasing tendency in advanced countries, and Taiwan even lacks overseas R&D investment (certain countries get 10 per cent from overseas R&D capital). Similarly, according to the performing of R&D, the proportion by enterprises also has a tendency to fall year by year. 3.2.
Education System and Human Capital
Taiwan had a total of 162 higher education establishments in the 2005 school year (from August 2005 to July 2006), 89 of which were universities. The student population of higher education for the same year was 938 648 students, 449 695 of whom belonged to the science and technology field, including 2165 doctoral students and 42 334 master’s students (Ministry of Education website: www.edu.tw, 2006). Compared with other nations, Taiwan has a higher registration ratio, with education and training in accord with national competitiveness requirements (Tzeng and Lee, 2001). This advanced educational achievement has been one of the primary factors in the vigorous development of the information industry.
294
Sectoral systems of innovation and production
Table 10.2 Item/year
Numbers of Taiwanese researchers, 1996–2004 1996
1997
1998
1999
2000
2001
2002
2003
2004
Numbers of 53 754 56 419 62 586 67 165 69 525 73 239 80 999 85 166 91 490 researchers (persons) Researchers/R&D NA NA NA 49.8% 50.5% 52.9% 53.9% 54.2% 54.3% personnel (%) Researchers per 2.5 2.6 2.9 3.0 3.1 3.3 3.6 3.8 4.0 1000 population Researchers per 5.8 6.0 6.6 7.0 7.1 7.4 8.1 8.5 8.9 1000 labour force
Source:
National Science Council (2005a, Table 1-19, 1–22).
An overview of the Taiwan educational system reveals that, in 2004, the government allotted 39.0 per cent of its budget to higher education. Public education enrolment rates reached 99.2 per cent in the 2004 school year, an achievement that compares favourably with that of other nations. Concerning the S&T indicators, Taiwan is rich in human resources. The number of researchers both per 10 000 of the population and per 10 000 of the labour force constantly increased over the four years up to 2004 (Table 10.2). R&D personnel include researchers, technicians and supporting personnel. A total of 168 524 people were engaged in R&D work in 2004, and maintained an increasing trend over the preceding years. The total for 2004 included 91 490 researchers (54.3 per cent), 59 583 technicians (35.4 per cent) and 17 451 supporting personnel (10.4 per cent). In terms of the distribution of researchers, 50 795 (47.3 per cent) worked in business enterprise, 17 020 (50.4 per cent), in the government sector and 22 781 (87.7 per cent) in higher education. Among these researchers, 23 306 (25.47 per cent) held Ph.D. degrees, and 38 912 (42.53 per cent) master’s degrees. Similarly, the percentage of researchers with master’s and Ph.D. degrees has also increased. The proportion of researchers among all R&D manpower has hovered around 56.3 per cent over the past ten years. 3.3.
Patents and Publications
In Taiwan, basic research is chiefly conducted at the Academia Sinica, the national laboratories, various research centres, and university departments and graduate schools. The number of academic papers published is a direct
The evolution of Taiwan ICT in a global economy
Table 10.3
Domestic patents applied for and granted, 2001–04
Item
Patents applied for
2001 2002 2003 2004 Source:
295
Patents granted
Total
Compatriot
Foreigner
Total
Compatriot
Foreigner
67 860 61 402 65 742 72 082
40 210 35 926 39 663 43 020
27 650 25 476 26 079 29 062
53 789 45 042 53 034 49 610
32 310 24 846 30 955 33 517
21 479 20 196 22 079 16 093
National and Science Council (2005a, Table 7-2).
indicator of basic research. The number of Taiwan’s papers cited in the SCI database has increased every year. In 2004, there were 14 989 articles cited by Taiwan’s authors in the SCI database, ranking at 19 in the world. Outside of universities and colleges, most of Taiwan’s engineering and applied research is conducted at the Industrial Technology Research Institute (ITRI) and other public or non-profit research institutes (such as III for ICT). Over the past decade, an excellent research record has been achieved in such areas as electronics, information, communications, materials science, biology, agriculture and food technology. There were 10 983 papers by Taiwan’s authors in the EI database in 2004, ranking 11 in the world. Another tangible result of research on science and technology has been the number of patents granted. Of the patents approved in 2004, 68 per cent were by Chinese nationals and 32 per cent by foreign nationals. This represents a substantial increase in the number filed by Chinese nationals (Table 10.3). As for innovation patents, in 2004 there were 41 919 invention patent applications, of which 20 454 were approved. The number of patents granted in the US to assignees in Taiwan has increased rapidly, as have patents granted in Taiwan, although a large share of the patents in Taiwan is granted to foreigners. Table 10.4 shows indicators of patents of the period 2000–04 (NSC, 2005:33-34). Table 10.5 shows indicators of research output of the period 2000–04 (NSC, 2005:33–34). 3.4.
The Role of Research Institutes
Since its inception in 1973, the ITRI has played a major role in upgrading Taiwan’s industrial technology. ITRI created Taiwan’s semiconductor industry from scratch, led the development of other high-tech industries, and helped traditional industries raise productivity. ITRI fosters young companies and new technologies until they are able to survive on their
296
Sectoral systems of innovation and production
Table 10.4
Patents, 2000–04
Item/year Invention patents applied Compatriot Foreigner Invention patents granted Compatriot Foreigner Patents granted in USPO
2000
2001
2002
2003
2004
28 451
33 392
31 616
35 823
41 919
6 830 21 621 15 657
9 170 24 222 24 429
9 638 21 978 23 036
13 049 22 774 25 134
16 747 25 172 20 454
3 834 11 823 4 667
6 477 17 952 5 371
5 683 17 353 5 431
6 399 18 735 5 298
7 521 12 933 5 938
Sources: 1. National Science Council (2005a, Table 7-3, 12). 2. Taiwan Intellectual Property Office, TIPO website, 2006.
own. Equally, since 1979, the Institute for Information Industry (III) has been a key technology contributor to Taiwan’s ICT industry. Its founding and continuing mission has been to increase Taiwan’s global competitiveness through the development of its IT infrastructure and industry. In order to support leading-edge research and speed up the pace of innovative breakthroughs, Taiwan has established a series of open-type national laboratories. The Taiwanese government has always promoted cooperation between industries and universities in recent years; however, several problems remain in the cooperation between industries and universities (this theme will be discussed in the following section), so the interaction between industries and universities is largely confined to the supply of talent. The share of enterprise investment in higher education R&D is much behind other countries. 3.5.
The Role of Universities
The Taiwanese government has always promoted cooperation between industries and universities in recent years; for example, the TDP for Academia which the Industrial Development Bureau in MOEA brought out is a best-policy action scheme. However, there are still several problems in the cooperation between industries and universities at the present stage: ● ●
The cognitive lag is great between industries and universities. Channels of communication between industries and universities are lacking.
297
110
103
100
100
101
GDP deflator
66 715
65 091
61 548
58 330
88 119
Cost (NT$m) 51 009 (579) 42 587 (730) 46 770 (760) 46 586 (716) 78 763 (1 210)
Papers 2 808 (32) 1 327 (23) 1 194 (19) 1 024 (16) 1 774 (27)
Patents 6 974 (79) 6 974 (120) 8 158 (133) 9 150 (141) 8 740 (131)
Technical reports 321 (4) 57 (1) 64 (1) 3 828 (59) 2 632 (39)
Copyrights 409 (5) 271 (5) 693 (11) 474 (7) 1 958 (29)
Technological innovations 70 (1) 60 (1) 342 (6) 142 (2) 32 (0.48)
Technology acquisitions
1 255 (14) 1 091 (19) 1 173 (19) 1 962 (30) 1 534 (23)
Technology transfers
30 295 (344) 32 990 (566) 53 333 (867) 61 770 (949) 23 615 (354)
Technical services
Source: National Science Council (2004a, p. 34).
Note: Figures in brackets are results indicators expressing the relative quantity of results obtained for each NT$1 billion of input. Since there is a one-year time lag between resource input and result output for the items of “papers”, “patents” and “technology transfers”, the results indicators for these items are consequently expressed as results (papers/items)/ cost during the previous year; the remaining items are calculated on the basis of cost during the current year.
73 213
61 579
2002
2004
58 330
2001
66 738
89 317
2000
2003
Funding (NT$m)
Research outputs indicators, 2000–04
Year
Table 10.5
298
Sectoral systems of innovation and production
Table 10.6
OECD USA Germany UK France Japan Korea Singapore Taiwan Note:
●
● ●
1998 %
1999 %
2000 %
6.0 6.1 10.5 7.3 3.4 2.3 13.1 4.1 3.0
6.1 6.2 11.3 7.3 3.4 2.3 10.8 5.3 4.8
6.2 6.0 11.6 7.1 2.7 2.5 15.9 6.0 4.1
2001 % 6.1 5.5 12.2 6.2 3.1 2.3 14.3 4.3 3.2
2002 %
2003 %
5.8 4.9 11.8 5.8 2.9 2.6 13.9 2.5 3.3
5.7 4.5 12.1 5.6 – 2.7 13.6 4.0 4.2
The sector classification is in accordance with the Frascati Manual, OECD.
Sources:
●
Proportion of R&D funds in higher education coming from enterprises
1. OECD (2005). 2. National Science Council (2004b).
The research results are hard to commercialize. Professors’ studies tend to be basic research, which does not fit the needs of industry, and the system of professorial promotion is inflexible. Incentive mechanisms are insufficient to encourage scholars in academic circles to engage in industry–university cooperation. No matter what the organizational scale or funds, the research in universities is insufficient.
Therefore, the interaction between industries and universities is largely confined to the supply of talent. The proportion of enterprise investment in higher education R&D (HERD) falls greatly behind that of other countries, which shows that the linkage is still not enough between industry and universities; this phenomenon will result in the insufficiency of innovation sources in industry, capacity being hard to promote, and the R&D results from universities unable to be commercialized (see Table 10.6).
4. 4.1.
MAJOR POLICIES AND GOVERNANCE Existing Industrial Policies
A dynamic, entrepreneurial and flexible private sector, made up largely of SMEs, has flourished in this surrounding of government encouragement,
The evolution of Taiwan ICT in a global economy
299
while both good governance and sound macroeconomic management by the government have earned the confidence of business. Next we will describe the related initiative, policies and its contents. 4.1.1. Tax credits for R&D and personnel training The MOEA is chiefly responsible for industrial technology research and its application. Apart from the MOEA’s directly subordinate research units, the R&D department of state-owned enterprises and independent research institutes undertaking commissioned projects are also engaged in industrial R&D and technology transfer. Research institutes are employing technology acquisition, joint research, foreign direct investments and strategic alliances to interact with foreign companies, and research institutes art as mechanisms to accelerate the industry’s technological development. In addition, the government relies on administrative measures like subsidies, matching grants and investment tax incentives to encourage industry to engage in R&D activities. 4.1.2. Small Business Innovation Research (SBIR) In accordance with the Knowledge Economics Development Act, the DoIT of MOEA launched Taiwan’s SBIR promoting programme, mostly referring to the US version of the SBIR, in November 1998, in order to enhance the private sector’s R&D competitiveness through promoting technological innovation and utilizing information technology on one side, and providing tax incentives and a subsidy of up to half of the cost of development and matching funds to resolve market failures and uncertainties of technology development on the other side. The types of research encouraged by this programme include: 1) developing a brand new idea, concept or new technology; 2) applying an existing technology to a new application; 3) applying a new technology or business model to an existing application; and 4) improving an existing technology or product in various aspects. By 2010, the SBIR promoting programme may assist in achieving the nationwide goal of Taiwan’s R&D rising to 3 per cent of GDP, and private sector R&D increases of up to 60 per cent, including 70 per cent from knowledgebased industries. 4.1.3. Technology Development Programs (TDP) The DoIT is Taiwan’s science and technology development flagship. Its primary mission is to promote industrial technology development to help create new national industries and help upgrade Taiwan’s existing industries. Therefore, in cooperation with the Executive Yuan’s promotion of the scientific and technical development scheme, the government began to
300
Sectoral systems of innovation and production
Table 10.7
TDP expenditures, 2001–04
Fields of R&D
2001
Telecom and optoelectronics Machinery and aerospace Materials and chemicals Biomedical Pioneer innovation programme
2002
2003
2004
37.48 (23.63%) 30.62 (19.3%) 22.86 (14.41%) 14.11 (8.89%) 32.52 (20.50%)
35.73 (20.78%) 26.10 (15.18%) 26.25 (15.27%) 18.14 (10.55%) 34.58 (20.12%)
137.59 (86.73%) 19.55 (12.32%) 1.5 0.95% 158.64
140.81 (81.93%) 27.35 (15.91%) 3.72 (2.16%) 171.88
34.85 (19.26%) 26.44 (14.61%) 26.85 (14.84%) 18.23 (10.07%) 19.81 (10.95%) 17.04 (9.42%) 143.22 (79.14%) 30.82 (17.03%) 6.94 (3.83%) 180.98
Others TDP for corporations TDP for the private sector TDP for academia Total Note: Source:
135.64 (89.4%) 16.08 (10.6%) 0 0% 151.73
Unit: NT$ million DoIT/MOEA (2004).
implement “the given-case program of MOEA scientific and technological research development” in 1979 (TDP for short). The DoIT takes charge of examining and allocating the subvention funds as the main overall promotion unit of scientific and technological given cases. In the initial stage of planning, the contracted research institutions (TDP-contracted research institutes for short) are authorized to assist in industrial innovation, introduce each perspective, and critical and compatible technology, and bring about the cooperation between manufacturing and studying in order to help industries upgrade and change their type, strengthen their innovative R&D abilities, and increase their international competitiveness. To sum up, TDP-contracted research institutes are to adjust domestic industries to R&D innovation and prospective technology. On the contrary, the non-profit research institutes, involving DoIT contracts with the private sector and academic organizations respectively, carry out and develop basic and pioneer technologies that are then licensed to Taiwan’s industries.
The evolution of Taiwan ICT in a global economy
4.2.
301
Major New Policies
The Executive Yuan’s current main tasks of policy implementation are to speed up the execution of major national development projects and to advance toward the goals set out in the Challenge 2008 Six-Year National Development Plan. 4.2.1. Deregulation of telecommunications Taiwan, in officially becoming a WTO member on 1 January 2002 aims to implement the accession commitments and continue forwarding telecommunications liberalization policies. The liberalization of telecommunications in Taiwan is an outgrowth of two policies, those regarding the Asia-Pacific Regional Operations Center and the National Information Infrastructure, and is opening up the island’s telecommunications market through a staged progression. In the first step toward liberalization, the ownership of terminal equipment by subscribers was opened up in 1987. Later, in 1989, the step taken was the opening of the market to value-added services so as to provide consumers with a diversity of such telecommunications services. The passage of three telecoms-related laws in 1996 led to the formal separation of the Directorate General of Telecommunications (DGT), which is in charge of telecommunications industry regulation, and the Chunghwa Telecom Co., which is responsible for operating the telecoms business. This separation more firmly established the policy directions for liberalization, and later further liberalization steps were taken, particularly in services of mobile telecommunications and satellite telecommunications. Since 1999, liberalization has continued in various fields of services, such as integrated fixed network telecommunications, international submarine cable leased circuit, local and long-distance leased-circuit cable, resale business, and third-generation mobile telecommunications (3G). The short-term objective of telecom liberalization is thus completed. After releasing 3G mobile telecommunication business to the public in 2002, the government released all telecommunication business, and Taiwan’s telecom market has moved to full liberalization. There is one thing worth a mention in passing. In view of the global development of digital convergence and the integration of the authorities regulating the telecommunications and broadcasting sectors, the government held its eighth Strategy Research Board (SRB) meeting in 1998 and proposed the establishment of one regulatory body to oversee telecommunications, information and broadcasting sectors within an integrated framework. The enactment of the Fundamental Communications Basic Act on 7 January 2004 and the National Communications Commission
302
Sectoral systems of innovation and production
Organization Act on 9 November 2005 enabled the establishment of the National Communications Commission (NCC) on 22 February 2006, thereby creating a governmental body for the regulation of the telecommunications, information and broadcasting sectors. In general, telecommunications liberalization policy in Taiwan has introduced competition mechanisms successfully, revitalizing the telecommunications industry structure and leading to the effective growth of the telecommunications business. However, the ratio of telecommunications revenues to GDP, though gradually increasing, is still below the world average of 3.4 per cent, indicating that the domestic telecommunications market is not yet fully expanded. 4.2.2. Projects A, B, C, D and E (FIND, 2004) Recognizing the importance of information technology towards the upgrading of Taiwan’s industrial competitiveness, in June 1999 the Executive Yuan expanded its existing industrial automation project into a new Industrial Automation and Electronic Business: iAeB Program. While continuing to promote automation in production, warehousing, transportation and sales, the MOEA was instructed to give priority to the establishment of B2B (business-to-business) e-commerce systems, in order to build model e-business systems for both the supply chain and the demand chain. In 1999 the DoIT formulated and began implementation of two pilot projects for promoting e-business in the IT sector – Projects A and B. After implementation of Projects A and B had been completed, in 2001 the MOEA began implementation of Projects C, D and E as a continuation of Projects A and B. The aim of these new projects was to ensure the provision of e-business services covering payment, accounts receivable management, online financing, global inventory management, delivery tracking and collaborative design services in order to maintain the competitive advantage of Taiwanese industry and meet industry’s evolving needs. The existing e-business supply chain system would be used as the foundation for further integration of cash flow, delivery systems and engineering collaboration, with the aim of strengthening the global logistics management capability of Taiwanese industry and its competitiveness in international markets (Figure 10.3). I would like to lay special emphasis on Project E, which helps manufacturers to transform themselves from OEM to ODM and to CDM (contract design manufacturer), and on strengthening collaborative design R&D management capability. The idea is to integrate different companies’ R&D capabilities at the product development stage, developing collaborative design business models that can meet the needs of the leading international IT vendors, to help to reduce lead time and time to market, and to facilitate
The evolution of Taiwan ICT in a global economy
303
Customers
Project E Project D
Delivery
Project A
Leading Domestic IT manufactures
Project E
Project C Bank
Project B Cash
Suppliers Source:
IT Applications Promotion Project, III, sponsored by DoIT, MOEA.
Figure 10.3
Relationship between Projects A, B, C, D and E
the smooth exchange and sharing of product information and the rapid solution of the problems. It is clear that actors and relationships involved in projects A, B, C, D and E include buyer–supplier and complementary companies (such as financial banks, logistics firms, etc.); knowledge flow among them is raised up to the period of “ideas design”, except for cash flow (see Figure 10.4). 4.2.3. Two Trillion Twin Stars programme Aiming to build Taiwan as a “green silicon island”, the Taiwan government brought out a six-year national development plan in mid-2002. To fulfil that goal, in 2002, the MOEA ran a four-year Two Trillion Twin Stars programme from 2002 to 2006. The programme drove the production value of Taiwan’s relatively mature semiconductor and flat-panel display (TFT-LCD in particular) industries to NT$1 trillion (US$29.6 billion) for each – hence the term “Two Trillion”. Actually the production value of Taiwan’s flat-panel display industry went beyond NT$1 trillion in October of 2006 and reached to NT$1.285 trillion by the end of 2006. Followed by the recession of the global economy in 2008–2009, however, the flat-panel industry has shown a poor performance. It also built the new digital content and biotechnology sectors into “star industries” – hence the name “Twin Stars”. The ‘Two Trillion’ (semiconductor and flat-panel display) are already relatively mature, while the ‘Twin Stars’ (biotechnology and digital content) are new and full of potential.
304
Sectoral systems of innovation and production
E Plan
AB Plan
(Collaboration design) (Aug. 2001–June 2004)
(e-Procurement) (Jan. 2000–Dec. 2001)
International Customers
• Product forecasts • Demand specifications
‘Center’ • Multi-location integration • Design changes manufacturers • Materials sharing • Module design
• Product forecasts • Electronic ordering • Customer order management • Materials management • Inventory management
Suppliers
D Plan
C Plan
(Delivery) (Cash flow) (Aug. 2001–Dec. 2003)
• International storage • Inventory management • Cargo tracking • Warehouse management • Transport and customs clearance
Logistics providers
• Bill presentation and payment • Capital allocation • Account consolidation • Multibank finance • Interbank finance • Financial instruments
Banks
Design
Source:
Purchase
Manufacture
Delivery
Billing
DoIT (2005).
Figure 10.4
Project architecture for Taiwan’s Projects A, B, C, D and E
According to Industrial Development Bureau, MOEA, the promotion of the Two Trillion Twin Stars programme begins with offering tax rewards, assisting in solving obstacles to investment such as land, water, electricity, environmental protection and so on, integrating the government and industrial resources, developing new products and new technology, and establishing the complete upstream and downstream industry system, as well as talent training, and in these ways improving the competitiveness of the industry overall. As for the industry development goals of “Two Trillion Twin Stars”, the semiconductor industry reached production values of NT$ 1.393 trillion, flat-panel display industries NT$ 1.285 trillion, the digital content industry NT$ 0.361 trillion, and the biotechnology industry NT$ 0.178 trillion in 2006. 4.2.4.
Policies for encouraging demand (e-Taiwan, M-Taiwan)
E-Taiwan: march on with a new vision To embrace the global e-trend and confront all the challenges that cloud the future of Taiwan’s IT industry – the impact of the global knowledge-based economy, outward-moving business and decreasing total revenues, etc. – the Convenor of NICI of the Executive Yuan, working with many chief officers from other government agencies, leading academies, research institutions, top enterprises and civil
The evolution of Taiwan ICT in a global economy
305
organizations, has formulated the e-Taiwan Program to counter all these issues. The e-Taiwan Program was formally approved by the Executive Yuan in June 2002 and combined with nine other plans to form the so-called Challenge 2008: The Six-Year National Development Plan. Not to overstate its importance, the e-Taiwan Program holds the key to the complete success of Challenge 2008. There are five integral parts in this plan, that is 15 million broadband users, e-Society, e-Industry, e-Government and e-Opportunity. The first part, “15 million broadband users”, is expected to deliver the following results by the end of 2007: (1) the broadband network is fully installed with implementation of IPv6 and wireless LAN environment; (2) small and medium enterprises are mostly brought online; (3) safety standards, regulation, strategy and legislation are properly installed and in full operation; (4) IC security enforcement is strictly observed and capable of fostering the related industries; and (5) CA cards have been successfully issued and commonly accepted as a primary means of identification. M-Taiwan (FIND, 2005) The third IT revolution aims to forge personal computers, the internet and mobile communications into a “ubiquitous network”. By utilizing this network, the government, entrepreneurs and end-users are able to get the information they need by any device, at any time and anywhere – more efficiently, more conveniently, and giving better quality of life. With the advantages of the world’s No.1 production value of WLAN products and mobile phone penetration rates, the Taiwan government has actively promoted mobile competitiveness. The NICI committee of the Executive Yuan (Cabinet), Ministry of the Interior (MOI) and MOEA coordinated to propose the M-Taiwan Program, with a budget of NT$37 billion in five years. The M-Taiwan Program is expected to build up the wireless networks, integrate mobile phone networks, set up optical-fibre backbones, and execute the Integrated Beyond 3rd Generation (iB3G) Double Network Integration Plan. It is also expected to shift Taiwan from an ‘e-nation’ to an ‘m-nation’, and to reach the vision of “mobile Taiwan, infinite application, and a brave new mobile world”. 4.3.
Mode of Governance
4.3.1. Governance for business – moving to ODM Since the 1970s when foreign-owned firms started to invest in Taiwan, many Taiwanese manufacturers became their original equipment manufacturer (OEM). Since the 1980s when production of the first personal
306
Sectoral systems of innovation and production
computers began, under the open system policy of IBM, some Taiwanese manufacturers produced IBM-compatible computers with their own brands, such as Multitech by Acer (renamed in 1987). Since then, the percentage of Taiwanese manufacturers producing products with their own brands has increased only slightly. As observed by Huang (1995), since 1989 when the US economy was in recession, the growth rate of sales for Taiwanese computers with their own brand in the US market decreased almost to zero. Especially in June 1992, when Compaq announced reduced prices of all its series products by up to 30–40 per cent, branded computer manufacturers in Taiwan were forced to give up their own brands. However, at the same time, many European and American computer enterprises started to look for the OEM that was able to control production costs efficiently. Thus, Taiwanese computer manufacturers returned to the mainstream computer market by way of the foundry. At the same time, Acer also announced it was giving up its own brand and becoming the OEM again. This indicated that, before the 1990s, the competitiveness of Taiwanese manufacturers was already recognized in terms of vertical specialization and internal management. Since then, Taiwanese manufacturers have started to globalize and conduct direct investment in South-East Asia and China. By exporting intermediate goods, equipment, technology and management knowledge to the Asia-Pacific area, Taiwanese enterprises started to export final goods to the global market. On the other hand, Taiwanese enterprises also put their resources into product design and became ODMs. In addition, since the mid-1990s, some big manufacturers with famous computer brand names, such as Dell and Compaq, started to utilize a strategy of “built to order”, which meant that, under their logistic information systems, customers could place an order and receive products directly from various global locations (Wu et al., 2002). In summary, since the early days of manufacturing IT equipment in Taiwan on an OEM basis, the percentage of Taiwanese manufacturers with their own brands has increased only slightly, contrary to the notion of an evolution from OEM to OBM (own-brand manufacture). On the other hand, Taiwanese enterprises put their resources into product design and became ODMs (own-design manufacturers) and CDMs (collaborativedesign manufacturers), while working to strengthen collaborative design R&D management capability. 4.3.2. Governance for government In the early stage, the flow of knowledge or competence building came through the way international companies invested in Taiwan. Afterwards, the Taiwanese government actively promoted the establishment of research
The evolution of Taiwan ICT in a global economy
307
institutions, not only introducing advanced technology and knowledge but also researching and developing by themselves, and then transferring technology from abroad, setting up spin-off companies, or floating talents, in these ways spreading knowledge and technology to the whole of industry little by little. The industries began to establish good connections with Hsinchu Science Park and Silicon Valley at the same time. In relevant governmental policies, the establishment of the Science Park and Projects A, B, C, D and E have both been quite successful, with an important influence on the industries. However, these comparatively successful plans are mostly an extension of past achievements in manufacturing and design. In addition, the liberalization of the telecommunication market has been quite successful, especially for wireless communication; nevertheless, relevant results still need to be observed. After the industries have gradually set up their R&D competences, the great progress in industrial technology means the whole innovative system has to be adjusted; for example, the research institution must transform its roles, the function of universities must be improved and so forth. The Taiwanese government has put more effort in this direction in recent years. 4.4.
Knowledge Flow and Interactions with Outside
The knowledge base and the learning processes have greatly affected organizational innovation activities. In fact, the introduction and upgrading of these ICT products and technologies mainly depend on the supply from other countries, not from domestic industrialists or research institutions. In the early stages (the 1970s), Taiwan’s technologies used in production were mainly transferred from industrialized countries. Therefore, technology transfer was adopted as an initial strategy for the quick development of the industry base at the very beginning. Research institutes are employing technology acquisition, joint research, foreign direct investments and strategic alliances to interact with foreign companies. Research institutes act as mechanisms to accelerate the industry’s technological development. Thus leverage and diffusion promotion were being exercised through acquisition of equipment and through the movement and transfer of skilled staff. Then technological R&D capacity gradually accumulated through imitating, copying or a limited improving of the existing foreign products, and later, around the 1990s, the IT industry gradually developed in-house R&D capacity through self-directed effort, direct alliances, and joint ventures with foreign companies. Meanwhile, domestic industrialists lean towards continuous innovation that raises values, successively strengthening the efficiency of the production lines. Besides, some major
308
Sectoral systems of innovation and production
Taiwanese IT companies have tried to find ways to differentiate their products, with branding and product design the two major strategies.
5.
OVERVIEW OF THE ICT SECTOR
5.1.
The ICT Evolution from a Historical Perspective
The development of the ICT industry and society can be divided into three stages generally (Figure 10.5). The first stage (1979–89) was the IT Awareness Promotion, which mainly popularized the idea and application of information. Therefore, in IT talent training, III (Institute for the Information Industry) cooperated with the National Youth Commission of the Executive Yuan to run various kinds of educational training, such as developing the professional information ability of those people without an IT background. At this stage, regarding the ICT industry, III developed the Chinese computer in 1983 in order to popularize the application of information technology. Besides, III formulated the Chinese ICT Standards in order to popularize the utilization of a Chinese-environment system for PCs. These measures were established as the most important foundation of Taiwan’s informatization. Besides, in 1988, supported by the MOEA, III started the SEED Plan (Software Engineering Environment Historical Perspective
1989
1979
1999 3M internet population In 3 years (1996–98)
Promote Info month (1980–2000)
Residential Registry System (1988–98)
Financial Data Exchange System (1989–92) First Bank Banking System (1980)
Custom EDI System (1987–96) Perform ITIS (1990–) Establish Sci& Tech Center (1996)
Establish MIC (1984)
Chinese Computer (1983)
Chinese & ICT Standard (1984–)
Complete SEED project (1988–92)
SDG II (1985)
SW 5 years plan (1994–2002)
Foster IT elite
Source:
Basic training (1979–92)
Plan e-Taiwan (2001) Info Security plan (2001–) e-Taiwan (2002–) Support e-Signature Law proposal (2002) CMMI (2002) Rosetta Net (2002)
e-Industry (1999–) SEEDnet ISP, Shinewave, StarBex E-toyou spin-off (1998) spin-off (2004)
Info industry (1995–)
Industry promotion
2005– Plan bridge digital divide (2001–)
Professional licence (1993–2002)
Digital content (2003–)
Communication (2003–) Develop e-learning platform (1999)
Info service (2004–) Integrate Domain (2003–)
Huang (2005).
Figure 10.5
Historical timeline of ICT development in Taiwan
WEF EIU OECD
Elevate national competitiveness
National eReadiness Development
ICT Society
IT Infrastructure Build-Up
ICT Industry
IT Awareness Promotion
The evolution of Taiwan ICT in a global economy
309
Development) for a duration of four years, whose main purpose was to create a SEED Net based on the internet. This was the first to adopt a TCP/IP internet communication agreement in the country before telecom liberalization; at the same time, it was free for enterprises to apply for open connecting software for a Chinese work environment, which promoted internet business applications in Taiwan (Ke, 2005). As for the social aspect of ICTs, III set up the first banking information system and ran the Information Month activity; the scale and content have increased year by year up to now. The second stage (1989–99) was that of IT Infrastructure Building Up. III was mainly to promote professional certificates and set up e-learning platforms in IT talent training. Taking the budding of internet business applications in 1996 to be a start, the Executive Yuan established the National Information Infrastructure (NII) and promoted projects of e-competitiveness. With respect to ICT promotion, the Ministry of Transportation and Communications announced liberalization of the fixed network telecom industry in 1997, the Ministry of Finance liberalized online banking, and online interbank transfers began; and, besides, Chunghwa Telecom launched a DSL service in 1999. From 1996 to 1998, Taiwan reached the landmark of “Three Million Onlines in Three Years”, which laid the foundation of flourishing internet development. The third stage (1999–2005) was that of National eReadiness Development. IT talent training included not only the e-learning platform as at the second stage but also the plan of cultivating ICT talent in integrating relevant fields. In relation to the ICT society, from 2001 to now, III assisted the government to propose a blueprint of Digital Taiwan, which is expected to take e-experience from urban areas to rural areas, from corporations to small businesses, from the hinterland to the world. In addition, III set up the safe mechanism plan of information infrastructure in 2001, promoted the Law of e-Signature in 2002, and planned the e-Taiwan Program, focused on accelerating the establishment of a digital society and leading Taiwan to become one of the most advanced e-nations in Asia. 5.2.
The Performance of the ICT Industry and Demand Conditions
During the past 20 years Taiwan has emerged as a leading producer of ICT products. The World Economic Forum (WEF) published the Global Information Technology Report 2005–2006, in which Taiwan ranked seventh in the Networked Readiness Index (NRI) out of 115 economies, gaining eight places on the previous year. In terms of all the main subindices covered in each dimension, Taiwan ranked in the top five worldwide, placed third for market environment, third for individual readiness
310
Sectoral systems of innovation and production
and fourth for government usage. Also noteworthy is the significant move up to tenth place from 23rd in the Environment Component Index, further evidence of the overall positive environment for ICT development in Taiwan. In the Usage Component Index (covering individual usage, business usage and government usage), Taiwan again leapfrogged into fifth place in the 2005–06 report, an improvement from the 11th place in 2004– 05. Indeed, technology-intensive industries, mostly in ICT, now make up over half of Taiwan’s economy, compared with less than a quarter in the late 1980s. Taiwanese manufacturers collectively produce well over half the global supply of the devices that make up the core of the worldwide ICT industry and infrastructure (Dutta and Lopez-Claros, 2005; Dutta et al., 2006). 5.2.1. Technology performance of the ICT industry Table 10.8 shows that the R&D investment of domestic industries mostly centres on the ICT industry; the proportion of ICT R&D expenditures to whole-enterprise innovation expenditure reached 70.3 per cent in 2004. However, the density of R&D investment in Taiwan’s high-tech manufacturing and ICT industries is relatively lower than that of other countries. This phenomenon may be related to the operational mode of Taiwan’s ICT industry, mostly OEM/ODM; though the operational scale of ICT-relevant industries is quite large, innovation investment is still relatively low. Information The mobile computing age is predicted to have three major impacts: digital content will grow explosively and diversely; integrated Table 10.8 R&D expenditure and personnel in the ICT industry, 1999–2004 Year
1999 2000 2001 2002 2003 2004
R&D expenditure (million NT$)
Percentage of BERD
72 128 78 483 85 893 94 914 104 555 118 032
59.0% 62.4% 65.9% 68.0% 69.5% 70.3%
R&D personnel
R&D researchers
Headcount
FTE
Headcount
FTE
43 867 49 107 51 069 57 328 63 882 69 421
36 861 40 484 43 034 46 789 51 895 57 686
22 163 24 044 26 609 30 265 32 940 36 410
20 358 21 858 24 992 27 157 30 002 33 385
Note: The range of ICT is based on the definition of the OECD Frascati Manual, 2002; FTE= full-time equivalents. Source:
National Science Council (2005a, p. 38, Table 2-2-7).
The evolution of Taiwan ICT in a global economy Funding (NT$ Million)
Manpower (Man-Year)
311
Funding/Manpower (NT$ Million)
1800 1600
5 1550.4
4.4
4.4
4
1400 1101.3
1200
3
1000 800
847.7 2.1
2.3
723.8
400
2.1
574.7
600 190.5
683.7
251.2
318.8
2 250.4
1
200 0
0 2000
Note: Source:
2001
2002
2003
2004
Funding is given as budget numbers. National Science Council (2005b).
Figure 10.6
Information R&D funding and manpower
digital convergence will enter into mainstream services; and services will continue to go online and become more user-friendly. These three trends will accelerate the development of user-friendly, smart, mobile and secure living environment technology. The development of various types of information, communications and video products will speed up the diffusion of the mobile computing age and user-centric applications and services. Recent performance in relation to R&D project funding and manpower is set out in Figure 10.6. Telecommunications Society is expected to rely on broadband networks to sustain a thriving knowledge economy in the coming age of deregulation, globalization and digitization. Besides maintaining a competitive telecommunications market and actively promoting the development of the broadband network infrastructure so as to develop a sound ICT development environment, the Taiwan government also aims to continue to pursue forward-looking telecommunications policies based on new ways of thinking. It had promoted the reform of existing controls, and intended to accelerate the integrated development of the telecom and broadcasting industries. In this way it is expected that Taiwan will be transformed into an internationally competitive high-tech island. Recent performance in relation to R&D project funding and manpower is set out in Figure 10.7.
312
Sectoral systems of innovation and production Funding (NT$ Million)
Manpower (Man-Year)
Funding/Manpower (NT$ Million)
100
2
80
79.8
1.6
69.1 55.4 1.2
60
75
1.1
47.7 40
29.7
5.4
0.7 0.5
0.3 0
0 2000
Note: Source:
1
46
20 20
1.5
57.9
2001
2002
2003
2004
Funding is given as budget numbers. National Science Council (2005b).
Figure 10.7
Telecommunications R&D funding and manpower
Electronics With support from the government, the nation’s semiconductor industry has achieved impressive results. Because the domestic electronics and information product industries have been able to obtain steady a supply of key parts and components, Taiwan has become one of the world’s leading information product suppliers. However, in the face of stiff cost competition and increasing emphasis on capital investment, the domestic semiconductor industry must boost its core competitiveness in the future by focusing on emerging areas characterized by innovation and value creation. As for the display industry – the other Two Trillion and Twin Stars industry – Taiwan is currently ranked second in the world in terms of flat-panel display (FPD) manufacturing capacity, and it has become a leading global high-tech product manufacturing and service centre. In order to increase Taiwan’s technological autonomy further in this area, the government has implemented an FPD promotion programme that should create a sound foundation for the development of the display industry. Recent performance in relation to R&D project funding and manpower is set out in Figure 10.8. 5.2.2. Industry performance of the ICT industry According to the National Science and Technology Program Implementation Regulations drafted and approved by the NSC in 1998 as a response to Taiwan’s major socioeconomic and employment needs, and in order
The evolution of Taiwan ICT in a global economy Funding (NT$ Million)
Manpower (Man-Year)
313
Funding/Manpower (NT$ Million)
2341.4
2500
8 2183 6.7
2000
6.7 6.2
7 6
1533.3 1500
5
4.9
4.7 1144.3
4
1008.1
1000
3 347.6
500
248.8
244.5
204.4
327.8
2 1 0
0 2000
Note: Source:
2001
2002
2003
2004
Funding is given as budget numbers. National Science Council (2005b).
Figure 10.8
Electronics R&D funding and manpower
to integrate up-, mid- and downstream R&D resources, there are three programmes related to ICT, respectively telecommunications, digital archives, and e-learning (see Table 10.9): 1.
2.
3.
The National Sci-Tech Program for Telecommunications was being implemented during the years of 1998–2003 and 2004–08 with total funding of NT$12.36 billion and NT$13.35 billion. Apart from the first phase’s areas of wireless communications and broadband internet, the second phase has added the category of application services in an effort to establish a full range of telecommunication systems technologies. The National Sci-Tech Program for Digital Archives was implemented during 2002–06 with total funding of NT$2.78 billion. The primary goals of this programme contain the digitization of the nation’s important artefacts and collections, and the use of a national digital archive to promote cultural, social, industrial and economic development. The National Sci-Tech Program for e-Learning was implemented during 2003–07 with total funding of NT$4.01 billion. The programme’s goals were to increase citizens’ opportunities for lifelong study, promote the development of e-learning industries, and encourage academic research on e-learning. This programme also emphasized social goals such as the promotion of e-learning at the level of citizens
314
Sectoral systems of innovation and production
Table 10.9
National science and technology programmes related to ICT
Programme
Fiscal year
Telecommunications
Total funding (NT$100m)
1998–2003 2004–08 2002–06 2003–07
Digital Archives e-Learning
FY2004 statutory budget Number (NT$100m)
123.6 133.5 27.8 40.1
20.2 4.5 6.1
Note: Total funding refers to the planned budget number after completion of project planning; annual budgets may be adjusted following review. Source:
National Science Council, (2005b, p. 25).
Table 10.10 Million US$ Notebook PC Desktop PC Motherboard Server LCD monitor CDT monitor DSC ODD Projector and others IT hardware total Growth rate Note: Source:
a
Shipment value of IT hardware in Taiwan, 2001–05 2001
2002
2003
2004
2005
12 239 6 866 5 647 1 040 3 131 5 240 1 132 2 107 N/A
13 922 6 933 5 636 1 324 5 646 4 544 1 003 3 146 285a
16 809 8 297 6 353 1 559 9 801 3 804 1 476 3 297 530a
21 831 9 404 7 639 1 837 14 402 3 492 1 972 3 544 5 479
30 301 10 080 7 985 2 060 15 726 2 059 2 777 3 700 5 600
42 750
48 435
57 101
69 600
80 036
–9.1%
13.3%
18.0%
21.9%
14.99%
= projector only. MIC website (accessed in December 2005).
and bridging the digital divide in Taiwan, as well as infrastructure support through the Network Science Park for e-learning. As can be seen from Table 10.10, the growth rate for whole IT hardware represents unstable although the shipment value steadily increased from 2001 to 2005. The notebook PC still has an impressive track record among IT hardware. However, complying with the trend of companies moving to China,
The evolution of Taiwan ICT in a global economy
Table 10.11
315
Shipment value of communication industry products, 2001–05
Million US$
2001
2002
2003
2004
2005
Mobile phone PDA Mobile handheld WLAN DSL VoIP Cable modem Datacom
858 491 1349 244 524 N/A 470 1238
1939 972 2911 359 677 N/A 455 1491
3121 1190 4311 606 750 N/A 299 1655
3644 1821 5465 1341 1179 254 482 3756
4374 3240 7541 1796 1148 340 727 4011
Source:
Advisory and Intelligence Service Program, III-MIC (December 2005).
Taiwan’s information and electronic industries have been investing in China since 1990. Worse than before, 79.5 per cent of the information hardware of Taiwan IT manufacturers was made in China, and it involved more than 80 per cent of the shipment value of Mainland China’s information hardware. This transformation will be considered in section 7. Table 10.11 shows that the shipment value of communication industry products increased gradually year by year, especially for mobile phones, PDAs and mobile handhelds, which were five to six times greater in value in 2005 than in 2001. 5.2.3. Demand conditions Taiwan, with medium-scale export competitiveness, has a manufacturing industry that also does well in the American and Japanese markets. In Taiwan, electronic components is the most competitive industry in terms of exports. Taiwan is also a leader in the adoption and widespread use of ICTs, in stimulating innovation and in demonstrating their effectiveness. As an export-oriented economy in a globalizing world, Taiwan demonstrates the advantages that long-term strategic vision combined with adaptive management can confer. The export proportion of high-tech industry accounts for approximately 42–45 per cent of all manufacturing industry (Table 10.12); however, the export proportion of computer-relevant products has dropped year by year in recent years. This phenomenon has partly to do with the fast decline of product cost caused by the operation pattern of accepting orders in Taiwan and manufacturing abroad and influenced by the global competitive environment and the rising of emerging markets (especially for China). Fuller discussion of this issue will be presented in section 7. Fortunately, there is a slight increase for the radio, television and
316
Sectoral systems of innovation and production
Table 10.12 %
Export shares of high-tech industry, 1996–2004 1996
Manufacturing 100 High-tech 34.28 Pharmaceuticals 0.20 Office, accounting 16.57 and computing machinery Radio, television 15.45 and communication equipment Medical, precision 2.06 and optical instruments Aircraft and spacecraft 0.00 Source:
1997
1998
1999
2000
2001
2002
2003
2004
100 100 100 100 100 100 100 100 36.51 38.94 41.84 45.46 43.20 43.18 43.03 42.53 0.19 0.18 0.17 0.12 0.13 0.13 0.13 0.13 18.27 20.06 20.57 19.88 19.72 18.14 14.64 11.24
15.90 16.51 18.64 22.61 20.52 21.28 22.89 24.27
2.14
2.16
2.42
2.77
2.73
3.55
5.30
6.82
0.01
0.03
0.04
0.08
0.11
0.07
0.06
0.07
Export/import data tape of ROC, 2004, calculated by TIER.
communication equipment sector. It means the communication industry has remained profitable in recent years. 5.3.
Demand Side of ICT
5.3.1. The infrastructure of ICT Challenge 2008: The National Development Plan was launched in 2002, by means of strategic actions of e-government, e-industry and the internet society, which were enlarged through informational applications for each department and informational education training. Those were a great help in improving the information environment and web services in Taiwan. Internet services could be divided into three parts: government, business and personal. The important achievements included setting up a government service web (GSN), a government certification mechanism, an official e-document exchange, an e-government service platform, an e-government portal, governmental website contents, thousands of official application forms, and several new government e-services including G2G, G2B and G2C. Although the internet services and applications in enterprise were mainly supported by business itself, the government also provided essential support, including Projects A, B, C, D and E, the Demonstrated IT Application Research Program, and the Technology Research Program for Innovative Services. Most of the support by government involved integrating standards from industries, developing job application platforms, and setting demonstrative applications, which amplified the leverage effects for industry.
The evolution of Taiwan ICT in a global economy
Table 10.13
The planning of internet construction in Taiwan, 2002–07
North–South Basebone for Island-wide (Gbps) International connection cable (Gbps) FTTC fixed line area fibre coverage rate (%) Broadband subscribers (10 000) Source:
317
2002
2003
850
950
1050
1150
1200
1250
150
200
200
200
250
250
82.0 205
85.5 300
2004
88.5 380
2005
91.0 460
2006
93.0 530
2007
95.0 600
Fan (2005).
5.3.1.1. Broadband infrastructure Broadband construction in Taiwan is divided into fixed line and wireless. For the fixed line, the main development plan is the “6 million broadband users” of e-Taiwan. The annual goals of this plan are set out in Table 10.13. In order to promote the application and development of wireless broadband, the second stage (five-year plan) of the National Science and Technology Program for Telecommunications provided by the NSC and the M-Taiwan Plan proposed by the MOEA, the government might provide help. Regarding the new relevant techniques, such as 3G IP core network techniques, W-CDMA, TD-SCDMA, Integration Heterogeneous Network etc., they are still in the lab stage instead of commercializing the market. Users could then have WLAN, GPRS and 3G, which could let at least 600 000 people have the convenience of internet surfing and of using web phones to connect in an environment of wireless networking. Besides, the goal of the M-Taiwan Program was to promote a ubiquitous network and e-services in Taiwan. 5.3.1.2. Information and telecommunication security In January 2001, the Executive Yuan set up the Contingency Centre of National Information and Communication, which is in charge of ICT infrastructure security and passed the National Information and Communication Infrastructure Security Mechanism Plan. Besides, the plan of “6 million broadband users” has a subordinate plan, “Setting up an ICT security environment”. The government also activated other information security programmes: 1.
2.
Public Key Infrastructure (PKI): set up a Promotion Task Force to promote a citizen’s digital certificate programme, corporation certificate programme, medical certificate programme and so on. Planned and set up government information exchange safety standards.
318
Sectoral systems of innovation and production
Table 10.14
Gross enrolment rate in higher education (18–21) Literacy rate for over age 15 R&D personnel R&D/GDP Internet users Internet penetration rate Broadband internet users Mobile internet users
Status of education, R&D and internet application in Taiwan, 1996–2004 Unit
1996
1997
1998
1999
2000
2001
2002
2003
2004
%
NA
NA
56.1
31.0
68.4
77.1
83.4
90.2
–
%
94.3
94.7
94.9
95.3
95.6
95.8
96.0
97.0
–
10 000 11.7 12.9 12.9 13.5 13.8 13.8 15.0 15.7 16.9 % 1.80 1.88 1.97 2.06 2.06 2.17 2.21 2.45 2.54 10 000 60 166 301 480 627 782 859 883 916 % 3 8 14 22 28 35 38 39 40
10 000
–
–
–
2
23
114
213
289
360
10 000
–
–
–
–
–
28
155
350
535
Notes: 1. Broadband internet users include xDSL, cable modem and fixed line with fibre connection. 2. Mobile internet users include WAP, GPRS, PHS and 3G. Source:
3. 4.
Ke (2005, p. 29).
Implemented related laws. Information safety technique research.
5.3.2. The current status of the Information Society in Taiwan After cultivating IT for so many years, Taiwan has made great progress in informational social readiness, as shown in education, R&D quality and internet penetration rate (Table 10.14). The internet connection can be divided into three segments: home, business and government (Table 10.15).
6.
POLICY IMPLICATIONS
After reviewing the policies and related performance of the past ten years, we might conclude that those policies played a role in supporting industry to follow the trends, especially Projects A, B, C, D and E and the TDP Program. TDP in research institutes also helped industries to
The evolution of Taiwan ICT in a global economy
Table 10.15
Internet connections of Taiwanese homes, business and government, 2001–04 2001
e-Government Government agencies broadband penetration rate Government agencies website penetration rate Government agencies online e-Business Corporate broadband penetration rate Corporate website penetration rate Corporate internet penetration rate e-Home Home PC penetration rate Home internet penetration rate Home broadband penetration rate Source:
319
– – – – – 44% – 39% 40%
2002
2003 2004
78% 100% 100% –
85%
85%
100% 100% 100% 80% 90% 96% 22% 62% 72% 53% 58%
27% 79% 71% 57% 73%
36% 81% 73% 61% 78%
Ke (2005, p. 30).
follow the emerging technological trends of ICT. Projects A, B, C, D and E helped the firms to connect more closely between Taiwan and international branding companies as well as between upstream and downstream ICT companies in Taiwan. This helped the OEM/ODM business model in which Taiwanese firms became more competitive than other areas. Besides manufacturing, the performance of product design has also shown some major achievements in international awards. When we reviewed the cases of companies and policies, we found that it was mainly because the capabilities were built up through two national awards (SOE and NAOE, since 1989), as well as the impact from Acer. Both awards woke companies in traditional industrial sectors to realize the importance of quality and product design, and they started to build up the capabilities. Then, when the ICT-related companies found that they were in a highly homogeneous product market, they invited experts in traditional industrial sectors to help and combined their product design into manufacturing processes. As for the demand side of ICT policies, we could see the impact of deregulation of the wireless service sector in the highest density of cellular mobile telephones in the world. However, the effect of other policies, like e-Taiwan and M-Taiwan, has not yet appeared in any macro-indicators, nor in the share of communication services in GDP. That means the demand side policies still need to be monitored.
320
6.1.
Sectoral systems of innovation and production
Review of Taiwan’s ICT: Policy Problems
6.1.1. The value-added rate of Taiwan’s ICT industry is at a low ebb Because Taiwan’s ICT manufacturing industry (computers and OA equipment; electronics and communication equipment; precision, optical and medical equipment) has a high degree of division of labour, the supply of information and electronic products exceeds its demand, and prices have plunged. In addition, Taiwan’s ICT industry is restrained by large MNC plants and threatened by catching-up countries; therefore the added value is lower than for advanced countries on the one side, and for South Korea too on the other side. 6.1.2.
Domestic large plants concentrate on the OEM stage of lower value added Different innovative patterns and the degrees of integration of value chain activity influence the profit rates of enterprises. The operations patterns of MNC large plants in ICT, like Microsoft, Intel or Nokia, contain many value-adding activities: R&D, design, engineering, manufacturing, logistics, marketing, sales and service, and so on. R&D intensity exceeds 13.5 per cent, and the operating profit margin reaches 17–40 per cent. Taiwan’s IC big plants such as USMC, UMC, Media Tek and so on are deeply committed to R&D and integrated with services; the operating profit margin also advances. In comparison, the Taiwanese IT system contract manufacturer focuses on the subcontract production mode, where the rewards are lowest in value; the R&D intensity and investment are relatively low, and the gross profit rate and operating profit margin are relatively low as well. 6.1.3. Lack of original and radical R&D ICT OEM enterprises mostly depend on the path of technology that the big international companies used. The depth of innovation is limited, which influences profit making. Because Taiwan’s ICT industry relies mainly on international OEM, its R&D and the direction of product development chiefly follow the main technological structure established by large foreign companies: the technology of mass production and cost control are much more important than the development of original products. Therefore it is difficult to create new and obvious added value (Chen and Liu, 2005). In addition, the result of an investigation in Taiwan (TIER, 2005) shows that the reasons why enterprises spend time and money on R&D but are still unable to improve the performance of management are mostly because they have no cost advantage (18.4 per cent), the product differentiation is insufficient for meeting the needs of the market (15.7 per
The evolution of Taiwan ICT in a global economy
321
cent), it is not easy to raise prices for products which are controlled by buyers (big foreign companies), and they fail to catch the most appropriate time to market (13.9 per cent). These factors reflect that domestic ICT industries are concentrated on OEM, the scale of enterprises is too small and the homogeneity is too high. When facing the lack of bargaining chips of negotiation with large international companies and the threat of replacement by catching-up country rivals, enterprises lean towards price competition, the pursuit of low defect rates, the advantages of production cost, and winning through quantity; these enterprises neglect to establish the diversity competencies of originality and uniqueness, so that profits become lower and lower. 6.2.
Policy Suggestions
In industrial policy, Mainland China has influenced demands in the local market for establishing infrastructure, drawing up technical standards (such as digital TV, TD-SCDMA, WAPI), installation fees and user fees for telecommunications (fixed line and mobile phones) and adjustments of household registration policy and so on in order to support domestic manufacturers and negotiate with multinational plants. Meanwhile, on the one hand, the government utilizes permissions to operate, the certification system, concessions for foreign investment and so on in cooperation with policies of foreign exchange control, quotas on domestic sales and so on, to lead the local plants of MNCs to expand their local production scale continuously; on the other, it utilizes the cluster effect of local assembly industries to encourage the component industry to move in progressively by means of tax credits, proportions of domestic purchase, alternative tariffs and so on. Therefore industrial policies play an important role in the development of ICT. Freer and more open attitudes are necessary for Taiwan nowadays. First, the government should cancel unnecessary restrictions on talent, funding and transportation in terms of the total planning of relevant preferential policies, actively introduce foreign technicians, and then open three direct cross-Strait links with moderate financial liberalization; these measures will encourage domestic and international enterprises to utilize Taiwan’s natural advantage of the Asian-Pacific Operation Centre to attract global resources and continuously improve the competitiveness of the industrial environment. In terms of active measures, as for the communication industry which has entered Mainland markets extensively, an interview with III adviser Chen Wen-Tang brought up the tactical thinking of “letting bulls into the wild, training calves and entrapping cows”, that is to say, when facing ICT
322
Sectoral systems of innovation and production
downstream assembly manufacturers that consider factors of cost and market (“bulls” such as the PC, monitor, notebook, cell phone, desktop, printer, scanner, keyboard, mouse, etc.), the government should adopt an attitude of open freedom, letting them advance in the Mainland and seizing the market there. For the ICT midstream industry, which refers to key components manufacturers for the needs of the “bull industries” (such as LCD, IC, battery, LED, RF, etc.), the key point for cultivating the “calves industry” depends on governmental policies. At present, except for fields or industries stipulated in the Statutes for Industrial Upgrading, there are no clear specific fields or industries setting foresight values. These midstream industries are just like growing calves; one should plant a lot of pasture and foster it actively. Accordingly, Taiwan’s government should strategically encourage the development of core industries that suit Taiwan with preferential policies: such as encouraging and promoting R&D alliances and R&D service industries, providing high-quality information services, encouraging enterprises to develop toward R&D design, brand marketing and logistic services and so on, increasing its added value, and keeping the “calves industry” in Taiwan. What the Taiwanese government recommends as “trapping” refers to the ICT upstream industry such as materials and equipment, which are much closer to encouraging foreign manufacturers or large domestic plants to set up R&D centres, in this way strengthening high-level R&D design. As for the communication industry, seeing that Mainland China discourages Taiwan manufacturers from setting up factories there out of considerations of national defence, the government should assist Taiwan manufacturers in developing new technological innovation and product integration ability, and grasping the business opportunities of the telecommunication service industry by helping Taiwan manufacturers cooperate with indigenous manufacturers. As for the consumer electronic industry, with enormous assembly capability in Mainland China, Taiwan can make the best of a weak key component industry and grasp relevant market chances such as digital TV and DVD through supplying the raw materials. Taiwan’s own information industry has developed rapidly in recent years, becoming one of the main motive forces in its economic development and making up, in a timely way, for the industrial gap left by the outmigration of labour-intensive industries. This has facilitated the smooth readjustment and transformation of Taiwan’s industrial structure and laid down a foundation for the development of the knowledge economy. Nevertheless, as said above, with the steady innovation of internet technology and the deepening of applications of information, the ability to develop the knowledge economy can no longer be assessed on production
The evolution of Taiwan ICT in a global economy
323
value alone but must also take into consideration the influence of information application capability on overall national competitiveness.
7.
THE OFFSHORE SITUATION FOR TAIWAN’S ICT/IT IN MAINLAND CHINA
7.1.
The Past, Present and Future of Taiwan IT Manufacturers Entering the West Market
All the time the process of heading west (which here means the Mainland China market) for Taiwan IT manufacturers is “Deeds speak louder than words”, in consideration of politics and the intervention of policies. Taiwan’s information and electronic industries have been investing in China since 1990; the range of investment began with early production activities, marketing activities after 1995 and then the fields of R&D. However, at present Taiwan has already become the largest source of trade deficit for Mainland China: 79.5 per cent of the information hardware of Taiwan IT manufacturers was made in China, and it involves more than 80 per cent of the shipment value of Mainland China’s information hardware. This study, following a research paper of III (Kao 2006/3), will explore the background of the industry, the modes of investment and the driving factors, and then analyse the current situation of Taiwan IT manufacturers’ production, marketing and R&D arrangements in China, in this way predicting some trends of future development. 7.1.1.
Industrial background
1990–95 In Taiwan’s electronic information industry, the manufacturers of desktop PC components (case, power, mouse, keyboard, etc.) with lower price and higher manpower demand, aiming at lowering production costs, engaged in the establishment of the production base in Mainland China in the early 1990s, being attracted by the plentiful and cheap production elements of Mainland China and relevant investment preferences. 1996–2000 As the low price tendency of PCs became more obvious day by day, manufacturers attempted to obtain economies of scale by largescale factories in order to lower costs and compete with those who had entered the market of South China beforehand, in this way making up for the inferiority of production costs as far as the latecomers were concerned. Meanwhile, under the political opposition of Taiwan and China,
324
Sectoral systems of innovation and production
the Taiwanese government strongly recommended a “southwards policy”, encouraging Taiwan IT manufacturers to invest in South-East Asia. 2001–05 1. The excessively optimistic sales forecast for the PC market in the first half of 2000 affected the large increase of stocks of mobile phone key components, the phenomenon of the dot.com bubble and the global IT market transforming from a growth period to a mature period. When facing the situation that the competition of prices of large international IT plants was becoming progressively keener, in order to maintain space for profit making, Taiwanese IT manufacturers who had expanded production excessively in China launched price negotiations using the scale of orders as chips on the one hand and on the other retracting the purchase right of components, discussing the scale and price of orders with the manufacturers of relevant components directly and then reducing the potential interests in ODM orders that Taiwan IT manufacturers accepted, which greatly reduced the profit making of Taiwan IT manufacturers. 2. At the same time, because there was not much space left for price competition, IT manufacturers gave up the strategies of contesting industrial ability and the competition via low prices, and then turned to focus on the R&D and marketing activities of market segmentation. Some components manufacturers strengthened progressively the recourse to inputs of their own brand in a situation where the profit-making space of ODM was limited. 3. In addition, after many international cases of merger and acquisition of local manufacturers in Mainland China, Taiwan’s IT manufacturer BenQ merged with the mobile telephone department of Siemens in 2005, through this encouraging the global market share of the Taiwan mobile phone industry to rise over 30 per cent from 10 per cent, and then obtaining the relevant intellectual property rights and international distribution. BenQ then began to have its own brand. 7.1.2.
Investment patterns
1990–95 In the early stage, because the scale of the market was not big in Mainland China and the management environment was strange, Taiwanese manufacturers of electronic information began production in ways that included importing materials, according to the strategy of reducing production cost. 1996–2000 Though there were many limitations on investment in Mainland China, when faced with cheap elements of production, similar language, relatively complete system of industry and so on, many Taiwanese IT manufacturers turned to investing in factories there in the
The evolution of Taiwan ICT in a global economy
325
name of large shareholders through tax concessions. In addition, under the policy of foreign exchange control in Mainland China, in order to avoid too many funds from restrictive utilization, manufacturers engaged in buying low and selling high for components and end products through trading companies in Hong Kong mostly, keeping a large amount of profit to use overseas. 2001–05 When facing the industrial policies in Mainland China (e.g. setting up factories in local areas, limitations on domestic sales proportions, requirements for domestic purchase) and the foreign exchange control and so on, in order to raise market share in the Chinese market, large MNC plants strongly asked Taiwanese IT manufacturers to raise shipment volumes from the Mainland, as the foundation of the internal marketing quota. On account of “customer requests”, Taiwanese IT manufacturers increased the proportion of production in Mainland China by a wide margin. The thinking behind entering Mainland China for Taiwan’s IT manufacturers changed from reducing production cost to the requirements of customers under the pressure of the market. 7.1.3.
The patterns of division of labour
1996–2000 After 1995, Taiwanese IT manufacturers shifted the base of production previously in South-East Asia progressively to Mainland China. Therefore, the proportion of production in Mainland China rose from 14 per cent in 1995 to 33 per cent in 1999, but dropped to 31.3 per cent in 2000. The products that Taiwan’s IT industry increased in the ratio of production in Mainland China mainly came from PC peripheral equipment such as monitors, CD-ROMs, cases, SPS and so on. 2001–05 As international IT manufacturers required Taiwanese manufacturers to increase shipment volumes from Mainland China after 2000, the proportion of production of Taiwan’s IT industry rose from 31.3 per cent in 2000 to 47.5 per cent in 2002 and to 79.5 per cent in 2005; however, the proportion of production in Taiwan fell from 49.1 per cent in 2000 to 35.7 per cent in 2002 and to as low as 6.8 per cent in 2005. Among the ways in which Taiwanese IT manufacturers divided the labour on both sides according to the mode of production, we find: 1.
Since 2000, products have shifted from desktop PC-related components and peripheral equipment of relatively low price to products that are relatively capital- and technology-intensive such as LCD displays, notebook computers, mobile phones, LCMs and so on.
326
2.
3.
Sectoral systems of innovation and production
The division of labour shifted from high- and low-level products before 2000 to test-production and mass production, which made Taiwan’s IT manufacturers increase the production rate substantially in Mainland China. Tracing its cause, the situation had something to do with the Mainland Chinese government being asked to improve operating efficiency by Taiwanese IT manufacturers and follow Taiwan’s experience to set up export processing zones and bonded factories, and by so doing improve import and export efficiency, and the inconvenient contact of R&D personnel on the two sides.
7.1.4.
Geographical distribution
1990–95 After Mainland China opened to reform, the government used relevant industry policies, such as the huge domestic-demand market, as a bait, with foreign exchange control and the proportional limitation on domestic sales in order to attract manufacturers to set up their factories there. Because of the undeveloped infrastructure and high transportation charges in Mainland China, most international enterprises set up their factories in the coastal areas (such as Shenzhen in South China) and meanwhile utilized the cheap essential factor of production in Mainland China and the international freight transport and financial function in Hong Kong. 1996–2000 Because the scale of investment was very large, in order to reduce operating risk and focus on the expansion of the local market in future, the PC peripheral manufacturers chose the East China area, where public security was good, China’s central government gave strong support, the government regulations were clearer and the geographical position was moderate. 2001–05 There were different effects of early or late steps of economic opening to the outside world in different regions, and the characteristics of the varied industries in the North, East and South part of Mainland China: 1.
North China area: Because the North China area is near the core of political power, it attracts international communications manufacturers, which are closely related to national defence security and manufacturers of local desktop PCs, which have to strive for government resources (including orders and funds). Except for Acer and DBTEL,
The evolution of Taiwan ICT in a global economy
2.
3.
327
there are few Taiwan manufacturers (like mobile phone components) which cooperate with Motorola. East China area: In the East China area, the regulations and management environment are relatively clear. Besides, the geographical position benefits the logistics of each domestic region in the domesticdemand market, and the actions to attract business investment are positive, so those latecoming IT and wireless communication manufacturers which have a large scale of investment and focus on the local market are attracted to enter this area. The production bases of Taiwan’s LCD display, notebook computer, mobile phone, LCM and semiconductor manufacturers are mostly concentrated here. South China area: Because the South China area opened to the outside world early and the management environment was flexible, this area attracted numerous desktop PC manufacturers and consumer electronic manufacturers who regarded lowering costs as necessary. Taiwan’s industries related to desktop computers were mainly concentrated here.
7.1.5.
Cluster patterns
1990–95 For Taiwanese IT manufacturers who set up factories in the South China area, their ways of investment tended to be more conservative at the stage of testing the local management environment. In order to keep the present trading relationship and to exchange relevant industrial information, the industrial clusters were relatively concentrated and, in addition, these manufacturers still maintained a cooperative relationship with Taiwanese manufacturers. 1996–2000 Taiwanese IT manufacturers who set up factories in the East China area had two different types of cluster. The first was setting up factories in the local industrial area solely by their own efforts; they aimed to persuade relevant manufacturers to set up factories in the East China area for nearby supply in order to improve purchasing efficiency, but did not require a cooperation relationship with the suppliers. The second type was of assembly manufacturers choosing a large amount of land to establish in the large-scale industrial areas and then ask the suppliers to be with them, maintaining the supply relationship under a long-term cooperative commitment. 2001–05 The scale of investment of Taiwan IT manufacturers became larger day by day, and each local government in China solicited business positively under considerations of the development of the local economy and over-investment in industrial areas. They obtained a large amount of
328
Sectoral systems of innovation and production
industrial land through collective negotiation. The mode of the ‘groupexclusive industrial area’ grew gradually. 7.1.6.
Marketing arrangements
1990–95 Because the scale of the Mainland IT market was still small, the investments of Taiwan IT manufacturers preferred the mode of importing materials; therefore most products were for sale overseas. 1996–2000 Because a high growth rate of 20–30 per cent appeared in the Mainland PC market, some Taiwanese manufacturers began to accept ODM orders from PC manufacturers in Mainland China. Though Taiwan’s manufacturers strove for a proportion of domestic sales, they still focused on the international market while selling. 2001–05 With the growth of the Mainland IT market, the proportion of products that Taiwan’s IT manufacturers sold to Mainland China also rose year by year, from 5.2 per cent in 2000 to 11 per cent in 2004. In 2005, because the American and European markets were enjoying booms, the amount of high-priced products, such as notebook computers and mobile phones, exported to Europe, the USA, Latin America and South-East Asia increased, so that the proportion that Taiwan’s IT hardware manufacturers exported to China fell to 7.4 per cent. 7.1.7.
R&D arrangements
1990–95 Taiwanese manufacturers focused on production activities in the initial stage, and were rarely involved in the activities of marketing and R&D. 1996–2000 With rising familiarity with the local environment and talent, Taiwanese IT manufacturers began to realize that, if they moved partial R&D activities to Mainland China, they could employ local project personnel who had basic science and engineering training for cheap wages, thus reducing manufacturing cost. In the early stage, Taiwan’s IT manufacturers adopted a division of labour mode for high/low product levels between Taiwan and Mainland China; in other words, the activities of R&D design were maintained in Taiwan, and the developing activities of process technology stayed in Mainland China. 2001–05 In consideration of costs and management, most Taiwanese IT manufacturers chose factory sites as the R&D centre; therefore most IT
The evolution of Taiwan ICT in a global economy
329
manufacturers distributed their factories across East and South China, and only a few game and software manufacturers chose to set up factories in North China, with fragmentary distribution. As the production rates of Taiwan’s IT manufacturers rose substantially in Mainland China, the patterns of division of labour of manufacture shifted to the labour-division mode of test- and mass production after 2000. In a situation where the two sides suppressed ‘Three Direct Links’ across the Taiwan Strait for political reasons, the R&D activities of Taiwanese IT manufacturers in Mainland China shifted quickly from the stage of accessing mass production to the stage of product development. In this stage, Taiwanese engineers came and went on both sides of the Strait. To sum up, the impact of Mainland China on Taiwan’s ICT sectors can be divided into four stages over the ten-year period: i) before 1997, mainly producing in Taiwan; ii) between 1997 and 2000, both Taiwan and Mainland China expanding; iii) between 2000 and 2002, expanding in Mainland China and staying the same size in Taiwan; and iv) after 2002, expanding in Mainland China and decreasing in Taiwan. Among the largest-scale 500 foreign-capital enterprises for import and export in Mainland China in 2004, 249 were IT manufacturers, and 70 came from Taiwan; hence the proportion of Taiwan manufacturers was 28 per cent. From the viewpoint of Taiwanese IT manufacturers’ contributions in Mainland China in 2004, the scale of imports and exports reached US$112.3 billion. The export value was US$62 billion and import value US$50.3 billion; the total favourable balance was up to US$11.7 billion, which represented 23.3 per cent of the import value. If we compare this with the same sample group of 2002, we find there were 37 manufacturers which were on the list for the first time, and 23 of them were listed first in the group. All the information shows that the production scale of Taiwanese IT manufacturers in Mainland China had a tendency to expand gradually. At the same time, if we analyse 33 Taiwanese IT manufacturers listed on board in 2002 and 2004, we find that, though the import value rose to US$18.9 billion, since the export value was US$30.1 billion, a US$3.89 billion of unfavourable balance (24.6 per cent of the import value) had already shifted to US$7.28 billion of favourable balance (20.9 per cent of the import value), which shows that, according to the trend where large MNC plants ask Taiwan IT manufacturers to raise production proportions in Mainland China, the orders for export which originally created foreign exchange for Taiwan had already been quickly transferred to China. The degree of dependence on Taiwan’s exportation to China had
330
Sectoral systems of innovation and production
Table 10.16
Degree of import/export dependence between Taiwan and Mainland China
Degree of import/ export dependence
The economic system Electronic components Manufacture Electronic equipment (radio, TV and communication) Source:
Export dependence of Taiwan on China
Import dependence of China on Taiwan
2003
2004
+/–
2003
2004
+/–
34.27 50.67 34.47 13.06
37.21 52.76 37.42 13.88
2.94 2.09 2.95 0.82
11.96 21.48 12.63 8.56
11.54 22.57 12.44 7.31
–0.42 1.09 –0.19 –1.25
Shen (2006).
reached 37.23 per cent by 2004, and the trade surplus exceeded $50 billion. The estimated amount of Taiwan’s investment in China was more than US$50 billion, which was more than 40 per cent of Taiwan’s total foreign investments. While the export dependence of Taiwan on China was getting deeper than before on the one hand, the import dependence of China on Taiwan was decreasing on the other hand (Table 10.16). For the Taiwanese electronic components industry in particular, the export dependence was 52.8 per cent in 2004, representing a 2.09 point increase over the previous year. In 1995, the percentage of production in Taiwan for information hardware manufacturing still remained 75 per cent, the percentage of investment in overseas production in Mainland China was 14 per cent, and in other foreign countries (mainly in South-East Asia) it was 11 per cent. In 2000, the percentage of production in Mainland China reached 31.3 per cent and 19.6 per cent in other foreign countries, and only 49.1 per cent production was left in Taiwan. The percentage of Taiwan’s information hardware industry production going abroad (especially to Mainland China) increased year by year, and reached 79.5 per cent in China and 13.7 per cent in other countries by 2005, with only 6.8 per cent remaining in Taiwan (see Figures 10.9a and 10.9b). In 1999, the percentages for power supply and casing among the products of information hardware produced in Mainland China reached over 60 per cent; motherboards, monitors and CD-Roms/DVD-Roms were around 35–45 per cent. It is worth noting that laptop PCs were mainly produced in Taiwan in 2000 (95.8 per cent); however, after that, the production percentage of laptop PCs in Mainland China rose rapidly to reach
The evolution of Taiwan ICT in a global economy
331
year 2000
China
1999
China
Taiwan Taiwan
China
1998
Taiwan
China
1997
Taiwan
China
1996 1995
Taiwan
China 0%
Source:
China Thailand Malaysia Others Taiwan
Taiwan
10%
20%
30%
40%
50%
60%
70%
80%
Overseas production by Taiwan’s information hardware manufacturing, 1995–2000
year 2005
Others
China
0%
10%
20%
Taiwan
Others
China
2000
Taiwan
Others
China
2001
30%
40%
China Others Taiwan
Taiwan
Others
China
2002
Taiwan
Others
China
2003
Taiwan
Taiwan
Others
China
2004
Source:
100%
III-MIC/ITRI (2006).
Figure 10.9a
Note:
90%
50%
60%
70%
80%
90%
100%
Data are based on the shipment value and exclude the projectors. III-MIC/ITRI (2006).
Figure 10.9b
Overseas production by Taiwan’s information hardware manufacturing, 2000–05
94 per cent by 2005. This also happened with other information hardware products. For instance, the percentage of CD-Roms/DVD-Roms, digital cameras and so on produced in China was over 90 per cent, and motherboards and LCD monitors were also 80 per cent (see Figures 10.10a and
332
Sectoral systems of innovation and production Product Mother board
China
Monitor
Others
China
CD Rom/ DVD Rom
Others
China
Taiwan
Others
Case
China
Scanner
Taiwan Others
China
Taiwan
Others
Laptop Others PC China
Others
Power
Taiwan
China 0%
Taiwan
Taiwan
Desktop PC
Note:
Taiwan
10%
20%
30%
OthersTaiwan
40%
50%
60%
70%
80%
90%
100%
1996 < > 1997 < > 1998 < > 1999 < >
Source:
III-MIC/ITRI (2006).
Figure 10.10a
Products produced overseas, 1996–2000
Product Mother board
China
Monitor
Others China
CD Rom/ DVD Rom
Taiwan Others
Taiwan
China
Case
Others Taiwan Others Taiwan
China
Scanner
Others
China
Laptop Others PC
China
Desktop PC
Others Taiwan China
Power PC
OthersTaiwan
China 0%
10%
20%
30%
Taiwan
Others 40%
50%
60%
70%
80%
Taiwan 90%
100%
Notes: 2005 motherboard sales volume includes types of pure motherboard, quasi-system, and full system, etc. 2000 < > 2002 < > 2004 < > 2005. < > Source:
III-MIC/ITRI (2006).
Figure 10.10b
Products produced overseas, 2000–05
The evolution of Taiwan ICT in a global economy
333
10.10b). Figures 10.10a and 10.10b also show the transformation of information hardware products, with some products that show up in Figure 10.10a for 1996–2000 not appearing again in Figure 10.10b, such as LCD monitors, LCM, digital cameras and so on.
NOTE 1. See below for many of these acronyms as well as further details.
BIBLIOGRAPHY Chen, Xin-Hong and Liu, Meng-chun (2005), ‘Industrial development models and economic outputs: a reflection on the “high tech, high value-added” proposition’, International Conference on 2005 Industrial Technology Innovation: New Value Creation Era, 25–26 August, Taipei. Council for Economic Planning and Development, Executive Yuan (2002), Challenge 2008: National Development Plan, http://www.cepd.gov.tw (accessed on 30 March 2006). DoIT/MoEA (2004), TDP Year Book 2004, Taipei: DoIT/MoEA. DoIT (2005), The Report of ABCDE Projects, Taipei: Department of Industrial Technology, Ministry of Economic Affairs, p. 12. Directorate General of Budget, Accounting and Statistics (DGBAS), Executive Yuan (2006), ROC National Income, DGBAS, Taipei. Directorate General of Budget, Accounting and Statistics (DGBAS) (2006), National Income in Taiwan, 1971–2004, DGBAS, Taipei. Dutta, Soumitra and Lopez-Claros, Augusto (eds), (2005), The Global Information Technology Report 2004–2005: Efficiency in an Increasingly Connected World, Palgrave Macmillan, Basingstoke. Dutta, Soumitra, Lopez-Claros, Augusto and Mia, Irene (eds), (2006), The Global Information Technology Report 2005–2006: Leveraging ICT for Development, Palgrave Macmillan, Basingstoke. Fan, Zhang-Kang (2005), ‘The essential ICT environment of information society’, Information Society Review, Initial Issue, III, Taipei. FIND (2004), Projects A, B, C, D and E, http://www.find.org.tw/eng/newsprint. asp?pos=0&subjectid=6&msgid=141 (accessed on 8 March 2006). FIND (2005), M-Taiwan Program, http://www.find.org.tw/eng/newsprint. asp?pos=0&subjectid=6&msgid=164 (accessed on 8 March 2006). Huang, C.Y. (1995), The Legend of Republic of Computer, Commonwealth Publishing, Taipei. Huang, Gwo-Jiunn (2005), ‘Enabling digital brilliance: industrial and social paradigm shifts in Taiwan’, Part of the contents of a speech in the National University of Kaohsiung, Kaohsiung, December. III-MIC/ITRI (2006), http://mic.iii.org.tw/NewMIC/Chinese/docdata/doc_img/ CDOC20060309007/3_800.gif (accessed on 20 March 2006). Kao, K.S. (2006), The Past, Current and Future of Taiwan’s IT Industry heading into
334
Sectoral systems of innovation and production
Mainland China, Advisory and Intelligence Service Program-Communications, Taipei: MIC. Ke, J.S. (2005), ‘Improving information society, strengthening e-competitiveness’, Information Society Review, Initial Issue, III, Taipei. Lin, X.-W. and Lin, H.-Y. (2005), ‘The compositeness of national innovation system and challenge of growth in Taiwan’, Paper presented at the International Conference on 2005 Industrial Technology Innovation: A New Value Creating Era, 25–26 August, Taipei. MIC (2005), ICT Overview, Advisory and Intelligence Service Program, MIC, Taipei. National Science Council (2002), Yearbook of Science and Technology, Republic of China, Taipei: NSC, p. 69. National Science Council (2004a), Central Government Scientific Technological R&D Performance, National Science Council, Taipei. National Science Council (2004b), Indicators of Science and Technology: Taiwan, National Science Council, Taipei. National Science Council (2005a), Indicators of Science and Technology: Taiwan, National Science Council, Executive Yuan, Taipei. National Science Council (2005b), Yearbook of Science and Technology: Taiwan, ROC, National Science Council, Executive Yuan, Taipei. Organisation for Economic Cooperation and Development (OECD) (2005), Main Science and Technology Indicators 2005, OECD, Paris. Science and Technology Advisory Group (2004), e-Taiwan 2004, e-Taiwan Project Office, Science and Technology Advisory Group, Executive Yuan, Taipei. Shen, Rong-Jin (2006), ‘The direction of Taiwan’s industrial development under the changing of the global economy and trade’, Speech, IDB/MOEA. TIER (2005), Going for New Value Creation and Growth: The Competitiveness Analysis of Taiwanese Innovation System, TIER, Taipei (funded by DoIT, 94-EC-2-A-17-0118-05). Tzeng, G.K. and Lee, M.Y. (2001), ‘Intellectual Capital in the Information Industry’, in Chang, C.V. and Yu, P.C. (eds), Made by Taiwan – Booming in the Information Technology Era, NJ: World Scientific. Wu, R.-I., Lin, X.-W. and Lin, H.-Y. (2002), ‘Moving from foreign technology to indigenous innovation: the case of Taiwan’, OECD-IPS Workshop Promoting Knowledge-Based Economies in Asia, 21–22 November, Singapore. Yu, T.S. (1999), The Story of Taiwan: Economy, Government Information Office, Taipei.
11.
Prospects for Jatropha biofuels in Tanzania: an analysis with strategic niche management Janske van Eijck and Henny Romijn*
1
1.
INTRODUCTION
Global energy supply is currently mainly based on fossil fuels, which have many disadvantages. It is now widely agreed that more sustainable alternative energy sources will need to be developed in the not so distant future. One potentially promising option is biofuels, since these are derived from biomass, have a closed carbon-cycle and do not contribute to the greenhouse effect. The biomass necessary for the production of biofuels can be derived from several sources. Oil-producing crops are prominent among these. Owing to relatively faster crop growth in the tropics as well as the substantial land requirements for large-scale production, developing countries could potentially play a significant role in the cultivation of such crops. Moreover, they could yield major potential economic and environmental benefits for these countries, helping to combat soil erosion, create additional income for the rural poor, and alleviate countries’ balance of payments constraints by lessening oil import dependency or even by yielding export revenue. A gradual transition of the dominant energy regime in these countries from fossil fuels towards biofuels could thus have many advantages. Researchers at Eindhoven University of Technology in the Netherlands recently have been exploring the potential of biofuels in Tanzania. Current initial activities have been directed towards the use of Jatropha curcas Linnaeus (henceforth abbreviated as Jatropha), an indigenous plant which requires little water and few nutrients and has a relatively high oil
* This article was published in Energy Policy, vol. 36, Janske van Eijck and Henny Romijin, ‘Prospects for Jatropha biofuels in Tanzania: An analysis with Strategic Niche Management’, 311–25, Copyright Elsevier (2008). 335
336
Sectoral systems of innovation and production
yield. Jatropha grows wild throughout Africa. Many developing countries in Asia and Latin America are known to have similar oil crops (e.g. Pongamia in India). The research in Tanzania thus promises to generate important lessons for other developing countries as well. This chapter reports on some important recent results from the research in Tanzania. The main questions addressed in the chapter are: To what extent does Tanzania have potential to develop Jatropha-based energy supply sources, and to what extent could these successfully substitute fossil-based energy supplies? What has been its progress in that direction? And what are the major obstacles in this respect? In order to answer these questions we map out the emerging national innovation system for Jatropha biofuels. We do this with an approach called strategic niche management (SNM), which is rooted in evolutionary economics.1 SNM has been designed specifically to study the introduction of radically new technologies in society that promise to contribute to more sustainable development patterns.2 Essentially, SNM conceptualizes the introduction of such technologies as the start of a broad and long-term transition process, in which widely used technologies with unsustainable characteristics are gradually replaced by new, cleaner technologies. SNM posits that this process encompasses a co-evolution of technology and societal factors such as culture, institutions, consumption patterns and preferences, economic regulation, and political governance systems. SNM comes with a set of concepts with which one can systematically document the initial activities and processes that should eventually lead to the adoption and broad diffusion of new technologies in society, and with which one takes stock of important stimulating and constraining factors in that process. Although we do not envisage that the currently emerging Jatropha activities in Tanzania will give rise to a wholesale transition towards a dominant biofuel-based energy regime in the country any time soon (or even at all), Tanzania is well placed to develop a more diversified energy supply system, in which the traditional dominance of fossil fuels is reduced and various renewables – also including hydro, solar and wind power – will play a more prominent role. Even in this less drastic scenario, SNM can still furnish a helpful framework for studying the complex issues surrounding the introduction of Jatropha as an unconventional energy source into an incumbent fossil energy regime, and for highlighting its likely prospects. The research on which this chapter is based involved substantial fieldwork in different parts of Tanzania during March–June 2005. Field data about all the key concepts used in SNM were gathered through interviews with all important actors involved in Jatropha-related activities. Existing literature was used as a secondary source of information. A comprehensive account of the research can be found in van Eijck (2006).
Prospects for Jatropha biofuels in Tanzania
337
The SNM approach is outlined in section 2. A brief introduction to Jatropha and its applications is given in section 3. The analysis is contained in sections 4 and 5, and section 6 contains conclusions and a policy discussion.
2.
THE STRATEGIC NICHE MANAGEMENT APPROACH
Central to SNM is the notion that the introduction of radical innovations that are socially, economically and environmentally sustainable is a complex and protracted process with a high likelihood of failure even if the new technologies appear to be promising. This is because technologies are always part of a much broader system – a socio-technological regime. This is defined as “the whole complex of scientific knowledge, engineering practices, production process technologies, product characteristics, skills and procedures, established user needs, regulatory requirements, institutions and infrastructures” (Hoogma et al., 2002, p. 19). The regime is thus the dominant or ‘normal’ way of doing things. A regime may span several sub-systems, which can be separate regimes in their own right. The emergence of radically new technologies may involve the development of entire new production chains that cut across sectoral boundaries. Some recent SNM studies take an explicit ‘multi-regime’ perspective (Raven, 2006, 2007). This chapter also adopts a multi-regime analysis. In turn, the regime is embedded in a wider context – the landscape, which consists of material and immaterial societal factors. Most of these, for example demographics, political culture, lifestyles and the economic system, can change only slowly over time (Raven, 2005, pp. 31–32). However, sudden or unexpected (often ‘global’) events also emanate from the landscape level. Examples are wars, nuclear disasters like Chernobyl, and the oil crisis of 1973. Mature, well-established technologies form an integral part of the dominant regime and fit into the overarching landscape. This is a result of a long process of mutual adjustment and adaptation of technological and societal factors, in which they gradually get attuned to one another. Innovations with radically new features, especially those that aim to improve environmental and social equity-related sustainability, do not rub well with extant socio-technical regime characteristics that reinforce the importance of short-term economic benefit alone. Simultaneous adaptations in all major parameters of the regime are thus required for such technologies to be successfully developed, introduced to
338
Sectoral systems of innovation and production
the market and adopted widely. However, there is a limit to the adaptability of the regime itself. Regime change is conditioned by landscape factors. This can result in powerful inertia, which can prevent new technologies from gaining ground, let alone unseating incumbent ones. To address these problems, the SNM approach advocates the creation of niches: temporarily protected spaces in which new technologies can incubate and become viable through gradual experimentation and learning by networks of actors. Important parties in these networks include manufacturers, users, researchers, civil society organizations, governmental organizations and possibly others, depending on the specific circumstances. This multi-level perspective – landscape, regime and niche – is adopted in SNM to analyse the potential of emerging transition processes and their dynamics. By identifying major characteristics and developments at each level and tracing their effects to the other levels, one gains insight into the constellation of major forces that push for, and hold back, the development of new technologies through niche formation. Developments at the landscape level are external to the regime (and thus also to underlying niches), but they do influence them, especially through their effect on regime stability. Scope for change in a regime may occur when important landscape conditions change over time, and when the effects of these changes begin to manifest themselves and start to impact on people’s lives. Such changes feed uncertainty, and tensions begin to appear between the main components of the extant regime. People also begin to perceive that problems are no longer solvable within the current regime itself. The extent to which upstart technologies can capitalize on the possibilities offered by this kind of situation is determined by three processes at niche level, where the experiments with the new technologies take place. These are: ‘network formation’, ‘learning’ and ‘stabilization and convergence of expectations’. High-quality niche processes are characterized by a wide and interconnected actor network, by extensive experimentation and learning – not merely about the new technology itself, but also about user acceptance, economic aspects, required infrastructure, etc. – and by expectations that are gradually stabilizing and becoming more specific. In the initial stage of niche formation, experiments with a new technology tend to be confined to a few isolated activities. The principal aim of the learning is to reduce uncertainty about future socio-technical development. Once this has been achieved and a promising technical design has been developed, a ‘technological niche’ is said to have been created. At this stage the innovative activities are still protected from regular regime selection pressures.
Prospects for Jatropha biofuels in Tanzania
339
When the niche processes continue to function well beyond this point, experiments begin to be joined up, learning becomes more widely shared, and its scope expands to include major societal factors. Gaining social and cultural acceptance, institutionalization, and achieving economic viability are of key importance in a regime that will still be rather unfavourable to emerging alternatives to its incumbent technologies. These broader societal adaptation and learning processes in turn induce and guide further improvement of the technology itself, which enhances its societal fitness. When this stage is completed successfully, a ‘market niche’ will have been created. This is the starting point for successful wider commercialization and diffusion processes that may ultimately culminate in a change of the regime itself (i.e. a transition). With this framework, we proceed to analyse the major recent developments with respect to Jatropha in Tanzania. After a brief introduction to Jatropha and its possible uses in section 3, we explore the landscape and regime conditions in section 4 and the niche dynamics in section 5. The information in sections 3 and 4 is largely based on secondary sources. The analysis of the niche dynamics is based on systematic interviews with all the major network actors.
3.
JATROPHA AND ITS APPLICATIONS
The Jatropha plant is easy to establish and drought resistant. It can grow up to 8 metres high and is not browsed by animals. Therefore it has been traditionally used as a hedge, and as markers for graves. It can live up to 50 years and can produce seeds up to three times per annum (Chachage, 2003). There are many possible uses for Jatropha. While it is beyond the scope of this chapter to explore all its uses in depth, we will briefly identify the existing and emerging uses in Tanzania in order to establish to what extent different groups and individuals in Tanzania are already familiar with the plant and have prior knowledge about its cultivation. Prior exposure could form a basis upon which the introduction of Jatropha as a nonconventional energy crop could be built. On the other hand, negative experiences already associated with these uses could constrain the uptake of Jatropha as an energy crop. For information about applications other than those identified by us in Tanzania, see Gübitz et al. (1999), Openshaw (2000), and Jones and Miller (1993). Figure 11.1 shows the main stages in the Jatropha production chain, from seed to end product. Under cultivation come the activities pertaining to growing the Jatropha plant and harvesting the seeds. In Tanzania, Jatropha is grown from
340
Sectoral systems of innovation and production Jatropha seeds
Nursery Cultivation
Seedlings/cuttings Cultivation Seeds Large amount of seeds
Production
Seedcake Oil-expelling facility Oil
Conversion to: – Soap
– Production of biogas – Fertilizer – Briquettes
Usage
Direct use in: – Cooking stoves – Oil lamps In diesel engines (either straight or converted into biodiesel)
Figure 11.1
Jatropha production chain
seeds in nurseries by some women’s groups. Villagers also use cuttings for propagation. Those take less time to establish, but the seed-grown plants are stronger because they develop a tap-root. Direct sowing of the seeds is also practised. Estimated seed yields in different countries and regions range from 0.1 to 15 tonnes per hectare per year (Jones and Miller, 1993; Heller, 1996; Daey Ouwens et al., 2007). People knowledgeable about the Tanzanian situation expect yields up to 10 t/ha/y in fertile locations with good rainfall, but harvests in less favourable conditions are more likely to lie somewhere between 3 and 5 tonnes per hectare per year. Yields depend on a range of factors such as water, soil conditions, altitude, sunlight and temperature. It is also important to maintain soil fertility. Although the plant can grow in poor soils, it is not nitrogen-fixing and requires a nitrogen-rich soil for continued good seed production (Openshaw, 2000). A good way to fertilize is to return (part of) the residue from the pressing of the seeds back to the plants (Gosh et al., 2007). Jatropha press cake has a nitrogen content of up to 6 per cent, similar to that of castor beans and chicken manure. Overall, however, there is still a dearth of
Prospects for Jatropha biofuels in Tanzania
341
research about the influence of various cultivation-related factors and their interactions.3 Seeds are harvested during the dry season, normally a quiet period for agricultural labour. They contain about 35 per cent oil. The oil and presscake contain toxic substances, including phorbol esters, trypsin inhibitors, lectins and phytates (Gübitz et al., 1999). The seeds themselves contain curcasin, also toxic. Curcasin is a strong purgative (Chachage, 2003). Production (or processing) involves pressing seeds to expel the oil, leaving seedcake. In Tanzania, oil is extracted with small manual rampresses and power-operated screw-presses. The extraction rate of the rampress is quite low; the left-over seedcake still contains some oil. About 5 kilograms of seed is needed for 1 litre of oil (Henning, 2004). The capacity is about 1.5 litres per hour. The ram-press is only suitable for processing small quantities, for example for lamp oil for local village use or for smallscale soap-making. The extraction rate of power-operated screw-presses is higher, and the cake residue is much dryer. The Sayari oil expeller, of German design, has a capacity of about 20 litres per hour (60 kilograms per hour) and can extract 1 litre of oil from 3 kilograms of seeds. It is manufactured in Tanzania by an NGO in Morogoro. A Chinese screw-press capable of processing 150 kilograms of seed per hour was installed by another NGO in 2005. Seed storage is important for continuous press operation, since the availability of Jatropha seeds is seasonal. Two options are bulk storage and bag storage. Only bag storage is practised in Tanzania. Storehouses should be well ventilated to prevent loss of seed quality. Location plays an important role, since it has a considerable impact on transport and storage costs (UNIDO, 1983). At the usage stage, the oil and seedcake are consumed or further processed to generate final products. A potentially major end-use from the perspective of this chapter is fuel for diesel-powered vehicles and electric generators. Since Jatropha oil is much more viscous than conventional diesel, using it pure in engines causes problems, despite some claims to the contrary (Heller, 1996). Problems encountered include premature wear of parts and clogging, and inability to start, especially in cool weather.4 Adequate solutions are being sought. Options include adaptation of the oil by mixing with methanol and caustic soda (Research Group IP, 2002); fitting vehicles with dual-fuel tank systems; making engine adaptations; and blending Jatropha with conventional diesel, which reportedly works well up to a proportion of 40–50 per cent Jatropha (Pramanik, 2002). The seedcake is also potentially valuable. In addition to being useful as a straight fertilizer, it can be used to produce biogas for cooking, or – in
342
Sectoral systems of innovation and production
briquette form – as fuel in ovens. Preliminary experiments indicate that Jatropha seedcake is a good feedstock for biogas production (Staubman, 1997; Karve, 2005; Visser and Adriaans, 2007). Moreover, the slurry that remains after digestion in the biogas plant can still be used as fertilizer, although it is debatable whether its quality is still the same as that of the raw seedcake. Experiments in this area are still in the early stages. Other applications for the oil include Jatropha-based soap, use in oil lamps and use in cooking stoves. Chachage (2003) identifies the current activities in Tanzania based on Jatropha oil as soap-making on a limited scale and use in oil lamps. The other applications listed in Figure 11.1 were started only within the past few years. The various activities in Figure 11.1 are linked in different ways. Some are so strongly complementary that one activity can hardly be expected to get off the ground without a simultaneous development of another. This is so for cultivation and processing, and again for processing and any significant type of end-use. The simultaneous initiation of experiments at each of the three stages in the chain would therefore seem to be vital for the emergence of a viable Jatropha chain as a whole. Relations of a more competitive nature are likely to prevail between some of the end-uses. For example, if utilization of the oil in diesel engines takes off, it may well begin to compete with the already existing utilization for small-scale soap production by poor women. Broadly speaking, then, in our view an effective initial constellation of experiments that could pave the way for the successful development of Jatropha as an energy source would need to exhibit the following features: ●
●
● ●
strong experimentation in each of the three stages in the production chain, and growing interconnections between the activities and actors in these stages; transport-related experiments that play at least some part in the end-use stage – it is simply impossible to conceive how a Jatrophabased energy regime can come about without the transport sector being weaned of its fossil fuel dependency; substantial experiments in cooking and lighting, the essential domestic energy services; experiments with the seedcake, since productive use of this by-product will enhance the financial feasibility of producing Jatropha oil.
The above considerations broadly determine the relative emphasis that will be devoted to the different Jatropha-related experiments discussed below.
Prospects for Jatropha biofuels in Tanzania
4.
343
POSSIBILITIES FOR JATROPHA-BASED ENERGY IN TANZANIA: LANDSCAPE AND REGIME ASSESSMENT
Landscape Influences A number of major trends at landscape level influence the scope for developing a Jatropha-based energy supply system. Worldwide trends as well as major Tanzania-specific factors play a role. Among the worldwide trends, the oil price has been a major factor. It has increased sharply during the last few years and is expected to remain high or to rise even further in the near future. In 2003 the benchmark Brent crude was under US$25 per barrel, rising to over US$60 in 2005 and US$63 in 2006 (Energy Information Administration, 2006). Closely related to this point, dependence on countries in the (unstable) Middle East is increasingly considered to be a risk. This is strengthening the movement towards development of sustainable energy sources. Then there is the issue of increased environmental awareness due to mounting evidence of global warming. Policies to promote renewables have mushroomed worldwide over the past few years. At least 43 countries, including ten developing ones, now have some type of renewable energy promotion policy. Although unconventional renewables accounted for only 2 per cent of global primary energy in 2004, more and more attention is being directed towards them. Worldwide production of biofuels exceeded 31 billion litres of bio-ethanol and 2.2 billions litres of biodiesel in 2004. This was 3 per cent of the worldwide petrol consumption of 1200 billion litres. About 900 000 people are active in this sector worldwide. The biodiesel sector grew by 25 per cent per annum between 2000 and 2004. Many other sources of renewable energy are also being stimulated (Renewable Energy Policy Network, 2005). Finally, after a prolonged period of neglect, the World Bank and other major development organizations are again interested in agriculture. Agricultural development is now viewed as crucial for achieving the Millennium Development Goals, especially for developing or low-income countries (World Bank, 2006). Among the Tanzania-specific landscape factors, Tanzania’s low level of development is salient. The country ranked 162 out of 177 on the UNDP Human Development Index. Tanzania’s GDP per capita in purchasing power terms was a mere US$674 in 2004 (UNDP, 2006). Of the country’s 37.6 million people (in 2004), about 36 per cent lived below the poverty line (National Bureau of Statistics, 2004; UNDP, 2006).5 Agriculture provided about half of GDP in 2002, and employed over 80 per cent of Tanzania’s
344
Sectoral systems of innovation and production
workforce (World Fact Book, 2006). In 2004, over 76 per cent of the population still lived in rural areas, where adequate energy services are lacking. Nationally, 94 per cent of energy consumption derives from biomass, mainly fuelwood. This creates serious problems, including soil erosion, deforestation and respiratory ailments. Tanzania’s energy problems have a major rural dimension. Rural energy consumption accounted for 85 per cent of total energy consumption in the country (UNDP, 2006). Another poverty-related aspect is scarcity of foreign exchange. Tanzania imports all its oil (mostly diesel). The official import of over 465 million litres of diesel in 2002 had a value of over TZS465 billion (US$423 million), or 4.7 per cent of GDP. This figure still does not take account of significant quantities of oil smuggled into the country, which have been estimated to constitute 30 per cent of the country’s total oil sales (MBendi: Information for Africa, 2005), representing annual lost tax revenue to the government to the tune of US$30–50 (Rwambali, 2000). These figures suggest that the commodity’s actual import value could be closer to TZS665 million, or approximately 6.5 per cent of GDP. Growing Jatropha could alleviate this problem substantially, while at the same time boosting growth. Substituting 665 million litres of diesel would require roughly 704 million litres of Jatropha oil,6 the production of which would require between 427 000 and 711 000 ha of land (using an estimated yield of 3–5 t/ha/y, yielding 990–1650 l/ha/y). Tanzania’s total surface area is about 90 million hectares, of which 85 million hectares are not arable or under permanent crops. This leaves 5 million hectares for Jatropha, which would be ample for self-sufficiency. Tanzania’s poor infrastructure can also work to the advantage of a Jatropha-based energy regime. The roads are hardly sufficient for lorries, especially in very remote areas. Electricity infrastructure is also very poor. Most electricity is generated in the south (through hydropower), and the power lines are unreliable. This results in regular power shutdowns. Jatropha biofuel can be produced in a dispersed manner, including in remote areas, creating a decentralized energy source independent of the current infrastructure. Despite these advantages, some structural weaknesses in Tanzanian policy-making capacity limit the emergence of a Jatropha-based energy system. The country’s current National Energy Policy (adopted in 2003) merely affirms the desirability of promoting “development and utilisation of appropriate new and renewable sources of energy” (Uisso, undated), without specifically mentioning biofuels. Policies to promote and regulate the production and use of biofuels are still under development. In 2005, the government began to set up a Rural Energy Agency (REA) which will be the responsible institution for rural
Prospects for Jatropha biofuels in Tanzania
345
energy development. REA will facilitate development of projects which will ultimately be owned and implemented by the private sector, NGOs and community-based organizations. A Rural Energy Fund (REF) will also be established to provide capital subsidies in order to reduce the risk to project developers. Both these inititatives are not specific to biofuels, however. The first specific biofuel-related government initiative started only in 2006 with the formation of a National Biofuel Taskforce that brings together several ministries and stakeholders. The taskforce has been charged with developing guidelines for the development and regulation of biofuel activities, which were ready in draft form in August 2008. Stakeholder meetings are currently being organized. The adoption of official guidelines should hopefully deal with the lack of a clear and fair biofuel tax regime. Investors interviewed for this research pinpointed this problem as the main bottleneck to the development of the sector. The recent government initiatives are positive developments in principle, but the pace of policy making and implementation is slow. Moreover, standards of governance will also need to improve for directives and institutions to work effectively. Large Jatropha plantation farmers interviewed for this research reported major problems in their dealings with the government. One noted that it is crucial to work with someone who knows his way around at the government level. Regime Dynamics From the perspective of SNM, the main focus of this section should be on Tanzania’s prevailing energy regime, because this affects the possibilities for using Jatropha as an energy source, and we are ultimately interested in uncovering the potential for a shift in this particular regime. However, there are other regimes whose features also influence the potential for a Jatropha-based energy supply shift in the country. In particular, Jatropha cultivation is influenced by the agricultural regime, oil production by the vegetable oil regime, and investment possibilities in Jatropha-related activities by the financial regime. We first discuss the energy regime and then go on to highlight key characteristics of the other three regimes in so far as they influence the scope for developing a Jatropha-based energy sub-sector. Some aspects of Tanzania’s extant energy regime have already been introduced. Of the total energy consumption in 2004, between 92 and 97 per cent derived from biomass,7 mainly fuelwood and charcoal. A smaller percentage is derived from agricultural and forestry waste and dung. Fewer than 8 per cent of households are connected to the electricity grid.
346
Sectoral systems of innovation and production
Table 11.1 Consumption of various fuels in Tanzania, in ’000 metric tons oil equivalent (MTOE) Fuel type Oil products Coal Hydro Combustibles, renewables and waste Natural gas Imported electricity Total
1999
2004
628 3 187 14 079
1 209 40 203 17 181
0 0 14 897
107 10 18 749
Average annual growth 1999–2004
4.7 %
Sources: Compiled from World Resources Institute (2006) (country profile 178) and International Energy Administration (2007).
Recent estimates of consumption of various fuels in Tanzania indicate a significant recent growth in energy demand of 4.7 per cent per year on average during 1999–2004 (see Table 11.1). Energy demand growth is unlikely to decelerate in the near to medium term. The potential for biofuels like Jatropha – as well as other renewables – can therefore be expected to be quite high. Several renewable energy technologies that avoid the problems related to use of traditional energy sources are available to help meet this demand. The main projects being promoted are biogas production for cooking, improved cooking stoves and kilns, solar thermal applications for water heating and cooking, and solar and wind technologies. However, few serious attempts have been made to utilize wind and solar energy. Schemes implemented to date have suffered from a number of problems, including: poor payment records; low awareness of, and lack of confidence in, these technologies; high investment requirements in relation to low purchasing power of target groups; weak institutional framework and infrastructure for effective promotion and support; and lack of appropriate credit mechanisms (AREED, 2001). While some of these problems could also be expected in programmes to utilize biofuels like Jatropha, the latter seem to have certain advantages over other sustainable energy sources. Initial investment requirements could be quite low, since many Jatropha activities can be started on a small scale, reducing the need for substantial loans. Another advantage is versatility. In principle, Jatropha oil can be used for all the main purposes for
Prospects for Jatropha biofuels in Tanzania
347
which energy is needed, that is for transport, electricity generation, direct lighting and cooking. Of course, switching from imported fossil fuels to Jatropha is unlikely to be problem-free. Diesel engines would have to be modified to be able to utilize pure plant oil (for example, by installing a two-tank system or a preheating system). The additional cost to vehicle owners would engender considerable resistance. The technical knowledge required for the modifications is not widespread either. However, there appear to be somewhat easier options, which are especially interesting for applications in the near future. When Jatropha oil is converted into biodiesel, vehicles require almost no modification (only the fuel hose needs to be resistant to biodiesel). The University of Dar es Salaam and Eindhoven University of Technology have begun to research the effects of biodiesel on vehicle engines. Jatropha oil could also be blended with conventional diesel and sold at petrol stations. People would not even know they were driving on biodiesel, so resistance from users would be low. Resistance to biofuel blends can, however, be expected from actors who currently dominate the existing energy regime, notably petrol distributors. Although the experience with Jatropha in Tanzania is still too recent to be able to gauge actual reactions by these parties locally, recent experiences in other countries are instructive in this respect. For example, in the UK most service stations are operated by (or leased from) the fossil oil industry, “which is pathologically opposed to going down the biofuel route. Even the supermarkets are loath to allow E85 [an 85–15 per cent fossil/bio-ethanol blend] pumps to take up valuable space on their forecourts. In fact, as yet just a handful of supermarkets in Somerset have agreed to sell E85, and only after being offered significant incentives” (Madslien, 2006). Car and tractor manufacturers are raising objections too. Some companies refuse to honour warranties when engines are run on higher than 5 per cent biodiesel blends, or they dictate special intensive maintenance regimes. Different manufacturers stipulate different conditions (Fone, 2006). This can only contribute to uncertainty and confusion, and an understandable reluctance to adopt biofuels among potential users. The comparatively high price of Jatropha is also still a drawback. In areas where fossil diesel is widely and easily available, consumers are unlikely to want to switch when the diesel pump price in Dar es Salaam and Arusha was TZS1100 (July 2005) and TZS1400 (2006), compared with TZS2000 for Jatropha oil (both in 2005 and in 2006). Only a small minority of environmentally aware urban consumers might be willing and able to pay, say, 30 to 50 per cent more for Jatropha-blended diesel than for conventional diesel. We conclude that fossil fuel prices would have
348
Sectoral systems of innovation and production
to rise much more for the biofuel alternative to become attractive to the majority of diesel users. However, this does not imply that there are no growth prospects for Jatropha-based energy at all until relative prices change. The situation is somewhat more favourable for Jatropha in more remote areas, where the diesel price is higher because of transportation costs. In Mtu Wa Mbu, an up-country town reachable by a good tarmac road from Arusha, the diesel pump price was TZS1300 (2005), while it was TZS1700 in Engaruka, a township approximately 50 kilometres from Mtu Wa Mbu reachable by an unpaved track. The first viable niches for Jatropha-based energy are likely to arise in areas like these, rather than in the major towns. A particularly interesting application for Jatropha could lie in the area of decentralized electricity provision. However, maintenance of the generating equipment requires some technical skills, while working out an equitable system of sharing the costs, responsibities and benefits of electrification can also be highly problematic in local communities. The viability of decentralized electrification systems is determined by many factors other than financial ones. It especially requires a degree of social cohesion, trust, good leadership and effective organization (Gunaratne, 2002). Another potential energy-related application for Jatropha could be as a source of direct lighting. Currently, kerosene is the light source of choice for more than 90 per cent of Tanzania’s low-income households; in rural areas, the figure is almost 100 per cent (Household Energy Network, undated). Kerosene is imported by private oil companies and sold at petrol stations. This facilitates its use by low-income urban and rural households (Household Energy Network, undated). Kerosene imports keep rising. In 1998, Tanzania imported over 122 million litres; in 2002 this had risen to 154 million litres (not counting unofficial imports). The price also keeps rising, from TZS426 per litre in 2002 to TZS850–980 in 2005.8 However, at TZS2000 per litre the Jatropha-based alternative is still far too expensive. Moreover, Jatropha needs a special oil lamp, which costs TZS1200. Although the Jatropha oil used in that lamp lasts for quite a long time compared to kerosene in kerosene lights, these (floating) lamps are not widely available. Kerosene prices will have to go much higher for this market niche to develop, except perhaps in some remote areas. Finally, Jatropha could be used as an alternative energy source for cooking. The dominant energy source for cooking is fuelwood. In 2000, households in sub-Saharan Africa consumed nearly 470 million tonnes of wood and charcoal. This is far more than on any other continent. Wood or crop residues are the primary source of energy for 94 per cent of rural households and 41 per cent of urban households in the region (Renewable
Prospects for Jatropha biofuels in Tanzania
349
Energy Policy Network, 2005). This would seem to be a highly favourable environment within which alternative supply sources like Jatropha could develop. Several NGOs are actively spreading more efficient cooking stoves, some for use with Jatropha. TaTEDO, for example, has provided several households with improved stoves. Kakute has been providing several groups with a biogas system that uses Jatropha seedcake. This indicates that some alternative energy technologies for cooking are available. However, our field observations indicate that they are not popular. The biogas system is used only when conventional biomass sources are unavailable (e.g. during the rainy season). Women interviewed for this research indicated that the biogas system increases cooking time. They also expressed misgivings about possibly poisonous fumes, a belief fed by their traditional knowledge that Jatropha nuts are poisonous when ingested. Unlike NGOs, they do not place much value on the positive health effects of decreased indoor smoke. Another major factor is cost. A Jatropha cooking stove costs TZS12 000–20 000,9 compared to TZS10 000 for a biogas cooker.10 Not surprisingly, fuelwood remains the cooking fuel of choice as long as it can be collected for nothing. The dominant cooking regime is strong. Thus far, alternative systems have not been able to meet people’s demands and priorities well enough. From the point of view of the energy regime as a whole, then, it is clear that there are prospects for Jatropha, but viable possibilities in the short term are likely to be restricted to small niche markets constituted by remote areas with high fossil diesel prices, and specific types of urban clients (like eco-safari companies). A host of technological and skill constraints and socio-economic problems also need to be overcome before Jatropha could eventually become a popular energy source in areas like transportation, electricity provision, cooking and direct lighting. In addition to analysing the conditions governing the appeal of different Jatropha energy-related uses, we need to determine how attractive it would be for people to grow Jatropha and/or to press the seeds. In order to shed light on the attractiveness of Jatropha cultivation, we examine prevailing conditions and trends in Tanzania’s agricultural regime. From our interviews with villagers some factors that are favourable to the introduction of Jatropha as a cash crop emerge. First of all, farmers indicate that cultivating Jatropha is not radically different from current practices in the agricultural regime. Farmers are already used to planting a fence around their land to protect their crops from wild animals. Many households are also already used to cultivating cash crops: 59 per cent of rural households in Tanzania sell food crops for onward processing (National Bureau of Statistics, 2000).
350
Table 11.2
Sectoral systems of innovation and production
Yields, prices and revenues for selected food crops and Jatropha in Tanzania
Crop
Average yield (kg/ha)a
Average price (TZS/kg)b
Average revenue per ha (TZS)
Maize Sorghum Finger millet Jatrophac
1205 717 674 3000 to 5000 (kg seeds)
120 to 271 (2004) 270 to 300 (2004) 260 to 280 (2004) 80 (2004–05)
144 600 to 326 600 193 600 to 215 100 175 240 to 188 720 240 000 to 400 000
Notes: a Yields for maize, sorghum and finger millet were calculated by averaging annual yields over the period 1996–2003, Arusha region. b Prices for maize, sorghum and finger millet are based on minimum and maximum average wholesale prices during three-quarters of 2004, Arusha region. c Assumed Jatropha seed yield of 3000–5000 t/ha/y and seed sales price of 80 TZS/kg. Most Jatropha fields were in the Arusha region. Sources: Tanzania Ministry of Agriculture and Food Security, http://www.agriculture. go.tz/MAFS-Services/Statistics.htm and http://www.agriculture.go.tz/AGSTATS/ Commodity-price.htm.
Moreover, farmers are actively looking for new crops because of recent low prices for most existing crops. Table 11.2 shows that cultivating Jatropha as a cash crop could be profitable compared with some major existing crops, especially in fertile areas with sufficient water, once a reliable market for seeds begins to develop. In less favourable areas unsuitable for food crop cultivation Jatropha might also be an attractive investment, because it can grow in difficult conditions (although yields would not be very high). The possibilities are still largely theoretical. There is no well-established market for Jatropha seeds yet, and this could hold back investments by farmers. Some NGOs and commercial companies (one commercial company on a regular basis, the other actors randomly) currently buy seeds from villagers on a small scale in the northern regions. Their system of local collection points and buying at weekly markets is comparable to the current system of private business persons buying agricultural produce from farmers. However, collecting the seeds in this way will become unwieldy as the supply of Jatropha seeds increases, especially in view of poor roads and inadequate transport facilities. Large-scale commercial Jatropha oil production will probably have to rely on more centrally located plantations, at least in addition to independent outgrowers. It remains to be seen to what extent independent seed growers, especially
Prospects for Jatropha biofuels in Tanzania
351
small-scale farmers, ultimately will be able to participate in such a system. Much will depend on the nature of the specific energy applications that will develop. For example, if Jatropha oil as a source for diesel engines for rural electrification would take off, local small-scale seed suppliers could get an excellent local market. In contrast, if the main application of Jatropha oil were to be as biodiesel for transportation (including for export), it would be much more difficult for small-scale remote farmers to market their seeds. One disadvantage of cultivating Jatropha compared to existing crops should also be noted, because it may impede investment. Jatropha is a multi-year crop which can be harvested only two to three years after planting. This requires more long-term thinking and financial reserves than cultivating annual and seasonal crops. Intercropping of Jatropha with other crops, or growing Jatropha as hedges around fields, that is, using it as a supplementary crop, could alleviate this problem. Additional barriers to the uptake of Jatropha are likely to emanate from farmers who are used to applying artificial fertilizers. It might be difficult to induce them to use Jatropha seedcake (or slurry from a Jatropha-biogas digester) as fertilizer instead. Artificial fertilizer is compact and therefore easy to transport. In contrast, raw seedcake and digested slurry are voluminous and difficult to handle. Farmers are also likely to be reluctant to change their habits, since they know that Jatropha seeds are poisonous. Applying Jatropha seedcake as fertilizer on food crops could constitute a psychological barrier. Overall, we can conclude that Jatropha seems to compare favourably to other major crops in terms of yield and price, but there are many nonfinancial barriers in the agricultural regime that still need to be ironed out. Compared to the energy regime and the agricultural regime, Tanzania’s vegetable oil regime is not an overriding constraint on the development of a Jatropha sector. Tanzania already produces substantial quantities of oil-seeds for edible purposes and for industrial use. Edible oil-seeds are generated from groundnuts, cashew and sunflower. An example of an industrial oil-seed is castor. It is not unusual in Tanzania for an oil facility to be owned by a local co-operative. Farmers have their seeds pressed there for a fee. So current practice is not very different from what would be required for production of Jatropha oil. The only problem emanates from the fact that Jatropha is poisonous. Processing firms are unwilling to use the same equipment to press edible seeds and a poisonous seed. It might be necessary to develop new oil-expelling facilities specifically dedicated to pressing Jatropha seeds. The fourth regime that exerts an influence on the prospects for Jatropha as an energy crop is the financial regime. Financial considerations are
352
Sectoral systems of innovation and production
all-important in decision making in poor countries like Tanzania. Any extra cost, for example for cooking fuel, will not be accepted easily. Many people are risk-averse. Especially in rural areas, interest on loans is often quite high and loans are not given easily. However, conditions are improving slowly. There are experiments with group-based savings and credit schemes by micro-credit organizations such as FINCA, which avoid the high overhead costs charged by banks. One Jatropha-NGO is linked to a micro-credit facility which has given loans to some farmers in Engaruka to the tune of TZS50 000–500 000. The year 2005 was declared ‘the international year of micro-credit’, an indication of growing awareness and increasing initiatives in this field, both worldwide and in Tanzania. The financial regime does not constitute the dominant constraint on the emergence and growth of a Jatropha energy sector. We conclude that the current energy regime and to a lesser extent the agricultural regime still present major obstacles for an emerging energy transition towards Jatropha, in spite of many favourable developments at landscape level. In the energy regime, cost considerations are of crucial importance. Uncooperative behaviour should also be expected from powerful parties in the dominant regime. Inexperience with implementation of small-scale decentralized electrification projects is also a factor to be reckoned with. In the agricultural regime, the apparent financial attractiveness of Jatropha relative to existing crops is counterbalanced by a mixture of barriers associated with organization, logistics and psychology.
5.
RECENT DEVELOPMENTS AT THE NICHE LEVEL
Fieldwork Methodology We tried to identify all significant socio-technical energy-related SNM experiments with Jatropha in Tanzania by talking to local people who were knowledgeable about the budding sector, primarily officials from the Ministry of Agriculture and Minerals and the National Biofuels Taskforce, NGO representatives, academics and private entrepreneurs. For identifying these key informants we relied on the snowball method, starting with a few known experts, and identifying others through these people. An “experiment” in this context should be understood as an activity undertaken by an individual or a group aimed at growing Jatropha, seed pressing, or developing one or more end-use applications for the oil or the seedcake. Most of the experiments took the form of development projects
Prospects for Jatropha biofuels in Tanzania
353
led by local NGOs and governmental agencies, but there were also a few for-profit ventures run by commercial companies. Some had foreign connections involving financial support or knowledge transfer, while others were purely local affairs. In total, 17 experiments were found, of which 16 were visited and one contacted through e-mail. Most experiments were situated in the Arusha and Kilimanjaro regions in the north-east. Others were situated in Morogoro, in Dar es Salaam and in Tunduru in the south. In total we conducted interviews with 28 participants in these 17 experiments. They included representatives of a local university, two local branches of transnational companies, seven NGOs and a micro-credit organization. The remainder were mainly individual farmers and farmer groups who participated in development projects run by the earlier-mentioned agencies or companies. Several government representatives were also visited, but since the government is not involved in actual Jatropha experiments these people did not form part of our survey about niche processes. The free-flowing interviews held with them predominantly focused on contextual information about Tanzania’s energy bottlenecks and strategy, and the government’s views on the role of biofuels. Participants located outside Tanzania (such as international donors and car manufacturers) were not interviewed. The interviews with the 28 survey participants were held face to face, with the help of a detailed standard checklist of open-ended questions. Each interview covered information about the goal, history and nature of the Jatropha activities undertaken. The respondents were requested to provide considerable details about the development trajectory of their Jatropha activities over time, in order to get a sense of the evolution of the sector. The three key SNM niche-formation processes were covered: actor network activities, people’s learning processes, and the dynamics of their expectations. Considering the complexity of the processes, the experimental nature of the research, and the low level of literacy and capacity for abstract thinking present among some respondents, we confined ourselves to gathering mostly qualitative information through informal discussions, loosely guided by our checklist. We did not ask respondents to rate issues on qualitative scales. We did, however, try to collect quantitative estimates from them about the costs and benefits of each major Jatropha-based activity. The discussion about the findings from the interviews is structured according to the different stages of the Jatropha value chain (Figure 11.1). For each stage, the findings about the three niche processes are discussed in the order of actor networking, learning, and dynamics of expectations.
354
Sectoral systems of innovation and production
Dynamic Niche Formation at the Cultivation Stage The actor network in the cultivation stage is expanding quite rapidly. Respondents indicated that more and more farmers are starting to plant Jatropha, expecting to make a considerable profit. This is happening mainly because they are now able to sell their seeds to a multinational company that started operations in 2005. The company pays a guaranteed fixed price (TZS80) for several years, reducing the risk of a price fall. Low and declining prices for existing crops acted as an additional push factor. The actor network is quite diverse. There is participation by NGOs, private farmers, farmer groups, individual larger farmers, and private companies (seed buyers). Only research organizations had not been involved, but one was beginning to undertake research at the time of our research. When we judge the quality of the network formation processes at the cultivation stage by the criteria discussed in van der Laak et al. (2007), it is possible to draw cautiously positive conclusions. Firstly, the network participants consistently reported evidence of continued expansion and increasing network variety. The experiments taking place at this stage are also connected to each other to some extent through what one could call “bridging organizations”. However, Tanzania’s poor infrastructure and large distances make it difficult for some actors in different regions (especially local farmers) to meet face to face. The inter-experiment network is not tight and unified. Respondents reported many learning processes in this part of the chain, mostly with regard to how Jatropha should be grown and managed (e.g. with respect to watering, fertilizing, intercropping and pests). One particularly important lesson reported by some farmers is that Jatropha establishes more successfully, grows considerably faster, and yields a lot more seeds when the plants are adequately watered. This suggests that the initial widespread belief, that Jatropha can do well in infertile and dry soils and that it therefore does not compete with food crops, may turn out to be too optimistic. If this emerging lesson were to be confirmed more widely, this could have major ethical ramifications, because the attractiveness of Jatropha as a source of biofuel in a poor country like Tanzania stems precisely from its supposed hardiness and its non-competition with food crops. However, possible implications of this nature were not yet being discussed locally. Lessons about institutional and psychological issues were also being learned. For example, the failure of an NGO to return to suppliers to collect and pay for seeds, as it had promised earlier, drove home to other buyers the importance of building trust and keeping promises with villagers if they are to turn into reliable seed suppliers. Another lesson relates to
Prospects for Jatropha biofuels in Tanzania
355
how local land ownership patterns can affect the ease with which villagers can enter into agreements with buyers. Buyers that had tried to persuade women in Maasai areas to start growing Jatropha had had disappointing results because, as it turned out, they do not have ownership over the local land. This problem is much less serious in areas with mixed tribes. Buyers were also learning about how local people’s cultural traditions can affect their behaviour. Some local people who are used to Jatropha being used as grave markers (the Swahili name for Jatropha is Mbono Kaburi, or graveyard tree) have misgivings about using the crop for commercial purposes. Some farmers had started to conduct systematic experiments for gathering specific bits of knowledge. They said that much knowledge about cultivation is still lacking, but that it is becoming clearer where the gaps are and how to fill them. These individuals were thus beginning to build learning capabilities: they were not merely learning new facts about Jatropha through practice, but also about the process of learning itself, about how they could accumulate new knowledge in a more systematic and effective manner. Still, we found little evidence of lessons being widely shared. At this stage, most lessons are still being learned by individual actors, and then sometimes being discussed with one or two others, who then sometimes conduct their own experiments for independent verification. The overall impression is one of widespread and varied learning but also one of high fragmentation. When viewing the learning processes against van der Laak et al.’s (2007) conditions for good-quality learning, this niche process was impressive in terms of breadth (which went well beyond technical issues), but we found little evidence about actors questioning underlying assumptions surrounding Jatropha cultivation (‘reflexiveness’), for example concerning the potentially sensitive issue of competition with food crops for fertile land and water. The expectations of actors involved in cultivation are predominantly high and positive, and in some cases rising even further (in response to higher than expected yields). However, the experience with the crop is still too brief for expectations to stabilize or to allow very specific conclusions. The positive expectations among growers are based mainly on short-term favourable profitability experiences combined with forecasts of a large future market for biofuels, which are fed by articles in the press, local seminars, statements by politicians and NGO workers, and so on. There are no tangible results from longer-term experiments yet. If the market for Jatropha oil turns out to be smaller or less profitable than anticipated, farmer prices will drop. One might also anticipate disappointing longer-term yields when plants are inadequately fertilized. Yet only one
356
Sectoral systems of innovation and production
respondent cautioned that the current expectations could be overblown, and that more research into cultivation practices, feasible yields and applications is urgently needed in order to avoid unrealistic expectations and their possible subsequent implosion. The current high expectations among cultivators may not be entirely justified, but they are understandable given their experiences with yields and seed sale prices in the short run. A tentative cost–benefit analysis for five Jatropha cultivation projects reveals real internal rates of return of well over 100 per cent for all except one very small loss-making project by a religious order that was not run efficienctly and that was in any case meant for self-sufficiency purposes only.11 In conclusion, when seen in the light of van der Laak et al.’s (2007) conditions for high-quality expectations formation, the performance of this niche process is rather mixed. On the one hand, it seems to have done quite well in terms of expectation sharing across actors. At the same time, one can point to a major weakness. At this early stage, people’s expectations are still mainly driven by subjective beliefs, wishful thinking and short-term results which may not be representative of longer-term conditions, rather than by tangible and more reliable results from longer-term experiments. Summing up the results for the three niche processes in the cultivation part, we see considerable growth and dynamism, but also still a lot of uncertainties, expectations partly based on ‘air’, as well as emerging ethical issues. Many knowledge gaps and uncertainties appear to be surmountable, however, because remaining barriers to growing Jatropha as a cash crop mainly have to do with local villagers’ lack of information on specific aspects of the cultivation regime. Conservative risk attitudes are partly culturally determined and would need time to change, but organizing demonstration plots, seminars, and field visits to successful projects can help. Overall, the niche formation processes seem to be proceeding reasonably well for the cultivation part of the chain. Fragmented Niche Dynamics at the Oil Pressing Stage The oil pressing part of the Jatropha chain shows a more mixed performance from an SNM point of view. With the involvement of an increasing variety of actors, including NGOs, women’s groups (press users), equipment producers and subcontractors, and even a foreign university, the network is quite substantial. However, a drawback is that most of the contacts run through one particular NGO, which is also known to be rather selective in the information it wants to share. There are few lateral links that could facilitate alignment among actors in the network.
Prospects for Jatropha biofuels in Tanzania
357
Given this state of affairs, it is not surprising that the learning processes in this part of the chain have been concentrated on technical lessons about the operation of the presses and the quality of the seeds, without extending substantially into social and institutional domains. Several respondents indicate their dissatisfaction with the manual ram-press because of its limited capacity and arduous operation, although the equipment is found to be reliable. Users are more satisfied with the Sayari expeller, although it was also observed by one respondent that this equipment is more prone to breakdowns, which might constitute a problem when using it in remote areas where spares cannot be procured easily. One problem that has emerged is that the equipment is designed for softer sunflower seeds, and has problems digesting hard Jatropha nuts. The respondent thinks that dedicated Jatropha presses will be made once the market begins to grow. There have been no broader learning processes in relation to infrastructure yet, about how best to set up a pressing facility or, for example, how best to store the seeds. Lessons in relation to user acceptance are even more limited. Some people report being reluctant to handle Jatropha seeds for pressing because of its supposed poisonous nature, but no scientific evidence is being generated with which to support or refute these claims. And we found no evidence about reflexive learning at all. From the perspective of van der Laak et al.’s (2007) criteria for a good-quality learning process, the quality and quantity of the processes observed by us have major limitations. A further sign of inadequacies in the niche formation process in the oil pressing stage is that the participants’ expectations vary widely. It was impossible for us to distil a consensus about the likely direction in which the Jatropha chain will evolve. Will small expelling units be installed in decentralized locations, or will there be a few big centralized ones? Will the main technology be smaller presses or larger expellers? Moreover, aspects such as transport, reliable and efficient equipment and its maintenance, and financial support are seen as important barriers and uncertainties by some. Yet, according to the data furnished by our respondents, the economics of oil pressing would be highly promising. On the basis of an eight-hour workday and a 50 per cent capacity utilization rate, we estimate a real internal rate of return of 1247 per cent for the ram-press, and 1387 per cent for the Sayari expeller.12 The pay-back period for both would be two years. The ram-press is highly labour-intensive to operate, but only 165 litres of oil need to be sold for the investment cost to be recovered. This is feasible even for a small-scale project. For the Sayari expeller, the breakeven point is 2000 litres, because its investment cost is much higher. Its profitability could increase further once a productive use for the seedcake
358
Sectoral systems of innovation and production
is found (for example, if it could be sold as fertilizer). However, data on actual quantities of oil sold is lacking, so we cannot report reliable profit estimates. We conclude the discussion about the niche processes in the oil pressing stage by saying that only in the field of equipment use do we see that people’s expectations are beginning to reflect tangible results of learningdriven experiments. Overall, the SNM processes in the pressing niche have not proceeded as well as in the cultivation stage. In particular, the network needs to develop more lateral relations for more effective and broader experimentation and learning, and for alignment of expectations to occur. Rudimentary Niche Formation at the End-Use Stage This brings us to the dynamics at the application stage. With respect to the use of Jatropha oil in diesel engines, there are mainly just positive expectations, but no actual lessons from experiments. The different potential options for oil use still remain to be explored. The actor network is quite limited and shows no signs of expansion. Just three actors – a transnational company, the University of Dar es Salaam (UDSM) and a development project – are pushing this application in Tanzania. Perhaps more actors will get moving when UDSM’s engine test results show no damage to the engine and get publicized. There are no learning processes on the user side yet. The only technical learning processes have been some experiments carried out by the transnational company in its home base in the Netherlands. Worldwide, of course, many more experiments are being carried out on the properties of Jatropha oil; these seem to point in a positive direction, especially about the possibilities for converted oil. However, some technical uncertainties remain, for example about long-term effects on engines. It is symptomatic of the embryonic stage of this part of the chain that we were unable to gather reliable data with which to conduct an economic cost–benefit analysis. All we can say is that conventional diesel is priced at TZS1100 per litre at the pump in cities, which is still low in comparison with the prices of about TZS2000 per litre quoted by Jatropha sellers. Expectations are positive but remain vague owing to lack of learning. As far as the utilization of the seedcake is concerned, niche formation processes are hardly present. Only one NGO has tried to experiment with a biogas installation so far, but – as reported earlier – the women users were dissatisfied with its performance. On the positive side, they noted that Jatropha biogas burns well and that the seedcake generates a lot of gas. Much more experimentation and learning about technical properties will be needed for this technology to take off. The seedcake could also
Prospects for Jatropha biofuels in Tanzania
359
be pressed into briquettes for use as fuel in wood stoves or ovens. This is practised in one remote village, Engaruka, but information about the experience is lacking. Using seedcake directly as fertilizer could be more promising. It has favourable nutritional qualities. This possibility was mentioned by several respondents, and some farmers apparently do put the cake back on their land. However, although it is generally known that the seedcake contains a good supply of nitrogen and is therefore a good fertilizer, systematic tests about its effects on the crop still remain to be carried out. Thus, there are no local learning processes yet, although expectations are slightly positive. Potential network actors in this domain are farmers who want to use the cake as fertilizer, and the oil pressing facilities, which generate Jatropha cake as a by-product. If a Jatropha-driven biogas application developed, then actors using such systems would also be important actors in this part of the network, since the digested feedstock could be re-used as fertilizer. It would appear to be highly important for the fertilizer-related niche to develop, because good long-term seed yields from Jatropha cultivation will require adequate fertilization. Use of Jatropha oil in lamps is not a well-developed application either. The network is rudimentary. Some young people were trained to produce Jatropha oil lamps, but did not use their knowledge. An NGO is the only current producer of the lamps. It claims to produce about 1000 lamps per year, an indication that there is some market. Households in remote Maasai areas such as Selela and Engaruka use them. Kerosene is relatively expensive there, and Jatropha is a common plant in their areas. Learning has been limited. A functional lamp has been developed, but this is a very simple adaptation of the ordinary kerosene lamp, with a thicker wick. Expectations about the future of this niche are unclear. There seem to be too many barriers to using Jatropha oil in cooking stoves to speak of a potentially viable niche. The stove that has been developed is not functioning properly, and users are unhappy. The emissions might even be dangerous. Early learning processes for this application were very good. The stove was developed and tested with the involvement of various actors. However, the experiments petered out, and many actors left when lessons were not widely shared, so a well-functioning network was not established. Expectations are already very negative. The prospects for this application are not good. Finally, small-scale Jatropha soap-making has been undertaken for several years by local women’s groups assisted by an NGO. This has generated many learning experiences for local women and their NGO partner with the cultivation, pressing and utilization of Jatropha. It has also contributed to network formation. The soap is an excellent product with
360
Sectoral systems of innovation and production
strong antiseptic qualities. There is a niche market for it, both in Tanzania and in neighbouring countries. It commands a high price compared with ordinary soaps, so only a minority of people can afford it. The expectation is that this market will not expand much beyond its current size. Overall, it is clear that the niche formation processes at the application stage are found wanting according to all the quality criteria listed by van der Laak et al. (2007). Expectations are highly unspecific and not based on tangible results from experiments, and networks are small and incomplete, with few signs of growth. Learning processes are limited and narrowly technical, and do not encompass social aspects surrounding the new technology. There is also no evidence of questioning of broader assumptions about the technology and its performance.
6.
CONCLUSIONS
The main conclusion has to be that the innovation system for Jatropha biofuels in Tanzania is still embryonic and that its future is still unclear. Despite the favourable constellation of many contextual ‘landscape’ factors, there remain prominent barriers within Tanzania’s existing energy regime and agricultural regime. These include: structural infrastructural and logistical problems; technical skill and knowledge gaps; a limited local research infrastructure; vested interests of powerful actors in the extant energy regime; cultural barriers associated with traditional uses of Jatropha; psychological obstacles emanating from known poisonous qualities of the crop; and a considerable price disadvantage for Jatropha oil except in remote locations. Moreover, the government’s role has not been facilitative enough. The niche analysis showed that the Jatropha activities in the different parts of the production chain still consist of a loose set of experiments – we cannot speak of a coherent actor network. There is no viable market niche for Jatropha yet, not even a well-developed technological niche. The niche processes in the cultivation part of the production chain have proceeded quite well. However, further downstream we see small, incomplete or dysfunctional actor networks, insufficient experimentation and learning, and extremely diverse expectations that are not showing any signs of convergence. All this clearly indicates that the dominant energy regime in Tanzania is still well entrenched. However, the many experiments that have been carried out so far have created much awareness and interest in the Jatropha plant, which future developments can capitalize on. Our analysis of the Jatropha innovation system with strategic niche management yielded several useful pointers for action by different actors
Prospects for Jatropha biofuels in Tanzania
361
who could foster a transition process. First of all, the Tanzanian government needs to become more pro-active and ensure effective, consistent and fair support for biofuels, and for renewables in general. One policy issue that still needs to be settled is to what extent Tanzania would want to get involved in Jatropha cultivation for export at this stage, as opposed to harnessing the resource for the domestic market, including for remote rural areas. Given the serious energy shortages in the country and the financial drain associated with oil imports, there is a clear case for encouraging domestic use. Currently the domestic tax regime does not favour adoption of biofuels, as happens in, for example, the EU. A useful task for the government would be to facilitate temporary protection of Jatropha niches, especially through a transparent system of taxes and subsidies that would make cultivation for domestic purposes more attractive. In a more favourable policy environment, organizations like NGOs, universities, farmers and large Jatropha investors can play complementary roles. The SNM analysis brought out that it is important for these actors to promote production-chain management: they should focus on stimulating simultaneous experiments at all the stages of the Jatropha chain with many different types of actors, building interconnections between these experiments, and disseminating the learning processes widely to all actors involved. Banks should make available micro-credit funds for (groups of) small-scale investors wishing to start Jatropha-based activities. Progressive farmers and active local women’s groups could become local niche champions, promoting niche formation at project (experiment) level through actor network building, stimulating learning and levelling expectations. A related policy dimension that needs attention is REA’s performance in relation to enhancing the participation and capabilities of local communities in rural energy projects. It has been noted that the majority of projects may not give sufficient attention to the needs and preferences of the rural poor (Arvidsen and Nordström, 2006; Shuma, 2006). There is indeed a danger that, if investment in Jatropha does begin to take off in earnest, the sector could be taken over by big commercial players interested in setting up large plantations. In this scenario, less glamorous but socially useful small-scale projects aimed at energy provision by and for local communities could lose out. An influx of large investors could also lead to undesirable competition with food crops. Although Jatropha can grow in hostile conditions, there is increasing evidence that seed yields are sensitive to soil fertility and water availability. Farmers could be induced to become outgrowers for large buyers, converting too much prime crop land to Jatropha cultivation. Poor villagers could also be induced to sell their land to large investors, while it is still unclear whether their short-term gains would
362
Sectoral systems of innovation and production
constitute adequate compensation for long-term loss of livelihoods and loss of land for food production. It has to be said that the Tanzanian government is taking a rather cautious stand in this respect, in the sense that it is not allowing unlimited access to prime land by large investors for Jatropha cultivation (for some recent evidence, see Edwin, 2007). Still, a more pro-active stance appears to be needed. There is an urgent need for the government to confront all the potentially major consequences of large-scale biofuel cultivation on its soil, rigorously debate the desirability of biofuels as a major energy source, and define its role in relation to other renewables and to non-renewables. Building on this, it needs to design and implement policies that would strike a reasonable balance between the interests of different groups. In particular, adequate supervision of the system of cultivating Jatropha is needed, so that large mono-cultures and their associated risks are avoided and many small-scale parties get opportunities to participate in an emerging Jatropha sector. Addressing the interface between Jatropha cultivation and the agriculture sector, particularly potential impact on land use patterns, crop prices and rural livelihoods, should be another particularly important policy priority. Although strategic niche management has been primarily designed for, and applied to, developed-country settings, it proved to be a useful instrument for analysing an innovation system in the context of a low-income country like Tanzania as well. It is an especially useful instrument for studying and explaining the emergence of a new sector, because, unlike in the national innovation systems approach, SNM systematically takes account of the contextual factors (the ‘landscape’ and the ‘regime’) that give rise to it. The most important specificity arising from applying the instrument to a developing-country context relates to the functioning of institutions. In the setting of Tanzania, it proved to be essential to include structural “governance” factors such as lack of pro-activeness and poor policy implementation capacity in the landscape analysis. In SNM studies done in developed countries, in contrast, the government is essentially treated as an actor (rather than a ‘factor’) alongside other actors in the innovation network. Another difference we encountered was that obtaining data for a comprehensive analysis proved to be quite difficult, mostly for logistical and cultural reasons. For example, Tanzanian people do not like to admit openly that some network relations do not function well or that they are having problems dealing with certain organizations. Low levels of formal education of some respondents also limited their ability to provide full information about aspects like costs and benefits of Jatropha experiments. The most important general limitation of SNM from our perspective is that all the activities undertaken in relation to a new technology are
Prospects for Jatropha biofuels in Tanzania
363
essentially viewed as useful techno-societal experiments; the more learning and the more networking among these experiments, the better the ultimate outcome is predicted to be. However, from the point of view of contributing to an emergent transition, it is clear that some experiments are likely to be more important than others. That is why we introduced the concept of a production chain, which shows all the logical relationships between the different activities and hence facilitates a better understanding of the importance of each. This shortcoming of SNM has also been recognised by Elzen et al. (2004), who state that, ‘from the wide variety of alternatives developed at niche level, it is necessary to select and focus, and to assess which of them should be stimulated in what way’. Our production chain instrument could be a simple aid for that task. Since the SNM methodology is still relatively new, some aspects are still under development. In that sense, it is encouraging that our study bore out the need and/or usefulness for certain extensions that have also been discussed in other recent SNM studies, namely the need for multiregime analysis (Raven, 2006; 2007) and the usefulness of specific criteria with which one can systematically assess the quality of niche processes (van der Laak et al., 2007). We hit upon the need for a multi-regime analysis ourselves and structured it with the help of the simple production chain concept, uninformed by the most recent advances in the SNM field at the time. The niche quality criteria were fitted retroactively to our fieldwork data. In hindsight, our fieldwork could have been more rigorous if we could have benefited from these useful new developments and insights.
NOTES *
1. 2. 3. 4. 5. 6.
An earlier version of this paper was presented at the Globelics Conference: Innovation Systems for Competitiveness and Shared Prosperity in Developing Countries, Trivandrum, India, 4–7 October 2006. The authors wish to thank Lex Lemmens for constructive comments during the research process, the Centre for Development Studies in Trivandrum for providing research facilities, Evie Kerkhof for research assistance, and Amin Kassam for editing services. Important SNM publications are: Weber et al. (1999); Hoogma et al. (2002); Raven (2005, 2006, 2007); and van der Laak et al. (2007). ‘Sustainable development’ in this chapter follows the definition in the Brundtland Commission Report of 1987: development that meets the needs of the present without compromising the ability of future generations to meet their needs. For details about recent research about yields and its determinants, see Daey Ouwens et al. (2007) at http://www.fact-fuels.org/en. For example, as shown by research done by TIRDO in Dar es Salaam. The poverty line estimate relates to the latest available year, 2002. One litre of Jatropha oil is equivalent to about 0.945 litres of fossil diesel oil owing to a difference in their calorific value (see Agarwal and Agarwal, 2007).
364 7. 8. 9. 10. 11. 12.
Sectoral systems of innovation and production The lower estimate is from International Energy Administration (2007); the higher estimate is from Tanzania’s National Bureau of Statistics, see http://www.nbs.go.tz/ publications/index.htm. Prices in Dar es Salaam on 28 July 2005 at different petrol stations. Information obtained from Green Garden Women’s Group, KIDT. Information obtained from Monduli Women’s Group. Project plot size varied between 1 and 200 hectares. We assumed a life span of five years and a seed sale price of TZS80 per kilogram. Full data are given in van Eijck (2006). We assumed that all seeds can be bought for TZS80 per kilogram and all produced oil can be sold at TZS2000 per litre. The diesel needed to operate the Sayari expeller costs 1100 TZS per litre. The estimated project life span is five years.
REFERENCES Agarwal, D. and Agarwal, A.K. (2007), ‘Performance and emissions characteristics of Jatropha oil (preheated and blends) in a direct injection compression ignition engine’, Applied Thermal Engineering 27, pp. 2314–2323. AREED (African Rural Energy Enterprise Development) (2001), Tanzania Report 2001, United Nations Environment Programme, http://www.areed.org/. Arvidsen, A. and Nordström, M. (2006), Energy Sector Policy Overview Paper, Stockholm Environment Institute, Stockholm. Chachage, B. (2003), ‘Jatropha oil as a renewable fuel for road transport: policy implications for technology transfer in Tanzania’, M.Sc. thesis, International Institute for Industrial Environmental Economics, Lund University, Lund. Daey Ouwens, K., Francis, G., Franken, Y.J., Rijssenbeek, W., Riedacker, A., Foidl, N., Jongschaap, R. and Bindraban, P. (2007), Position Paper on Jatropha Curcas: State of the Art, Small and Large Scale Project Development, Fact Foundation, http://www.fact-fuels.org/media_en/Position_Paper_on_Jatropha_Curcas. Edwin, W. (2007), ‘UK firm invests $20m in biofuel farm’, East African, 7 August. Eijck, J.A.J. van (2006), ‘Transition towards Jatropha biofuels in Tanzania? An analysis with strategic niche management’, M.Sc. thesis, Eindhoven University of Technology, Eindhoven. Elzen, B., Geels, F.W. and Green, K. (2004), ‘Conclusion’, in Elzen, B., Geels, F.W. and Green, K. (eds), System Innovation and the Transition to Sustainability: Theory, Evidence and Policy, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA, USA, pp. 282–300. Energy Information Administration (2006), Short Term Energy Outlook, http:// www.eia.doe.gov/emeu/steo/pub/contents.html. Fone, N. (2006), ‘Run on rape: which tractor models are up for biodiesel?’, Farmers Weekly, 12 December. Gosh, A., Patolia, J.S., Chaudhary, D.R., Chikura, J., Rao, S.N., Kumar, D., Boricha, G.N. and Zala, A. (2007), ‘Response of Jatropha curcas under different spacing to Jatropha deoiled cake’, Jatropha Curcas FACT Expert Seminar, Wageningen University, Wageningen, 26–28 March. Gübitz, G.M., Mittelbach, M. and Trabi, M. (1999), ‘Exploitation of the tropical oil seed plant Jatropha curcas L.’, Bioresource Technology 67, pp. 73–82. Gunaratne, L. (2002), ‘Rural energy services: best practices’, Study prepared for USAID under the South Asia Regional Initiative for Energy. Heller, J. (1996), Physic Nut Jatropha Curcas L.: Promoting the Conservation
Prospects for Jatropha biofuels in Tanzania
365
and Use of Underutilized and Neglected Crops, Institute of Plant Genetics and Crop Plant Research, Gatersleben, http://www.bioversityinternational.org/ publications/Pdf/161.pdf. Henning, R. (2004), ‘The Jatropha system: integrated rural development by utilisation of Jatropha curcas L. as raw material and as renewable energy’, Paper presented at Studientag Möglichkeiten und Grenzen erneuerbarer Energien in Tansania: Erfahrungen in der Partnerschaftsarbeit, Hamburg, 24 April, http:// www.tanzania-network.de/?Themen/Energien. Hoogma, R., Kemp, R., Schot, J. and Truffer, B. (2002), Experimenting for Sustainable Transport: The Approach of Strategic Niche Management, Spon Press, London and New York. Household Energy Network (undated), Tanzania Country Report Synthesis, http:// www.hedon.info/goto.php/TanzaniaCountrySynthesis. International Energy Administration (2007), IEA Energy Statistics: Energy Balances for Tanzania, http://www.iea.org/Textbase/stats/balancetable. asp?COUNTRY_CODE=TZ. Jones, N. and Miller, J.H. (1993), Jatropha Curcas: A Multipurpose Species for Problematic Sites, World Bank, Washington, DC. Karve, A.D. (2005), ‘Compact biogas plant: compact, low-cost digester for biogas from waste starch’, Mimeo, http://www.bioenergylists.org/en/compactbiogas. Laak, W.W.M. van der, Raven, R.P.J.M. and Verbong, G.P.J. (2007), ‘Strategic niche management for biofuels: analysing past experiments for developing new biofuel policies’, Energy Policy 35, pp. 3213–3225. Madslien, J. (2006), ‘Biofuel raises global dilemmas’, http://news.bbc.co.uk/2/hi/ business/4603272.stm. MBendi: Information for Africa (2005), Tanzania – Oil and Gas: Petroleum and Petroleum Products Wholesalers, http://www.mbendi.co.za/indy/oilg/ogds/af/ta/ p0005.htm. National Bureau of Statistics (2000), Integrated Agricultural Survey Report 1998/98, Rural Tanzania Mainland, Government of Tanzania, Dar es Salaam, http://www.tzonline.org/pdf/integratedagriculturalsurvey.pdf. National Bureau of Statistics (2004), Tanzania National Profile Data in Brief, Government of Tanzania, Dar es Salaam. Openshaw, K. (2000), ‘A review of Jatropha curcas: an oil plant of unfulfilled promise’, Biomass and Bioenergy 19, pp. 1–15. Pramanik, K. (2002), ‘Properties and use of Jatropha curcas oil and diesel fuel blends in compression ignition engine’, Renewable Energy 28, pp. 239–48. Raven, R. (2005), ‘Strategic niche management for biomass: a comparative study on the experimental introduction of bioenergy technologies in the Netherlands and Denmark’, Ph.D. thesis, Technische Universiteit Eindhoven, Eindhoven. Raven, R. (2006), ‘Towards alternative trajectories? Reconfigurations in the Dutch electricity regime’, Energy Policy 35, pp. 581–595. Raven, R. (2007), ‘Co-evolution of waste and electricity regimes: multi-regime dynamics in the Netherlands’, Energy Policy 35, pp. 2197–2208. Renewable Energy Policy Network (2005), Renewables 2005 Global Status Report, Worldwatch Institute, Washington, DC, http://www.mvo.nl/biobrandstoffen/ download/renewables%202005%20global%20status%20report.pdf. Research Group IP (International Programs) (2002), An Industry and Market Study on Six Plant Products in South Africa: Jatropha or Physic Nut, Washington State University, Pullman.
366
Sectoral systems of innovation and production
Rwambali, F. (2000), ‘Dar tells oil firms: stick to import regulations’, East African, 1 May, http://nationaudio.com/News/EastAfrican/current/. Shuma, J. (2006), ‘Rural Energy Agency: opportunities and challenges in Tanzania’, Access 3, February, pp. 17–19, http://oregonstate.edu/international/outreach/ rlc/resources/Jatropha.pdf. Staubman, R. (1997), ‘Biogas production from Jatropha curcas press-cake’, Applied Biochemistry and Biotechnology – Part A, Enzyme Engineering and Biotechnology, pp. 63–65, 457–467. Uisso, J.P. (undated), ‘Government policy on rural energy development’, Mimeo, Ministry of Energy and Minerals, Government of Tanzania. UNDP (2006), Human Development Report, United Nations, New York. UNIDO (1983), ‘The vegetable oil industries development sector: a review of policies and approaches to appropriate vegetable oil industrialisation’, Paper prepared by the UNIDO secretariat for the 50th International Fair, Novi Sad. Visser, J. and Adriaans, T. (2007), ‘Anaerobic digestion of Jatropha curcas press cake’, Report produced for FACT, Ingenia Consultants and Engineers, Eindhoven. Weber, M., Hoogma, R., Lane, B. and Schot, J. (1999), Experimenting with Sustainable Transport Innovations: A Workbook for Strategic Niche Management, Twente University, Enschede. World Bank (2006), Agriculture and Achieving the Millennium Development Goals, Agriculture and Rural Development Department, World Bank, Washington, DC, http://siteresources.worldbank.org/INTARD/Resources/Ag_ MDGs_Complete.pdf. World Fact Book (2006), http://www.cia.gov/cia/publications/factbook/geos/ tz.html. World Resources Institute (2006), Earth Trends: The Environmental Information Portal, http://earthtrends.wri.org/text/energy-resources/country-profile-178.html.
Index 3G IP core network techniques 317 3G mobile telecommunications 301, 317 absorptive capacity 17, 29, 140, 141, 151, 247 ABTCP (Associação Brasileira Técnica de Celulosa e Papel: Brazilian Pulp and Paper Technical Association) 100, 101, 102, 103, 104, 109, 114, 116, 117, 118, 119, 120, 121 Academia Sinica 294 accessibility of knowledge 9 accountability 88 Acer 291, 306, 319, 326 Acha, B. 57, 61, 89, 232, 233, 234 Act for Encouraging the Machine Parts Industry (Korea, 2001) 276 active learning system 160, 181 Adriaans, T. 341 Aeromot 168 aeronautical sector, Brazilian 15–16 deregulation of (1978) 168, 191 distance from global innovation frontier 192 exports 157, 173–4 system of innovation 166–74 following loose characteristic of national system of innovation 193 general characteristics of Brazilian civil aircraft market 172–4 learning strategy of Embraer 170–72 post-privatization period 169, 191 pre-foundation of Embraer 166–7 seeking international market phase 168–9, 190–91 starting-up phase 167–8 technology capability of suppliers 16, 157–8
analytical framework for studying 164–5 characteristics of foreign suppliers located in Brazil 177–9 characteristics of local SME supplier firms 175–7 government policies for upgrading 193 linkages for each group of SME suppliers 184–92 results on technological capacity accumulation by local SME suppliers 179–90 selection of sample 174–5 see also Embraer Aerospace Technical Centre (CTA) 167, 168, 169, 188, 190 African Rural Energy Enterprise Development (AREED) 346 Agarwal, D. 363 Agarwal, K. 363 agricultural regime, Tanzanian 349–51, 352, 360 AIAB (Associação das Industrias Aerospaciais do Brasil) 173 air conditioning 46 Airbus 179 aircraft system integrators 175 airframe structural components 176, 177 airframe structural sections 176, 177, 178, 179 airframe structures and related systems 175, 176, 177, 178 Albu, M. 89 Alcatel 45 Alvial, A. 246 AMA (Applied Material Inc.) 273 American Air Force 166 Amesse, F. 60 Amsden, A. 208
367
368
Sectoral systems of innovation and production
AMT 600 aircraft 181 AMX aircraft 170, 178 analysts/programmers 146 Andersen, B. 6 ANOVA test 74, 76, 96 antitrust regulations 8 Antonelli, C. 63, 266 Antunes, Azevedo 110 application knowledge 137, 138 Applied Material Inc. (AMA) 273 appropriability of knowledge 9, 143 Aquanotica 249 Aracruz Celulose 108–9, 110, 111–12, 113, 114–15, 118 architectural innovation 17, 235, 244, 251 AREED (African Rural Energy Enterprise Development) 346 Argentina literacy rates in 136, 137 market for software products and services in 135, 136 participation in education in 136 Ariffin, N. 163 Arocena, R. 145, 152 Arrow, K.J. 63 Artech 143, 144 Arthur, B. 11 Arthur, M.B. 89 Arusha 347, 352 Arvidsen, A. 361 ASEAN (Association of Southeast Asian Nations) 208–9 asexual reproduction 111 Asian financial crisis (1997) 262 Asian-Pacific Regional Operations Center 301, 321 Associação Brasileira de Normas Técnicas (ABNT) 106, 120, 121 Associação das Industrias Aerospaciais do Brasil) (AIAB) 173 Association of Salmon and Trout Producers of Chile see Association of the Salmon Industry in Chile (SalmonChile), formerly Association of Salmon and Trout Producers of Chile Association of Southeast Asian Nations (ASEAN) 208–9
Association of the Salmon Industry in Chile (SalmonChile), formerly Association of Salmon and Trout Producers of Chile 18, 238, 241, 242, 244, 246, 247, 248, 258 collective capability and role of 248–51 Audretch, D. 151 AUO 275 Australian pulp and paper industry 104 automobile industry, Korean 276 automotive engineering courses 216 Automotive Experts Dispatching Programme (2003–05), Thailand 215 avionics components 176, 177 B-testers 143 B2B (business-to-business) e-commerce systems 302 Bacha, C.J.C. 100, 102, 103, 106 Bae 172 Bahia Sul 114, 115 Baker, A. 104 Banco Nacional de Desenvolvimento Econômico e Social (BNDES) 105–6, 109, 110, 114, 120, 121, 123 Bandeirante aircraft 167, 170, 173 Bangalore software cluster 133 banking information systems 309 Barras, R. 234 Bell, M. 57, 59, 60–61, 65, 89, 161, 163, 165, 201, 208, 235–6 Belo Oriente, Minas Gerais 109 BenQ 324 Bernardes, R. 157, 160, 164, 167, 168, 169, 172, 174, 194 biofuels 335, 343, 345, 347, 361 see also Jatropha biofuels in Tanzania biogas 341, 345, 348–9, 358, 359 biosafety 115, 116, 120 biotechnology firms located in Zonamérica 148 forestry-related 99, 111, 112, 113, 116, 123 impact on high- and low-tech sectors 232 salmon-farming related 244 biotechnology sector, Taiwan 303, 304
Index Birkinshaw, J. 58 Bitzer, J. 134 Bjorndal, T. 238 BNDES (Banco Nacional de Desenvolvimento Econômico e Social) 105–6, 109, 110, 114, 120, 121, 123 BNDESPAR 114, 120 Board for Industrial and Financial Reconstruction (BFIR) 40 Bolivia literacy rates in 137 participation in education in 136 Bombardier–Canadair (Bombardier Aerospace) 172 bonded factories 326 Boonchukosol, K. 216 Borgatti, S.P. 72, 84 Borges, I. de C. 116 Borrás, S. 99 Botucatú 168 bounded rationality 5 Boyd, W. 101 BRACELPA (Brazilian Association of Pulp and Paper) 117, 118, 121, 130 Brasil 71 Brasilia EMB 120° aircraft 168, 169, 170, 173, 177 Brazil aeronautical sector in see aeronautical sector, Brazilian; Embraer demand for software products in 136 federal system of government in 123 ICT sector in see ICT sector, Brazilian industrialization policies of 104, 156–7, 160, 172, 173 liberalization policy adopted by 160, 173 literacy rates in 137 market for software products and services in 135, 136 military dictatorship in 106, 108, 156 offset programmes of 190 national innovation system of 160, 193 participation in education in 136 pulp and paper industry in see pulp and paper industry, Brazilian
369
selected economic sectors in relation to their distance from the global innovation frontier 192–3 Brazilian Air Force 166–7, 169, 177, 189 Brazilian Association of Pulp and Paper (BRACELPA) 117, 118, 121 Brazilian Embassy 191 Brazilian Eucalyptus Genome Network 116 Brazilian Genome Project 116, 120 Brazilian ICT law 58, 67–74, 95 Brazilian Institute for Geography and Statistics (IBGE) 174 Brazilian Ministry of Aeronautics 167, 173 Brazilian Ministry of Agriculture, Livestock and Supply (MAPA) 118, 121, 123 Brazilian Ministry of Defence 168, 169, 170, 178 Brazilian Ministry of Development, Industry and Foreign Trade (MDIC) 118, 121, 123 Brazilian Ministry of Environment (MMA) 118, 121, 123 Brazilian Ministry of Science and Technology (MCT) 67, 116, 118, 120, 121, 122, 123 Brazilian Pulp and Paper Technical Association (ABTCP) 100, 101, 102, 103, 104, 109, 114, 116, 117, 118, 119, 120, 121 Breschi, S. 65, 159 BRIC group 193, 277 bridging organizations 354 broadband development in Taiwan 305, 311, 317, 318 broadband wireless products 45, 305, 317 Brown, J.S. 66 Brundtland Commission Report (1987) 363 Brusoni, S. 63, 175 BSNL 45, 52 ‘built to order’ strategy 306 C-DOT 42–52 index of innovation capability 47–50 spillover effects of 50–52
370
Sectoral systems of innovation and production
C100 motorcycles 211, 218 Callon, M. 6, 264 Campinas 83 Campinhos, Edgard 112 Campos, W. de O. 112, 113 Canadair 172 Cantwell, J. 236 capital goods industry, Korean 18, 259–86 exports 262–3, 277, 278 ICT use in 265–6, 277 importance of the capital goods industry and the Korean experience 260–64 imports from Japan 259, 261–4, 268, 276 opportunities for catch-up 275–7 sectoral system of innovation and the machine tools industry 264–7 demand conditions or market regimes 266–7 regimes of knowledge and technologies 265–6 roles of actors 267 summary and conclusions 277–9 three barriers to catch-up in 267–75 dumping pricing by incumbent firms 270–72, 282–6 filing IPR lawsuits against catching-up firms 273–5 weak demand and weak R&D 267–70, 271 capital goods industry, Thailand 213 Carlsson, B. 6, 159, 264 cash crops 349 Cassiolato, J.R. 83, 159, 160, 174 Catholic University, Uruguay 146 CBA 123 aircraft 170 CD-Rs 272, 285, 330, 331, 332 Celma 179 Celmar 115 Celulose Nipo-Brasileira (Cenibra) 109, 110, 111, 113, 115, 118, 122 Central Drug Research Institute (CDRI), India 33, 41, 42 Centre for Development of Telematics see C-DOT
Centro Nacional de Pesquisa Florestal (CNPF) 112, 120, 121 CESAER 2001 174 Cessna 172 Chachage, B. 339, 340, 341 Chaebols 261 Chagas disease 35 Challenge 2008: the National Development Plan 288, 301, 305, 316 Champion Paper Company 105, 106, 115 Chandler, A.D. 64 Chaudhuri, S. 30, 35, 38, 39, 40, 41 chemical pulp processes 99, 101–5, 107, 108, 117 chemistry (plastics) 242 Chen, Dung-Shen 214 Chen, Wen-Tang 321 Chen, Xin-Hong 320 Chernobyl disaster 337 Chiang Tung Bank 291 Chile literacy rates in 136, 137 participation in education in 136 salmon farming in see salmon farming China competition to Thai and Vietnamese motorcycle industries from 17, 209, 218–19 extent of impact of 219–22 reasons for differences in responses to 226–8 transformation of sectoral innovation systems as a result of 222–6 FDI in Taiwan 291 foreign exchange control in 325 Korean machine tool exports to 262–4, 277, 278 Taiwanese investment in 306, 314–15, 323–33 telecommunications industry in 42–3 threats and opportunities to ASEAN countries from 208–9 trade balance between Taiwan and 323, 329–30 transformation of innovation system in 55
Index China Automotive Industry League 209 China Automotive Technology and Research Center 209 Chinese computer 308 Chinese ICT standards 308, 321 Chinfon Group 229 Chulalongkorn University 216 Chunghwa Telecom Co. 301, 309 Cia Suzano de Papel 105, 114, 115, 116 Cia Vale do Rio Doce 114 circuit switching 42 Clapp, J. 237 Clark, K. 234, 235, 236 Cleaner Production Agreement (Acuerdo de Produccion Limpia: APL) 247, 248, 249, 258 clinical trials 36 cloning 111–13, 114, 116 clustering of innovative activity Bangalore cluster 133 Brazilian ICT sector 83 Chilean salmon farming industry 17–18, 242, 251 Silicon Valley cluster 133, 307 Taiwanese IT manufacturers in mainland China 327–8 Uruguayan software sector 15, 131–53 CMM certification 85 CMMI (Capability Maturity Model Integration) project 149 CNPF (Centro Nacional de Pesquisa Florestal) 112, 120, 121 Coase, R.H. 60, 62 ‘code of good practice’ (codigo de buenas practicas) 246, 258 codified knowledge 59 Código Florestal (Law No. 4771, 15 September 1965) 106–7 coevolution 5, 11, 21 of knowledge base and organization network in Brazilian ICT sector 57–98 of technology and societal factors 20, 336 Cohen, W. 10, 140 collaborative agreements in the Brazilian aeronautical sector 189, 190, 191
371
collaborative-design manufacturers (CDMs) 302, 306 collaborative factor 248 collective capability 232, 235, 237, 250 importance for low-tech sectors 252–3 and role of the Association of the Salmon Industry in Chile 248–9 influence on external standardsetting 249–51 collective efficiency 250 Colombia literacy rates in 137 market for software products and services in 135 participation in education in 136 Comércio, J. 105 Commander, S. 134 Commonwealth Scientific and Industrial Research Organization 107 communality concept 169 communities of practice 14, 65–6, 83–5, 87 Companhia Melhoramento 102 Companhia Paulista de Estradas de Ferro 101 Companhia Vale do Rio Doce 109 Compaq 306 comparative advantages 57, 63, 64–5, 81, 83, 193, 266, 287 complementarities 10–11, 132 completely build units (CBUs) 211, 216, 224 completely knock-down (CKD) units 211 conferences and exhibitions, knowledge flows from 143, 144, 145, 150 conflict of interests 65–6 consultants, knowledge flows from 144 Contingency Centre of National Information and Communication 317 contracts 8 Cooke, P. 6, 158, 159 core competence 235, 236 Costa, I. 160 Council of Scientific and Industrial Research (CSIR) 41, 42, 50 Coutinho, L. 167, 168
372
Sectoral systems of innovation and production
CTA (Aerospace Technical Centre) 167, 168, 169, 188, 190 cultural traditions 354, 360 Cummings, J.L. 66 cumulativeness of knowledge 9–10, 143 Cusmano, L. 57, 61, 89 custom-made motorcycle parts 211 customers, knowledge flows from 143, 144, 145, 184, 188–92, 260, 265, 266, 279 customs duties 45, 50, 55, 148 Customs Tariff Law (Japan, 2003) 275 Cusumano, M. 134 CUTI (Uruguayan Business Association of Information Technologies) 131, 133, 148–9, 150 Cutler, C. 237, 245 CVD (chemical vapour deposition) machines 273 Daewoo Electronics 275 Daey Ouwens, K. 339, 363 Dagnino, R. 157, 166, 167, 168, 172 Dahlman, C.J. 100, 138, 160, 192 Dahmen, E. 10 Damiani, J.H.S. 174 Dar es Salaam 347, 352, 363 see also University of Dar es Salaam David, P. 11 Davies, A. 57 DBTEL 326 De Ferranti, D.M. 159, 232 De Havilland 172 DeBresson, C. 60 decentralized electricity provision 348, 352 De Fillipi, R.J. 89 deforestation 100 delayed payment 269, 270 Dell 306 demand 5, 7, 10, 11, 132 for Chilean farmed salmon 238–9 for Korean machine tools 266–70, 277 for software products and services in Latin America 135–8, 152 for Taiwan’s ICT products 309–10, 315–18, 319 for Thai and Vietnamese motorcycles 218–19
Demonstrated IT Application Research Program, Taiwan 316 dengue fever 35 Department of Industrial Technology, Ministry of Economic Affairs (DoIT/MoEA), Taiwan 288, 299–300, 302 Department of Science and Technology, India 36 Department of Telecommunications, India 44, 45–6, 55 developed networks 79–80, 86 developing networks 79, 86 development blocks 10 Development Fund of Executive Yuan 291 DGBAS (Directorate General of Budget, Accounting and Statistics), Taiwan 290, 291 diesel engines, use of Jatropha oil in 341, 346–7, 358 digital archives 313, 314 digital content sector, Taiwan 303, 304 Digital Taiwan 309 Directorate General of Budget, Accounting and Statistics (DGBAS), Taiwan 290, 291 Directorate General of Telecommunications (DGT), Taiwan 301 DiStar 213 distributed innovation systems 6 diversification of products 241, 242, 251 division of labour 260, 320, 325–6, 328–9 Docpharma 41 Doctorate degrees 294 dominant design 11 Donângelo, A. 172 Dornier 172 Dosi, G. 6, 8, 9, 58, 61 dot.com bubble 324 Doughty, R.W. 101, 102, 104, 107, 111, 112, 113 Draft National Pharmaceuticals Policy (2006,2), India 31–2 Drug Price Control Order (DPCO), India 33, 35
Index Drugs and Cosmetics Act (India, 1940) 36 Drugs and Cosmetics Rules (DCR), India 36 DSL service 309 Duguid, P. 66 Dumont, Santos 166 dumping pricing 270–72, 282–6 Dunning, J.H. 65 Duran, J. 232 Duratex 116 Dutta, S. 310 Dyer, J.H. 66 dynamic capabilities 61, 161, 163 dynamic complementarities 10–11, 132 e-business 288, 302–3 E-Business Standard Research Plan 288 e-learning 309, 313–14 e-Taiwan program 288, 304–5, 309, 317, 319 E85 fuel 347 E&E of Taiwan 275 Eagle, J. 237 economic geography literature 151 economies of scale 64, 108, 110, 111, 138, 169, 241, 323 economies of scope 64, 138 Edquist, C. 6, 264 education participation in Latin American countries 136 in Taiwan 293–4, 318 educational establishments role in knowledge networks in Brazilian ICT sector 14, 71, 72, 73, 75, 82, 83, 86 see also university–industry linkages Edwin, W. 361 EI database 295 Eindhoven University of Technology 335, 347 Eleb 175 electricity provision 348, 352 Elzen, B. 362 EMB 110 Bandeirante aircraft 167, 170, 173 EMB 120 Brasilia aircraft 168, 169, 170, 173, 177
373
EMB 121 Xingú aircraft 169, 173 Embraer Brazilian aeronautic sector before founding of 166–7 changes implemented (1996–2005) 174 economic performance of 173–4 founded (1969) 16, 157, 166, 167 learning strategy of 170–72 main civil aircraft models manufactured by 172–3 main competitors of 172 monopoly granted to 172 post-privatization period 169, 191 privatization (1994) 16, 157, 169, 173, 191 seeking international market phase 168–9, 190–91 starting-up phase 167–8 technological capability of suppliers 16, 157–8 analytical framework for studying 164–5 characteristics of foreign suppliers located in Brazil 177–9 characteristics of local SME supplier firms 175–7 government policies for upgrading 193 linkages for each group of SME suppliers 184–92 results on technological capacity accumulation by local SME suppliers 179–90 selection of sample 174–5 EMBRAPA (Empresa Brasileira de Pesquisa Agropecuária) 112, 119, 120, 121, 122, 123 embroidery machines 273 Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA) 112, 119, 120, 121, 122, 123 EMR (Exclusive Marketing Rights) 34 enabling networks 78–9, 86 ENAER 171 Energy Information Administration, Tanzania 343 energy regime, Tanzanian 345–9, 352, 360 Engaruka 347, 351, 358, 359
374
Sectoral systems of innovation and production
engineering projects and consulting 176, 177 Enos, D. 161 enterprise resource planning (ERP) 136–7, 177, 182 Environment Component Index 310 environmental and social responsibility 109, 110, 113, 116, 124 environmental standards 245, 247, 248, 249, 258 ERJ 135 aircraft 173 ERJ 140 aircraft 173 ERJ 145 aircraft 169, 170, 171, 173, 177, 178 ERJ 170 aircraft 173, 177, 178 ERJ 190 aircraft 169, 173 ERJ 195 aircraft 173 Ernst, D. 64, 228, 236 ERP (enterprise resource planning) 136–7, 177, 182 Espírito Santo 110 EU Community Innovation Survey 133 eucalyptus increasing investment in R&D of eucalyptus forestry and pulp process 104–7 introduction into Brazil 99, 100–103 new techniques in production of 111–16 responsibility for research into 119, 120, 122–3 take-off in production of eucalyptus pulp 108–10 Eucalyptus World Congress 107, 119, 120 European Union 193, 360 evolutionary theory 5–6, 7, 9, 233 innovation projects central in 57–8 strategic niche management (SNM) rooted in 336 and technological capability accumulation 161 Exclusive Marketing Rights (EMR) 34 Executive Yuan, Taiwan 291, 299–300, 301, 302, 304, 305, 308, 309, 317 exhibitions and conferences, knowledge flows from 143, 144, 145, 150
expectations of actors, dynamics of 338 Jatropha biofuel production in Tanzania 355–6, 357, 358, 359, 360, 361 experts, hiring of 184, 189, 215 export processing zones 326 exports Brazilian aeronautical sector 157, 173–4 Brazilian pulp and paper industry 14, 108, 110 Chilean farmed salmon 237, 238, 251 Chinese motorcycle industry 17, 209, 219–22, 224–5 farmed salmon by major countries 238 Japanese machine tools to Korea 259, 261–4, 268, 276 Korean machine tool industry 262–3, 277, 278 Taiwanese high-tech industry 315–16 Taiwanese IT manufacturers in mainland China 329–30 Uruguayan software sector 131, 148–9 Failache, C. 138, 147 Fairchild 172 Fairchild–Dornier 172 ‘family’ concept 169 Fan, Z.-K. 317 FAO (Food and Agriculture Organization) 107, 118, 119, 121 FAPESP (State of São Paulo Research Foundations) 116 Faust, K. 60 Federal Rural University of Rio de Janeiro (UFFRJ) 117, 119, 121 Federal University of Parana (UFPR) 119, 121 Federal University of Rio Grande do Sul (UFRGS) 119, 121, 122 Federal University of Viçosa (UFV) 106, 117, 119, 121, 122, 123 Federation of Korean Industries (FKI) 268, 269, 271 Feldman, M.P. 151 Ferraz, J.C. 167, 168
Index Ferreira, M. 111, 112, 113 fertilizers 341, 351, 358–9 Fforde, A. 218 Figueiredo, P.N. 161, 162, 163 filière 10 financial regime, Tanzanian 351–2 FINCA 351 FIND 287, 302, 305 FINEP 174 flat-panel display industry, Taiwan 303, 304, 312 flexible specialization 235 flight controls 178, 179 Fokker 172 Fombrun, C.J. 65 Fone, N. 347 Fonseca, M. da G. de 115 Food and Agriculture Organization (FAO) 107, 118, 119, 121 Fordist mass production 235 foreign exchange control 321, 325 foreign technology agreements 33 Forest Eucalyptus Genome Sequencing Project Consortium 116, 120 forestry research institutes 106 forestry schools 106, 117, 119 Foss, N.J. 58 Fransman, M. 161, 164 Freeman, C. 6, 10, 158, 159, 264 Freitas, I. 245, 254 Frischtak, C. 100, 160, 164, 167, 168, 169, 170, 192 Fundacion Chile 246, 249, 258 Fundamental Communications Basic Act (Taiwan, 2004) 301–2 fuselage parts and components 176, 177 Gamesa 171 Gann, D.M. 64 Garcia, C.H. 100, 101, 102, 103, 106 Gargiulo, M. 63 GEAE 171, 178, 179 general purpose machine tools sector, Korean 277 General Statistics Office, Vietnam 209, 218 genetics 105, 112, 113 Genexus 143 Genolyptus Project 116, 120, 122
375
genomic eucalyptus research 115–16, 122, 123 geographical proximity 145, 151, 268 Germany, Korean machine tool exports to 278 Geuna, A. 63 Giddens, A. 61 Giuliani, E. 57, 61 Glenmark 41 global production networks 228 global sourcing strategies 212, 217 global warming 343 Globelics India Conference (Trivandrum, 2006) 4 GLP certification 36, 37–8 GM Daewoo 272 gmelina 110 GMI (Graduate School of Management and Innovation) 212, 213, 216 GMN Automobile and Motorcycle Parts Manufacture JV Co., Ltd. 214 Goldstein, A. 108 Good Laboratory Practice (GLP) Compliance Monitoring Authority 36 Gordiho Braune & Cia 102 Gosh, A. 340 governance in the global market 235–7 government e-services, Taiwan 287, 316 government research institutes see public research institutes, interaction with GPRS 317 Graduate School of Management and Innovation (GMI) 212, 213, 216 Granovetter, M.S. 61 Granstrand, O. 69 Grattapaglia, D. 116, 122–3 Green, R.D. 169, 172 Gübitz, G.M. 339, 340 Guess, G.M. 109 Gulati, R. 63 Gunaratne, L. 348 HACCP (Hazard Analysis and Critical Control Point) 246 HACCP-CC 246, 258 HACCP-PP 246, 258
376
Sectoral systems of innovation and production
Hamel, G. 235 Hanoi University of Technology 217 Hansen, M.T. 66 hardware product development knowledge network 14, 59, 69, 70, 73, 76, 77, 78, 79, 81, 82, 83, 84, 86, 95, 97, 98 Hart, S. 109 Heeks, R. 135 Heller, J. 339, 341 Henderson, R. 234, 235, 236 Henning, R. 341 HG products 241 Hilgemberg, E.M. 100 Hindustan Antibiotics Limited (HAL) 39–40 Hirsch-Kreinsen, H. 232, 233, 234, 235, 236 HIV/AIDS 35 Ho Chi Minh City University of Technology 217 Hobday, M. 57, 64, 89, 229, 234 Hodgson, G.M. 158 Hoge, W. 110 Honda 211, 212, 213, 214, 217, 218, 221, 222, 227 Honda Supercub motorcycle 211, 229 Honeywell 171 Hong Kong 325, 326 Hoogma, R. 337, 363 horizontal collaborations, knowledge flows from 143–4, 151 Household Energy Network 348 Hsinchu Science Park 290, 307 Huang, C.Y. 306 Huang, Gwo-Jiunn 308 Huawei 44 Hughes, T.P. 6, 264 Hunter, H. 100 Hwanak 272 Hwang, Chul-Ju 273 hydraulic systems 176, 178, 179 hydro power 336 Hyundai Heavy Industry 272, 286 Hyundai Motors Kia Motors 272 Iammarino, S. 236 IBGE (Brazilian Institute for Geography and Statistics) 174 IBM 306
ICT emergence of 232 evolution in Taiwan see under Taiwan use in Korean capital goods industry 265–6, 277 see also ICT sector, Brazilian; software sector, Montevideo, Uruguay ICT law, Brazil 58, 67–74 passim, 95 ICT sector, Brazilian knowledge networks in 14, 57–98 analysis and implications 85–9 boundaries between firms and technological partners 14, 59, 62–4, 74, 76–80, 86, 95–6 definition and use of knowledge network 60–62 formation of channels for knowledge flows 14, 59, 65–6, 75–6, 83–5, 86, 87, 98 general characteristics of database 66–74 specialization in different governance mechanisms 14, 59, 64–5, 74–5, 81–3, 86–7, 97 liberalization of 58, 83 total investments in R&D in telecommunications and computers sectors 71 total outsourcing of R&D in telecommunications and computers sectors 71–4 IDE (Institute of Developing Economies) 210 Ideasoft 143 IFI (Institute for Development and Coordination of the Aerospace Industry) 168, 169, 174 III (Institute for Information Industry) 19, 288, 292, 295, 296, 308–9, 315, 321, 323, 331, 332 Iizuka, M. 245, 254 import-source diversification policy 276 import substitution in Brazil 58, 104, 156, 160, 172, 173 in Korea 275, 276 in Thailand 211
Index in transitional ASEAN economies 209 in Vietnam 211, 213 import tariffs, Vietnamese motorcycle industry 221, 224 imports Brazilian aeronautical sector dependent on 16, 157, 174 controls in Indian telecommunications industry 45 controls in Korean capital goods industry 275, 276 Indian telecommunications industry dependent on 13, 27, 29, 44 machine tools into Korea from Japan 259, 261–4, 268, 276 motorcycles and parts into Thailand and Vietnam 17, 219–22, 224–5 controls by Thai government 221 controls by Vietnamese government 213, 215, 216, 221, 224, 226 oil imports into Tanzania 343–4, 348 pulp imports into Brazil 107–8 by Taiwanese IT manufacturers in mainland China 329–30 Impuesto a la Renta de Industria y Comercio (IRIC) 147 IMS Health–ACNielsen (formerly ORG) 36 incubation programmes in Uruguayan software sector 141, 147 India market for software products and services in 135 pharmaceutical industry in see pharmaceutical industry, Indian telecommunications industry in see telecommunications industry, Indian Indian Drugs and Pharmaceuticals Limited (IDPL) 39–40 Indian Patents Act (1970) 34, 54 Indian Pharmaceutical Policy (1994) 31 individual (tailor-made) software 134, 137 Industrial Automation and Electronic Business: iAeB Program, Taiwan 302 industrial district literature 151
377
Industrial Machinery and Instruments (IMI) Holding 217–18 industrial robots 270–72, 286 Industrial, Technology and Trade Policies (Brazil, 2004) 156 Industrial Technology Research Institute (ITRI) 19, 292, 295–6, 331, 332 industry associations, Brazilian pulp and paper industry 119, 120, 121, 122 industry–university linkages see university–industry linkages informatics 137 Information Month 309 Ingenio incubation programme 141, 147 innovation intensity 232, 235, 236 innovation projects, knowledge networks formed by, in Brazilian ICT sector 14, 57–98 analysis and implications 85–9 boundaries between firms and technological partners 14, 59, 62–4, 74, 76–80, 86, 95–6 definition and use of knowledge network 60–62 formation of channels for knowledge flows 14, 59, 65–6, 75–6, 83–5, 86, 87, 98 general characteristics of database 66–74 specialization in different governance mechanisms 14, 59, 65–6, 74–5, 81–3, 86–7, 97 innovation system literature 6, 264 innovative capability 162, 163, 165, 202 in Brazilian aeronautical sector 16, 157–8, 179–81, 183–8, 190, 191 in Uruguayan software sector 138–40, 141, 151 Institute for Development and Coordination of the Aerospace Industry (IFI) 168, 169, 174 Institute for Information Industry (III) 19, 288, 292, 295, 296, 308–9, 315, 321, 323, 331, 332 Institute of Developing Economies (IDE) 210
378
Sectoral systems of innovation and production
institutions 5, 7–8, 29, 132, 233 new institutions in Brazil 105–7 strategic niche management (SNM) and 362 supporting Chilean salmon farming industry 245–50 supporting Thai and Vietnamese motorcycle industries 219, 225–6 supporting Uruguayan software sector 133, 144, 145, 147–50, 152 Instituto de Pesquisas e Estudos Florestais (IPEF) 106, 107, 111, 112, 115, 117, 120, 121, 122, 123 Instituto Universitario Autónomo del Sur 146 Intarakumnerd, P. 212, 215 Integrated Beyond 3rd Generation (iB3G) Double Network Integration Plan 305 Integration Heterogeneous Network 317 Integro 148, 149 Intel 320 Intellectual Property Institute 215, 223 interaction processes 5, 11, 132 in Uruguayan software sector 132–3, 140–46 interactive learning spaces 152 Inter-American Development Bank 135, 147, 148 intermediate materials or parts industry, anti-dumping appeals by firms in 272 International Energy Administration 346, 363 International Finance Corporation (IFC) 114, 120, 121 International Paper Company 115 international partnerships in the aeronautical industry 170 international seed bank 107 internet penetration rate, Taiwan 318 investment capability 162, 163 IPEF (Instituto de Pesquisas e Estudos Florestais) 106, 107, 111, 112, 115, 117, 120, 121, 122, 123 IPR lawsuits against catching-up firms 273–5
ISDN facilities 46, 50 ISO 9000 standards 182–3, 245, 247, 248, 258 ISO 14000 standards 245, 247, 248, 258 ISO 14001 standard 247 Israel, market for software products and services in 135 ITA (Technological Institute of Aeronautics) 167 ITI Ltd 44, 46, 52 ITRI (Industrial Technology Research Institute) 19, 292, 295–6, 331, 332 Jacobson, S. 159 Jaffe, A.B. 151, 153 Japan IPR protection by 274–5 machine tool exports to Korea 259, 261–4, 268, 276 machine tool imports from Korea 277, 278 Japan External Trade Organization (JETRO) 228, 229 Jari project 110, 115 Jatropha biofuels in Tanzania 20, 335–63 conclusions and policy issues 360–63 internal rates of return for 355, 357 Jatropha production chain 339–42, 361, 362, 363 landscape assessment 342–5 recent developments at the niche level 352–9 dynamic niche formation at the cultivation stage 353–6 fieldwork methodology 352–3 fragmented niche dynamics at the oil pressing stage 356–7 rudimentary niche formation at the end-use stage 358–9 regime assessment 345–52 agricultural regime 349–51, 352, 360 energy regime 345–9, 352, 360 financial regime 351–2 vegetable oil regime 351 relative price of 347, 348, 349, 353, 358, 359, 360 strategic niche management approach described 337–9
Index uses biogas 341, 345, 348–9, 358, 359 briquettes as oven fuel 341, 358 decentralized electricity provision 348, 352 in diesel engines 341, 346–7, 358 fertilizers 341, 351, 358–9 oil in cooking stoves 341, 359 oil lamps 341, 348, 359 soap-making 341–2, 359 Jessop, B. 236 jigs and tools 176, 177 Johnson, B. 236 joint ventures Brazilian aeronautical sector 175 Brazilian pulp and paper industry 114–15 Korean capital goods industry 276 Taiwanese ICT sector 307 Thai and Vietnamese motorcycle industries 209, 212, 213, 214, 224, 228 Jones, N. 339 Joosung Engineering 273–4 Jou, Sue-Ching 214 Jubilant Organosys 41 Jundiaí, São Paulo 101 just-in-time delivery 174 Jusung Engineering Co. 269 Juvenal, T.L. 100, 105, 106, 108, 109, 110, 113, 115 Kao, K.S. 323 Karve, A.D. 341 Kasetsart University 223 Katikarn, C. 223 Katz, J.M. 138, 160, 161, 251 Katz, R. 66 Kawasaki 212, 213, 272 Ke, J.S. 309, 318, 319 kerosene 348, 359 Kesidou, E. 139, 142, 147 Kilimanjaro 352 Kim, L. 64, 162–3, 228, 236 Kim, S.-R. 89 King Mongkut’s University of Technology Thonburi (KMUTT) 209 Kiryung Electronics 275 Klabin 115
379
Klein, P.G. 58 Klepper, S. 11 Knight, L. 66 Knight, P. 114, 192 knowledge, dimensions of 9–10 knowledge base 5, 8–11, 21, 29, 132, 233 in low-tech sectors 232, 252 type of, and knowledge networks in the Brazilian ICT sector see knowledge networks in the Brazilian ICT sector of Uruguayan software sector 134–5, 151 knowledge-based economy 9, 192 Knowledge Economics Development Act (Taiwan) 299 knowledge networks in the Brazilian ICT sector 14, 57–98 analysis and implications 85–9 boundaries between firms and technological partners 14, 59, 62–4, 74, 76–80, 86, 95–6 definition and use of knowledge network 60–62 formation of channels for knowledge flows 14, 59, 65–6, 75–6, 83–5, 86, 87, 98 general characteristics of database 66–74 specialization in different governance mechanisms 14, 59, 64–5, 74–5, 81–3, 86–7, 97 knowledge spillovers 10, 65, 88 definition of 153 in Uruguayan software sector 140–46, 149, 151 knowledge transactions in the Uruguayan software sector 140–45, 151 Korea automobile industry in 276 capital goods industry in see capital goods industry, Korean industrialization policies in 156–7, 275–6 main export products of 262 national innovation system of 160 technological capability accumulation in 162–3
380
Sectoral systems of innovation and production
Korea International Trade Association 262, 263 Korean Federation of Small and Medium Businesses (KFSB) 269, 270 Korean Ministry of Commerce, Industry and Energy 278 Korean Trade Commission (KTC) 272, 286 Kosnik, J. 211 KPMG 41 kraft pulp process 104 Kubitschek, Juscelino 104 Küppers, G. 60 Kuwayama, A. 232 L1 price 46 Laboratorio Tecnológico del Uruguay (LATU) 147 laboratory and equipment infrastructure knowledge network 59, 69, 70, 72, 77, 78, 81, 82, 83, 84, 86, 87, 95, 97, 98 labour mobility in Taiwanese ICT sector 307 in Uruguayan software sector 15, 140, 142, 147, 151 labour productivity in Taiwan 290 Lall, S. 139, 162, 163, 164, 228, 235 Lam, A. 89 Lamberg, J.-A. 101 land ownership 354 landing gear 175, 176, 177, 181 landscape 337, 338 influences on development of Jatropha biofuels in Tanzania 342–5 Lane, P.J. 66 Lastres, H. 159, 160, 174 Latecoere 171 Latin American Development Bank 118, 120 LATU (Laboratorio Tecnológico del Uruguay) 147 Law of Biosafety (Brazil, 2004) 116, 120 Law of e-Signature (Taiwan, 2002) 309 Law of Similars (Brazil, 1967) 172 LCD (liquid crystal displays) machine market 273–4
Leão, R.M. 100, 104, 106, 107, 111, 112 learning at niche level 338 Jatropha biofuel production in Tanzania 354–5, 356–7, 358, 359, 360, 361 Lee, K. 260, 264, 266, 267, 268 Lee, M.Y. 293 leishmaniasis 35 Leon Feffer & Co. 103 leprosy 35 Levin, R.C. 8 Levinthal, D. 10, 140 LG Electronics of Korea 275 licensing 260, 265 Liebherr Aerospace 175 Lifan Motorcycle Manufacturing JV Co. 214 Likert scale 140, 141, 145 Lilja, K. 101 Lim, C. 260, 264, 266, 267, 268, 277 Lin, H.-Y. 290 Lin, X.-W. 290 linkages capability 162 links and complementarities 10–11, 132 Lissoni, F. 65 literacy rates 136, 137, 318 lithium batteries 272, 282, 283, 286 Liu, Meng-chun 320 local content ratio 214, 221, 224 local knowledge, role of 236 Lopez-Claros, A. 310 Lorentzen, Erling 108–9 low-tech sectors changes at macro level 232 development strategy for developing countries 234–5, 236, 251–3 innovation in 17–18, 233–4, 236, 237, 250, 253 see also salmon farming Lubatkin, M. 66 Lucent 52 Ludwig, Daniel 110 Lundvall, B.A. 6, 99, 159, 160, 236, 264 M-Taiwan Program 288, 305, 317, 319 machine tools industry, Korean see capital goods industry, Korean
Index machinery, basic 244 Madslien, J. 347 malaria 35 Malerba, F. 4, 5, 8, 9, 28, 29, 57, 60, 61, 66, 131, 132, 143, 159, 160, 207, 233, 252, 259, 264, 279 Manaus Free Trade Zone 67 Mani, S. 30, 42, 44, 47, 50, 54, 55, 83, 264 Manning, S. 89 Mansell, R. 134 manufacturing and design outsourcing 175 Marin, A. 65 market hyper-segmentation 234 market niche, creation of 234–5, 236, 253, 338–9, 360 market segmentation 324 Marques, R. 160, 164, 166, 192 Marshall, A. 59, 145 Martin, R. 151 Martini, A.J. 101, 102 Mashelkar, R.A. 33 Mashelkar Committee 33, 36–8 Maskell, P. 233 Massachusetts Institute of Technology (MIT) 167 master’s degrees 294 Matrix Labs 41 Mattos, R.L.G. 100, 105, 106, 108, 109, 110, 113, 114, 115 MBendi: Information for Africa 344 McKendrick, D.G. 99 mechatronics 277 Media Tek 320 Mejía, T. 137–8, 146 Mello, Helládio do Amaral 106 Mandonça, S. 232, 233, 234, 254 mergers and acquisitions aeronautical industry 172 Brazilian pulp and paper industry 114, 115 Chilean salmon farming industry 239, 241 Indian pharmaceutical industry 40–41 Taiwanese IT industry 324 metallurgy 242 Metcalfe, S. 6, 11 Metso 117
381
Mexico market for software products and services in 135 software clusters in 135 Meyer, A.D. 62 Meyer, M. 58, 89 micro-credit 351–2, 361 Microsoft 320 middleware (system) product development knowledge network 14, 59, 69, 70, 73, 76, 77, 78, 79–80, 81, 82, 84, 85, 86, 87, 95, 97, 98 Millennium Development Goals 343 Miller, D. 64 Miller, J.H. 339 Minas Gerais 106, 109, 118, 121 mobile communications industry India 42, 44, 45, 46, 51–2 Taiwan 301, 305, 315, 319, 321 operations in mainland China 324, 325, 327, 328 Mogi-Guaçu, São Paulo 105 Montero, C. 243 Montevideo software cluster see software sector, Montevideo, Uruguay Mora, A.L. 100, 101, 102, 103, 106 Morogoro 341, 352 Mortimore, M. 228 motorcycle industry, Thai and Vietnamese 16–17 competition from China 17, 209, 218–19 extent of impact of 219–22 reasons for differences in responses to 226–8 transformation of sectoral innovation systems as a result of 222–6 demand conditions 218–19 firms 212–14 government policies 214–16 institutions 219, 225–6 inter-firm linkages 225, 227 methodology for study of 209–10 products 210–12 supporting knowledge-producing agents 216–18, 225 motorcycle repair shops 223
382
Sectoral systems of innovation and production
Motorola 52, 327 Mowery, D. 173 Mtu Wa Mbu 347 Mu, Q. 264 multi-regime analysis 345, 363 multi-technology firms 69 Nadvi, K. 237 NAFTA (North American Free Trade Agreement) 193 Najji 272 Nanya Technology 275 NAOE award 319 Narula, R. 65 NASA 167 National Aquaculture Policy,Chile (Politica Nacional de Acuicultura en Chile: PNAC) 248 National Biofuels Taskforce, Tanzania 344, 352 National Bureau of Statistics, Tanzania 343, 349, 363 National Commission for Aquaculture, Chile (Comision Nacional de Acuicultura) 248 National Communications Commission (NCC), Taiwan 302 National Communications Commission Organization Act (Taiwan, 2005) 301–2 National Council of Research, Brazil 116, 118, 120, 121, 123 National Fishery Service, Chile (Servico Nacional de Pesca: SERNAP, later SERNAPESCA) 246, 249, 258 National Information and Communication Infrastructure Security Mechanism Plan (Taiwan) 317 National Information and Communications Initiative Committee (NICI), Taiwan 288, 304, 305 National Information Infrastructure (NII), Taiwan 309 National Metal and Materials Centre (MTEC), Thailand 217 National Pharmaceutical Pricing Authority (NPPA), India 36
National Pine Institute, Brazil 102 National Science and Technology Development Agency (NSTDA), Thailand 209, 217, 228 National Science and Technology Policy Committee, Thailand 215 National Science and Technology Program for Digital Archives (Taiwan) 313, 314 National Science and Technology Program for e-learning (Taiwan) 313–14 National Science and Technology Program for Telecommunications (Taiwan) 313, 314, 317 National Science and Technology Program Implementation Regulations (Taiwan) 312 National Science and Technology Strategic Plan (2004–13), Thailand 215 National Science Council (NSC), Taiwan 290, 293, 294, 295, 296, 297, 298, 310, 311, 312, 313, 314, 317 national systems of innovation 6, 264 Brazilian 160, 193 differences between sectoral innovation systems and 11–12, 158 Korean 160 Taiwan’s 19, 288, 292–8 National Youth Commission of the Executive Yuan 308 Navarro de Andrade, Edmundo 101, 102, 103 NCC (National Communications Commission), Taiwan 302 neglected diseases, development of drugs for 13, 28, 35 Neiva 168 Nelson, R.R. 6, 7, 9, 11, 59, 63, 158, 159, 161, 264 Network Science Park 314 Networked Readiness Index (NRI) 309–10 networks 5, 7, 21–2, 233 developed networks 79–80, 86 developing networks 79, 86 enabling networks 78–9, 86
Index formation at niche level 338 Jatropha biofuel production in Tanzania 353–4, 356, 358–9, 360, 361 Indian pharmaceutical industry 54 Indian telecommunications industry 54 knowledge networks in Brazilian ICT sector 14, 57–98 analysis and implications 85–9 boundaries between firms and technological partners 14, 59, 62–4, 74, 76–80, 86, 95–6 definition and use of knowledge network 60–62 formation of channels for knowledge flows 14, 59, 65–6, 75–6, 83–5, 86, 87, 98 general characteristics of database 66–74 specialization in different governance mechanisms 14, 59, 64–5, 74–5, 81–3, 86–7, 97 in low-tech sectors 17, 233–4, 237, 251, 252 new chemical entities (NCEs) 35 new drug discovery research (NDDR) 35 New Technology Enterprises (NTEs) 55 New Zealand, pulp and paper industry in 104 niches, creation of 234–5, 236, 253 conclusions and policy issues for Jatropha biofuels 360–63 description of SNM approach 337–9 potential niches for Jatropha biofuels 347–9 recent developments for Jatropha biofuels 352–9 dynamic niche formation at the cultivation stage 353–6 fieldwork methodology 352–3 fragmented niche dynamics at the oil processing stage 356–7 rudimentary niche formation at the end-use stage 358–9 Nicholas Piramal 41 NICI (National Information and Communications Initiative Committee), Taiwan 288, 304, 305
383
NII (National Information Infrastructure), Taiwan 309 Nokia 320 non-profit organizations 139 Nonaka, I. 89 Nordström, M. 361 North American Free Trade Agreement (NAFTA) 193 North American pulp and paper industry 100, 101, 104 notebook computers 314, 322, 325, 327, 328 Nunes, P. 104 OECD 36, 253, 288, 298, 310 OEM–ODM–OBM model 234 see also original equipment manufacturers (OEMs); ownbrand manufacturers (OBMs); own-design manufacturers (ODMs) Ohara, M. 211, 222 OHSAS 18001 standard 247 oil crisis (1973) 337 oil lamps 341–2, 348, 359 oil price 343 Oliveira, L.G. 160, 161, 164 Olmstead, A.L. 100 online banking 309 open-type national laboratories 19, 296 openness factor 247 Openshaw, K. 339, 340 operating profit margin 320 optical-fibre backbones 305 Organization for Economic Cooperation and Development (OECD) 36, 253, 288, 298, 310 Organization for Nucleotide Sequencing and Analysis Network 116, 120 organizational-centred capability 165, 202–3 in Brazilian aeronautical sector 179–81, 184, 187 organizational contingencies literature 64–5 organizational learning 22, 63, 87, 88, 89 organizational process changes 164–5, 201
384
Sectoral systems of innovation and production
in Brazilian aeronautical sector 182–4, 187 organizational structures 6, 7 in Brazilian pulp and paper industry 115 original equipment manufacturers (OEMs) 212, 225, 302, 305, 306, 310, 319, 320–21 orphaned drugs, development of 33 Orsenigo, L. 9 Otahara, J. 229 Owen-Smith, J. 57 own-brand manufacturers (OBMs) 212–13, 223–4, 226, 227, 306 own-design manufacturers (ODMs) 212, 302, 306, 310, 319, 324, 328 packet switching 42 Padgett, J. 172 Panamericana Têxtil 105 Papel Simão 105 Paraguay literacy rates in 137 participation in education in 136 Paraná 118 Parker Hannifin 171 passive learning system 160, 179 Patel, P. 58, 65, 159 patent lawsuits 273–5 patent licence fees 274 patent licensing 260, 265 patent system 8 Indian pharmaceutical industry 13, 28, 34–5, 54 TRIPS compliance 13, 28, 34–5 patents granted to assignees in Taiwan 295, 296, 297 Indian pharmaceutical industry 27, 30, 41, 42 Indian telecommunications industry 27, 30 path-dependency 9, 11, 61, 88 Patton, M.Q. 174 Pavitt, K. 58, 59, 60–61, 63, 64, 159, 163, 165, 201, 208, 236, 265 PECVD (plasma enhanced chemical vapour deposition) 273 Penrose, E.T. 63 Perez, C. 234–5, 236 Perez-Aleman, P. 246
Pérez Casas, A. 147 performance indicator, aircraft 173 Perrow, C. 65 Perry, G.E. 159 Peru literacy rates in 137 participation in education in 136 pharmaceutical industry, Indian government research institutes 13, 28, 41, 42 innovative performance of 12–13, 27, 28, 30–31, 53–4 manufacturing enterprises 36–41 mergers and acquisitions 40–41 orphaned drugs development 33 overall policy framework 31–3 patent regime 13, 28, 34–5, 54 patents granted 27, 30, 41, 42 Pharmaceutical Research and Development Support Fund (PRDSF) 33 price regulations 32, 35–6 private sector enterprises 399–40 product and quality regulations 36, 37–8 public policy support for 31–6 public sector enterprises 39–40 R&D expenditure 30, 35 R&D-intensive companies 33 sectoral system of innovation 13, 28, 31–41, 54 tax incentives 33 trade balance 30 Pharmaceutical Research and Development Support Fund (PRDSF) 33 Pietrobelli, C. 232 PINTEC (Brazilian innovation survey) 71 Piore, M. 235 Piper 172, 191 Pisano, G. 6 Planning Commission 53 Ponsse 122 Ponte, S. 237 Pontifica Universidade Católica (PUC) 119, 121 Porter, M. 260 Porto Alegre 168 Portocel 110
Index Portuguese pulp and paper industry 104 POS (Procedimento Operacion de Saneamiento: Sanitary Operation Procedure) 246 poverty in Tanzania 343–4 Powell, W.W. 57, 60, 65 Pöyry Group 117 Prahalad, C.K. 235 Pramanik, K. 341 Prebisch, R. 232 predatory pricing 270–72, 282–6 Prencipe, A. 175 pre-sensitized printing plates 272, 282, 283, 284, 285, 286 price competition 222, 227, 321, 324 price discounts 269, 270 price index 36 price regulations in the Indian pharmaceuticals industry 32, 35–6 private research institutes, interaction with Brazilian pulp and paper industry 122 in knowledge networks in the Brazilian ICT sector 14, 71, 75, 81, 82, 83, 85, 86 Vietnamese motorcycle industry 217 private sector firms in Brazilian pulp and paper industry 15, 106, 108–25 passim in Indian pharmaceutical industry 13, 28, 38–40, 54 in Taiwanese ICT sector 19 privatization of Embraer (1994) 16, 157, 169, 173, 191 process technology knowledge network 14, 59, 69, 70, 73, 76, 77, 78, 79, 81, 82, 84, 85, 87, 95, 97, 98 product-centred capability 165, 202–3 in Brazilian aeronautical sector 179–81, 184, 185 product changes 164–5, 201 in Brazilian aeronautical sector 183–5 product standards see standards production capability 162, 163, 165, 202–3 in Brazilian aeronautical sector 179–82, 184–7 in developing countries 207, 228
385
production capacity 161, 209 production-chain management 361 production cycle 174 production (process and equipment) changes 164–5, 201 in Brazilian aeronautical sector 182–7 Proença, D.J. 157, 166, 167, 168 Programa de Apoyo al Sector del Software 148–9 Programa de Desarrollo Tecnológico (PDT) 147 project-based revealed technological advantage (PRTA) index 59, 75, 81, 82, 86, 97 Projects A, B, C, D and E 288, 302–3, 304, 307, 316, 318–19 property rights 8 propulsion systems 175, 176, 178 public goods 63 Public Key Infrastructure(PKI) 317 public–private partnerships in Brazilian aeronautical industry 169 in Brazilian pulp and paper industry 106, 120, 122 public research institutes, interaction with Brazilian aeronautical sector 166–7 Brazilian pulp and paper industry 122–3 Indian pharmaceutical industry 13, 28, 41, 42 Indian telecommunications industry 13, 28–9, 42–4, 45, 46–52, 54 in knowledge networks in Brazilian ICT sector 14, 71, 72, 73, 75, 81, 82, 86 Taiwan’s ICT sector 19, 295–6, 307, 308–9 Thai motorcycle industry 216–17, 223, 225, 227 Vietnamese motorcycle industry 217–18, 227 public technology procurement in the Indian telecommunications industry 28, 44, 45–6, 48–50, 54 Puerto Montt 254 pulp and paper industry, Australian 104
386
Sectoral systems of innovation and production
pulp and paper industry, Brazilian 14–15, 99–125 beginnings of comprehensive government innovation policy (1955–1970) 103–7 catch-up learning dynamics and the second-generation innovation system 113–16 conclusion 124–5 establishment of learning network (1900–1955) 100–103 exports 14, 108, 110 innovation, industrial growth and the culture of entrepreneurship (1970–1985) 107–10 new institutions in 105–7 production statistics for long- and short-fibre pulp (1950–2005) 105, 108, 109, 113, 128–30 research policy 123, 124 sectoral innovation system today 116–24 shifting learning dynamics of the sectoral innovation system (1967–1990) 111–13 world ranking in production 14, 99 pulp and paper industry, New Zealand 104 pulp and paper industry, North American 100, 101, 104 pulp and paper industry, Portuguese 104 pulp and paper industry, Scandinavian 101, 104 PVC plates 272, 286 Pyka, A. 60 Quadratic Assignment Procedure (QAP) 59, 75–6, 84, 86 Quadros, R. 160 ‘quality seal’ standards (sello de calidad) 246, 249 quality standards see standards Queiroz, S.R.R. 160 R&D expenditure Brazilian telecommunications and computers sectors 71 Indian pharmaceutical industry 30, 35
Indian telecommunications industry 30 Taiwan’s ICT industry 310, 311, 313, 314, 318 Taiwan’s total 292–3, 297 R&D intensity in the Taiwanese ICT industry 320 R&D-intensive companies (gold standard companies) 33 R&D outsourcing Brazilian telecommunications and computers sectors 71–4 Indian telecommunications industry 44–5 R&D personnel, Taiwan 294 R&D quality systems knowledge network 59, 69, 70, 72, 77, 79, 81, 82, 84, 85, 97, 98 Rabellotti, R. 135, 232 railways, Brazilian 100, 101, 103, 110 Rajya Sabha 49 ram-press 341, 356, 357 RAMA (2001) 247, 248, 249 Ramani, S.V. 34 Ranbaxy 40–41 rate contract 45–6 Raven, R. 337, 363 Raynolds, L. 232 Reed, G.M. 101 reflexive learning 355, 357, 359 reforestation 102, 106–7, 113 regime change of 339 influences on development of Jatropha biofuels in Tanzania 345–52 agricultural regime 349–51, 352, 360 energy regime 345–9, 352, 360 financial regime 351–2 vegetable oil regime 351 regional/local innovation systems 6 Reier, S. 110 renewable Energy Policy Network 343, 348 renewable energy sources 336, 343, 345–6, 361 see also biofuels; Jatropha biofuels in Tanzania
Index replacement equipment manufacturers (REMs) 212 replication 6 RESA (2002) 247, 248, 249 research activities knowledge network 14, 59, 69, 70, 72, 76, 77, 78, 80, 81, 82, 83, 84, 85, 86, 87, 95, 97, 98 Research and Development of Key Parts, Components and Products Program, Taiwan 290 Research Group IP (International Programs) 341 Research Institute of Technology for Machinery (RITM) 217 resource-based view 63, 80 retrofitting capabilities 47 revealed technology advantage (RTA) index 75 reverse engineering 28, 34, 35, 163 reverse product cycle 234 Rhode, P. 100 Rhodes, R. 236 Rieiro, M. 138, 146 Rio Claro, São Paulo 101 Rio Grande do Sul 112, 168 see also Federal University of Rio Grande do Sul (UFRGS) Ripasa 115, 116 risk sharing/co-development partnerships 171–2, 177, 188, 189 robots 270–72, 286 Rocha, M.G. de B. 112, 113 Rodrik, D. 261, 275 Rodríquez-Clare, A. 260, 261 Romer, P. 260 Romijn, H. 138 Rosenberg, N. 7, 9, 10, 158, 159, 173 Rothwell, R. 65 routine capability 163, 165, 202 in Brazilian aeronautical sector 179–88 royalties and fees 273 RPG (Aventis) 40–41 RTA index 75 Rugman, A.M. 64 Rural Energy Area (REA) Tanzania 344 Rural Energy Fund (REF), Tanzania 344 Rwambali, F. 344
387
S/A Industries Reunidas Francisco Matarazzo 103 Saab 172 Sabel, C. 235 Safe Quality Food (SQF-SOTA) standards 249, 250, 258 salmon farming Chilean 17–18, 237 concentration of industries 240–41 conclusions drawn from 250–53 diverse variety of technologies needed for 242–4 exports 237, 238, 251 FDI in 240–41 global position of industry 238–9 increase in value added in products 241–4, 251 policy implications 252 prices 239 production volume 238 skill development 244 standards developed in 246–51, 258 suppliers to 241–4, 251 growth worldwide 237 production and exports by country 238 SalmonChile see Association of the Salmon Industry in Chile (SalmonChile), formerly Association of Salmon and Trout Producers of Chile Salter, A.J. 64 Samsung Electronics 275 Samsung SDI 275 Santos, P.T. dos 113 Sanyang Motors 213, 214, 221 São Paulo 101, 102, 105, 116, 118, 121, 168 see also University of São Paulo São Paulo Research Foundations (FAPESP) 120, 121 satellite telecommunications 301 Saxenian, A. 133, 151 Sayari oil expeller 341, 356–7 SBIR (Small Business Innovation Research), Taiwan 299 Scandinavian pulp and paper industry 101, 104
388
Sectoral systems of innovation and production
Schmitz, H. 250 Schumpeter, J. 160 SCI database 295 science parks 290, 292, 307, 314 scientific publications 295, 297 Second Pulp and Paper plan (II PNPC), Brazil 114 Second World War 104, 166, 232 sector, definition of 5, 264 sectoral system framework 5–12, 160–61, 233 aggregation issue 12 boundaries of 9, 10–11, 233 broad, open and flexible framework 12 in context of developing countries 207–8, 234–7 differences between national innovation system perspective and 11–12, 158 elements of 5, 6–11, 29, 132–2, 207, 233, 264 evolutionary theory and 5–6, 7, 9, 57–8, 233 innovation system literature and 6, 264 Sedjo, R. 116 Seed Net 309 SEED Plan (Software Engineering Environment Development), Taiwan 308–9 selection 5, 6, 132 in Brazilian pulp and paper industry 100, 101 in Uruguayan software industry 133 Selela 359 semiconductor industry, Taiwan 19, 295, 303, 304, 312, 327 semiconductor production equipment 262, 273, 274 semiconductors product development knowledge network 14, 59, 69, 70, 73, 76, 77, 78, 79, 81, 82, 84, 95, 97, 98 Seoul National University 274 SEPIN 67, 71, 72 SERNAP see National Fishery Service, Chile (Servico Nacional de Pesca: SERNAP, later SERNAPESCA)
SERNAPESCA see National Fishery Service, Chile (Servico Nacional de Pesca: SERNAP, later SERNAPESCA) Serviço Florestal do Brasil 101 Shen, R.-J. 330 Shenzhen 326 Shuma, J. 361 Siemens 52, 324 SIGes (Sistema Integrada de Gestions: Integrated Management System) 247, 249, 258 Silicon Plaza 148 Silicon Valley 133, 307 Silveira, J.M. da 116 silviculture 99, 105, 111–12 Simmie, J. 146 Simon, H.A. 61 Singer, H.W. 232 Singh, H. 66 Singh, Manmohan 55 Sjolander, S. 69 skilled workers, availability of in Brazil 117 in Chile 244 in India 35 in Taiwan 287, 293–4 in Uruguay 15, 136, 141 small and medium-sized firms enterprise resource planning (ERP) products for 136 making a technological catch-up in the Korean capital goods industry see capital goods industry, Korean technological capability of local SMEs in the Brazilian aeronautical sector 16, 157–8 analytical framework for studying 164–5 characteristics of local SME suppler firms 175–7 government policies for upgrading 193 linkages for each group of SME suppliers 184–92 results on technological capacity accumulation by local SME suppliers 179–90 selection of sample 174–5
Index in Uruguayan software sector 15, 137, 138, 139, 141, 147, 150, 152 Small Business Innovation Research (SBIR), Taiwan 299 SME 007 Plus 213, 223, 225, 227, 228 Smith, Adam 260 Smith, S.E. 39–40 smuggling 216, 344 Snoeck, M. 131 soap-making 341–2, 359 social network analysis 60 social responsibility 109, 110, 113, 116, 124 Sociedade de Investigaçoes Florestais 120, 121, 123 SOE award 319 software consulting services 138, 139 software development, firms engaged in 138, 139 software product development knowledge network 14, 59, 69, 70, 73, 76, 77, 78, 79–80, 81, 82, 83, 84, 85, 86, 87, 95, 97, 98 software sector, Montevideo, Uruguay 15, 131–53 conclusions and policy implications 150–52 emergence and development of 132, 135–8 exports 131, 148–9 key characteristics of sample of firms 138, 139 knowledge base of 134–5, 151 learning processes and innovation outcomes 132–3, 138–46 primary data collection 133 role of public policy 146–7 sectoral systems of innovation (SSI) framework for analysing 132–3 size and growth of 134 supporting institutions and private networks 133, 144, 145, 147–50, 152 solar power 336, 346 Sonaca 171 SOTA (Salmon of the Americas) 249
389
‘special quality’ 235 specific tools sector, Korean 277 spin-offs Embraer founded as 167 in Uruguayan software sector 15, 140, 141–2, 145–6, 151 SS7 Intelligent Network signalling systems 46, 47 stage-skipping catch-up 277 standardized software 134, 138 standards in Brazilian aeronautical sector 182–3 in Chilean salmon industry 246–51 collective capabilities and role of Association of the Salmon Industry 248–9 details of 258 factors that make firms comply 247–8 historical perspective 246–7 influence on external standardsetting 249–51 Chinese ICT standards 308, 321 for developing countries in the global context 237, 245–6 in Indian telecommunications industry 46 in Thai and Vietnamese motorcycle industries 214, 215, 219, 224, 226 Stankiewitz, R. 6, 264 Stanturf, J.A. 100, 101 State of São Paulo Research Foundations (FAPESP) 116 Staubman, R. 341 Stefanuto, G.N. 83 Stigler, G. 260 Stoker, G. 236 Stolovich, L. 133, 138 StoraEnso 114–15 Storper, M. 151 strategic alliances Brazilian pulp and paper industry 118, 122 BRIC countries 193 risk sharing/co-development partnerships in Brazilian aeronautical industry 171–2, 177, 188, 189
390
Sectoral systems of innovation and production
strategic niche management (SNM) 20, 336 applicability for developing countries 362 description of SNM approach 337–9 extensions to 362–3 limitations of 362 niche analysis for Jatropha biofuels 352–9 dynamic niche formation at the cultivation stage 353–6 fieldwork methodology 352–3 fragmented niche dynamics at the oil processing stage 356–7 rudimentary niche formation at the end-use stage 358–9 subcontracting 269, 320 Subramanian, D. 44 Subsecretaria de Pesca 248 sulphate pulp process 101, 104–5, 107, 108 sulphite pulp process 101, 102, 103, 104 Sunley, P. 151 Sunstar Co. 273 suppliers to Chilean salmon farming industry 241–4, 251 knowledge flows from 144, 184, 188–90 technological capability of see technological capability Survey of National Science and Technology Activity, Taiwan 292 sustainable development 336 Sutz, J. 145, 152 Suzuki 212, 213, 214, 221 Swan, J.A. 66 system of innovation approach 158–61 Szapiro, M. 83 tacit knowledge in communities of practice 66 converted into explicit knowledge through software 266 of Embraer suppliers 191 innovation project approach and 59, 60 from interface between producers and customers 260, 265, 266, 279
Taiwan growth drivers since 1990 289–91 industrial structure change 289–91 trends in productivity 290 trends in venture capital and FDI 290–91 ICT industry in Challenge 2008 Program 288, 301, 305, 316 Chinese standards adopted in 308, 321 current status of information society 318 demand conditions 315–16 deregulation of telecommunications 291, 301–2, 307, 309, 319 e-Taiwan Program 288, 304–5, 309, 317, 319 evolution from a historical perspective 308–9 expenditure to GDP ratio 287 exports 315–16 FDI in China 306, 314–15, 323–33 hardware sector output 287 industry performance of 312–15 infrastructure of 316–18 interaction with research institutes 19, 295–6, 307, 308–9 interaction with universities 296–7 internet penetration rate 318 internet subscribers 287, 318 M-Taiwan Program 288, 305, 317, 319 methodology and analytical framework for study of 288–9 motivation and objectives 287–8 performance and demand conditions 309–10 policy implications 318–23 product differentiation in 308, 320 Projects A, B, C, D and E 288, 302–3, 304, 307, 316, 318–19 software sector output 287 technology performance of 310–12, 313 Two Trillion Twin Stars programme 287, 303–4 major policies and governance 298–308
Index existing industrial policies 298–300 knowledge flow and interactions with outside 307–8 major new policies 301–5 mode of governance 305–7 transformation of innovation system 292–8 education systems and human capital 293–4 patents and publications 294–5, 296, 297 role of research institutes 295–6 role of universities 296–7 Taiwan Intellectual Property Office (TIPO) 296 Taiwan Ministry of Economic Affairs (MoEA) 19, 288, 289–90, 292, 296, 299, 300, 302, 304, 305, 308, 317 see also Department of Industrial Technology, Ministry of Economic Affairs (DoIT/ MoEA), Taiwan Taiwan Ministry of Education 293 Taiwan Ministry of Finance 309 Taiwan Ministry of the Interior 305 Taiwan Ministry of Transportation and Communications 309 Takeuchi, H. 89 Taller de Informática 146 Tanzania energy policy in 344–5 poverty-related indicators for 343–4 total surface area 344 see also Jatropha biofuels in Tanzania Tanzanian Ministry of Agriculture and Minerals 352 Tata Consultancy Services (TCS) 148 tax incentives Brazilian aeronautical sector 168, 190 Brazilian forestry sector 102, 107 Brazilian ICT sector 58, 67, 69 Indian pharmaceutical industry 33 for Jatropha niches 360 in mainland China 325 for R&D and personnel training in Taiwan 299, 304, 321 Uruguayan software sector 147, 148
391
TCP/IP internet communication agreement 309 TD-SCDMA 317 TDP-contracted research institutes 300 technological capability catch-up in the Korean capital goods industry see capital goods industry, Korean definitions of 161–2, 165 domains of 162–3, 165, 202–3 evolutionary perspective and 161 levels of see innovative capability; routine capability obtaining in developing countries 207–8, 228, 235–7 see also collective capability relationship between technological change and 164 of suppliers in Brazilian aeronautical sector 16, 157–8 analytical framework for studying 164–5 characteristics of foreign suppliers located in Brazil 177–9 characteristics of local SME supplier firms 175–7 government policies for upgrading 193 linkages for each group of SME suppliers 184–92 results on technological capability accumulation by local SME suppliers 179–90 selection of sample 174–5 of suppliers in Thai and Vietnamese motorcycle industries 212–13, 215, 218, 223–8 technological change levels and domains 164–5, 201 implemented by SMEs in Embraer supply chain 182–7 linkages contributing to implementation in Embraer supply chain 184–92 relationship between technological capability accumulation and 164 and system of innovation literature 159
392
Sectoral systems of innovation and production
Technological Institute of Aeronautics (ITA) 167 technological intensity 232, 234, 235 technological niche, creation of 338, 360 technological opportunities 10, 63, 66, 79, 80, 87 technological services knowledge network 14, 59, 69, 70, 72, 76, 77, 78, 80, 81, 82, 84, 85, 86, 95, 97, 98 technological systems 6, 264 technological training knowledge network 14, 59, 69, 70, 72, 76, 77, 78, 80, 81, 82, 83, 84, 86, 87, 95, 97, 98 technology deterministic perspective 65 Technology Development Programs (TDP), Taiwan 299–300, 318–19 Technology Information and Forecasting Assessment Council, India 52 technology licensing 260, 265 Technology Research Program for Innovative Services, Taiwan 316 technology transfer 19, 163, 181, 189, 190, 223, 228, 297, 307 Technopress 240, 242 Teece, D. 6, 59, 63, 66 Teeselink, M. 153 telecommunications industry, Chinese 42–3 telecommunications industry, Indian dependent on imports 13, 27, 29, 44 domination by MNCs 13, 27, 29, 30, 44, 52 government research institute 13, 28–9, 42–4, 45, 46–52, 54 innovative performance of 12–13, 27, 30–31, 53–4 manufacturing enterprises 52–3 outsourcing deals between foreign MNCs and Indian contract research organizations 44–5 patents granted 27, 30 public technology procurement 28, 44, 45–6, 48–50, 54 R&D expenditure 30 sectoral system of innovation 13, 28–9, 41–53, 54
technical standards 46 trade balance 30 telecommunications industry, Taiwan 288–9 deregulation of 291, 301–2, 307, 309, 319 exports 315–16 further development of 311 National Science and Technology Program for 313, 314, 317 policy suggestions for 322 R&D project funding and manpower 312 revenues to GDP ratio 302 shipment value of products 315 see also mobile communications industry Teng, B.S. 66 Thai Automotive Institute 209, 215 Thai German Institute 216–17 Thai Ministry of Commerce 223 Thai Ministry of Industry 215, 223 Thailand economic growth rate 218 GNI per capita 218 motorcycle industry in see motorcycle industry, Thai and Vietnamese Thaksin administration 215 Thompson, J.D. 58, 65 Tidd, J. 57, 58, 62 TIER 320 Tiger Motorcycle 209, 212–13, 223, 225, 227–8 TIRDO 363 Toivanen, H. 100, 101, 102, 104 Tokai 273 Torch Programme 55 tourism 208 Toyota 216 trade balance between Taiwan and mainland China 323, 329–30 Indian pharmaceutical industry 30 Indian telecommunications industry 30 Korean machine tool industry 259, 261–4, 276 transaction cost theories of the firm 62–3
Index transaction costs 246 transportation costs 347 TRIPS compliance 13, 28, 34–5 trust building 354 tuberculosis 35 Tunduru 352 turbines 179 Tushman, M.L. 66 Two Trillion Twin Stars programme 287, 303–4 Tzeng, G.K. 293 Uisso, J.P. 344 UMC 320 UNCTAD 135, 228 UNDP Human Development Index 343 UNIDO 341 United Kingdom biofuel blends in 347 Korean machine tool exports to 278 United States, Korean machine tool exports to 277, 278 Universidad de la República 146 university–industry linkages Brazilian aeronautical sector 188–92 Brazilian ICT sector 68 Brazilian pulp and paper industry 106, 117, 119, 120, 122 literature on 89 Taiwanese industry generally 296–7 Thai motorcycle industry 216, 223, 225, 227 Uruguayan software sector 141, 144, 145, 146 Vietnamese motorcycle industry 217, 227 University of Dar es Salaam 347, 358 University of São Paulo 106, 112, 117, 119, 121, 122 University ORT of Uruguay 146, 147 Uruguay literacy rates in 136, 137 participation in education in 136 real GDP growth 136 software sector in see software sector, Montevideo, Uruguay Uruguayan Business Association of Information Technologies (CUTI) 131, 133, 148–9, 150
393
Uruguayan Ministry of Education and Culture 147 US Forest Products Laboratory, Madison, Wisconsin 102 Usage Component Index 310 user fees for telecommunications 321 USMC 320 Valença, A.C. de V. 114 Valle, C.F. do 114 value added increases in Chilean salmon farming industry 241–4, 251 increases in Thai motorcycle industry 222, 227 of Taiwan’s ICT industry 320 van der Laak, W.W.M. 354, 355, 356, 357, 359, 363 van Eijck, J.A.J. 336, 363 Vandergeest, P. 237 variety creation 5, 6, 89, 132, 133 Vega, M. 65 vegetable oil regime, Tanzanian 351 Veloso, F. 136 venture capital 290–91 Veracel Celulose 114–15, 118 Vietnam GNI per capita 218 motorcycle industry in see motorcycle industry, Thai and Vietnamese Vietnam Academy of Social Science (VASS) 210 Vietnam Engine and Agricultural Machinery Corporation 217 Vietnam Institute of Economics 210 Vietnam Manufacture and Export Processing Co., Ltd. (VMEP) 213, 214, 221 Vietnamese Ministry of Industry 217, 229 Vietnamese Ministry of Transport 229 Viotti, E. 159–60 Visser, J. 341 Voith 117–22 von Hippel, E. 58 von Tunzelmann, N. 60, 66, 89, 232, 233, 234, 236, 254 Votorantim Celulose e Papel 114, 115, 116, 118
394
Sectoral systems of innovation and production
W-CDMA 317 wages 109, 328 Wal-Mart 249, 258 Walker, G. 66 Waltring, F. 237 wars 337 Wasserman, S. 60 Wave a motorcycle 222 Weber, M. 363 Wehn, U. 134 Weiszflog, Hasso 102 Wenger, E. 66 Westphal, L.E. 162 Whitelaw, J.A. 148 Williamson, O.E. 60, 62 WiMAX 55 wind power 336, 346 Winter, S.G. 6, 158, 161 WITSA 135 WLAN 317 Woodward, J. 58, 65 World Bank 114, 118, 136, 137, 153, 218, 343
World Economic Forum (WEF) Global Information Technology Report 2005–2006 309–10 World Fact Book 343 World Resources Institute 346 World Trade Atlas 220 World Trade Organization (WTO) 34, 172, 301 Wright, G. 100 Wu, R.-I. 291, 306 Yamaha 212, 213, 214, 221, 222, 227 Yaskawa 272 Yin, K.K. 174 Young, A. 260 Young, J. 109 Yu, T.S. 289 Zanfei, A. 65 Zeidan, R.M. 115 Zonamerica Business and Technology Park 133, 148 ZTE 44
