GLOBAL PRODUCTION NETWORKING AND TECHNOLOGICAL CHANGE INEAST ASIA EDITORS SHAHID YUSUF M. ANJUM ALTAF KAORU NABESHIMA GLOBAL PRODUCTION NETWORKING AND TECHNOLOGICAL CHANGE IN EAST ASIA SHAHID YUSUF M. ANJUM ALTAF KAORU NABESHIMA EDITORS A copublication of THE WORLD BANK the World Bank and Washington, D.C. Oxford University Press © 2004 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. First printing June 2004 1 2 3 4 07 06 05 04 The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the govern- ments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522- 2422, e-mail pubrights@worldbank.org. Library of Congress Cataloging-in-Publication Data Global production networking and technological change in East Asia / edited by Shahid Yusuf, M. Anjum Altaf, Kaoru Nabeshima. p. cm Includes bibliographical references and index. ISBN 0-8213-5618-6 1. Manufacturing industries--East Asia. 2. Business networks--East Asia. 3. Production management--East Asia. 4. Technological innovations--East Asia. 5. Electronic industries--East Asia. 6. International business enterprises--East Asia. 7. Globalization--Economic aspects--East Asia. I. Yusuf, Shahid, 1949­ II. Altaf, M. Anjum, 1950­ III. Nabeshima, Kaoru. HD9736.E18G55 2004 338.8'895--dc22 2004044087 Cover design by Debra Naylor of Naylor Design, Inc. CONTENTS Preface v 1 Competitiveness Through Technological Advances Under Global Production Networking 1 Shahid Yusuf 2 The New Global Supply-Base: New Challenges for Local Suppliers in East Asia 35 Timothy J. Sturgeon and Richard K. Lester 3 Global Production Networks in East Asia's Electronics Industry and Upgrading Perspectives in Malaysia 89 Dieter Ernst 4 Production Networks in East Asia's Auto Parts Industry 159 Richard F. Doner, Gregory W. Noble and John Ravenhill 5 The Global Strategies of Japanese Vehicle Assemblers and the Implications for the Thai Automobile Industry 209 Ken'ichi Takayasu and Minako Mori 6 Chinese Enterprise Development and the Challenge of Global Integration 255 Edward S. Steinfeld 7 Logistics in East Asia 297 Trevor D. Heaver iii iv CONTENTS 8Technology and Innovation in Developing East Asia 353 Hal Hill 9Technology Transfer in East Asia: A Survey 395 Kaoru Nabeshima 10 An Investigation of Firm-Level R&D Capabilities in East Asia 435 Gary H. Jefferson and Zhong Kaifeng Annex: Comparison of Technology Parks 477 Index 481 About the Editors 489 PREFACE This is the third volume in a series of publications from a study co-spon- sored by the Government of Japan and the World Bank to examine the sources of economic growth in East Asia. The study was initiated in 1999 with the objective of identifying the most promising path to development in the light of global and regional changes. The first volume, Can East Asia Compete?, published in 2002, provided a compact overview of the relevant strategic issues and future policy di- rections. Innovative East Asia, the second volume, analyzed each of main issues and consequent policy choices drawing comprehensively upon re- cent empirical research and the findings of firm surveys conducted for the study. Its principal message is that sustained economic growth in East Asia will rest on retaining the strengths of the past--stability, open- ness, investment, and human capital development--on overcoming the sources of current weaknesses--in the financial sectors, corporate, judi- cial, and social sectors--and on implementing the changes required by the evolving economic environment, particularly with regards to tech- nology development. This book, Global Production Networking and Technological Change in East Asia, is the first of two volumes of papers commissioned for the East Asia study. It provides detailed information, analysis and case studies fur- ther illuminating some of the topics covered in the earlier volumes. The contributors rigorously examine the effects of the changing global geog- raphy of production for the growth prospects of East Asian economies. They conclude that in the face of a global environment, economies in East Asia need to adapt to the changing character of global production networks and to nurture and develop technological capabilities in order to sustain their growth prospects. The companion volume, Global Change v vi PREFACE and East Asian Policy Initiatives, will include a second set of papers that ex- plore regional and institutional policy options that East Asian economies face in a similarly in depth manner. Both volumes complement Innovative East Asia and are addressed to researchers, students and policymakers. The financial backing of the government of Japan through its Policy and Human Resources Development Fund provided vital support for this project, as did senior public officials who gave generously of their time. We are deeply grateful to Haruhiko Kuroda, Naoko Ishii, Masahiro Kawai, Kiyoshi Kodera, Rintaro Tamaki, Junichi Maruyama and Takatoshi Ito. The staff of the World Bank's Tokyo office facilitated the reviews and sem- inars, and we greatly appreciate the assistance provided by Yukio Yoshi- mura, Shuzo Nakamura, Mika Iwasaki, Tomoko Hirai, and Hitomi Sasaki. We also deeply appreciate the support we received from Deepak Bhattasali and Jianqing Chen at the World Bank's Beijing office. We owe special thanks to K. Migara De Silva for his enthusiastic and tireless support in or- ganizing and participating in seminars in Beijing and Tokyo. The papers in this volume were presented at seminars and workshops in Beijing; Cambridge, Mass; Tokyo, and Washington, DC. The com- ments received helped the authors in revising their drafts. We would like to thank all of those who participated in these seminars, along with the many reviewers of the entire manuscript and, in particular, Jose Luis Guasch. At the World Bank, the Development Research Group has for several years offered a base for the study. In addition, we are grateful for the sup- port provided by East Asia and Pacific Region. We are especially indebt- ed to Jemal-ud-din Kassum and Homi Kharas for their guidance and strong encouragement. The study team was ably supported by the research skills of Soumya Chattopadhyay, Farhan Hameed, and Yifan Hu. The manuscript was pre- pared by Paulina Flewitt, Marc Sanford Shotten, and Rebecca Sugui; and we thank Susan Graham, Patricia Katayama, and Ilma Kramer of the Of- fice of the Publisher for their expert management of the editorial and print production of the volume. CHAPTER 1 COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES UNDER GLOBAL PRODUCTION NETWORKING Shahid Yusuf S tarting in the early 1990s, the emerging economies of East Asia ac- celerated the pace of deregulation and integration with the global economy.1 Although the crisis of 1997­98 resulted in a brief hiatus, making countries wary of the risks from open capital accounts and skeptical of the gains, the process has continued into the early twenty-first century, spearheaded by China's entry into the World Trade Organization (WTO).2 Regional and bilateral free trade areas, the continuing evolution 1. Since the late 1990s, globalization has been analyzed, chronicled, and debated in obsessive de- tail, with frequent backward glances at the earlier round of globalization that commenced around the turn of the nineteenth century. While the gains from trade and the sharing of technology and ideas have been acknowledged with qualifications (Bhagwati 2004), a legion of skeptics worry about cultural imperialism, the risks from volatile capital flows, the loss of jobs as a result of rapid shifts in comparative advantage, and the declining role of the nation-state. For a compact review of the trends signifying globalization and their implications for developing nations, see World Bank (2000). Micklethwait and Woolridge (2000) provide an engaging account for the general reader, while a close look at globalization in historical perspective can be found in O'Rourke and Williamson (1999) and Held and others (1999). Some of the tensions generated by globalization are brought out in Stiglitz (2002), and its impact on markets and on inequality is discussed by Bour- guignon and others (2002). The interactions between globalization and the business sector are explored in Cairncross (2002) and Dicken (2003), while Yeung and Olds (2000) show how Chinese business groups are participating in and contributing to globalization. 2. A provisional summing-up of the considerable research on the effects of capital account openness would be as follows: middle-income countries with strong market and financial institutions are likely to benefit over the medium term from an open capital account. Low-income countries are less likely to derive much benefit, and their institutional gaps might make them prone to crises. Countries can contain the effects of crises by way of capital controls, although their utility, even in the cases of China, Chile, and Malaysia, is by no means rigorously established. See Arteta, Eichengreen, and Wyplosz (2001); Edison and others (2002); Eichengreen and Leblang (2003); Henry (2003), Klein (2003); Wyplosz (2001). 2 GLOBAL PRODUCTION NETWORKING IN EAST ASIA of international production networks, and substantial flows of foreign di- rect investment (FDI) within and to the region have tightened the eco- nomic strands binding East Asian firms to one another and to the developed nations of the Organisation for Economic Co-operation and Development (OECD). Together, these factors are defining the parame- ters impinging on the growth and profitability of firms. This process of global integration is having five major consequences for East Asia. First, it has substantially widened the opportunities for firms, as tariff and nontariff barriers are stripped away, information cir- culates more freely, and transport costs decline.3 The downside is that competitive pressures have also increased, and firms that do not meet the market test are quickly weeded out. Second, many more firms are now linked to global production networks, the outgrowth of ongoing efforts by many large multinational corporations (MNCs) to establish foreign subsidiaries, to deverticalize their activities, to outsource certain func- tions, and to subcontract the production of numerous components.4 The rise of such global production networks has intensified both trade within industries--to almost a third of total world merchandise trade--and trade among affiliates of multinational companies--approximately 45 percent of total trade or about $3.2 trillion in 1999.5 It is also differentiating the roles of firms that are original equipment manufacturers (OEMs),6 firms that have risen into the ranks of original design manufacturers (ODMs), and firms that have become original brand manufacturers (OBMs) and stand at the apex of their global production networks. As discussed in this chapter, the prospects of OEMs are starting to diverge from those of ODMs and OBMs.7 3. For example, regional tariffs in the ASEAN (Association of South East Asian Nations) Free Trade Area fell to the 0­5 percent range for a wide range of goods, including auto parts ("ASEAN: Free Trade Fears" 2002; "Autos and AFTA" 2002). 4. The global production network, according to Ernst, provides the lead firm "with quick and low- cost access to resources, capabilities, and knowledge that complement its core competencies" (chap- ter 3 of this volume). For an application of the global production network--that is, the setting-up of decentralized worldwide production networks by firms that are either lead buyers or lead producers--to the apparel industry, see Gereffi and Memedovic (2003). 5. See also the analysis by Feinberg and Keane (2003) of the growth of trade related to multi- national companies. 6. OEMs are firms that serve as subcontractors and produce according to the specifications of buyers. See Hanson, Mataloni, and Slaughter (2003); "Multinationals and Globalization" (2001); World Bank (2000). 7. Most East Asian firms from the industrializing countries fall into the ODM categories. Very few have progressed to the OBM category. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 3 Third, whereas price competitiveness remains an important determinant of success in the marketplace, firms of all stripes are finding that the capac- ity to innovate is the key to productivity, growth, and greater profitability. This applies to OEM suppliers, to firms seeking to achieve greater inde- pendence and to enlarge their market share and profitability, using the ad- ditional leverage provided by innovative design, and to firms trying to establish a brand in the international marketplace.8 Fourth, the progressive leveling of trade barriers and the emphasis on lean production and frugal networking practices, supported by just-in-time de- livery of components, are bringing about the geographic consolidation of certain industries such as automobiles, engineering, electronics, and white goods in closer proximity to their principal markets. For example, the pro- duction of one-ton trucks and, to a lesser extent, sedans for the Southeast Asian market is shifting to Thailand, and leading auto companies are likely to choose China and the Republic of Korea as the hub for their regional ac- tivities. Multinational companies favor China and Thailand because they have the lowest labor and overhead costs in the region. This and the quest for optimal scale is also leading to a shakeout of firms in major subsectors, such as hard disk drives (120 companies in the early 1980s--less than 10 re- main out of 120) and personal computers, and to a coalescence of market power in the hands of relatively large firms with a regional, if not a global, footprint. Fifth, auto assemblers and manufacturers of consumer electronics are streamlining their product lines to realize the economies from higher- volume production runs. They are also seeking to reduce the number of components to lessen the complexity of products, lower costs, and minimize the need for numerous suppliers. STAGES OF EAST ASIAN DEVELOPMENT How East Asian manufacturing firms respond to the changing market environment, harness the potential of global production networks, and buttress price competitiveness by building technological capability will strongly influence their growth prospects. The purpose of this introduc- tory chapter is to underscore the significance of the global production net- works and to indicate how their evolution requires East Asian firms to redouble their efforts to build technological capability, improve logistics, and harness foreign direct investment to increase competitiveness. 8. Hallward-Driemeier, Iarossi, and Sokoloff (2003) find that firms participating in export mar- kets have higher productivity, although research reported by Tybout (2000) suggests that the causality runs from higher productivity to participation in trade rather than vice versa. 4 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Labor-Intensive Light Manufacturing Viewed through the lens of trade and technological change, development in East Asia can be divided into four stages. Starting in the 1960s, Hong Kong (China), Republic of Korea, Singapore, and Taiwan (China) began in- vesting in light, labor-intensive manufacturing industries whose products were quickly able to penetrate the lower and most price-sensitive end of markets in Western countries. The American involvement in Vietnam at that time and the interest of the United States in strengthening the economies of its allies in the region led to policies that eased access to U.S. markets. Access was facilitated further by the efficiency of the retail sector and the capacity of American retail chains to source products from East Asia. Having established a beachhead, the East Asian exporters con- solidated their market position by improving quality and broadening the range of manufactured exports. Within a matter of years, they established a reputation not just for competitiveness and quality but also for on-time de- livery, flexibility, and readiness to enter new product lines. Upgrading by Newly Industrialized Economies, Enter the Southeast Asians This first stage of export-led growth in developing East Asia was followed by a second, extending from the early 1970s through the late 1980s, during which four newly industrialized economies were joined by several Southeast Asian countries and, after the early 1980s, by China.9 The newly industrial- ized economies shifted into more capital- and skill-intensive manufacturing industries as well as into producer services, enabling newcomers to enter the subsectors they were vacating. East Asia's gathering reputation as a politi- cally stable region with a demonstrated track record of industrial perform- ance and export competitiveness attracted a rising flow of foreign direct investment from Japan, the United States, and some of the European coun- tries. Having determined that East Asia had the manufacturing capability, the supplies of educated and trainable workers, and the policy environment conducive to rapid industrialization, Western and Japanese corporations be- gan investing in local firms and establishing wholly owned facilities to lever- age the advantages of lower labor and overhead costs.10 Major corporations 9.Thishasbeencalledtheflying-geesepatternofdevelopment,initiatedbyJapan(Ozawa2001).Each groupofcountriesthathasmoveduptheproduction-technologyladderinthewakeofJapanhasmade wayforagroupofsuccessfuleconomiesloweronthescaleofdevelopment.SeeOzawa(2003).Mean- while, access to the U.S. market has stimulated export-led growth (Cutler, Berri, and Ozawa 2003). 10. Foreign direct investment from Japan to the East Asian economies increased as cost differ- entials widened following the Louvre Accord in 1985 that led to appreciation of the yen. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 5 that were capable of managing geographically dispersed production facil- ities transferred components and machinery to overseas subsidiaries and subcontractors that engaged in labor-intensive assembly operations and ex- ported the final product back to the parent company or to a third party. The cost advantages of parceling out production of parts and components to numerous specialized producers throughout East Asia increased as local manufacturing and intermediaries gained experience, tariff barriers fell, and logistics improved. This, in turn, led to rising trade flows within indus- tries and within production networks led by MNCs. Starting with items such as garments, footwear, and consumer electronics, these patterns of pro- duction and trade spread to machinery, electronics, chemicals, and transport industries. Production Networking Gathers Momentum By the end of the 1980s, East Asia was being drawn into a web of relation- ships created by intertwined flows of trade and FDI. Hatch (2003, p. 29) states, "This export of Japanese capital and technology helped weave to- gether the economies of Asia." He also cites the work of Tamara, who has remarked, "Japanese multinational companies are building a regional divi- sion of labor that emphasizes technology-intensive prototype production in Japan and mass production of standardized products in Asia" (Hatch 2003, p. 31). This third stage lasted through much of the 1990s and wit- nessed an unusually rapid expansion of trade. Between 1990 and 2000, world trade grew at an annual average rate of 6.6 percent. In the East Asian region, it grew nearly 14 percent annually. These years also saw the flow- ering of global production networks that were budding in the previous decade. Market deregulation, which had been gathering momentum, be- came widespread in conjunction with WTO-sanctioned trade liberaliza- tion, and it dramatically ratcheted up the level of market competition. An added twist was imparted by the intensifying of technological change, most notably, in the many manufacturing and service industries related to the provision, distribution, and use of information. Moreover, by simplifying the management of dispersed production and enhancing the efficiency of supply chains, information technology further encouraged firms to dever- ticalize and to outsource not just a variety of production tasks but also a multitude of back-office functions. During the space of a few years, the combined forces of trade, FDI, and organizational restructuring elaborated a system of global net- working that made the tradable segments of the light manufacturing sectors in East Asia's emerging economies almost coextensive with those 6 GLOBAL PRODUCTION NETWORKING IN EAST ASIA of Japan, the United States, and, to a lesser extent, Europe. Large MNCs from these countries moved their labor-cost-sensitive activities to East Asia and, in addition, invested in other production facilities to access pro- tected domestic markets. In industries such as autos, electronics, office equipment, and optical instruments, these moves by the large firms were matched by many of their local parts suppliers, which also set up sub- sidiaries in East Asia. At the same time, the big retail chains in the Western countries and pro- ducers of commodity items began sourcing entire categories of products from East Asia, extending from men's shirts to personal computers. Some of these companies, such as Nike, retained control of design and research, while others limited themselves to brand management and quality control, leaving their overseas suppliers to take the initiative in submitting new de- signs and fresh products. Technological Deepening By about the mid-1990s, the manufacturing subsectors of the industrializing East Asian economies were linked to the economies of leading OECD coun- tries through trade that was intermediated by buyer- and producer-driven supply chains and through a host of relationships arising from FDI. How- ever, even as integration between the East Asian region and the OECD countries appeared to be tightening, the dynamics of continuing trade lib- eralization, intensifying competition from China, both in product markets and for FDI, and the apparent advantages of firm size in a globalizing world were ushering in a fourth stage of development in which competitiveness is predicated on technological deepening. A discussion of this most recent stage and what it demands of firms takes up the balance of this chapter. Its various facets are examined in detail by the contributors to this volume. There is no better exemplar of the leverage provided by steady technolog- ical capability than China, which has emerged in the short space of 10 years as the world's fourth largest trading nation in 2003 and second largest pro- ducer of information technology hardware, with close to 14 percent of the world market ("Link by Link" 2003). SUSTAINING COMPETITIVENESS: THE ISSUES How the East Asian economies respond to the challenges posed by this fourth stage will decisively influence their growth in the early twenty-first century. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 7 Four issues differentiate this new and difficult phase of development from preceding phases. By exploring these issues, economies in the region can elucidate their major strategic choices. The issues revolve around changes in the global market brought by shifting dynamic comparative advantages; various strategies adopted by MNCs and the emergence of contract manu- facturers; the characteristics and capacities of firms to face these changes; and government policies supporting firms' efforts. Direction of Dynamic Comparative Advantage The rapid industrialization of East Asian countries and their success at ex- porting have transformed the market for a host of manufactures. As the ef- ficiency of production has risen and firms have expanded capacity, prices have dropped steadily and products once aimed at a limited number of high- income consumers have become commodities now sold on mass markets. For products ranging from bicycles to computer memory, entry barriers have fallen, supply elasticity has risen substantially, and, as a consequence, profit margins have narrowed.11 Because wage costs are rising in the middle- income economies of East Asia, the pressure from lower-income countries in the region and in South Asia is growing. This is forcing producers in middle-income economies either to upgrade their offerings and differen- tiate their existing range of products, so as to accommodate higher costs and at the same time enhance earnings, or to diversify into other products and services where the competition is less severe. One widely pursued strat- egy, discussed by Dieter Ernst (chapter 3 in this volume), is for economies to upgrade to the assembly of high-technology products such as mobile handsets or digital cameras. However, most of the East Asian economies are already adept at assembling electronics equipment, and a shift to high- technology items does not necessarily increase local value added by much. Moreover, earnings only increase when the growth of supply for a more so- phisticated line of products lags behind demand. A large and continuous increase in value added that is more likely to result in higher earnings is be- coming tied to steady technological advance in certain commodity prod- ucts and in complex products or services (Hobday 1998)12 These include 11. By some accounts, excess capacity in the manufacturing sector is widespread because of high levels of investment in the 1990s, scale economies with associated falling marginal costs, and the high price of exit for firms with capital-intensive assets (see Crotty 2002). 12. For example, the profitability in the flat-panel display business has come to depend on the ability to move from one generation of manufacturing facilities to the next generation with greater production capacity and the technology to produce panels of larger size. 8 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the machinery needed to produce the parts and components and to assem- ble the manufactures pouring out of East Asian factories, the conceptual- ization, design, integration, and maintenance of complex products such as plant equipment, and the supply of or high value­added services, most of which are now tradable and can command global markets. For more than a decade, upgrading has been the mantra of East Asian middle-income economies. As they have collectively struggled to reach the next rung of the ladder, economies in the region have often only succeeded in extending the embrace of commodification. Upgrading can be achieved through incremental product or process innovation that can be the source of modest rents until the competition catches up, which can occur in a mat- ter of months. However, when upgrading is based on significant design and product innovation protected by patent rights or the products themselves are difficult to imitate, substantial rents can accrue over several years. Building a reliable base of technological capability is becoming increas- ingly urgent for all those East Asian economies seeking to avoid the trap of low-level growth associated with commodity production. The risk has cer- tainly risen over the past five years, as low-cost producers in China have dramatically expanded capacity and flooded the markets for manufactured commodities. Commodification of many mainstay tradable manufacturers is one reason for building technological capability, but there are others as well, including the changing requirements for participating in global production networks and the emerging threat from multinational contract manufacturers. Evolution of Global Production Networks During the third stage of East Asian development, global production net- works offered efficient and enterprising firms a shortcut to international markets. A firm that was prepared to meet the product specifications and delivery schedules demanded by MNCs and buyers from the industrial countries gained overseas market access and some market intelligence. The firm also stood to benefit from steady demand for its products as well as a variety of technical assistance that could be obtained from client firms to boost skills, quality, production efficiency, and logistics.13 Many firms gravitated toward membership in global production networks for these reasons. If they wanted to penetrate foreign markets and did not have the resources to engage in direct marketing, global production networks offered 13. See, for instance, Humphrey and Schmitz (2002). On the advances in and the importance of logistics, especially for time-sensitive products, see chapter 7. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 9 an attractive intermediate alternative: the opportunity to export at an af- fordable transaction cost. Moreover, links with global production net- works generated two-way flows of information that could be a means of inducing FDI in East Asian firms seeking a dose of capital, a partnership, or a transfer of ownership. Through the 1990s global production networks mediated a large increase in trade within industry and trade among MNCs, their subsidiaries, and affiliates. Since the late 1990s, the criteria for participating in global production networks have become more exacting, especially for first-tier suppliers, because of pressures on the MNCs and the deepening imprint of infor- mation technology on transactions and business processes. In a remark- ably short period of time, the large MNCs have assimilated and harnessed information technology and have become increasingly comfortable in using the new technologies to coordinate dispersed activities, manage the supply chain, synchronize production and marketing, and establish mech- anisms for gathering market and technological intelligence and for shar- ing this information with the constituent parts of the corporation as well as with affiliated firms. Familiarity with information technology and a bet- ter understanding of its potential are inducing firms to seek leaner, often much flatter, organizational structures, to reconfigure and streamline inter-firm relationships, and to embark on innovative trading, process, and information-sharing arrangements with a wide range of business partners. Indeed, information technology is being woven into virtually every element of corporate strategy. Information technology has served to reinforce the efforts of firms to en- hance their competitiveness in a globalizing market environment by stream- lining their own operations, retaining only those promising the highest returns, many of which can be overseas.14 As firms subcontract with spe- cialized suppliers, they are seeking to minimize the number of first-tier sup- pliers and to establish close long-term relationships with them, which can involve acquiring an equity stake or entering into long-term and binding contractual undertakings (Whittaker 2003). Toyota enlarged its equity in its three large parts suppliers--Aishin Seiki, Denso, and Toyota Gosei--in the late 1990s (Hatch 2003). By reducing the number of components used and purchasing from a smaller number of first-tier suppliers, assemblers can lower the cost of production and coordination in addition to simplifying supply chain management. More important, in a marketplace where compe- tition is paced by technological advances, even the largest companies are 14. The global diversification of U.S. companies rose substantially between 1984 and 1997 (Denis, Denis, and Yost 2002). 10 GLOBAL PRODUCTION NETWORKING IN EAST ASIA finding it necessary to share the burden of technology development with suppliers or buyers.15 Moreover, if technological advance is to be a shared activity with significant mutual exchanges of proprietary information among firms, lead corporations must choose and vet their potential partners with some care. Every relationship that assigns a major role in the development of a technology or the supply of a key component (or module) is fraught with risks and with reciprocal obligations. Hence, even as MNCs have deverti- calized and sought to purchase what they once produced in-house from specialized suppliers in competition with one another, they are seeking de- pendable suppliers that can directly support their global operations. MNCs in the automotive sector, for example, are inclined to recruit first-tier sup- pliers from among their traditional suppliers, such as Bosch, Johnson Con- trols, Lear, or TRW, former subsidiaries now hived off as fully fledged independent companies, such as Delphi and Visteon, or entities in which they have acquired a sizable, if not a majority, stake. Most of the first-tier suppliers in Thailand are foreign owned, and this appears to be the trend in Korea as well.16 The imperative of finding trustworthy suppliers able to fulfill exacting requirements is continually sharpened by the worry that technological dependence and sharing with partners and suppliers can erode a firm's tech- nological edge. A lead firm that has transferred technology to a supplier exposes itself to the risk that this supplier will permit a competitor access to the technology as well (the clients of wafer fabrication companies have expressed this concern). Even more threatening is the possibility that a sub- contractor could acquire both the technological and the manufacturing capability to displace the lead firm in its principal product market (a possi- bility that worried Acer's clients). In fact, the mixed experience with technology-sharing arrangements as well as with the ability of subcontractors to meet the worldwide needs of lead firms has underscored the advantage of selective reverticalization. Sony and Toyota have both moved in this direction in an effort to preserve vital 15. Knowledge-sharing networks, technology development partnerships, and research consortia have also proliferated as firms find it increasingly necessary to leverage the research and resources of other firms in order to arrive at desired outcomes. See Buchel and Raub (2002); Dyer, Kale, and Singh (2001); chapter 9 by Nabeshima. Dyer, Kale, and Singh (2001, p. 37) note, "Strategic alliances have become an important tool for achieving sustainable competitive advantage . . . . Currently, the top 500 global businesses have an average of 60 strategic alliances each." 16. These first-tier suppliers have gradually been enlarging their own roles. Johnson Controls, for example, has widened its functions to include cockpit engineering and added new items to those traditionally located in the cockpit, such as cell phones incorporated in the roof liners. Be- yond that, the company also serves as the intermediary that deals with the second-tier suppliers providing parts for its modules ("How to Move" 2001). COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 11 skills. In countries where the local supply base is weak, it also argues for the acquisition of local firms. The most promising avenues, not infre- quently, entail the acquisition of a start-up with an innovative technology or even an established firm with good local production, marketing, or brand assets. As firms introduce these changes, they are altering their organizational structures and networking relationships. Some firms are deverticalizing and outsourcing components, subsystems, or services. Others are adding new production activities and entering new fields frequently related to their core business. These dynamics are continuously heightening the tempo of competition in East Asian markets. But, at the same time, com- petition and the growth of markets are presenting fresh opportunities. Local firms are struggling hard to become first-tier suppliers to the MNCs and thereby earn the higher profits accruing from close longer- term relationships and joint development of technologies. East Asian firms are also competing with foreign component suppliers that have fol- lowed lead MNCs to the region and set up production facilities in close proximity to those of the lead firms (Hatch 2003). Because MNCs are in- clined to channel technology selectively to their own subsidiaries and af- filiates and to demand just-in-time delivery, domestic firms are finding that they must either redouble their efforts at building technological ca- pability (which is a stepping stone to fruitful research alliances) and exploit advances in logistics or risk being squeezed out of the global production networks.17 Inevitably, the more agile East Asian manufacturers, such as Haier and Hyundai, are laying the groundwork for their own global production net- works. Such firms are first establishing a regional presence to solidify a brand image before venturing farther afield. Emergence of Contract Manufacturers Competition has emerged from another quarter as well. Coinciding with the deverticalization of leading MNCs is the growth of multi- national contract manufacturers. Five firms, all based in North Amer- ica (Celestica, Flextronics, Jabil Circuit, Sanmina/SCI, and Solectron), account for the lion's share of total output.18 They have taken over 17. A firm's stock of patents can be the currency needed to acquire technology from others and to enter into partnerships. 18. The contract manufacturer phenomenon is discussed in detail by Sturgeon and Lester (chap- ter 2) and by Ernst (chapter 3). 12 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the tasks of component sourcing, development, engineering, final assem- bly, and, to a limited extent, design for many of the major electronics and telecommunications firms. Contract manufacturers are responsible for myriad products from Microsoft's X-Box to Sony-Ericsson's mobile hand- sets to printers for Hewlett-Packard based on prototypes and design spec- ifications provided by their clients. In many instances, they handle the entire supply chain, culminating in final assembly and delivery to either the client's warehouses or wholesale outlets. Contract manufacturers have grown by acquiring the production facili- ties disposed of by the lead MNCs, by taking over firms in the industrializ- ing economies, and by investing in new plants. Flextronics, the largest, had 87 plants in 27 countries and a turnover of $146 billion in 2002. Their com- petitive strength derives from economies of scale, scope, and geographic reach. Contract manufacturers can quote the lowest prices because of their high turnover; because they are now able to offer a wide range of elec- tronics items, they can switch production from one category of manufac- turers to another; they can pool the inventories of several customers and thereby cut total inventories; and, by maintaining production in close prox- imity to the major markets in North America, Europe, and East Asia, they can spread production over multiple countries and promise timely, often just-in-time, delivery (Lakenan, Boyd, and Frey 2001). More recently, contract manufacturers are venturing into research and design activities with higher added value, possibly in response to the slowing world economy after 2001. This move could tighten their links with some of their clients, which might wish to focus even more narrowly on downstream marketing and brand management. Such capabilities also position the con- tract manufacturers to launch new products, possibly under their own brand name, although such a move risks alienating their key clients. Moreover, in the lead up to the telecommunications bubble in 2000, firms such as Cisco discovered that subcontracting to contract manufac- turers had substantially reduced their flexibility to boost production when demand for certain products surged. Their objectives of quick market pen- etration and growth of market share are likely to be at odds with those of contract manufacturers, which must manage inventory and costs with the utmost care to sustain what are often meager profit margins (Chapter 3 and Lakenan, Boyd, and Frey 2001). In 2002 the multinational contract manufacturers, which have been joined by key East Asian ODMs such as Hon Hai Precision and Quanta, ac- counted for 20 percent of the market for electronics contract manufactur- ing. As they better integrate their expanding global production systems and strengthen their design capacity, the contract manufacturers are likely to be COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 13 a force to be reckoned with.19 The small- and medium-size firms in some of the Southeast Asian economies that lack technological capability and are faced with rising wage costs will need to devise strategies to cope with pres- sure from three sides: from the push by MNCs to reduce the number of first-tier suppliers and to enlarge their respective roles; from the continuing influx of foreign firms that subcontract with the multinational companies and are relocating some of their facilities to reduce costs and service their client operations; and from the regional and multinational contract manu- facturers that are enlarging their economies of scale and scope and their links with MNCs. To meet these challenges, firms in East Asia need to make concerted efforts to build technological capabilities and to grow in size. In this they might be assisted by the size and unwieldiness of the contract man- ufacturers, which are frequently operating 50 to 70 plants in a score or more countries. Firms in East Asia might also have a little time to prepare for the challenge, because, as Dieter Ernst notes in chapter 3, relatively few of the plants operated by contract manufacturers are currently located in East Asia. Building Technological Capability Although, in the medium run, purchase of capital equipment embodying new technology and licensing can be the means of raising the technological threshold, over the longer term, there is no viable alternative to investment in research and development (R&D) by the firm. R&D that facilitates technology transfer and some innovation is becom- ing a necessary stepping stone to enter the first rank of suppliers to MNCs, which now expect such firms to partner with them on the upgrading of ex- isting products and to contribute to the next generation of technologies. Be- yond this, investment in R&D can enable a firm to graduate from the status of an OEM supplier and credibly train its sights on design-based manufac- turing as a prelude to launching its own branded products. The most likely escape from the treadmill of cutthroat competition in markets for com- modities is via cultivating design skills and innovation. By enabling firms to differentiate their products or, in rare instances, create demand for entirely new products with a disruptive technology, research and design can gener- ate rents, providing the financial wherewithal for growth. R&D expenditure is high and rising throughout much of industrializing EastAsia(relativetootherindustrializingregions),andmoreofitisnowcon- centrated in the corporate sector, reflecting a trend observed in the advanced countries as well. The largest gains registered during 1991­2000 were by 19. Many risks associated with the contract manufacturer strategy arise from global market volatility in a low-margin business. These are discussed in chapter 3. 14 GLOBAL PRODUCTION NETWORKING IN EAST ASIA China, which raised R&D spending to more than 1 percent of GDP from 0.6percentandwhosepurchasingpowerparity­adjustedspendingisnowthe third highest in the world after the United States and Japan. China has also been able to increase its scientific publications, while the number of patent applications in China rose 26 percent annually between 1994 and 1999. The incentives offered and the lure of China's market have drawn most of the top 500 MNCs to China and induced them to establish 400 research centers ("China, R&D Spending" 2003; OECD 2003). On a smaller scale, China's efforts are being matched by those of Malaysia and Singapore, which are de- terminedtobecomeinnovativeeconomies.EstimatedsocialreturnstoR&D range from 20 to 40 percent for the OECD countries to 60 percent for middle-income countries and to almost 100 percent for low-income ones ("Does Science Drive" 2000; Lederman and Maloney 2003). This invest- ment is already yielding rich dividends in Korea and Taiwan (China), which rankfifthandthirdamongeconomieswithrespecttoU.S.utilitypatentsreg- istered between 1995 and 2001. These two economies also rank high when measured by the significance of the technologies represented by the patents. However, two-thirds of all patents registered by Korean entities were by the top five corporations, led by Samsung. In Taiwan (China), the five leading corporations registered one-quarter of all patents. Much of the R&D in Japan, Korea, and Taiwan (China) is by the larger firms (Whittaker 2003). Firms that seek to compete and to grow on the basis of innovation need to invest heavily in acquiring research capability and, beyond that, to spot promising innovations, develop them, and find ways to realize and profit from their full market potential. Serial innovation by small companies is rel- atively infrequent. Most small firms are lucky to have a single innovation to their credit, and significant innovations--or truly disruptive innovations-- are rare.20 Hence, where technological capability is underwritten by wide- spread innovation, entry barriers to new firms are low, and market and financial institutions support the growth of the more dynamic firms. Technological dynamism has contributed to the global ambitions of Hyundai, LG, and Samsung and to the acquisition of a favorable brand image. Taiwanese (Chinese) firms have moved to the forefront of ODMs in the electronics subsector and have acquired a regional brand image. A few companies in China, Singapore (for example, Creative Technologies and 20. Geroski, Van Reenen, and Samiei (1996) find that few of the firms they studied in the United States and United Kingdom registered more than one patent or had more than one innovation to their credit. Research on 10 OECD countries shows that between 20 and 40 percent of the growth of total factor productivity is because of gains from the entry and exit of firms, with new innovation and productivity growth generated by new firms offsetting or more than offsetting the effects of exit (Scarpetta and others 2002). On the nature and potential of disruptive innova- tions that give rise to entirely new markets, see Christensen, Johnson, and Rigby (2002). COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 15 Chartered Semiconductor), and Thailand are also attempting to establish themselves through their technological prowess, but these countries still lack sufficient numbers of large firms in a position to sustain sizable re- search programs and to capitalize on their findings. Becoming Bigger, Maybe Better The global market environment now appears to favor firms that can oper- ate on international scale. Such firms can combine the advantages of scale economies and greater learning by doing, with product or process inno- vation by way of in-house R&D combined with inter-firm collaboration. Larger firms are also better positioned to mobilize the finances, whether from their internal or market sources, to introduce process innovations and develop, test, and market new products. Writing on industrial upgrading in Taiwan (China), Amsden and Chu (2003, pp. 2­3) observe, Small-scale firms may be the first in a latecomer country to introduce a world-class foreign technology. But typically, small firms take a backseat in upgrading to national large-scale firms. Whereas small firms in advanced countries may be technological pioneers, small firms in latecomer countries do not fulfill this role because they do not yet have technology at the world frontier. Many are [also] backward in terms of management skills . . . . Both small and large firms in Taiwan (China) moved their manufacturing opera- tions to China, but only large firms had enough resources to maintain prod- uct development facilities in Taiwan, enabling them to upgrade concurrently. Both large and small firms in the electronics sector have R&D, but the con- tent differs. Small firms dominate in the non-electrical machinery industry, but this industry has stagnated in terms of its share of GNP. Even the small suppliers of parts and components to electronics assemblers have been sub- ject to one-stop shopping rationalization as each assembler has demanded a greater number of parts from a smaller number of suppliers, thereby push- ing the importance of bigness down the supply chain. By the second or third generation of industrial development, most start-ups or promising small- scale firms are tied in one way or the other to an existing largefirm.21 In an integrating world economy, the most profitable and growing firms are generally those able to market their products in several of the major in- ternational markets. Such firms are generally among a handful that domi- nate the global markets for a product or service, whether it is the software for mobile handsets (for example, ARM of Cambridge, United Kingdom), 21. Japanese small- and medium-scale enterprises were a less prolific source of innovation from the 1950s through the 1970s. Of 34 major innovations, they were responsible for only two (Hatch 2003). 16 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the machinery for making frozen pizza, or the equipment for producing flat-panel displays (for example, Applied Materials). Large firms that are active on a global scale can leverage the advantages of size and international operations by creating their own global production networks; by acquiring promising innovative firms worldwide to augment their R&D capability and product range; and by investing in overseas subsidiaries that reap the benefits of proximity to important markets. This is the path taken by vir- tually all the leading Western and Japanese firms. It is a model being emu- lated with some success by a small number of Chinese, Korean, and Taiwanese (Chinese) firms.22 In a world where the dynamics of competition are being redefined by innovation, information technology, and business relationships favoring firms that have honed their research and design skills and have the capacity to operate on an international scale, small, nimble firms can certainly survive, but larger firms are more likely to prosper. Firms such as Hyundai, LG, and Samsung from Korea; Haier, Huawei, Legend, and ZTE from China; and Acer, Hon Hai, Mostek, Quanta, Asustek, and Vitelic fromTaiwan (China) are beginning to respond to the changing global environment by seeking an international presence and aiming for the sta- tus and profitability of OBMs. Nevertheless, firms from China and Korea, even the medium-size ones, remain too highly diversified and vertically in- tegrated compared with their rivals from the Western countries. Vertical integration is a means of containing transaction costs when market institu- tions are weak, and diversification is often the only means of growing when competition mounts in existing product lines. But there are limits to such strategies, and it remains to be seen whether East Asian firms can grow and compete without changing their structures and acquiring greater focus ("Samsung Way" 2003). Government Policies To date, the East Asian governments have initiated and supported techno- logical advance through six broad policies: 1. Human capital deepening 2. Creation of publicly financed research institutes 3. Grants, subsidies, and tax incentives for private R&D activities and gov- ernment contracts 22. On balance, the size of even the largest Chinese firms is still relatively small compared with their foreign competition, a point that Nolan and Zhang (2002) and Nolan (2002) have carefully docu- mented. However, a few firms, such as Sinochem, Baosteel, and Shougang Steel, are approaching the size of the multinational companies ("Beijing Fails to Dent" 2003; "Steel Magnolia" 2003). COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 17 4. Technology licensing policies and technology transfer arrangements through FDI in high-tech industries 5. R&D by public sector firms 6. Incentives for information technology. These are discussed in more detail below. Circulation and Creation of Technology Complementing R&D spending is the availability of knowledge workers with the requisite skills and experience. This is vital for the productivity of research endeavors and for the commercialization of findings, a point stressed by Hill (chapter 8), Nabeshima (chapter 9), and Jefferson and Zhong (chapter 10) in this volume. Governments in East Asia have emphasized in- vestment in primary and secondary education for decades. More recently, they have increased the outlay on tertiary-level education and are taking steps to raise its quality. Governments in Japan, Korea, and Taiwan (China) have also come to recognize the desirability of orienting the leading uni- versities toward research, promoting university-business links, and encour- aging universities to patent research findings. Universities in the OECD countries account for close to half of all basic and applied research, whereas in East Asia, the share is lower, with universities in China, for example, accounting for just a quarter of the country's applied research and a third of its basic research (OECD 2002).23 Starting with Japan in the 1950s (and earlier), overseas training has been viewed as a way of supplementing the domestic supply of trained workers, enhancing the quality, and acquir- ing relevant experience. Now several East Asian economies have large diasporas of highly skilled workers, many of whom have started to return as opportunities in their home economy have multiplied. Looking forward, the higher-income countries will need to sharpen their focus on the qual- ity of graduates from tertiary-level institutions and to impart a stronger re- search orientation to the leading universities. Middle-income countries such as Malaysia and Thailand will also need to shift attention and re- sources to building those types of skills that will contribute to innovation and not just manufacturing capability. High- and middle-income East Asian economies have already insti- tuted tax incentives to encourage research by the private sector.24 These 23. By giving researchers higher royalties, U.S. universities have induced an increase in investment (Lach and Schankerman 2003). 24. Wang and Tsai's (2003) research on Taiwan (China) shows that R&D has contributed sub- stantially to the productivity of manufacturing firms. 18 GLOBAL PRODUCTION NETWORKING IN EAST ASIA are supplemented by grants and subsidies for specified activities and by government-financed (and some privately financed) venture capital for high- tech start-ups (Kenney 2004). Singapore, for instance, has earmarked close to $3 billion for the development of its biotech industry. Singapore and Taiwan (China) have, among others, also used government-funded research institutes to absorb new technologies and to serve as the nucleus for an in- dustrial subsector.25 The Industrial Technology Research Institute (ITRI) in Taiwan (China) is an oft-cited example of a public research entity that helped to launch the microelectronics industry and whose expert midwifery was responsible for the birth of both TSMC (Taiwan Semiconductor Man- ufacturing Corporation) and UMC (United Micro-Electronics Corpora- tion), which are among the two leading foundries in the region (see Mathews and Cho 2000). However, public research institutes in China, Korea, and Japan have been less productive in generating both basic research and com- mercially viable innovations. It is likely that basic research might continue to require public funding, but whether it goes to public institutes or is chan- neled to universities and private research centers is a matter of debate. Com- mercially oriented research might be left more profitably to private industry, with the state providing suitable financial inducement. The examples of early successes such as ITRI in Taiwan (China),26 the Korea Advanced Institute of Science and Technology (KAIST) in Korea, and the Karolinska Institute in Sweden are few in number, and the experience was not repeated in the 1990s. In fact, the trend in Europe and the United States is for the state to scale down its funding for public research institutes, which can be small and iso- lated, in favor of support for the private sector and the university system. Avenues for technology transfer through licensing and trade are well marked, and firms in East Asia are savvy enough to seek out the technolo- gies they require and to negotiate reasonable terms, whether through trade or licenses. Arguably, there is less need and also less scope for governments to involve themselves with specific technologies and with the details of con- tractual terms, as happened in the 1960s and 1970s. Markets are more com- petitive, participants are more experienced, and the rules of the game have changed. Debate continues to swirl around the modalities of technology trans- fer through FDI. Most researchers would agree that FDI generally leads to technology transfer through a variety of channels, more in some 25. As Frantzen (2000) and others have shown, for non-G-7 countries, assimilating research done elsewhere initially can be more important for technological advance than domestic research. 26. According to Hill (chapter 8 of this book), ITRI has working relations with 20,000 companies. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 19 industries than in others. But researchers differ on the degree to which FDI results in horizontal spillovers to other firms in a subsector versus vertical transmission of technology, mainly to subsidiaries and affiliates of foreign firms. Blalock and Gertler (2003) show that vertical trans- mission dominates, and their findings are supported by other research. However, evidence of horizontal spillovers--in particular, through the cir- culation of workers--is not lacking, and we have certainly not heard the last word on this.27 What can be stated with some assurance is that FDI will remain advantageous for East Asian economies in the future, just as it was in the past, especially for Malaysia, Singapore, and Taiwan (China), because it is a vehicle for transferring a range of technologies and capital. Moreover, it offers avenues for linking firms to global production net- works. Even in countries, such as China, where the aggregate supply of capital more than matches demand, financial market distortions impede the cross-provincial flow of resources and constrain the access to funding. For many firms, FDI is often the only means of overcoming the dearth of financing (Huang 2003). Recognizing the gains from FDI--in particular, its potential contri- bution to move up the value chain--East Asian economies are continu- ing to engage in contests to attract FDI. They are offering relatively closely matched fiscal incentives, together with the infrastructure and fa- cilities of special economic zones and increasingly liberal trading arrange- ments that facilitate intra-industry trade and transactions that sustain global production networks. Even Korea, which historically viewed FDI with a skeptical eye, is actively wooing foreign investors and preparing to carve out special economic zones. In exchange, East Asian economies can demand technology transfer, set thresholds for local content, and require that some fraction of the output of joint ventures or foreign-owned firms be exported. Market Institutions and Incentives Most East Asian countries have progressed well beyond the basics of technology policy. They have a sophisticated understanding of the rel- ative importance of human capital, the role of R&D in private compa- nies and universities, and the role of FDI and trade. Countries have learned from their own experiences, from that of their neighbors and 27. See Sjöholm (1997), who notes that technology transfer to local Indonesian firms tends to be greatest when markets are highly competitive. 20 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the advanced countries, and from numerous contacts with the inter- national business community. To a greater or lesser extent, they are dis- engaging from old-style industrial policies and concentrating on augmenting the supply of skills, research, foreign capital flows, technology transfers, and trade. The direction being taken is broadly appropriate given the current state of knowledge. But the early returns still lag behind expectations, sug- gesting that a new mix of policies might be needed. It is far from obvious that development can be accelerated beyond a point through greater outlay of resources. Instead, East Asian economies that are eager to enhance their technological capability may need to combine invest- ment in technological capacity with greater emphasis on reforming regula- tory policies, including crafting and enforcing competition policies that will eventually be harmonized across the region; building intellectual property rights regimes that are credible and promote innovation; permitting firms to engage in mergers and acquisitions (M&A) subject to the rules defined by competition policy; and encouraging international research collaboration through links among companies, universities, and research institutes. In ad- ditiontonationalpolicies,thereisamajorroleformunicipalpoliciesthatcan influence the location of firms--in particular, firms producing high-value- added goods and services, which are the largest employers of knowledge workers. Most East Asian countries have legislated such policies but have yet to put much energy into implementing them. Familiarity with information technology and its use by businesses and for research has risen quickly throughout East Asia in response to demands of the leading electronics and automotive companies and aided by government incentives. This trend could be reinforced by actions on two fronts. First, firms need to be more proactive in acquiring infrastructure and building in- formation technology skills. This can only enhance their competitiveness as members of global production networks; a strong base of information tech- nology, continuously upgraded through investment, is essential for a firm with the ambition to become a global ODM or OBM. Second, govern- ment regulatory policies toward the telecommunications sector and In- ternet access affect the supply of services, cost, quality, technological advance, and access, as analyzed by Crandall and Litan (2002) and by Yusuf and Evenett (2002). Competition can be a powerful spur to innovation, while an active mar- ket for corporate control through M&A not only reinforces the competitive pressures emanating from the market and trade in goods but also enables successful firms to grow and reap the benefits of size. Research on compe- tition policy and its harmonization across a region suggests that there are significant longer-term gains to be realized from defining and policing rules COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 21 governing firm behavior. Harmonization, in turn, can facilitate cross- border transactions, including M&A, and diminish transaction costs. Glob- alization has enhanced the attractiveness of institutional uniformity in these areas, and whether or not the trend toward global integration persists, a more limited goal of creating common regional institutions itself could be worthwhile. Until recently, there was little demand from the business community for a competition policy and scant interest on the part of governments in in- troducing or pursuing such a policy. Similarly, the business culture through- out East Asia is resistant to the idea of unbridled M&A, as owners can be possessive about the companies they have set up, worry about a loss of face if their firm is taken over, and are often at pains to avoid upsetting patterns of market coexistence and imperiling settled relationships. This has influ- enced growth in the size of East Asian firms from middle-income countries, limited the consolidation of firms within industrial subsectors into more optimally sized units, and encouraged growth through diversification. As the region integrates with the global economy and faces stern com- petition, reducing the institutional obstacles to M&A and reconsidering the resistance to foreign M&A might initiate the shakeout of industries and the emergence of a few large firms able to challenge MNCs on the global market.28 At the same time, competition policy should seek to ensure that the barriers to entry are kept low so as to induce a steady flow of start-ups that bring promising new producers, innovations, and entrepreneurial en- ergy to the industrial sector. In China, as Steinfeld points out in chapter 6 of this volume, provincial and county authorities routinely prevent M&A of a local firm by a firm from another province so that their administrative control over industry is not diluted. These very same authorities on occa- sion force quite irrational mergers among firms under their jurisdiction and also curtail the entry of firms from other parts of the country that might pose a competitive threat. Parallel actions pertaining to intellectual property and the formation of industrial clusters in urban areas can reinforce such institutional stops. All of the middle- and high-income economies of East Asia are signatories of the WTO-TRIPS (Trade-Related Aspects of Intellectual Property Rights) agreement and are committed to the rules governing intellectual property. As these countries are now seeking to buttress their future competitive posi- tion by relying more extensively on innovation, especially in the high-tech 28. On December 2, 2002, China introduced new regulations making it easier to take over listed companies ("China, M&A Rules" 2002). Similarly, attempts to deregulate markets are under way in Japan and Korea ("Japan Talks Tough" 2003). 22 GLOBAL PRODUCTION NETWORKING IN EAST ASIA areas, defining and securing intellectual property take on far greater impor- tance than was the case when they were firmly in the imitative phase of de- velopment. Research in the United States indicates that, by protecting intellectualproperty,governmentsprovideincentivestoinnovationsthrough a perceived patent premium (Arora, Ceccagnoli, and Cohen 2003). This is especially valuable in those areas where research and testing are costly, as in biotechnology and pharmaceuticals. Securing intellectual property rights is also of increasing importance for the "creative industries,"29 including me- dia, publishing, and software, as the copying and mass production of music, video games, and computer programs have become easier, putting great pressureonthecompaniesresponsiblefortheoriginalcontent.Nevertheless, stronger intellectual property rights alone will not galvanize innovation--in fact, Korea's experience suggests that competition more than the intellec- tual property regime is largely responsible for the rise in innovation (Luthria and Maskus 2004). However, as East Asian countries move to strengthen the rules governing intellectual property and their legal enforcement,30 these steps, in conjunction with measures to increase basic and applied research efforts, could deepen technological capability and permit East Asian coun- tries to earn higher returns by differentiating manufactured products and entering new and more profitable product niches. Clusters and Urban Policy The innovative impulse is likely to be stimulated further if East Asian economies also redouble their efforts to create an urban environment con- ducive to the clustering of firms. Global production networks flourish when the affiliated firms can cluster in urban areas richly supplied with producer services (see Bresnahan and others 2001; Yusuf with others 2003; chapters 3 and 10 of this volume). Such clustering can be especially ad- vantageous for high-tech and creative industries that derive benefits from local and specialized markets, from spillovers of tacit knowledge, from technological collaboration, from just-in-time delivery, and from the abil- ity to access the services of financial, accounting, legal, and logistics firms among other providers. In the developed countries, the dynamic clusters of innovative firms are found in a relatively small number of cities that are able to attract the tal- 29. The rising significance of these industries for East Asian cities is discussed in Yusuf and Nabeshima (2004). 30. The parallel institutional strengths of the legal system will be of critical significance, as em- phasized by Steinfeld in chapter 6 and by Yusuf with others (2003). COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 23 ented knowledge workers and the type of firms that rely on their skills.31 These cities offer an attractive physical environment, efficient infrastruc- ture, convenient transport links with other countries, and reliable utilities. Beyond that, cities that are hubs of innovation have--by dint of regulation, investment in physical assets, and civic action--put in place cultural and so- cial amenities that greatly enhance the appeal of what otherwise would be just efficient cities. As an example of what can be done to improve efficiency, Singapore's investment in port infrastructure, a skilled work force, customs organization, and information systems has made it possible for importers of electronics components to receive delivery six hours after the product arrives at a port or an airport. Similar efforts are under way in Shanghai. As East Asian economies seek to improve their innovation capability in the interests of competitiveness, they will have to struggle to retain foot- loose high-tech firms. The urban dimension of policymaking will be no less critical than the rest. A credible regime of intellectual property rights will not contribute to the government's objectives if the leading cities within the country cannot attract and retain firms in the sectors with the most prom- ise for growth. Hence, policies are needed to draw firms to specific urban locations in the hope of creating rapidly growing clusters of dynamic firms (Florida 2002). Innovation policy is thus becoming more complex. In a glob- alizing world, how cities evolve can determine whether and where pools of skilled workers collect and can influence the location decisions of firms that will decide tomorrow's economic performance. OVERVIEW OF THE VOLUME The contributors to this volume take us deep into the dynamics sketched above, starting with global production networking, then moving to the closely linked transformation sweeping the logistics sector, and then coming to technological change, which is the force now determining the competi- tiveness of firms, defining production relationships, and, increasingly, shap- ing the industrial geography of the East Asian region. As Timothy Sturgeon and Richard Lester show in chapter 2, with reference to the electronics and automobile industries, the lead firms dominating international production networks are adopting a global approach to their business operations and have begun to consolidate manufacturing in fewer locations, generally in 31. Contract manufacturers such as Flextronics, much like the auto assemblers, prefer to have their suppliers co-locate in clusters or business parks. Hill observes in chapter 8 that a quarter of Singapore's labor force is foreign and an even higher proportion is foreign born. 24 GLOBAL PRODUCTION NETWORKING IN EAST ASIA countries where potential market growth is greatest. The lead firms have focused their own activities on areas such as research, design system architec- ture, supply chain management, integration, and marketing. Such reorien- tation is proceeding hand in hand with the outsourcing of the production and often the design of entire modules to a small number of trusted suppli- ers. These suppliers, in turn, purchase their components from many second- tier producers, which are forced to compete fiercely for market share. Sturgeon and Lester observe that commodification of products and sharp competition are intensifying pressures on East Asian manufacturers and driving down profits. Moreover, the emergence of North American­based contract manufactures with globally distributed facilities and substantial economies of scale and scope threatens to squeeze the smaller regional pro- ducers even further. If East Asian firms in the electronics and auto industries are to grow and enhance their profitability, they will need to vie for the role of first-tier suppliers for lead firms, to operate on a global scale, and, in cer- tain instances, to co-locate plants near the facilities of lead assemblers. De- veloping research and original design capability can further strengthen competitiveness, enable firms to take responsibility for entire modules, and eventually make a transition to original brand manufacturing on a regional or global scale. Sturgeon and Lester note that some East Asian firms have entered the ranks of first-tier electronics suppliers with ODM capacity. A select few, such as Samsung, have emerged as OBMs, but the vast majority remain second- and third-tier producers of standardized commodities. What is required for such firms to upgrade is the central theme of Dieter Ernst in chapter 3. He discusses the increasing role of contract manufac- turers in the global production networks, the integration of Chinese and Malaysian companies into these networks, and how digital information sys- tems have facilitated the process. But Ernst also underscores the challenges faced by the contract manufacturers as well as gaps in technological capa- bility between East Asian firms and their foreign competitors, a gap that is highlighted by the continuing heavy dependence on imported inputs. Al- though investment by the Malaysian authorities in information technology infrastructure can assist in the upgrading process, the principal bottleneck is the shortage of skilled workers and research personnel. Ernst concludes that, by easing skill shortages, the state can shift industry along the value chain and heighten technological capability. Although industrial upgrading can help firms in countries such as Malaysia to remain a part of global production networks and possibly even improve their ranking as suppliers of components, the impending regional consolidation of the automobile and electronics industries will be painful for some producers. In chapter 4, Richard Doner, Gregory Nobel, and John Ravenhill describe the dispersed development of the auto industry across COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 25 East Asia behind tariff barriers that are now being dismantled. The freeing of trade is reinforcing the efforts of multinational auto companies to ration- alizetheirproduction,tocutcostsbyconcentratingonahandfulofplatforms to be marketed worldwide, and to rely on a few hundred first-tier suppliers for parts and modules. Doner, Nobel, and Ravenhill visualize an increasing concentration of the East Asian auto industry in three hubs. Korea's auto sec- tor, the most advanced in East Asia after Japan, will benefit from strong do- mestic demand, expanding exports, and investment by multinational companies that have entered the business of assembling and producing auto parts, in the process buying up many local firms. A second hub is emerging in Thailand, which has used fiscal incentives to encourage the manufacture of light trucks and has since attracted investment in the assembly of sedans and the production of auto parts. Major auto companies are likely to use Thailand as the base from which to export automobiles and parts to other countries in Southeast Asia. This is likely to affect the development of the auto industry in Indonesia, Malaysia, and the Philippines, each of which has built up substantial production capacity. Malaysia, in particular, has aggres- sively used tariff and other incentives to promote a domestically owned auto industry that will have difficulty competing against imports once the tariff barriers are scaled down. A third hub is taking shape in China, driven by the rapid expansion of the domestic market, massive investment in capacity by leading automakers as well as by local producers, and the efforts by the Chinese authorities to localize parts and production and to assimilate foreign technology. With de- mand in China likely to grow over the foreseeable future, automakers such as General Motors and Toyota will be relying on their sales in China and ex- ports from their factories located there to sustain their profitability. Thus the new geography of regional production and marketing is likely to favor firms from the Northeast Asian countries and Thailand that are part of automotive global production networks. Chapter 5 by Ken'ichi Takayasu and Minako Mori and chapter 6 by Edward Steinfeld explore the implications for Chinese and Thai firms against the backdrop of industrial changes occurring in these two countries. Takayasu and Mori note how the major Japanese and Western assemblers have been developing a base of suppliers to reduce their costs and ensure just-in-time delivery. Most of these suppliers, especially first-tier ones, are subsidiaries of foreign firms that work closely with lead assemblers in their home country. Thai firms figure among the second-tier suppliers and have been slow to acquire the technology and design skills that would enable them to undertake the manufacture of higher-value parts and to work more closely with assemblers and first-tier suppliers in R&D and in design. Thai producers, much like firms in Indonesia and Malaysia, are constrained by 26 GLOBAL PRODUCTION NETWORKING IN EAST ASIA a shortage of workers with the technical and managerial skills that could as- sist with the assimilation of design and production technologies. These scarcities have been apparent for some time, and it will require concerted action by both government and industry to build a stronger base of Thai auto component suppliers that can participate in the rapid expansion of the auto sector in Thailand and--through global production networks-- in other countries as well. However, skills are not the only bottleneck. Thai firms have also been slow to acquire expertise in information technology and the infrastructure that is now essential for effective participation in the finely tuned business of auto manufacturing. For all the leading assemblers, a careful orchestration of the supply chain is vital to achieve efficient and profitable production and avoid excess outlay on inventories. In some respects, the situation in China is not dissimilar. Over the past decade, Chinese firms have engaged in trade on a vastly larger scale than their counterparts in Thailand as a result of FDI in export-oriented indus- try and through the initiative of local enterprises. This has multiplied their links with global production networks for a wide range of products. How- ever, most Chinese firms remain relatively small, and typically they work with codified technologies to produce standardized commodities with the minimum of design and R&D inputs. Larger firms, such as Huawei, Legend, and ZTE, have moved up the value chain and are beginning to penetrate international markets under their own brand name. But even these firms continue to operate in the medium-tech range, producing standardized products or modules and selling these for minimal profits in highly competitive markets. In order to grow and to increase their profits, the larger Chinese firms are still relying on vertical integration and product diversification. This has yieldeddecentshort-termdividends;however,Steinfeldconcludesthatinthe longer run such firms will have to imitate their foreign competitors, develop proprietary technologies that generate higher rents, and be able to market productsorservicesstraddlinganumberofindustries.Smallfirms,andlarger ones, will need to enhance their technological capability in the interests of competitiveness and profitability not just in products and processes, but also in logistics and supply chain management. These topics are discussed in the latter part of the volume, starting with chapter 7 on logistics. The massive increase in East Asian trade and production networking has led to a gap in logistics capacity across the region, although more so in China and Vietnam than in Southeast Asia. As Trevor Heaver shows in chapter 7, it is not only the physical infrastructure that needs to be augmented but also the institutions, skills, procedures, and technologies. Producers and buyers order more frequently in smaller lots and expect to track their shipments so COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 27 that they can synchronize deliveries with their own production schedules and with a minimum of warehousing. They also expect to conduct and set- tle more of their transactions electronically. Thus the logistics system in East Asia must cope with an enormous increase in volume, while at the same time it must adapt to the demand for speed in the delivery of time-sensitive items often shipped in small lots. Economies such as Singapore have responded by expanding their ports, airports, and information technology facilities to accommodate the new traffic and to facilitate electronic clearance of goods through customs. In Hong Kong (China), Malaysia, and Singapore, the interfaces between dif- ferent transport modes are becoming easier to negotiate, greatly facilitating multimodal transport. Shipping has improved as a result of efficiencies in customs clearance and freight forwarding and also the emergence of firms that monitor, consolidate, and manage freight movement for their clients. Firms such as APL Logistics and Maersk Logistics now provide integrated logistics services that are further reducing the transaction costs for firms and helping to multiply as well as cement networking arrangements within and beyond the East Asian region. Progress has been slower in China, although port and airport capacity has been rapidly augmented in the principal gateway cities such as Shang- hai and Tianjin and in the Pearl River Delta area. China's deficiencies are most serious in three areas. First, multimodal transport, especially the in- land surface transport facilities, is inadequate, and the interfaces between transport modes confront shippers with numerous problems. Second, for- warding, consolidation, and other logistics-related services are also defi- cient, and Chinese customs clearance falls well short of the standards and technology attained by Hong Kong (China) and Singapore. Alongside these weaknesses, China's information technology infrastructure and the skills to effectively assimilate information technology with logistics are still relatively underdeveloped. Nevertheless, an amazing amount of progress has been made during the past 15 years in China and throughout the re- gion in building physical infrastructure and modernizing logistics services to support the growth of commerce. In Southeast Asia continuing invest- ment and stronger competition should further improve services, but com- petition will also put pressure on providers to pare costs and continuously upgrade technology. In China and in Vietnam the ground to be covered is greater, but the mo- mentum, as Heaver shows, is strong. Infrastructure--both soft and hard-- is being created, and the use of information technology for production, networking, and logistics is widening steadily. Moreover, with accession to the WTO, China is increasing access to foreign logistics suppliers, and this, 28 GLOBAL PRODUCTION NETWORKING IN EAST ASIA coupled with policy reform and the emergence of domestic firms, should bring about a progressive upgrading of the logistics system. The centrality of innovation for industrial progress under conditions of globalization is the theme of the remaining three chapters. In chapter 8, Hal Hill reviews the building of innovation capacity in Indonesia, Korea, Malaysia, Singapore, and Taiwan (China) and draws a number of lessons. He indicates that economies, such as Singapore and Taiwan (China), that are at the forefront with respect to the development of domestic innovation systems and have integrated most tightly with global production networks have adopted similar policies. The two economies pursued macroeconomic stability, openness, and the building of a modern transport and communi- cations infrastructure. This was combined with aggressive efforts to attract investment by MNCs, which created a base of export-oriented industries and offered access to new products and process technologies. In both economies, the assimilation and leveraging of technology to move up the value chain were promoted by the elastic supply of skills and by well-funded research institutes. These were the outcome of industrial policies that fo- cused increasingly on innovation systems rather than on targeting manu- facturing industries, as was the case in Indonesia, Korea, and, to a lesser extent, Malaysia. Hill observes that Korea sacrificed some innovation capa- bility by discouraging FDI, a move that also limited the integration of Korean firms into global production networks. Both Indonesia and Malaysia poured substantial resources into major industrial projects that have yielded few technological gains. As the competitive pressures from Chinese ex- porters intensify, other East Asian economies will need to strengthen their innovation capability by building technical skills, stimulating re- search effort, and pulling FDI into higher-technology activities with greater value added. The complementarity between the supply of skilled workers, FDI, and technology transfer is the principal message of Kaoru Nabeshima's wide- ranging survey in chapter 9. After carefully examining the various conduits for technology transfer such as capital investment, licensing, and FDI, Nabeshima concludes that investment in human capital and R&D by lo- cal firms--although geared more toward learning--are necessary to ab- sorb technology from overseas, to maximize the spillover benefits from FDI, and to start the cycle of domestic innovation. In the final chapter in the volume, Gary Jefferson and Zhong Kaifeng fo- cus on the determinants of innovation at the level of the firm based on data fromasurveyof1,826infirmsin11citiesinsevenEastAsianeconomies.The chapter also assesses the relative innovativeness of individual cities. Jefferson COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 29 and Zhong's empirical estimates validate many of the findings reported in earlier chapters. For example, they show that R&D capability and associ- ated firm performance are explained by human capital intensity and the institutional setting, especially the level of competition and the degree of international exposure. The survey findings further show that firms benefit from clustering in cities where this improves access to information tech- nology and physical infrastructure and increases networking relationships. Among the cities surveyed, Seoul has the highest human capital intensity, but also the lowest degree of international exposure and FDI, which is consistent with Hill's observations. Kuala Lumpur, Singapore, and Manila are ranked lower in terms of human capital but are more open and enjoy the advantages of higher levels of FDI. Among Chinese cities, Guangzhou and Shanghai lead the field in terms of firm performance and openness, with Chengdu and Tianjin trailing behind. The performance of firms in Shanghai reinforces the key theme of this volume: in an integrat- ing world environment, firms in more open economies that benefit from FDI and are linked to global production networks have an edge over other firms. Beyond this, Jefferson and Zhong's research also reinforces the point made repeatedly in earlier chapters that the winners from inter- national networking are the firms that rise to the uppermost tiers by har- nessing research, design, and production skills and acquiring innovation capability. For firms that develop such capability, global production net- works can provide the route to wider markets and larger profits. In sum, the message of this volume is a simple one. The freeing of world trade, the commodification of many manufactured products, and the chang- ing relations between lead firms and others participating in global produc- tion networks are forcing firms to compete more on the basis of innovation, the effective harnessing of logistics, and the assimilation of information tech- nology. Success ultimately rests on the initiative of firms themselves, while government policies strongly influence the supply and quality of human capital, the dynamism of the institutional environment, and the efficiency of the physical infrastructure. Export-led growth throughout much of industrializing East Asia was spearheaded by small- and medium-size firms, and these will remain a key source of industrial vigor. However, the experience of a number of eco- nomies such as Finland, Korea, Sweden, Switzerland, and Taiwan (China), for example, points insistently toward the contribution of large firms to innovation as well as to the branding and marketing of products on a global scale. An environment that is conducive to the growth of national firms that can compete against the MNCs on world markets, match their 30 GLOBAL PRODUCTION NETWORKING IN EAST ASIA innovative capability, and vie with them in creating global production networks would be part and parcel of a development strategy pegged to technological advance. Although competitiveness will be determined by the performance of firms, as Doner, Nobel, and Ravenhill note in chapter 4, converting "strate- gic options into reality [will depend] on policies, institutions, and politics." As described above, the supply of human capital, which is the main input into the production of technology, is determined in East Asia largely by gov- ernmentpolicy.GovernmentpolicyalsoinfluencestheexpenditureonR&D by firms, universities, and research institutes. Regulatory policies of the state determine the access to information technology and its assimilation into the mainstream of business activities. Moreover, public regulatory policies-- central and municipal--have a large hand in determining the growth of dynamic industrial clusters in key urban areas. Many layers of regulatory policies complement significant institutional measures. In this chapter, I have noted three sets of institutions that govern the innovation of firms: M&A activity, intellectual property rights, and venture capital. They are by no means the only ones. Indeed, one can think of a host of others that im- pinge on innovation, albeit more indirectly and with less force. However, these three certainly deserve prominence and are clearly at the center of on- going efforts to develop technology in East Asia. In the coming decades, East Asian countries will confront a far more de- manding global market environment. Their ability to capitalize on the op- portunitiesinherentinthisenvironmentwillrestinnosmallpartonacquired technological capability and on how East Asian firms enlarge the gains from participating in global production networks. The East Asian firms have shown that they are fast learners. If they retain this knack, their future could be as bright as it has been for the past three decades. REFERENCES The word processed describes informally reproduced works that may not be commonly avail- able through libraries. Amsden, Alice H., and Wan-Wen Chu. 2003. Beyond Late Development. Cambridge, Mass.: MIT Press. Arora, Ashish, Marco Ceccagnoli, and Wesley M. Cohen. 2003. "R&D and the Patent Premium." NBER Working Paper 9431. National Bureau of Economic Research, Cambridge, Mass. Processed. Arteta, Carlos, Barry Eichengreen, and Charles Wyplosz. 2001. "When Does Capital Account Liberalization Help More Than It Hurts?" NBER Working Paper 8414. National Bureau of Economic Research, Cambridge, Mass. Processed. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 31 "ASEAN: Free Trade Fears." 2002. Oxford Analytica, August 13. "Autos and AFTA." 2002. Business Asia, August 26. "Beijing Fails to Dent Steel Industry's Rapid Expansion." 2003. Financial Times, October 30. Bhagwati, Jagdish. 2004. In Defense of Globalization. New York: Oxford University Press. Blalock, Garrick, and Paul Gertler. 2003. "Technology Acquisition in Indonesian Manu- facturing." Paper prepared for East Asian Prospects Study. World Bank, Washington, D.C. Processed. Bourguignon, François, and others. 2002. Making Sense of Globalization. CEPR Policy Pa- per 8. Paris: Centre for Economic Policy Research for the European Commission Group of Policy Advisors. Bresnahan, Timothy, Alfonso Gambardella, Anna-Lee Saxenian, and Scott Wallsten. 2001. "Old Economy Inputs for New Economy Outcomes: Cluster Formation in the New Silicon Valley." SIEPR Policy Paper 00-43. Palo Alto, Calif.: Stanford Insti- tute for Economic Policy Research. Buchel, Bettina, and Steffen Raub. 2002. "Building Knowledge--Creating Value Net- works." European Management Journal 20(6, December):587­96. Cairncross, Francis. 2002. The Company of the Future. Boston, Mass.: Harvard Business School Press. "China, M&A Rules to Entice Investors." 2002. Oxford Analytica, October 31. "China, R&D Spending." 2003. Oxford Analytica, November 20. Christensen, Clayton M., Mark W. Johnson, and Darrell K. Rigby. 2002. "Foundations for Growth: How to Identify and Build Disruptive New Businesses." MIT Sloan Manage- ment Review 43(3, Spring):22­31. Crandall, Robert, and Robert Litan. 2002. "The Internet, Telecommunications, and Eco- nomic Growth in East Asia." Paper prepared for East Asian Prospects Study. World Bank, Washington, DC. Processed. Crotty, James. 2002. "Why There Is Chronic Excess Capacity." Challenge 45(6):21­44. Cutler, Harvey, David J. Berri, and Terutomo Ozawa. 2003. "Market Recycling in Labor- Intensive Goods, Flying-Geese Style: An Empirical Analysis of East Asian Exports to the U.S." Journal of Asian Economics 14:35­50. Denis, David J., Diane K. Denis, and Keven Yost. 2002. "Global Diversification, Industrial Diversification, and Firm Value." Journal of Finance 57(5):1951­79. Dicken, Peter. 2003. Global Shift: Reshaping the Global Economic Map in the 21st Century. London: Guildford Press. "Does Science Drive the Productivity Train?" 2000. Science, August 25. Dyer, Jeffrey H., Prashant Kale, and Harbir Singh. 2001. "How to Make Strategic Alliances Work." MIT Sloan Management Review 42(4):37­43. Edison, Hali J., Michael W. Kleing, Luca Ricci, and Torsten Sloek. 2002. "Capital Account Liberalization and Economic Performance: Survey and Synthesis." NBER Working Paper 9100. National Bureau of Economic Research, Cambridge, Mass. Processed. Eichengreen, Barry, and David Leblang. 2003. "Capital Account Liberalization and Growth: Was Mr. Mahathir Right?" NBER Working Paper 9427. National Bureau of Economic Research, Cambridge, Mass. Processed. Feinberg, Susan E., and Michael P. Keane. 2003. "Accounting for the Growth of MNC- based Trade Using a Structural Model of U.S. MNCs." University of Maryland, 32 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Robert H. Smith School of Business; Yale University, Department of Economics. Processed. Florida, Richard. 2002. The Rise of the Creative Class: And How It Is Affecting Work, Leisure, Community, and Everyday Life. New York: Basic Books. Frantzen, Dirk. 2000. "Innovation, International Technological Diffusion, and the Changing Influence of R&D on Productivity." Cambridge Journal of Economics 24:193­210. Gereffi, Gary, and Olga Memedovic. 2003. "The Global Apparel Value Chain: What Prospects for Upgrading by Developing Countries?" United Nations Industrial Devel- opment Organization, Vienna. Available at www.unido.org. Processed. Geroski, Paul A., John Van Reenen, and Hossein Samiei. 1996. How Persistently Do Firms Innovate? Discussion Paper 1433. London: Centre for Economic Policy Research. Hallward-Driemeier, Mary, Giuseppe Iarossi, and Kenneth Sokoloff. 2003. "Exports and Manufacturing Productivity in East Asia: A Comparative Analysis of Firm-Level Data." NBER Working Paper 8894. National Bureau of Economic Research, Cam- bridge, Mass. Processed. Hanson, Gordon H., Raymond J. Mataloni Jr., and Matthew Slaughter. 2003. "Vertical Pro- duction Networks in Multinational Firms." NBER Working Paper 9723. National Bureau of Economic Research, Cambridge, Mass. Processed. Hatch, William. 2003. "Japanese Production Networks in Asia: Extending the Status Quo." In William W. Keller and Richard J. Samuels, eds., Crisis and Innovation in Asian Tech- nology. Cambridge, U.K.: Cambridge University Press. Held, David, Anthony McGrew, David Goldblatt, and Jonathan Perraton. 1999. Global Transformations. Palo Alto, Calif.: Stanford University Press. Henry, Peter Blair. 2003. "Capital Account Liberalization, the Cost of Capital, and Eco- nomic Growth." NBER Working Paper 9488. National Bureau of Economic Research, Cambridge, Mass. Processed. Hobday, Michael. 1998. "Product Complexity, Innovation, and Industrial Organization." Research Policy 26(6):689­710. "How to Move Gently Downstream." 2001. Financial Times, October 16. Huang, Yasheng. 2003. Selling China: Foreign Direct Investment during the Reform Era. New York: Cambridge University Press. Humphrey, John, and Hubert Schmitz. 2002. "How Does Insertion in Global Value Chains Affect Upgrading in Industrial Clusters?" Regional Studies 36(9):1017­27. "Japan Talks Tough about Closing the Gap." 2003. Financial Times, November 25. Kenney, Martin. 2004. "Venture Capital Industries." In Shahid Yusuf, M. Anjum Altaf, and Kaoru Nabeshima, eds., Global Change and East Asian Policy Initiatives. New York: Ox- ford University Press. Klein, Michael W. 2003. "Capital Account Openness and the Varieties of Growth Experi- ence." NBER Working Paper 9500. National Bureau of Economic Research, Cam- bridge, Mass. Processed. Lach, Saul, and Mark Schankerman. 2003. "Incentives and Invention in Universities." NBER Working Paper 9727. National Bureau of Economic Research, Cambridge, Mass. Processed. Lakenan, Bill, Darren Boyd, and Ed Frey. 2001. "Why Cisco Fell: Outsourcing and Its Peril." Strategy+Business 24:54­65. COMPETITIVENESS THROUGH TECHNOLOGICAL ADVANCES 33 Lederman, Daniel, and William F. Maloney. 2003. "R&D and Development." World Bank, Washington, D.C. Processed. "Link by Link." 2003. Business China, May 26. Luthria, Manjula, and Keith Maskus. 2004. "Protecting Industrial Inventions, Author's Rights, and Traditional Knowledge: Relevance, Lessons, and Unresolved Issues." In Kathie Krumm and Homi Kharas, eds., East Asia Integrates, pp. 139­62. Washington, D.C.: World Bank. Mathews, John A., and Dong-Sung Cho. 2000. Tiger Technology: The Creation of a Semicon- ductor Industry in East Asia. Cambridge, U.K.: Cambridge University Press. Micklethwait, John, and Adrian Woolridge. 2000. A Future Perfect. New York: Crown Publishers. "Multinationals and Globalization." 2001. Oxford Analytica, June 14. Nolan, Peter. 2002. "China and the Global Business Revolution." Cambridge Journal of Eco- nomics 26(1):119­37. Nolan, Peter, and Jin Zhang. 2002. "The Challenge of Globalization for Large Chinese Firms." World Development 30(12):2089­107. OECD (Organisation for Economic Co-operation and Development). 2002. OECD Science, Technology, and Industry Outlook 2002. Paris. ------. 2003. OECD Science, Technology, and Industry Scoreboard 2003. Paris. O'Rourke, Kevin H., and Jeffrey G. Williamson. 1999. Globalization and History. Cambridge, Mass.: MIT Press. Ozawa, Terutomo. 2001. "The `Hidden' Side of the `Flying-Geese' Catch-up Model: Japan's Dirigiste Institutional Setup and a Deepening Financial Morass." Journal of Asian Economics 12:471­91. ------. 2003. "Pax Americana-led Macro-clustering and Flying Geese­Style Catch-up in East Asia: Mechanisms of Regionalized Endogenous Growth." Journal of Asian Eco- nomics 13:699­713. "The Samsung Way." 2003. Business Week, June 16. Scarpetta, Stefano, Phillip Hemmings, Thierry Tressel, and Jaejoon Woo. 2002. "The Role of Policy Institutions for Productivity and Firm Dynamics: Evidence from Micro and Industry Data." Working Paper 329. Organisation for Economic Co-operation and Development, Paris. Processed. Sjöholm, Fredrik. 1997. "Technology Gap, Competition, and Spillovers from Direct For- eign Investment: Evidence from Established Data." Working Paper 38. European Institute of Japanese Studies, Stockholm. Processed. "Steel Magnolia." 2003. Economist, November 22. Stiglitz, Joseph E. 2002. Globalization and Its Discontents. New York: Norton. Tybout, James R. 2000. "Manufacturing Firms in Developing Countries: How Well Do They Do, and Why?" Journal of Economic Literature 38(1, March):11­44. Wang, Jiann-Chyuan, and Kuen-Hung Tsai. 2003. "Productivity Growth and R&D Ex- penditure in Taiwan's Manufacturing Firms." NBER Working Paper 9724. National Bureau of Economic Research, Cambridge, Mass. Processed. Whittaker, D. Hugh. 2003. "Crisis and Innovation in Japan: A New Future through Techno-Entrepreneurship." In William W. Keller and Richard J. Samuels, eds., Crisis and Innovation in Asian Technology. Cambridge, U.K.: Cambridge University Press. 34 GLOBAL PRODUCTION NETWORKING IN EAST ASIA World Bank. 2000. Entering the 21st Century: World Development Report 1999­2000. New York: Oxford University Press. Wyplosz, Charles. 2001. "How Risky Is Financial Liberalization in the Developing Coun- tries?" CEPR Discussion Paper 2724. London: Centre for Economic Policy Research. Yeung, Henry Wai-chung, and Kris Olds. 2000. "Globalizing Chinese Business Firms: Where Are They Coming from, Where Are They Heading?" In Henry Wai-chung Yeung and Kris Olds, eds., The Globalisation of Chinese Business Firms, pp. 1­28. London: Macmillan. Yusuf, Shahid, with M. Anjum Altaf, Barry Eichengreen, Sudarshan Gooptu, Kaoru Nabeshima, Charles Kenny, Dwight H. Perkins, and Marc Shotten. 2003. Innovative East Asia: The Future of Growth. New York: Oxford University Press for the World Bank. Yusuf, Shahid, and Simon J. Evenett. 2002. Can East Asia Compete? Innovation for Global Markets. New York: Oxford University Press for the World Bank. Yusuf, Shahid, and Kaoru Nabeshima. 2004. "Creative Cities in East Asia." Cities. Forth- coming. CHAPTER 2 THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS IN EAST ASIA Timothy J. Sturgeon and Richard K. Lester I n the 1970s and 1980s the newly industrialized economies (NIEs) of East Asia moved along high-growth trajectories. A combination of pru- dent and stable macroeconomic policies, the targeting of specific sectors for development, and country-specific industrial and social structures enabled the absorption of key technologies and the accumulation of the skills required to achieve industrial upgrading and growth (Amsden 1989; Evans 1995; MacIntyre 1994; Wade 1990; World Bank 1993). In the Re- public of Korea, development was driven by the chaebol, large-scale con- glomerates with privileged access to state-provided capital; in Hong Kong The preparation of this chapter benefited greatly from the ongoing work of the Globalization Study team at the Massachusetts Institute of Technology's Industrial Performance Center. The authors are grateful to Sara Jane McCaffrey for providing invaluable research assistance. Qualitative field research was conducted in Southeast Asia by Raphael Bonoan, Douglas Fuller, and Vincent Sawansawat. Suzanne Berger made many important contributions throughout. The authors would like also to thank the other participants in the Industrial Performance Center's Globalization Study, including Akintunde Akinwande, Dan Breznitz, Brian Hanson, Donald Lessard, Richard Locke, Teresa Lynch, Mike Piore, Charles Sodini, and Edward Steinfeld for valuable discussions. Shahid Yusuf and Simon Evenett of the World Bank were helpful and patient during the paper's preparation. Helpful comments were also provided by the participants in the East Asia's Future Economy Conference, organized jointly by the Development Economic Research Group of the World Bank and the Asia Pacific Policy Program of the Kennedy School of Government at Harvard University and held on October 1­2, 2001. The comments of Dennis Encarnation were especially valuable. 36 GLOBAL PRODUCTION NETWORKING IN EAST ASIA (China) and Taiwan (China) it was driven by small- and medium-scale enter- prises that were financed largely with private or family capital; in Singapore the state played a key role in encouraging foreign direct investment (FDI) by multinational corporations (MNCs) and in fostering the development of small- and medium-scale enterprises to serve them. Industries in these economies upgraded by building links to international markets and to the necessary sources of technology, expertise, managerial experience, and capital in the advanced countries. As Guillèn (2001) has shown, country- specific links between domestic structures and national policies produced different growth trajectories among the Asian NIEs. Less well understood is that each of these growth trajectories was also in- fluenced by the industrial structures of the advanced countries with which the NIEs were interacting. It is commonly observed that the rapid growth of the NIEs required trade openness in the West. This is indeed a necessary condition for export-oriented development. But the development paths pur- sued in the Asian NIEs were also influenced and enabled by the competitive strategies of American, European, and Japa-nese firms, which involved es- tablishing local operations in the NIEs, identifying local firms as suppliers, transferring skills and technologies to them, investing in them, and buying from them. The NIEs exploited these strategies to establish and upgrade a critical set of domestic technological and industrial capabilities. In this chap- ter we refer to this process as supplier-oriented industrial upgrading--de- veloping a supply base tuned to serving advanced-economy lead firms as a key mechanism of industrial upgrading. Both firms and government policy- makers tend to view supplier-oriented industrial upgrading as a stepwise learning process, beginning with manufacturing services only, perhaps in an export processing zone, then gradually progressing to manufacturing plus design, and culminating with developing-country firms having the capabil- ity to design, manufacture, and sell their own branded products on world markets. Today, however, lead firms based in the West are reorganizing pro- duction networks in ways that still seek to engage with the newly indus- trialized and emerging economies of East Asia, but with different understandings about how global production networks should be or- ganized and what roles local firms should play in them. The central ar- gument of this chapter is that changes emanating from the advanced industrial economies, particularly the United States, have begun to alter substantially the prospects for supplier-oriented industrial upgrading in East Asia. A key finding of our research is that American and European THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 37 lead firms have recently become more dependent on a set of "global sup- pliers" based in the West, even as they have increased their direct involvement in Asian production and Asian markets. These lead firms are increasingly relying on large suppliers and contract manufacturers from within their own societies to support their global operations. These global suppliers have, in turn, experienced rapid growth and global expansion and have become influential global actors. In the electronics sector, where the global suppliers are most highly developed, industry consultants estimate that the $90 billion of business that went to contract manufacturers in 2000 accounted for 13 percent of the total market for world circuit board assembly, product-level electronics man- ufacturing, and associated services. We argue that, in the key manufac- turing sectors of electronics and motor vehicles, the rapid rise of global suppliers based in Europe and the United States presents an important new challenge to the supplier-oriented industrial upgrading paths that Asian economies have pursued in the past. Not only do these global suppliers offer stiff competition to established Asian suppliers and allow lead firms to bypass the painstaking process of bringing new sup- pliers on stream, but they also provide lead firms with a strategic alter- native to relying on a group of Asian suppliers that appear to be focused on becoming, sooner or later, competitors to the lead firms in product markets. The chapter is organized as follows. We begin by introducing the no- tion of supplier-oriented industrial upgrading and then document the rise of global suppliers in the electronics and motor vehicle industries. Next, we draw on evidence from our field research to argue that the threshold re- quirements for suppliers to participate in GPN rose sharply in the 1990s. For East Asian firms, an especially challenging requirement is for par- ticipating suppliers to expand the geographic scope of their operations. We then outline some of the risks and uncertainties introduced by the emer- gence of the global supplier model for both lead firms and global suppliers and briefly discuss the changing architecture of Japanese production net- works in East Asia and their impact on development. Finally, we explore some policy implications of the evidence presented for the economies of East Asia. Some of the empirical evidence presented here is drawn from more than 400 field interviews conducted during the period 1999­2003 by a team of researchers at the Industrial Performance Center (IPC) of the Massachu- setts Institute of Technology (MIT) as part of its ongoing Globalization 38 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Study.1 Other data were collected during the course of the World Bank's Project on East Asia's Economic Future. A NEW FOCUS ON SUPPLIER-ORIENTED DEVELOPMENT STRATEGIES FOR INDUSTRIAL UPGRADING IN EAST ASIA By the late 1990s the debate over the relative roles of macroeconomic (that is, fiscal and monetary) policies and micro-level industrial policies in driv- ing development in East Asia, summarized in the World Bank's 1993 vol- ume The East Asian Miracle, had begun to run out of steam. There were two main reasons for this. The first was the careful work of authors such as Evans (1995) and Guillèn (2001) showing that both had been important and that macroeconomic and industrial policies had, in fact, been effectively com- bined and coordinated in a variety of ways in different countries. The most important lesson of this work was less about the effectiveness of one set of policies relative to the other and more about the need to build up the capac- ity of states to act effectively in a variety of realms through the nurturing of a competent, professionalized, politically insulated bureaucracy with effec- tive but transparent links to a country's business elite. The second factor was the growing strength of the World Trade Organization (WTO). The tradi- tional debate about whether development strategies such as infant industry protection and technical assistance, followed by trade liberalization and ex- 1. For the past four years (1999­2003), a team of 24 researchers from the IPC has been investi- gating the confluence of globalization and industry reorganization in several sectors, including electronics, motor vehicles, software, textile, and apparel. Field research for the IPC Globaliza- tion Study has consisted of semi-structured qualitative interviews with company personnel and relevant individuals from government agencies, labor unions, and academia. In-person interviews and plant tours have been conducted in Canada, mainland China, France, Germany, Indonesia, Ireland, Israel, Italy, Japan, Korea, Malaysia, Mexico, the Philippines, Singapore, Spain, Taiwan (China), Thailand, Romania, and the United States. By the end of 2002, more than 350 inter- views had been conducted. Of these, 108 were conducted in the electronics industry, 72 were con- ducted in the textile and apparel industry, and 61 were conducted in the motor vehicle industry. Sponsors of the IPC Globalization Study include the Chinese National Federation of Industries, the Fujitsu Research Institute, the Volkswagen Foundation, and the Alfred P. Sloan Foundation. For more information about the IPC Globalization Study, including our methodology and de- tailed research findings, see http://globalization.mit.edu. The direct quotes included in this chap- ter are intended to provide a window into the attitudes and desires of industry participants that might not be otherwise available. They are not meant as statements of fact or accurate predictors of outcomes. Many of the statements are highly controversial and are meant to highlight impor- tant issues and areas of tension. These are individual opinions with which the authors may or may not agree. Manager and company names have been excluded to protect the confidentiality of the research subjects. Unless otherwise noted, all interviews were conducted by IPC researchers. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 39 port promotion, were effective or not was becoming less relevant as more of these policies became "actionable" (illegal) under WTO rules. Attention accordingly began to shift to the role of global production networks in stimulating and sustaining industrial upgrading in Asian soci- eties (Dolan and Humphrey 2000; Encarnation 1999; Gereffi 1994, 1999; Kaplinsky 2000; Lee and Chen 2000; Shimokawa 1999; Tachiki 1999). The demands made by advanced-economy firms on local enterprises tended to be above and beyond what was required for the local market, and this gap stimulated the rapid advance of supply-base capabilities in East Asia (Keesing and Lall 1992). Through their roles as suppliers of parts and products and as purchasers of specialized process equipment, local firms gained access to important product and process know-how without violating WTO rules. As the case of hard disk drives in Singapore illustrates (McKendrick, Doner, and Haggard 2000), East Asian economic development has been strongly influenced by the strategies of advanced-economy firms in specific industries. Borrus, Ernst, and Haggard (2000), Dedrick and Kraemer (1998), Ernst (1997), Ravenhill (1995), and others note that production networks emanating from different advanced economies have had different characteristics and, in particular, that the "openness" of U.S. networks-- that is, the willingness of lead firms from the United States to install local management and ratchet up their demands on local firms--speeded the up- grading process in East Asian locations such as Hong Kong (China), Korea, Singapore, and Taiwan (China). These insights led Dedrick and Kraemer (1998) to argue that encouraging and facilitating the participation of local firms in global production networks was the only effective policy tool left for developing countries. The intensifying global trend toward economic liberalization and the de- clining appeal of autarkic development policies that sought to wall off do- mestic industries until "national champions" were strong enough to compete withforeignrivalshavebroughtnewattentiontopoliciesthatseektodevelop the capabilities of local firms as suppliers to lead firms from advanced economies. But advocates of the "supplier-oriented" approach to industrial upgrading often fail to account fully for the fundamental organizational changes taking place in the industries that have been key to the creation of cross-border production networks, the new actors that have arisen within them, and the new demands being placed on suppliers as a result. In the conventional supplier-oriented model of economic develop- ment, domestic suppliers continuously upgrade their capabilities either by serving the needs of the local affiliates of MNCs or by supplying lead firms in advanced countries from a distance. In both cases, if the model is extended further, the expectation is that the local firms will leverage their 40 GLOBAL PRODUCTION NETWORKING IN EAST ASIA experience by building up design competencies of their own. These design capabilities not only provide new sources of revenue, they also enable the firms eventually to develop their own branded product lines and perhaps even to emerge as direct competitors to advanced-economy lead firms. The upgrading process can proceed in stepwise fashion, beginning with simple assembly, where labor is applied to components and designs supplied by foreign buyers; followed by the supply of complete products with locally sourced components manufactured to specifications provided by foreign buyers, the so-called original equipment manufacturing (OEM) relation- ship; followed by the addition of post-conceptual design services to the man- ufacturing function, a combination known as original design manufacturing (ODM). Once design competencies are well established, the supplier can begin to conceptualize, develop, and manufacture finished products, for sale first under the brand labels of its customers and later under its own brand name. At that point, the local firm becomes what is sometimes referred to as an original brand manufacturer (OBM). In this fully blown version of the supplier-oriented upgrading path, the local firm eventually steps fully out of the supplier role to become a lead firm in its own right. The supplier-oriented upgrading path, although straightforward in con- cept, is often far from smooth in practice, especially if it is to be followed to its ultimate conclusion. At each stage along the way, difficult problems have to be addressed by the companies involved. One price of entry into cross- border networks, especially for suppliers whose operations are tightly cou- pled with those of customers in advanced economies, has been heightened vulnerability to rapid changes in customer strategy. Thus, for example, in 1997, when Compaq adopted a new strategy of marketing a complete desk- top personal computer system for under $1,000--a strategy that competi- tors quickly imitated--many small manufacturers of personal computers in Taiwan (China) were forced to close. Those that survived did so only by rap- idly shifting production to China. The development of in-house product design capabilities allows firms to capture more of the final price of the product and can also provide some protection against uncertainties in the business environment by enabling more rapid response to shifts in demand. In fact, many Asian firms have suc- cessfully made the shift from OEM to ODM status (although, as table 2.1 suggests, many more have not done so). The transition from ODM to brand-name production has been more problematic. There are certainly success stories among Asian brand-name producers. Firms like Fang Brothers and VTech in Hong Kong (China), Samsung in Korea, Creative Labs in Singapore, and Acer in Taiwan (China) have developed brands for Western as well as for Asian markets. The OBM THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 41 65)= 14 5 China (n ehiclesV 12)= 29 52 ASEAN-4 (n China 142)= 17 01 (n onics 2000, Electr 23)= 24 42 ASEAN-4 (n Sector by 40)= 16 6 goods China (n Suppliers, 4)= 40 40 and Consumer (n ASEAN-4 63)= 13 4 Customers el China (n eign Appar For 18)= 29 01 ASEAN-4 (n with ms Fir China 310)= 15 7 (n sectors 2001. All 58)= 28 19 early in ASEAN-4 (n Manufacturing Bank to to Asian orld W East the services of in-house by R&D and total) designed conducted Interaction of interaction customers customers parts design 2.1 Survey of cent eign eign ce: ableT (per ypeT Supply for Supply for Sour 42 GLOBAL PRODUCTION NETWORKING IN EAST ASIA model also has been influential throughout the region, especially in Taiwan (China), where Acer, the flagship firm of the Taiwanese (Chinese) com- puter industry, apparently has made the transition successfully (Dedrick and Kraemer 1998). The head of a firm in Hong Kong (China) that produces both brand-name goods and OEM products for clients explained some of the challenges and attractions of brand-name production (Berger and Lester 1997, p. 39): With a label, you take on a series of challenges. It's a baby to continuously enhance. If you work only OEM, if you work only for others, you're taking commands from them. You're on the hand-me-down side. You're not in the decision making seat. But in the course of our interviews we encountered numerous firms that have retreated from a brand-name strategy back to an ODM or OEM focus, and our fieldwork in Taiwan (China) and elsewhere suggests that for many firms the OEMODMOBM upgrading path has stalled at the ODM phase. Even Acer struggled and ultimately failed to establish a significant presence in the U.S. market, and Fang Brothers has sold off its chain of retail outlets in the United States. These problems have several causes. For would-be OBM manufacturers, the importing, sales, marketing, and distribution functions associated with brand-name products are entirely new and very different from their deep competencies in manufacturing and design. In our interviews, many Asian newly industrialized countries (NICs) firms mentioned the high cost of pen- etrating advanced-economy markets, especially at the retail level. Second, lead firms from the West continue to enjoy substantial design and marketing advantagesresultingfromtheirproximitytoadvanced-economymarketsand lead users. While some Asian manufacturers have gained a great deal of ex- pertise in post-architectural, detailed design, advanced-economy lead firms appear to have a strong advantage in the realm of new product creation and conceptual design, which allows them to continue to set product strategy and drive the broad trajectory of market development. Third, Asian brand-name manufacturers that have retained OEM and ODM operations have had dif- ficulty reassuring customers that the customer's intellectual property and the manufacturer's quality of service will not be compromised as the manufac- turer develops its own branded products; as a result, firms that have achieved a measure of success with their own brands sometimes have seen their status as preferred supplier slip away. Acer chairman Stan Shih recently acknowl- edged the problems that Acer's own-brand operations created for its contract manufacturing business: "In many cases we were in the final list [as a con- tract manufacturer], but . . . when they made a decision, they picked our THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 43 competitors in Taiwan [China]" ("Acer Plots Path" 2001). The "pure-play" OEM-ODM suppliers that had not developed their own-brand business, such as Hon Hai and Quanta, were seen as less of a competitive threat to lead firms, and their business grew accordingly. Like other companies that have struggled with the OBM strategy, Acer appears to be refocusing on its OEM and ODM businesses. (According to its latest corporate restructuring plan, the company intends to divest its con- tract manufacturing operations from its brand-name business; "Acer Plots Path" 2001.) However, the OEM-ODM strategy, important as it has been for industrial upgrading in the region, now faces new questions. Indeed, one of the biggest challenges to the supplier-oriented upgrading approach in East Asia may turn out to be the emergence of a new class of highly sophis- ticated pure-play suppliers, mostly U.S.-based, capable of supporting the manufacturing needs of European and U.S. lead firms in Asia and around the world. The supplier-oriented model of economic development has generally as- sumed that lead firms will continue to seek out or develop distinct supply bases in multiple locations and that to achieve these objectives they will sup- port the upgrading efforts of local firms. But our interviews with managers at dozens of lead firms and suppliers in the electronics and motor vehicle in- dustries have cast doubt on both assumptions. This research has revealed a growing propensity on the part of lead firms to expect considerably more from their suppliers, both functionally and in terms of geographic scope. The new requirements go well beyond excellent manufacturing perfor- mance and low costs, which are perceived as being widely available and com- modified. Today, suppliers must provide a capability for independent process development and an ability to perform a wide range of value adding func- tions associated with the manufacturing process, including help with prod- uct and component design, component sourcing, inventory management, testing, packaging, and outbound logistics. Lead firms are also demanding that suppliers have the ability to support the lead firm's operations and market-serving activities around the world. Getting the right part or process in the right place at the right time--as well as at the right cost and with a minimum of inventories in process and in transit--is critical. Even slightly out-of-date components, products, and processes quickly lose their utility and value. As the market and operational scope of many lead firms has become truly global, suppliers are being se- lected on the basis of their ability to provide global support. Such suppli- ers have the scale and scope to coordinate component sourcing and inventory management at a global level. Finally, as noted, lead firms have become less willing to use actual or potential competitors as suppliers, 44 GLOBAL PRODUCTION NETWORKING IN EAST ASIA especially as globally operating pure-play suppliers have appeared on the scene as an alternative. As we show in the following sections, these new requirements pose dif- ficult challenges for the larger, established suppliers in the Asian NIEs. The new requirements are also raising the barriers to market entry, throughout the region, for smaller and younger suppliers that are seeking to participate in GPNs. On a larger scale, the emergence of a global sup- ply base serving lead firms in the West raises questions about the long- term viability of the supplier-oriented upgrading model and suggests a need to rethink at least some of the economic development policies that countries in the region have been pursuing. We take up these questions in the following sections. ELECTRONICS: THE RISE OF GLOBAL CONTRACT MANUFACTURERS Historically, most U.S. electronics firms purchased electronic components and assembled them in-house into subsystems and final products of their own design. MNCs used a mix of onshore and offshore assembly plants. Since manufacturing processes were quite labor-intensive, plants assem- bling high-volume, price-sensitive products were often located in areas with low labor costs, especially Mexico's northern border region and East Asia. In the 1960s Japanese firms began licensing local firms in Hong Kong (China), Korea, and Taiwan (China), initially to produce transistor radios and, later, hand-held calculators. Beginning in the 1960s American (and, later, European) semiconductor firms moved labor-intensive chip assembly processes to low-wage locations in East Asia, especially Hong Kong (China),Malaysia,Singapore,andThailand(Scott1987).Theattraction was low labor costs and, for American firms, tariff exemptions that allowed them to pay duties only on the value that had been added through the as- sembly process, not on the semiconductor chips themselves. Once the semiconductors were assembled in Asia, they were shipped back to Eu- rope and the United States, where they were sold to brand-name electron- ics companies and assembled into final products. For a time, nearly all semiconductor wafer fabrication, circuit board assembly, and product-level assembly stayed in Japan, Northern Europe, and the United States. For cir- cuit board assembly, domestic contract manufacturers were used as "shock absorbers" during times of peak demand, when the internal capacity of brand-name companies was fully utilized, but not as a serious alternative to in-house manufacturing. In electronics, the typical contract manufacturing THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 45 arrangement during the 1970s was for the brand-name firms to provide laborcontractors--thenknownas"boardstuffers"--withkitsofcomponents from their own inventories on a consignment basis. The contractors then supplied the labor needed to assemble the customer-supplied kits. Some overseas firms, in locations that included Hong Kong (China), Singapore, and Taiwan (China), also stepped into the role of board stuffer. In the 1980s American brand-name electronics firms operating in high- volume, price-sensitive market segments, such as disk drives and personal computers, began following the semiconductor industry offshore as a way to tap low-cost labor in Asia, either by establishing local subsidiaries or by tapping the growing capability of Asian producers, especially in Taiwan (China). In the late 1980s as progress toward European unification contin- ued, American electronics companies moved the assembly of some prod- ucts destined for the European market to Scotland and Wales in an effort to increase value added within Europe. Scotland and Wales were chosen most often because of their relatively low labor and engineering costs, a well-trained English-speaking work force, and aggressive efforts by local economic development agencies to attract FDI in electronics. IBM, for ex- ample, set up its largest European personal computer assembly facility in Greenock, Scotland. Over time significant numbers of local contract manufacturers assem- bling circuit boards for the affiliates of American brand-name electronics firms emerged, especially in Mexico, Scotland, Singapore, Taiwan (China), the United States, and Wales. Drawing on components from nearby semi- conductor assembly plants, both brand-name firms and their contract man- ufacturers, especially those located in Asia, began sourcing the bulk of their components locally. The local availability of components allowed brand- name firms to ask their Asian contractors to begin buying components on a turnkey basis, that is, with financing provided by the supplier. In the early 1990s some brand-name electronics firms in the United States moved beyond the tactical use of their contractors as providers of overflow capacity and began to use the most capable of them for more strate- gic purposes. The advantages included manufacturing close to end markets or with low-cost labor, subjecting internal operations to market forces, keeping abreast of fast-moving assembly technologies, and focusing their own activities on increasingly challenging "core competencies," such as product definition, design, sales, and marketing. As the contract manufac- turers grew in size, additional cost advantages accrued from scale economies derived from the pooling of manufacturing capacity and component pur- chasing. Lead firms also discovered dynamic advantages to moving manu- facturing further out of house: they could ramp production levels up and 46 GLOBAL PRODUCTION NETWORKING IN EAST ASIA down more quickly and with less cost at contractor facilities than with in- house manufacturing. This proved to be an important asset in the face of a highly volatile and contentious market environment. Today, production outsourcing in electronics has become a widely accepted practice for both large and small brand-name electronics firms based in the United States and, increasingly, Europe. More recently, globally operating lead firms have been consolidating their contract manufacturing relationships by giving a larger share of their manufacturing to a smaller group of large, technologically sophisticated con- tract manufacturers, nearly all of them of North American origin. Brand- name electronics firms are demanding that their contractors have a "global presence" as a way of streamlining the management of their outsourcing relationships. As a result, North American contract manufacturers have themselves been aggressively internationalizing their operations since the mid-1990s. During the past decade, global suppliers in the electronics industry ex- perienced rapid revenue growth, consolidation at the top levels, and geo- graphic expansion. In the 1990s established North American electronics firms in the computer and networking sectors, such as Apple Computer, Hewlett-Packard, IBM, Lucent, Maxtor, Nortel, and 3Com, rapidly moved toward outsourcing their circuit board and product-level assembly, notably by selling off much of their domestic and offshore production facilities to the five largest contract manufacturers. Many newer North American elec- tronics companies such as Cisco Systems, EMC, JDS Uniphase, Juniper Networks, Network Appliance, Sun Microsystems, and Sycamore Net- works outsourced most of their production from the outset, and their rapid growth during the late 1990s fueled growth of the largest electronics contract manufacturers. All of the top five contract manufacturers are based in North Amer- ica. They consist of Celestica, based in Toronto, Canada; Flextronics In- ternational, incorporated in Singapore, but managed from its San Jose, California, headquarters; Jabil Circuit, based in St. Petersburg, Florida; Sanmina/SCI, based in San Jose; and Solectron, based in Milpitas, Cal- ifornia. These five firms collectively grew at an average annual rate of 36 percent a year between 1994 and 2001 (see table 2.2). In the latter part of the 1990s, the outsourcing trend began to spread to most of the major European suppliers of communications infrastructure as well, especially Alcatel, Ericsson, and, to a lesser degree, Nokia and Siemens. In 1997 Ericsson made a decisive series of moves, first by outsourcing pro- duction to Flextronics, SCI, and Solectron and then by selling its principal domestic production facilities in Karlskorna, Sweden, to Flextronics and a THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 47 Table 2.2 Revenue at the Top Five Electronics Contract Manufacturers, 1994 and 2001 (millions of U.S. dollars) Average annual growth Company 1994 2001 rate (percent) Celestica 1,989,000a 10,004,000 26 Flextronics 210,700 12,110,000 78 Jabil Circuit 404,056 4,331,000 40 Sanmina/SCI 2,363,581 11,248,651 25 Solectron 1,641,617 18,692,000 42 Total 6,608,954 56,385,651 36 a. In 1994 all of Celestica's revenues were from IBM. Source: Company annual and quarterly reports. plant in Brazil to Solectron (Dunn 1997). Solectron established a local pres- ence in Sweden as well but shifted the bulk of Ericsson's circuit board as- sembly to its existing network of plants in France, Germany, and Scotland (Jonas 1997). In 2000 Ericsson shifted the remainder of its cell phone pro- duction to these American contract manufacturers and sold its U.S. pro- duction facilities to SCI ("Ericsson Shifts Phone Production" 2000).2 In the past two years even a few Japanese electronics firms have tested the waters. In December 2000 NEC, whose cell phone handset business was do- ing poorly in the midst of a fierce shakeout, announced that it was selling its cell phone production facilities in England and Mexico to Solectron, while keeping facilities in China and Japan ("NEC to Sell Cell Phone Plants" 2000). In October 2000 Sony announced that it was selling two underutilized Asian facilities to Solectron, one in Miyagi, Japan, and a second in Kaoh- siung, Taiwan, China (Levine 2000). In January 2002 NEC announced the sale of two of its advanced manufacturing facilities in Miyagi and Yamanashi, Japan, to Celestica.3 About 1,200 highly skilled NEC manufacturing spe- cialists and related support staff became Celestica employees. As part of the deal, Celestica assumed management of the supply chain as well as responsi- bility for subassembly, final assembly, integration, and testing of a broad range of NEC's optical backbone and broadband access equipment. The companies expect the deal to generate revenue of approximately $2.5 billion for Celestica over a five-year period. According to Kaoru Yano, senior vice president of NEC and company deputy president of NEC Networks, 2. In 2001 SCI was acquired by Sanmina. 3. The design and development functions currently performed by NEC Miyagi will remain with NEC, and NEC Miyagi will continue as a developer of optical transmission systems. The devel- opment and manufacturing of optical devices and optical submarine cable systems at NEC Yamanashi's Otsuki plant will also remain with NEC. 48 GLOBAL PRODUCTION NETWORKING IN EAST ASIA NEC's growing partnership with Celestica will allow us to improve our com- petitive positioning by further leveraging our leading-edge R&D [research and development], product development, and manufacturing expertise with Celestica's global manufacturing capabilities and supply chain management expertise. Through the alliance with Celestica, NEC intends to improve price competitiveness, production lead times, and supply chain flexibility to optimize overall manufacturing efficiency. NEC also chose to work with Celestica based on its reputation for providing global, advanced manufactur- ing capabilities and cost-effective supply chain solutions for the world's best communications and information technology companies. As mentioned, most of the growth in electronics contract manufacturing has taken place in the very top tier of firms. Electronic Trend Publications (2000) estimated that the top five contract manufacturers had captured 38 percent of the electronics contract manufacturing market by 1999 and projected this share to grow to almost two-thirds in 2003. This rapid expan- sion, fueled by the acquisition of competitors and customer facilities as well as organic expansion in existing and newly established facilities, was aided by the U.S. stock market run-up in the late 1990s, which concentrated a sig- nificant share of market capitalization in contract manufacturing in the top five firms. Each of the largest electronics contract manufacturers has established a global network of plants that consists of (a) low-product-mix, high-volume production sites, mostly in Asia, East Europe, and Mexico; (b) high-product- mix, medium- to high-volume production sites in Canada, the United States, Western Europe, and now Japan; (c) engineering-heavy "new product introduction" centers, often located near an important customer's design activities; and (d) facilities that perform final assembly and product config- uration to order or that provide after-sales repair service, often located near major transportation hubs, such as Amsterdam and Memphis, Tennessee. All have large-scale investments for high-volume production in East Asia, especially in Southeast Asia and, increasingly, in China. The rapid geographic expansion of these firms is worth noting in some detail. Celestica, which spun off from IBM in 1996, began with two pro- duction locations, a large complex near Toronto, Canada, and a small facility in upstate New York, since closed. Today, after completing 29 ac- quisitions, Celestica operates nearly 50 facilities in Asia, Europe, North America, and South America and generates annual revenues in excess of $11 billion (see figure 2.1). In Asia the company established or acquired facilities in Dongguan, Suzhou (two), Shanghai, Xiamen, and Hong Kong in China; Johor Baru (three), Kulim, and Parit Buntar in Malaysia; Batam and Bintan in Indonesia; and Singapore (three). Other high-volume produc- tion sites were established in Kladno and Rajecko in the Czech Republic; THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 49 Figure 2.1 Celestica's Global Footprint, 2001 Source: Celestica. Source: Celestica. Guadalajara, Monterrey (two), and Querétaro in Mexico; and Guarulhos and Hortolândia in Brazil. Another striking example of rapid geographic expansion is the largest electronics contract manufacturer, Solectron, which was concentrated in a single campus in Silicon Valley until 1991, when its key customers, includ- ing Hewlett Packard, IBM, and Sun Microsystems, began to demand global manufacturing and process engineering support. Within 10 years, the company's footprint had expanded to nearly 50 facilities worldwide (see table 2.3). Today this network consists of global and regional headquarters, both high- and low-mix manufacturing facilities, materials purchasing and management centers, new product introduction centers, after-sales repair service centers for products manufactured by Solectron and others, and technology centers that develop advanced process and component packag- ing technologies. According to estimates by Technology Forecasters, contract manufac- turers penetrated roughly 17 percent of the total available market for cir- cuit board and product-level electronics manufacturing in 2001. A recent Bear Stearns survey of brand-name electronics firms (Levine 2001) con- cluded that the rate and size of outsourcing agreements will continue to increase, with 85 percent of the firms interviewed planning on further increases in production outsourcing. As a group, the branded electronics firms in the survey expected to outsource 73 percent of total production 50 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Table 2.3 Solectron's Global Locations and Functions, 2001 Materials New After- Global purchasing product sales and Manufac- and intro- repair Process regional turing management duction service technology Location headquarters facilities centers centers centers centers Americas Brazil São José dos · · · · Campos Hortolândia · · Canada Vaughn · Calgary · Mexico Guadalajara · · Monterrey · · · United States Milpitas, Calif. · · · · · Fremont, Calif. · · · Austin, Tex. · · · · Charlotte, N.C. · · · · Columbia, S.C. · · · San Jose, Calif. · · · Atlanta, Ga. · · · · Westborough, · · · Mass. Suwanee, Ga. · · · · Fremont, Calif. · · · Everett, Wash. · · Raleigh, N.C. · · Aguadilla, P.R. · Aguada, P.R. · Los Angeles, · Calif. Austin, Tex. · Memphis, Tenn. · Louisville, Ken. · Europe and Middle East Bordeaux, France · · · · Germany Herrenberg · · · Munich · · · · Dublin, Ireland · · (continued) THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 51 Table 2.3 Solectron's Global Locations and Functions, 2001 (continued) Materials New After- Global purchasing product sales and Manufac- and intro- repair Process regional turing management duction service technology Location headquarters facilities centers centers centers centers Carrickfergus, · · Northern Ireland Timisoara, · · Romania Scotland Dunfermline · · East Kilbride · · Östersund, Sweden · · · · Istanbul, Turkey · · · · Reading, United · Kingdom Australia Wangaratta · Liverpool · Asia China Suzhou · · Bangalore, India · Japan Tokyo · · Kanagawa · · · Malaysia Johor Baru · · · Penang · · · · Penang · · Singapore · · · Taipei, Taiwan · Source: Solectron. needs on average, and 40 percent intended ultimately to outsource 90­ 100 percent of final product manufacturing. The chief executive officer of Flextronics has stated publicly that he expects annual revenues at his com- pany to reach the $100 billion range in the next 5­10 years. In some instances, the expansion of North American contract manufac- turers in East Asia has come about through the acquisition of established regional firms, such as Solectron's acquisition of NatSteel Electronics (Singapore) and Ocean Electronics (Hong Kong, China) in the late 1990s. In the future, the North American contract manufacturers may compete 52 GLOBAL PRODUCTION NETWORKING IN EAST ASIA directly with the largest indigenous electronics contract manufacturers in East Asia, and this trend may already be constraining the growth of Asian- owned firms. Given the importance of the electronics industry in East Asia's industrial upgrading, the question of how much business indigenous East Asian electronics suppliers will capture in the ongoing shift from in-house to outsourced production is critical for the future of economic development in the region. Chapter 3 explores this question further and, in particular, examines not only the strengths of the contract manufacturers but also the risks they face. MOTOR VEHICLES: THE RISE OF THE MEGA SUPPLIERS The interest in producing motor vehicles close to final markets has always been strong, although the motivations have changed over time. Prior to the advent of mass production in 1910, automobiles were luxury products, and the need for customization required proximity to pools of well-heeled cus- tomers. When mass production lowered the cost of the automobile to the point where mass markets developed, final assembly moved close to final markets to reduce transport costs. In the 1930s overcoming trade barriers erected by national governments become the main motivation for offshore production.Automakerswereforcedtoestablishlocalproductionortoforgo participation in the most promising emerging markets of the day. By 1928 Ford and General Motors were assembling vehicles in 24 countries, includ- ing Brazil, India, Japan, and Malaysia. A decade later both companies were operating large-scale integrated "transplant" facilities in Europe. When trade barriers were extended to automotive parts, automakers moved to in- tegrate offshore production and source parts locally to the extent possible. Both tariff and nontariff barriers to trade in finished vehicles--or the threat thereof--continue to be a key motivation for the growth of offshore pro- duction today. The number of automakers willing and able to compete in the arena of international production increased markedly in the 1980s, but American automakershadbeguntofacecompetitioninthedevelopingworldinthelate 1950s, when European producers had recovered sufficiently from World War II to begin investing in Australia, Latin America, and South Africa. During the 1960s and 1970s a regional pattern emerged. Most new offshore assembly plants established by American and European automakers were located in Latin America, and most offshore plants established by Japanese firms were located in Asia. There were exceptions to this pattern, namely the investments of Ford and General Motors in Taiwan (China) and a few THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 53 small Japanese investments in Brazil, Ecuador, and Peru.4 In the 1980s the remarkable successes of the Japanese automakers' export strategy in the United States at the direct expense of the American Big Three sparked a political backlash, which resulted in the setting of "voluntary" limits to con- tinued expansion of market share via exports. In response to these quotas, JapaneseautomakersembarkedonawaveofplantconstructionintheUnited States (Kenney and Florida 1993). By 1995 Japanese automakers were locally manufacturing two-thirds of the passenger vehicles they sold in the United States. A similar dynamic led to a wave of Japanese transplants in Europe, beginning with Nissan's plant in the United Kingdom in 1986. By 1995 Japanese automakers were locally manufacturing nearly one-third of the passenger vehicles they sold in Europe. In the 1980s American and European automakers, under increasing pressure from the sudden appearance of fierce new competition from Asia, began importing finished vehicles into their home markets from operations in lower-cost peripheral locations (for example, Canada, Eastern Europe, Mexico, and Spain) within the context of regional trade agreements such as the European Union and the North American Free Trade Agreement (NAFTA). During the 1990s there was a wave of new assembly and supplier plant construction in emerging markets such as Brazil, China, Eastern Europe, India, Mexico, Thailand, and Vietnam. These new investments were driven by increased competition and market saturation at home, the opening of new spaces for investment following the end of the cold war, host-country requirements for local content and production, and an effort by automakers to cut costs within the context of regional trade arrangements such as the European Union and NAFTA. The idea that emerging markets, particularly in Asia, would be the locus of rapid economic growth in the medium term stimulated a huge wave of new investment in Asia during the 1990s, particularly in China and the Association of South East Asian Nations (ASEAN). Industry consultants projected that Asia and Eastern Europe would have the highest rates of pro- 4. Investments by Japanese automakers, however, tended to be of a very different character than those of American and European firms. American and European firms tended to build larger, more integrated plants, whereas Japanese firms built plants that relied heavily on the assembly of vehicle "kits" sourced from home factories. Japanese investments were highly conservative, in that assembly plant investments remained scaled to the actual, not potential, size of the local market-- something that is still true today. Still, in places where Japanese automakers faced no competi- tion from more aggressive investors, such as the ASEAN countries of Indonesia, Malaysia, and Thailand, they were able to capture the lion's share of these markets, especially in countries where local content rules became more stringent over time (Doner 1991). 54 GLOBAL PRODUCTION NETWORKING IN EAST ASIA duction growth, with Asia outside of Japan generating 55 percent of the world's new production. Developing countries, taken together, were pro- jected to account for 80 percent of all new production (Sturgeon and Florida 2003). The Asian financial crisis brought the investment boom to an abrupt end, and many of the motor vehicle investment projects in Asia were scaled back or put on hold. Still, the long-term projections for the development of the motor vehicle market in large Asian countries such as China and India remain very positive. In an interview conducted in 2000, the manager of a German automotive parts supplier in China asserted that 50 million cars would be sold annually in China by 2050, 3.5 times the size of the current U.S. market. As the number of production locations multiplied during the 1990s, auto- makers sought to streamline operations on a global scale, particularly in the area of vehicle design and component sourcing. Most automakers today are seeking to place a greater number of car models on fewer underbody plat- forms, allowing for greater commonalization and reusability of parts, while retaining the ability to adapt specific models to local tastes and driving con- ditions. Such strategies call for global sourcing, tighter coordination of worldwide design efforts, and in cases where platform design activities have become geographically dispersed over time (that is, in American firms), con- solidation of project management in core locations and the formation of in- ternational design teams. At the same time, the need to respond to unique market requirements has created pressure to localize body design, prompt- ing highly centralized automakers (that is, Japanese firms) to set up regional design studios to cater to local tastes. Since the benefits of global platforms can only be reaped when they are used and reused across a broad product line, there has been a wave of consolidation in the industry, as large players acquire small, specialty producers. Efforts to create global platforms--often thought to have begun with the "world car" strategies of the 1980s--in fact long predate this and can be traced as far back as Ford's failed "1928 Plan," which aimed to supply Model As to the world from three giant River Rouge­style plants in Canada, Detroit, and England. The current trend, however, includes an entirely new feature: the formation of a global supply base. The recent round of globalization in the motor vehicle industry has helped to change the nature of relationships between automakers and their largest suppliers. First-tier suppliers are moving to module design, second- tier component sourcing, and the provision of local content in the context of emerging markets. The growing need to provide automakers with mod- ules on a worldwide basis is driving a wave of consolidation and geographic expansion among first-tier suppliers, just as it has in the electronics indus- try. For suppliers that serve multiple automakers, the geographic scale of THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 55 operations can surpass that of any single customer. In the long run it may well be suppliers, not automakers, that generate the vast majority of the in- dustry's future FDI--and the associated economic and social benefits (such as employment). As in electronics production, the outsourcing trend in the motor vehicle sector has been strongest among American firms. The structure of the mo- tor vehicle industry in the United States--and the characteristics of the jobs within it--have changed radically since 1986, when long-time rough em- ployment parity between the assembly and parts sectors began to diverge (see figure 2.2). Since the 1980s the supply sector has been the main source of job growth in the U.S. automotive industry, adding 220,900 jobs since 1982, compared with only 25,300 jobs in assembly. In the realm of vehicle manufacturing, automakers are performing far fewer functions within their assembly facilities than they have in the past. Vehicle assembly lines have been streamlined; integrated "feeder" lines that build up subassemblies such as seats, cockpits, and climate control systems within vehicle assembly facilities have all but disappeared. Assembly work- ers now bolt together a greater number of large subassemblies of individual components, known as "modules," that have been preassembled off-site by Figure 2.2 Employment in Assembly and Parts in the U.S. Motor Vehicle Industry, 1958­2000 (Thousands of jobs) 600 assembly 550 parts 500 450 400 350 300 250 200 1958 1964 1970 1976 1982 1988 1994 2000 Note: Assembly includes SIC 3711 (motor vehicles and car bodies), and parts includes SIC 3714 (motor vehicle parts and accessories). Source: U.S. Bureau of Labor Statistics (various years). 56 GLOBAL PRODUCTION NETWORKING IN EAST ASIA suppliers. Modules arrive fully assembled on the loading docks of final as- sembly plants, ready to be attached to vehicles as they move down the line. The result is more production workers in supplier plants and fewer in final assembly plants. Modularity Modularity has already been mentioned in the context of final assembly, but it has important consequences for the scope of supplier activities as well. As automakers do less within their assembly plants, suppliers have an opportunity to do more. For example, vehicle doors can be delivered with the glass, fabric, interior panels, handles, and mirrors preassembled. Dash- boards can be delivered complete with polymers, wood, displays, lights, and switches. One manager interviewed during our research estimated that 75 percent of vehicle value can be accounted for by only 15 modules. Im- portant modules are suspension (supplied as "corners"); doors; headliners (which can come with grip handles, lighting, wiring, sunroof, sun visors, and trim preassembled); heating, ventilation, and air conditioning (HVAC) units; seats; dashboards; and drive trains (that is, engines, transmissions, and axles). A continuation of the trend toward modules would mean that suppliers will provide automakers with groups of related modules, in what might be called "module systems." For example, seats, interior trim, the headliner, dashboard, and cockpit module could be supplied as a complete "interior system." Figure 2.3 provides a graphic representation of the trend from discrete parts to modules and then module systems.5 The drive toward modularity is associated with supply-base consolida- tion, as first-tier suppliers buy second-tier suppliers to gain the broader range of capabilities needed to supply modules and systems. TRW's recent 5. Some modules comprise contiguous subassemblies, while others do not. For example, seats and HVAC units comprise physically contiguous subassemblies, while vehicle electronics or occupant safety can consist of a variety of physically discrete components that work together to make up a functional system. Contiguous subassemblies provide the key benefit of assembly-line simplifica- tion, while noncontiguous systems do not. Sourcing noncontiguous modules from a single supplier is a way for automakers to pass the responsibility for system integration to suppliers. For example, an electronics supplier such as Bosch, Delphi, Siemens Automotive, or Visteon can take responsi- bility for ensuring that engine controls work properly with temperature, pressure, revolutions per minute, and other sensors that provide information to the control unit. In other instances, sourc- ing noncontiguous modules is a way for automakers to pass warranty responsibility for entire as- pects of vehicle quality--such as engine and transmission sealing or occupant restraint--on to suppliers. This is not to imply that industry nomenclature in the area of modules and systems has been standardized; some automakers refer to contiguous subassemblies as "modules" and func- tionally related noncontiguous parts as "systems." THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 57 Industry Body system Chassis system ehicleV rear- Motor chassis train and the Time modules in Module Skin Finish Trim Drive Rolling Front- end Modularity trim mirrors tires rear Part Stamping Hardware Fasteners Primer Paint Overcoat Exterior Undercoat Lenses, Fluids Gaskets Engine Axles Transmission Front, suspension Brakes Wheels, Bumpers Trim Radiator Fan Component of rajectoryT Interior system Electrical and electronic system System: to trim module electrical electrical Module Time to Module Seat Interior Cockpit Ignition Chassis Interior Part om Fr panels wheel control brakes 2.3 plugs, lights e frame harness Figur Part Fabric Foam Seat Headliner Interior Trim Dashboard Gauges Shifter Steering Trim Engine Wires, distributor Alternator Transmission Wire Antilock Audio Interior 58 GLOBAL PRODUCTION NETWORKING IN EAST ASIA acquisitions, for example, have given the company the capability to deliver all aspects of occupant restraint systems. Modularity creates natural break- points in the value chain and makes the outsourcing or relocation of mod- ule design and production more feasible. But modularity can also be pursued largely as an in-house strategy, as at Volkswagen, where internal subsidiaries have moved beyond the manufacture of parts to the assembly of modules and systems that are delivered to streamlined assembly plants. Because larger modules are more difficult and expensive to ship over long distances and are more likely to be sequenced (that is, coordinated tightly with the final assembly process), the adoption of modular assembly processes is associated with the co-location of assembly plants and supply plants. In- line sequencing has accelerated the adoption of "just-in time" parts delivery, where modules are delivered according to the sequence of cars moving down the assembly line. A key motivation for in-line-sequencing is color match- ing. Mirrors, interior panels, seats, dashboards, carpets, door handles, and bumpers all have to match or accent the body color and thus must be tightly sequenced with the colors of vehicles on the final assembly line. Globalization is occurring at the same time as increased outsourcing and the move to sourcing modules and systems, and so many suppliers are taking a larger role in the globalization process. Companies like Bosch, Johnson Controls, Lear, Magna, Siemens Automotive, TRW, Yazaki, and others have become the preferred suppliers for automakers around the world. Many first-tier suppliers have responded by embarking on a wave of vertical integration (through mergers, acquisitions, and joint ventures) and geographic expansion to gain the ability to deliver parts and modules on a global basis. The entry of Ford's and General Motors' former component divisions into the merchant market for vehicle components, modules, and systems has, almost overnight, created the world's two largest, most diver- sified, and geographically extensive automotive suppliers. Unlike the global contract manufacturers in electronics, global suppliers in the motor vehicle industry have emerged from Europe and Japan as well as North America.6 Still, the trends toward rapid growth, geographic ex- pansion, and consolidation are most pronounced among suppliers based in North America. Bosch and Siemens Automotive, both based in Germany, have tended to remain more focused on their core activities, but since their 6. The integral nature of product architecture in motor vehicles makes it harder for lead firms in the industry to switch suppliers, so they tend to bring their suppliers with them when they set up international operations, while the more standard interfaces and production processes in elec- tronics have allowed contract manufacturers from North America to surge ahead and more easily win business from lead firms of many nationalities. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 59 focus has long been electrical and electronic systems for vehicles, their growth has been in part due to the increased electronic content in vehicles and in part to increased sales to Japanese and especially American auto- makers. Major European suppliers are experimenting with a modular ap- proach, however, especially with their American customers. In 2000 Siemens Automotive acquired another German firm, VDO, which added cockpit instrumentation capability to Siemens' climate control and interior plas- tics capability. This has allowed the firm to bid on completely built-up dash- board modules. To illustrate the radical pace of change in the motor vehicle supply base, we consider the case the German tire manufacturer Continental AG in some detail. Continental has long specialized in tires for the retail market and had already established a global manufacturing presence in tires by the early 1990s.7 As late as 1995, Continental was still concentrated in the retail tires market and ranked 52nd in the world in direct sales to automakers. That year, the company established Continental Automotive Systems Group and began acquiring automotive suppliers with a wide variety of competencies and geographic attributes, such as TBA Belting (United Kingdom) and ITT Brake and Chassis (United States). The latter acquisi- tion, which was valued at nearly $2 billion, added 23 plants and 10,000 em- ployees. To round out the company's global footprint, Continental made a series of additional acquisitions in Argentina, Brazil, Mexico, Romania, Slovakia, and South Africa. In 2001 Continental acquired Temic Micro- electronic GmbH, a medium-size ($900 million in revenues) German auto- motive electronics firm with 3,000 employees and nine manufacturing facilities in Germany and also a small, newly established global footprint, including 2,800 workers at factories in Brazil, China, Hungary, Mexico, and the Philippines, and two technical centers, one at its headquarters in Germany and a new center in Auburn Hills, Michigan, just north of Detroit. Continental's product strategy is threefold: to leverage competence in syn- thetic rubber by entering markets for power transmission belts and other rubber parts for motor vehicles; to develop integrated modules from the tire 7. Continental established a global manufacturing presence in tires largely through acquisition. In 1979 the company acquired the European assets of Uniroyal (United States), a deal that in- cluded plants in Belgium, France, Germany, and the United Kingdom. In 1985 Semperit Reifen AG (Germany) was acquired, adding plants in Austria and Ireland (since closed). In 1987 Conti- nental acquired General Tire (United States), including four plants in the United States, two in Mexico (since sold), and a series of joint venture operations in Asia, Africa, and South America. In 1991 a joint venture agreement was signed with Yokohama (Japan) and Toyo (Japan) to make commercialtiresfortheU.S.market(http://www.conti-online.com/).In1992thecompanyacquired the Swedish tire producer Nivis Tyre. 60 GLOBAL PRODUCTION NETWORKING IN EAST ASIA inward, including assembled wheels, brakes, and suspension parts; and to enter the high-growth area of vehicle electronics. Today Continental's auto- motive divisions operate 140 facilities in 36 countries and employ 64,000 workers. Development centers are located in Detroit and Germany. By 2000 Continental had jumped to number 12 in the global ranking of sales to auto- makers (see table 2.4.) As tables 2.4 and 2.5 show, most of the largest and most rapidly growing suppliers providing auto parts and modules are based in North America. Consider the example of Lear. The company's focus is on interior modules and systems, which are used in vehicles bearing the nameplates of Audi, BMW, Chrysler, Daewoo, Fiat, Ford, General Motors, Honda, Hyundai, Isuzu,Jaguar,Mazda,Mercedes,Mitsubishi,Nissan,Opel,Peugeot,Porsche, Renault, Rover, Saab, Subaru, Suzuki, Toyota, Volkswagen, and Volvo, Table 2.4 Top Fifteen Motor Vehicle Parts Suppliers, by World Rank, Sales, and Compound Annual Growth Rate, 1995­2000 Compound World OEM annual sales (millions growth rate Home country World rank of U.S. dollars) (percent) and region Company 1995 2000 1995 2000 1995­2000 North America United States Delphi 1 1 26,400 29,100 2 United States Visteon 6 3 9,200 19,500 16 United States Lear 13 5 4,707 14,100 25 United States Johnson Controls 15 6 4,420 11,869 22 United States TRW 7 7 6,100 11,000 13 Canada Magna 19 8 3,223 10,099 26 United States Arvin Industries 32 13 1,792 5,153 24 United States DuPont Auto 18 14 3,500 5,100 8 Average 7,418 13,240 17 Japan Japan Denso 2 4 15,000 16,392 2 Japan Aisin World 5 9 11,587 8,301 -6 Japan Yazaki 10 11 5,000 6,000 4 Average 10,529 10,231 0 Europe Germany Robert Bosch 3 2 14,200 20,550 8 Germany Continental 52 12 800 5,500 47 France Valeo 11 10 5,000 8,200 10 Average 6,667 11,417 22 Average for the 7,924 12,205 14 top 15 Source: Automotive News (1996); Crain's Detroit Business (2001). THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 61 Table 2.5 Top 15 Motor Vehicle Parts Suppliers, by Share of Sales in North America, 1995 and 2000 Share of sales OEM sales in in North North America American (millions of market U.S. dollars) (percent) Home country Company 1995 2000 1995 2000 North America United States Delphi 21,800 23,600 83 81 United States Visteon 8,140 14,400 88 74 United States Lear 3,373 8,600 72 61 United States Johnson Controls 3,257 7,596 74 64 United States TRW 3,300 5,610 54 51 Canada Magna 2,579 6,111 80 61 United States Arvin Industries 892 3,252 50 63 United States DuPont Auto 2,500 2,550 71 50 Average 5,730 8,965 72 63 Japan Japan Denso 2,300 3,803 15 23 Japan Aisin World 563 664 5 8 Japan Yazaki 1,600 2,400 32 40 Average 1,488 2,289 17 24 Europe Germany Robert Bosch 1,576 6,200 11 30 Germany Continental 350 1,650 44 30 France Valeo 600 2,246 12 27 Average 842 3,365 22 29 Source: Automotive News (1996); Crain's Detroit Business (2001). among others. Headquartered in Southfield, Michigan, Lear has grown to 120,000 employees working at more than 200 locations in 33 countries. Lear rose from the world's 13th largest automotive supplier in 1995 to the 5th largest in 2000, with record sales of $14.1 billion. As mentioned, the spin-off of the internal parts divisions of Ford and General Motors in the late 1990s created the world's two largest and most diversified automotive parts suppliers, with capabilities to supply complete modules and with global operations from the outset. For example, Visteon has system and module capabilities in chassis, climate, electronics, glass and lighting, interior, exterior trim, and power train. The company cur- rently operates 42 facilities in Canada and the United States; 29 in western Europe; 22 in Asia, 9 in Mexico, 6 in eastern Europe, and 4 in South Amer- ica. In East Asia Visteon operates plants in Changchun, Nanchang, and Shanghai (3) in China; Japan (5); the Philippines; Republic of Korea (5); 62 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Taiwan (China); and Thailand (2). (Three other plants are located in In- dia.) System and module engineering work is carried out in plants in Eng- land (3), Germany (3), Japan, and the United States (4). Although outsourcing is an industry-wide phenomenon, our research has also identified significant variations in the speed, extent, and nature of deverticalization among automakers. Ford and General Motors, long among the most vertically integrated automakers, have been aggressively outsourcing to cut costs and reduce overhead, both by increasing their use of outside suppliers and, as noted, by moving to spin off their internal parts subsidiaries as independent "merchant" firms. Even so, sourcing is still fairly traditional at Ford and General Motors, which have globally centralized and notoriously predatory purchasing organizations. In the resulting atmo- sphere of price pressure and mistrust, suppliers are only slowly and irregu- larly gaining influence over design. There is some experimentation with preselection of suppliers and involvement prior to project approval where suppliers are asked to bid on the parts they would like to design and pro- duce, but the drive toward lowest-cost sourcing and ongoing cost reduction is still very strong. As a result, there is tension between the purchasing or- ganization, which pushes for lower costs, and manufacturing, which pushes for modularity, local content, and co-location. DaimlerChrysler's Chrysler Division, by contrast, has long sourced as much as 70 percent of the value of its vehicles from outside suppliers. The relationship of Chrysler with sup- pliers is far more consultative than that of Ford or General Motors, and the company has asked suppliers to perform a significant amount of module design and engineering work. Japanese automakers are well known for their extensive reliance on multi- tiered supplier networks and high levels of outsourcing. The nature of Japan- ese supplier networks tends to be more "captive" than those that have been developedbyAmericanandEuropeanfirms--thatis,Japanesesupplierstend to be more dominated by their largest customer. For example, Japan's largest supplier, Denso, a Toyota Group company, generated half of its rev- enues from Toyota in 1997 and none from Toyota's arch rival, Nissan. On the one hand, such captive relationships inhibit the build-up of external scale economies and engender financial and technological dependence of sup- pliers on their largest customers. On the other hand, the integral nature of product architecture in motor vehicles may well favor this kind of tight inte- gration between lead firms and their suppliers, as the continued superior per- formance of firms such as Toyota may attest. In this hierarchical system, it is not surprising that the trends toward modularity and the outsourcing of component design and engineering are much weaker among Japanese automakers. As mentioned, Volkswagen has pursued modularity and final assembly plant simplification largely as an in-house strategy, although its THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 63 plants outside of Germany appear to be making much greater use of exter- nal suppliers. Premium European brands such as BMW and Mercedes have outsourced very little of their component design and engineering. Prospects for Local Suppliers in East Asia Global suppliers in the motor vehicle industry most often establish pro- duction facilities in developing countries at the behest of their customers who have set up final assembly plants and are trying to meet local con- tent requirements. While suppliers are hard pressed to establish plants in all of the locations where their customers operate, they are often willing to invest in certain countries, especially those where multiple automak- ers are present and where lower operating costs raise the possibility of large-scale exporting. China is seen as such a location, as are smaller countries with dynamic vehicle markets and the possibility of raising quality to world standards without too much trouble, such as Thailand. What is clear is that having a plant in each East Asian country makes no economic sense to suppliers and that trade liberalization will very likely lead to dramatic consolidation. The rise of suppliers with global reach and the technical sophistication necessary to design and produce complex modules and subsystems is mak- ing supplier-oriented industrial upgrading in the motor vehicle industry increasingly difficult. To paraphrase a manager of a Taiwanese (Chinese) automotive parts producer, Despite lower operating costs than Western suppliers, we cannot win business in ASEAN outside of our home economy because we are not present in the industry's design centers. In China, we suffer from the ad- ditional problem of not being part of the local supply base. We have not been asked to follow our American customers to new locations such as In- dia. They prefer to work with American suppliers worldwide. In general, it is difficult for Taiwanese [Chinese] firms to get into the American sup- plier networks because all Big Three design is done in Detroit. To supply Ford, a company needs to have at least an office in Detroit and, in some cases, manufacturing facilities there too. If we were in Detroit, we could get assembly drawings and make bids on parts. We have more or less given up on building business with the Big Three. This is a critical problem be- cause the local market is slow growing. Only by expanding regionally, es- pecially into China, will we be able to attain better scale economies. But the [vehicle] platforms must be the same. The ongoing globalization of the motor vehicle industry and the new role of suppliers in supporting the increasingly integrated global operations of automakers have important implications for supplier-oriented development in East Asia. The potential for local sourcing in the motor vehicle sector is 64 GLOBAL PRODUCTION NETWORKING IN EAST ASIA high because of the large number, size, and weight of components and ma- terials. With tariffs on imported parts, locally operating suppliers have a huge advantage. For example, a manager at a global automotive parts sup- plier in China stated a willingness to pay local suppliers a price equivalent to the worldwide price plus tariffs and transportation costs. What is clear is that local content, when it is provided, will come from the affiliates of global suppliers; a lack of technical competence has generally confined locally owned suppliers to simple, standardized, and slow-changing components, such as bearings, where there is a wider market that supports adequate scale economies. Local suppliers in Asian NIEs have gained little responsibility for design. As a manager of a German automotive parts producer in China put it, "De- sign responsibility remains in Germany." According to a local supplier in- terviewed in Taiwan (China), "Ford does not allow changes in the `hot' or critical points of cars, such as gear boxes, engine, chassis but only `soft point' changes, for example, front- and back-end styling." The result is a some- what loose technological relationship between actors in the offshore sup- ply chain, while tight linkages are forged and maintained in the advanced economies. This puts a ceiling on the engineering capabilities that can be developed at suppliers based in emerging economies. According to a man- ager at an American-owned auto parts firm in China, "The auto industry in China is a `veneered industry'--while it appears that Chinese firms are ma- jor suppliers to assemblers, all are propped up by huge amounts of imports." Another manager at a foreign-invested manufacturer of fuel system com- ponents and wire harnesses in China stated, We use only five or six local suppliers, and local content is only around 2 per- cent. Even for relatively simple products like wire harnesses 85 percent of ma- terialsisimportedfromplaceslikeJapanandKorea.Becauseofcustomsduties, transportation costs, "harassment" at customs, etc., these materials are very ex- pensive. Automakers do not even buy all their steel in China because they can- not get the necessary quantities at the right quality. [The lack of] true local sourcing is the biggest impediment to growth of the industry in China. China hasn't done enough to allow its basic industries to survive. If the raw materials sectors do not develop, Chinese firms that do not have access to a global sup- ply system will die after WTO is implemented. First-tier suppliers are pres- suring their material suppliers to move to China. We have given many of our suppliers an ultimatum: by a certain year, they must either establish facilities in China or provide the company with imported material at a "domestic" price--the price they could achieve if they were located in China and not pay- ing customs duties, transport costs, and so forth. Material suppliers, though, are reluctant to move to China because they are not sure they can reach criti- cal economies of scale and because there is a shortage of basic commodity in- THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 65 puts in China. Material suppliers are particularly sensitive to commodity prices and quality because as one moves further up the supply chain in wire harnesses, materials become far more important than labor in determining total costs. The result of these tensions is that the supply chains that are emerging in developing countries are increasingly foreign owned and very "thin." Core design activities remain concentrated in advanced economies, and many parts and materials continue to be imported. Trade, market, and in- vestment liberalization will cause a further thinning as specific activities consolidate in particular places and export either regionally or globally. Some of the respondents interviewed in the Chinese motor vehicle indus- try believed that foreign automakers have no intention of manufacturing locally when China joins the WTO and drops its import restrictions. This view sees current operations as loss leaders that provide foreign automakers with a means to develop brand recognition and distribution and service networks for vehicles that they will later import. Thun (2001) suggests two alternative strategies for local motor vehicle parts suppliers facing these difficulties. The first calls for local suppliers to join forces with global suppliers, either through mergers, joint ventures, or alliances. Local suppliers can offer global suppliers additional production locations and help to create "complementarity schemes," where vehicle components are exchanged within a region to offset tariffs and concentrate assembly so as to enhance scale economies. The second strategy requires local suppliers to serve global suppliers at a lower tier, supplying local con- tent. But as local content rules are phased out, many parts will be produced on a regional basis to increase scale economies, undermining the rationale for regional complementarity schemes and putting further pressure on lo- cal firms. This pressure may force local suppliers from smaller Asian coun- tries to invest in regional production centers, where they will either come into direct competition with global suppliers or, if the second-tier option is followed, with local suppliers following the same strategy. A third option, of course, is to exit the business entirely. OLD AND NEW REQUIREMENTS FOR SUPPLIERS IN GLOBAL PRODUCTION NETWORKS: DRIVERS, RESPONSES, AND OUTCOMES The evidence presented in the previous two sections points to the emergence of a new organizational structure underpinning the GPN in two key manufacturing industries. The rise of this global supply base can best be understood in terms of drivers, responses, and outcomes. 66 GLOBAL PRODUCTION NETWORKING IN EAST ASIA First, the "deverticalization" of value chains has been driven by increased levels of international competition, rising competence in the supply base, and a belief among lead firm managers in the West in the doctrine of "core competence" (Prahalad and Hamel 1990). Second, the ascendancy of the WTO and the strengthening of regional trading blocs in Europe and North America is driving high-volume production to low-cost sites within the blocs (that is, Eastern Europe and Mexico) and to the largest countries of Asia (that is, China and India). This trend toward the regional organization of production has been reinforced by the rising demand in end-user markets for rapid-response, build-to-order, and configure-to-order performance on the supply side. At the same time, deverticalization has been enabled by the increased use of information and communications technologies through- out the supply chain, but particularly in the design process and in supply chain management. These drivers have led to the emergence of lead firms with little if any in-house manufacturing and the rapid expansion, growing financial strength, and increasing competence of the largest external suppliers of core manufacturing services, which in a self-reinforcing dynamic of "in- dustry co-evolution" has fueled further outsourcing by lead firms and more consolidation within the supply sector (Sturgeon and Lee 2001). An impor- tant outcome of this process has been the growing requirement for greater involvement by suppliers early in the design process, which has strength- ened the competitive position of suppliers that are able to co-locate and co- ordinate engineering and production activities with those of their customers at a global level. These changes are raising the threshold of performance for supplier firms in several areas, including design and engineering, sourc- ing, the effective use of information and communications technology, and the ability to operate in and coordinate between multiple locations. In the following paragraphs we discuss several of these developments in more detail, with an emphasis on the new requirements they raise for suppli- ers, paying particular attention to the implications for supplier firms in East Asia. Deverticalization The deverticalization of firm structures is at least partly a response to the increased volatility of the markets in which the firms are operating. Shorter product life cycles, the complexity and high cost of introducing new prod- ucts, and unforgiving end-user and capital markets are combining to impose new pressures on all areas of the firm, from research and development to manufacturing to marketing and sales. In response to these pressures, and THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 67 to offload risk, firms in a wide range of sectors and locations have sought to outsource non-core functions, especially those, like manufacturing, that are associated with large-scale fixed capital. Outsourcing has been especially prominent in competitive and industries like electronics, motor vehicles, and textiles and apparel (Fine 1999; Quinn and Hilmer 1994). Lead firms have focused on the areas and functions that they believe to be essential to the creation and maintenance of competitive advantage, es- pecially product innovation, marketing, and other activities related to brand development, and have increasingly come to rely on specialized suppliers to provide non-core functions. The belief is that by divesting non-core func- tions, lead firms can reap value more quickly from innovations while spread- ing risk in volatile markets (Venkatesan 1992). Firms that develop, market, and sell electronic hardware, cars, and clothing have turned to suppliers for production and, increasingly, post-architectural (that is, detailed) design services. By tapping the competencies of suppliers, lead firms are able to maintain substantial market presence without the fixed costs and risks of building and supporting a vertically integrated corporate organization (Sturgeon 2000, 2002). Among the advantages claimed for these production networks are that they are more adaptable than an integrated firm would be and are capable of providing better economic performance in highly com- petitive and volatile markets (Cooke and Morgan 1993; Powell 1990). Design and Engineering Lead firms in the advanced economies are asking their suppliers to take on more responsibility for the design and engineering of products and sub- systems. Suppliers are increasingly being chosen and brought into the de- velopment process before products are fully designed. By doing so, the lead firm is able to spread risk and reduce costs. When suppliers participate in the development of prototypes, moreover, they typically improve their per- formance in design for manufacturability and in the implementation of sub- sequent engineering change orders. Product redesign for different markets is also easier and quicker if suppliers are actively involved from the outset. One of the keys to achieving heavy supplier involvement in early devel- opment activity is the promise of future business. Small numbers of suppli- ers or even sole sourcing has become common for specific product models and generic product platforms. Conversely, situations in which a new sup- plier is brought on board after volume production has begun are becoming increasingly rare. In industries such as motor vehicles, where models remain in production for two to five years, business is won and lost by suppliers in large blocks and for long periods. The lead firms try to reduce supplier mar- 68 GLOBAL PRODUCTION NETWORKING IN EAST ASIA ket power in various ways (by using different suppliers for different models and types, for example), but a supplier for a successful model can gain a great deal of leverage. Many Asian suppliers have aggressively pursued design and engineering competencies (although, as table 2.1 shows, many more remain at the OEM stage). Leading Asian suppliers in the electronics, motor vehicle, and textile- apparel industries now offer a full range of design and engineering services. Advanced Taiwanese (Chinese) notebook computer ODMs and Hong Kong (Chinese) garment producers, for example, provide their customers with completely designed finished products as well as products that are made-to- specification and co-designed. However, few Asian suppliers are yet able to provide design services that are co-located with those of their customers. One manager at an American- based global motor vehicle parts supplier observed, "Co-location of design with our customers' vehicle development is important because the design interface is very complex and technology changes very quickly. Face-to-face meetings are still needed to resolve design issues." Continuing advances in broadband communications and information technologies may make long- distance concurrent engineering between lead firms and suppliers a more viable alternative to co-location, but to the extent that co-location remains important or becomes more so, suppliers with design facilities located near their customers will have an advantage. Information and Communications Technology The increasing use of advanced information and communications technol- ogy (ICT) has accelerated the possibilities for effective integration in geo- graphically and organizationally dispersed value chains. Firms carrying out functions at different points along a value chain have a greater ability to ex- change data so as to achieve high degrees of conformance with specifica- tions and tight coordination of productive activities. In the past, achieving such conformance and coordination required firms to locate the relevant functions within their own vertically integrated organizations or within tightly controlled networks of subordinate suppliers. The promise of ICT is that codifiable specifications and standard interfaces will make it possi- ble to coordinate activities through market-based exchanges among auto- nomous organizations. So far, the most important contribution of ICT has been in manage- ment of the supply chain and in the design and manufacturing of products and components. Specific applications in the area of supply chain manage- ment include enterprise resource planning systems, business-to-business THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 69 e-commerce marketplaces, and electronic data interchange. The application of digital technology to the product and component design process involves tools such as electronic design automation, computer-aided engineer- ing, and computer-aided design. These technologies, when combined with computer-aided manufacturing systems embedded in numerically controlled and robotic production equipment, allow complex product specifications to be handed off to outside suppliers. Using these tools, suppliers can cre- ate fully computer-integrated manufacturing environments to track prod- uct quality and inventory and shorten design and production cycles. New ICTs enable lead firms to ask much more of established suppli- ers in terms of rapid response, design collaboration, lower costs, and close monitoring. It is not unusual for lead firms to exert great pressure on their suppliers to adopt the latest ICT to improve quality, facilitate the tracking of in-process inventory, and streamline the order and reorder process. Whether new ICT systems raise or lower barriers to entry is an open question. Leading-edge ICT systems are expensive and often must be adapted to the specific requirements of the firms that use them. Successful adoption of ICT requires competent vendors and specialized personnel to build and operate the new systems and frequently also requires user firms to adapt their organizational routines to accommodate them. ICT systems typically have short life cycles and must be continually upgraded to remain compatible with customer systems and changing industry standards. Stay- ing on this treadmill of ongoing capital spending, learning, and organiza- tional change is only possible at considerable effort and expense. Leading Asian merchant manufacturers and suppliers such as Taiwan Semiconduc- tor and Li and Fung are among the most sophisticated users of ICT to be found anywhere, and in our field research we came across many firms that had adopted the latest information technologies. Examples include semi- conductor foundries, producers of personal computers and peripherals, and textile-apparel suppliers based in Taiwan (China); electronic component suppliers based in Singapore; and textile-apparel suppliers in Hong Kong (China). But many smaller firms in East Asia, especially from the de- veloping economies in the region, are struggling to keep up. One indica- tion of the general situation with respect to the adoption of ICT is provided by table 2.6, which reports the (very low) rates of Internet usage by manu- facturers in China and the four ASEAN economies of Indonesia, Malaysia, the Philippines, and Thailand.8 8. About 24 percent of the ASEAN firms also reported that they participate in an electronic data interchange network. 70 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Table 2.6 Internet Use by ASEAN-4 and Chinese Manufacturing Firms (percent) Type of use ASEAN-4 China Communicate with clients via Internet 24 6 Communicate with suppliers via Internet 15 5 Place orders via Internet or e-mail 1999 10 2 2000 14 4 2005 (estimated) 23 6 Source: Survey conducted by the World Bank in early 2001. ICT systems, especially for supply chain management, are relatively new and still rapidly evolving. Vendors have emerged from various fields (man- ufacturing automation, enterprise computing, and so forth), business mod- els are often untested and experimental, battles over standards are in full swing, and merger and acquisition activity among information technology vendors is proceeding at a rapid pace--and, since the bursting of the dot- com bubble in the United States, some have gone out of business. More- over, the vendors and standards used in different industries are often very different. Our interviews suggest that, given the rapidly changing technol- ogy and standards, user firms may derive advantage from being located near the nexus of information technology innovation. In places such as Silicon Valley, users and vendors have set up informal and formal working groups intended to set standards and develop better applications (Sturgeon 2003). Although all have sales offices around the world, the leading vendors of information technology in the field of supply chain management are based in California (Ariba, Oracle, and PeopleSoft), Colorado (J. D. Edwards), Germany (SAP), and the Netherlands (BAAN). None of the major vendors is based in Asia, which may be a disadvantage for Asian suppliers. Global Suppliers As lead firms have outsourced more of the manufacturing, sourcing, and lo- gistics functions that were previously carried out in-house, their preference for dealing with suppliers with international reach has grown. The reasons are several. First, the lead firms are in many cases marketing their products globally and require engineering, manufacturing, and logistics support in multiple locations. Second, the lead firms often seek to economize on development costs by creating global product platforms that share and re- use many common parts, modules, and subsystems. Partnering with a small number of suppliers, or even a single supplier, enables lead firms to exploit THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 71 these economies of scope more fully, while avoiding the cost of requalifying new suppliers for each new market. Third, cost pressures require purchas- ing organizations to scan the world for low-cost, high-quality parts, and, to the degree that suppliers are taking on these responsibilities, they too must have global sourcing capabilities. Fourth, suppliers based in protected final markets can combine global sourcing with local sourcing and subassembly to help lead firms meet local content requirements. Fifth, the preference for key suppliers to take on a more active role early in the development process requires these suppliers to be able to co-locate at least some of their own de- sign activities with the design facilities of their customers. Some lead firms have given their key suppliers an ultimatum: provide support on a global basis or lose the business entirely. Managers at three global automotive sup- pliers made essentially the same point during separate interviews with IPC researchers: The industry began to change 5­10 years ago. Today it is a requirement to serve platforms--it is part of the bid. If a supplier doesn't have a global strat- egy, it can't bid. New projects are no longer seen as an opportunity to expand globally--instead, a supplier must have a global base in place to even make a bid. This forces suppliers to have a global supply system in place. Suppliers must support assemblers as a sole source for global product lines to support commonalization. We must supply the same part, with the same quality and price, in every location. If [the automaker] says to go to Argentina, we must go or lose existing, not just potential, business. Logistics are becom- ing a key competitive advantage; we must have the ability to move production to where customer's facilities are. We want our plants to be present where vehicles are produced. Sometimes customers ask us to locate near offshore assembly plants to provide local con- tent . . . . We will follow our customer's strategy by establishing local engi- neering operations in large emerging markets only, such as Brazil, Korea, and Mexico. Providingthiskindofsupportinvolvescoordinatingflowsofcomponents, subassemblies, and products across production networks that often span sev- eral countries or even continents. It also requires setting up design opera- tions close to the design centers of the lead firms. As a result of the popularity of outsourcing among American and European lead firms, the great major- ity of the suppliers that have risen to this challenge so far have originated from Europe and North America, where the lead firms have the bulk of their design activities, where there is a deep pool of management talent with long experience in international operations, and where capital is available to finance global expansion. The pressure to expand has been met partly by internal growth but even more by aggressive merger and acquisition activity. 72 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Acquisitions of competitors in similar lines of business have yielded sud- den jumps in geographic coverage. Acquisition of firms with upstream or downstream capabilities has broadened the range of products and services on offer (frequently, acquisitions have served both purposes simultane- ously). Finally, acquisitions of customer facilities have also helped suppli- ers to win new business while expanding their geographic and functional scope.9 The rise of regional production systems under trade arrangements such as the European Union and NAFTA has strengthened the hand of global suppliers in several ways. First, the drive to serve advanced economies from proximate low-labor-cost locations such as Eastern Europe and Mexico, for both tariff and speed-of-response reasons, has reduced the competitiveness of suppliers serving advanced-economy markets solely from Asian locations. Some Asian suppliers in the electronics and textile-apparel sector have re- sponded by opening plants in places like Central America and Mexico, but most have not. For global lead firms and suppliers, fragmented national pro- duction systems are being replaced by regional systems that allow increases in scale economies in plants serving regional markets. This regional strat- egy is nested within a global strategy that seeks to commonalize and reuse as many design elements as possible among regions. Suppliers that do not or cannot put a global or regional system in place to match the strategy of the lead firms in their industry, because either they are too small or are con- centrated in a single region, may find long-term success elusive. For Asia the increased popularity of the regional or global model has meant the consol- idation of production in China, and more global suppliers are scaling back existing investments in Southeast Asia and building up very large-scale op- erations in China. In the words of the president of Manufacturing Services Limited, a medium-size electronics contract manufacturer based in Con- cord, Massachusetts, "We want to consolidate [the Malaysia and Singapore operations] in China . . . we want to supersize our China operation. They have a great pool of available labor" (Serant 2001). 9. For example, Sanmina, an electronics contract manufacturer based in San Jose acquired Nortel's Wireless Electro-Mechanical Subsystem Assembly in August 1999. This deal included inventory, production equipment, and 230 employees in manufacturing facilities in Calgary (Canada) and Châteaudun (France) as well as a small engineering design group located in Guyancourt (France) outside Paris. Both the Calgary and Chateaudun facilities were located adjacent to Nortel's wireless system design houses, allowing the production facilities, now owned by Sanmina, to remain poised to bring new products quickly into production. The acquisition provided Sanmina with new loca- tions in Canada and France, locations that can support not only Nortel but its other customers, as well as new expertise in radio-frequency electronics manufacturing. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 73 For established East Asian manufacturing firms, the requirement for global reach may be the most challenging of all the performance require- ments associated with the emerging supply base. According to a manager at one global motor vehicle parts supplier, "Supplier logistics need to be syn- chronized with [lead firms], and many local suppliers do not have the capa- bility to do this." Except for the Japanese, few East Asian suppliers have expanded geographically outside East Asia. Most have lacked access to the capital that would enable the aggressive merger and acquisition strategies needed to build an international position. Many East Asian firms are still controlled by their founders or the founders' family members. Other, pub- licly held, firms are listed on thinly capitalized local exchanges that lack the liquidity of U.S. and European public equity markets. Most of these firms, too, have lacked access to the managerial expertise needed to operate suc- cessfully outside Asia. Of 100 ASEAN-4 manufacturing firms responding to the World Bank's 2001 survey, only 16 percent reported employing man- agers with any foreign work experience. RISKS AND UNCERTAINTIES OF THE GLOBAL SUPPLIER MODEL The organizational and technological changes we have described in the pre- vious sections are still emerging, and many questions have been raised about their sustainability. Some of the strongest doubts about the global supplier strategy were expressed during our interviews with Japanese lead firms in the electronics sector, but our research at lead firms in Europe and the United States also revealed conflicts and confusion over the desirability of increased outsourcing. Some of these concerns are of long standing but have grown more acute as the pace of outsourcing and supply base consolidation has ac- celerated. There are fears of the loss of innovative capacity as production- related activities migrate to global suppliers, of the leakage of proprietary information to competitors through shared suppliers, of the creation of com- petitors if suppliers choose to move into the production of final products, and, more generally, of the increased market power of the largest suppliers. There are also concerns about the feasibility of product design modulariza- tion in cases where unanticipated or complex interdependences between the components of the product lead to much higher coordination and transac- tion costs. There are also examples of outsourcing leading to additional ef- fort. In some lead firms we observed the emergence of "shadow" engineering organizations, whose ostensible role in monitoring cost and quality and maintaining the ability to switch suppliers has expanded into large-scale du- plication of supplier engineering work. 74 GLOBAL PRODUCTION NETWORKING IN EAST ASIA The reliance of lead firms on one or just a few external suppliers, or sometimes even a single plant, when coupled with the reduced inventories associated with just-in-time delivery practices, has significantly increased the vulnerability of their operations to disruption from external work stop- pages, accidents, or natural disasters. When key supply plants go off-line, the whole system can go down. Other questions have been raised about the sustainability of the suppliers' business model itself. Some global suppliers have encountered difficulties in finding a way to get lead firms to pay for the additional engineering, com- ponent purchasing, and logistics support they are providing, leading to very low or negative profit margins. Moreover, the rapid international expansion of supplier firms, often driven by acquisition of companies in different insti- tutional settings with dissimilar corporate governance systems, industrial re- lations practices, product and process technologies, and ICT systems, may be re-creating, or even exacerbating, the operational inefficiencies that out- sourcing was (in part) intended to remedy in the first place. The separation of product strategy and design from manufacturing also raises concerns--especially at Japanese electronics firms, where there is a strong tradition of using what is learned on the shop floor to inform and im- prove product design--that outsourcing will undermine innovation in the long run. Whether product innovation can be sustained in the face of high levels of outsourcing remains to be seen. Similarly, there are concerns that process innovation will suffer when so much production is in supplier firms with fewer capabilities in and funding for research and development. The recent global economic downturn presents one more challenge to these global supply networks, which owe much of their growth to the long, uninterrupted economic expansion of the 1990s and whose stability in re- cessionary conditions is only now being tested. It also remains to be seen whether this pattern of deverticalization and global-scale organization will be affected by the September 11 terrorist attacks and their aftermath. The pattern is real and has considerable momentum. Even so, a caution against assumptions of irreversibility and inevitability is appropriate, especially at the present juncture. Research on these questions is continuing at the MIT Industrial Performance Center. JAPANESE SUPPLY CHAINS IN EAST ASIA Japanese firms have had a major impact on industrial upgrading through- out East Asia. The flows of technology, investment, and training re- sources from Japan to other East Asian economies have been substantial, THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 75 and today Japanese firms are sourcing throughout the region on a large scale. Supplier-oriented upgrading in East Asia based on linkages with Japanese firms has taken two distinct, though sometimes overlapping, forms. The first has occurred through the acquisition by developing-country suppliers, often through licensing or joint ventures, of process technologies forthemanufactureofinputsandcomponentsthathavebecomehighlycom- petitive, standardized, and unprofitable for Japanese producers. Japanese firms, pursuing the "flying geese" strategy of letting older, less profitable in- dustries migrate to their developing neighbors, have systematically devel- oped suppliers in East Asia for inputs that are becoming commodities or are subject to great demand volatility. The logic of the flying geese strategy is to create a "shock absorbing" supply base outside Japan capable of meeting the variable portions of world demand. This is achieved by off-loading pro- cess technologies for product categories where commoditization results in intense price competition and low profit margins. With an alternative source of supply for low-value items in place, higher-value components and inputs, and many of the final products using these components, can continue to be produced in Japan.10 In the 1970s and 1980s Japanese firms began to transfer synthetic tex- tile and steel production technology to firms in Korea and Taiwan (China). In the 1980s process technology for the production of computer memory chips, or DRAMs, was transferred to Korean firms.11 In the mid-1990s Japanese firms began to transfer process technology for flat panel displays to Korean and Taiwanese (Chinese) firms (Akinwande, Fuller, and Sodini 2001). In this way, an important path to industrial upgrading in Korea and Taiwan (China) served simultaneously (and, from the Japanese perspective, more importantly) as a strategy for promoting industrial transformation in Japan. Technical progress in developing countries based on the flying geese model has been widely documented (see, for example, Encarnation 1999). Amsden (1989) refers to it as "apprenticeship," the acquisition of foreign technology through licensing and technical assistance programs, and dif- ferentiates it from "imitation," the copying of foreign technology through 10. American and European firms have also licensed process technology to firms in East Asia, but these moves have often been motivated by short-term tactical considerations (usually the gener- ation of cash in times of trouble) rather than strategic ones. In some cases, however, Western firms have taken a more strategic approach. The development of a supply base in Taiwan (China) for low-cost desktop and notebook personal computers by American computer firms such as Com- paq and Dell is a case in point. 11. Taiwanese (Chinese) firms acquired DRAM technology from the Japanese and also from European producers but have met with very limited success. 76 GLOBAL PRODUCTION NETWORKING IN EAST ASIA tactics such as reverse engineering, a method that has also been widely used in the Asian NIEs.12 For developing economies, the risk of upgrading via the flying geese model comes down to one of low profitability due to technological lag. Since the technology that has been transferred has invariably been one or more generations behind the leading edge, countries that have relied on this strat- egy for development find themselves on the "bleeding edge" of many mar- kets, stuck in low-profit and volatile sectors such as low-cost personal computers, computer monitors, scanners, power suppliers, batteries, key- boards, DRAMs, and mass market apparel and footwear products. More profitable product segments, such as high-end computers and servers, com- munications equipment, software, logic semiconductors, and high-fashion apparel and footwear, continue to be pursued primarily by American, Eu- ropean, and Japanese firms working closely with advanced users, who are concentrated in the advanced economies. The second variant of industrial upgrading that has been based on linkages with Japanese firms is that of FDI. Japanese firms have a long history of FDI in East Asia, and Japanese investment has in the past been widely dispersed throughout the region--much more so than U.S.-originated FDI (Mason and Encarnation 1995). Until the mid- 1980s these investments, which were concentrated in consumer and white goods, consumer electronics, and automobiles, were established mainly to serve protected local markets. Over time, Japanese firms gained a strong foothold in all but the most protected markets in East Asia. In East Asian countries other than Korea, for example, Japanese motor vehicle firms have a market share that exceeds 90 percent (Doner 1991; Sturgeon and Florida 1999). The appreciation of the yen in the mid-1980s accelerated Japanese investment in Southeast Asia, and some production platforms were upgraded and expanded to serve ex- 12. Amsden (1989, p. 110) argues that the acquisition of advanced production equipment has been a major avenue for industrial upgrading in East Asia. Productivity increases are drawn from the world "technology shelf" through imports of foreign technology embedded in advanced production equipment. Operating advanced production equipment at scale economies sufficient to minimize unit costs, and learning to use it more efficiently than foreign rivals, enhances competitiveness on world markets. As long as profits from increased output are invested in new equipment that embodies the latest technology, growth can be maintained. Increased output results in greater scale economies and expands the opportunities for learning by doing, improving efficiency and increasing wages, and driving growth in the domestic market. Although the state can do a lot to initiate this process, effective application of new technology depends on what happens on the shop floor, which according to Amsden helps to explain the importance of managers over entrepreneurs in countries, such as Korea, that have pursued this manufacturing-led approach to industrial upgrading. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 77 port markets in the West.13 But the cross-border networks led by Japanese firms have tended to rely on internal subsidiaries and on Japanese suppli- ers, and, given the scale of investment, the opportunities for local firms to upgrade their competencies via supply relationships have consequently been very limited.14 To the extent that local firms were included, it was usu- ally in the smallest and least promising markets such as the Philippines and Vietnam, where they were used as final assemblers and retailers who would work under license, source all components from Japanese firms, and absorb the risk of distributing and selling finished products in small and uncertain markets. In a few cases where the network relationships were of long stand- ing and the local supplier's capabilities were high, local firms were tapped to serve the export market. But such firms were never part of the core of the production network and were particularly vulnerable to changes in the lead firm's strategy. The experience of a Philippines-based independent consumer electron- ics contract manufacturer is a case in point. The firm started out in 1965 as a television distributor for a major Japanese consumer electronics company, and, following the introduction of tariff barriers for fully assembled sets, began assembling televisions for the local market under license in 1971. By 1994 production at this dedicated facility had increased to 12,000 units per month. In 1977 the company began assembling televisions under license for a second Japanese consumer electronics firm, initially for the local mar- ket and, then, beginning in 1993, for export. In 1994 the company built a 1 million unit per year plant in an export-processing zone. Production at this plant quickly rose to 40,000 units a month. The license agreements re- quired the company to source components from its Japanese partner's sub- sidiaries, some of which were located in Japan and Singapore, and also allowed the company to provide after-sales service, marketing, and distri- bution in the Philippines. As the company grew, it added a plastic injection molding division, which was spun off as an independent company in 1994. Its strategy was eventu- ally to develop its own original brand of products in addition to its contract 13. Japanese export platforms in Southeast Asia were not as extensive as they might have been, because Japanese firms began at the same time to make very large investments within or adjacent to Western markets (Abo 1994; Curry 2000). 14. Some analysts have contrasted the closed Japanese production networks in Asia with the rel- atively open networks led by North American--and to a lesser extent European--firms, which they argue have created more opportunities for East Asian firms to upgrade their capabilities (see Borrus, Ernst, and Haggard 2000; Ernst 1997; Ravenhill 1995). On the captive, hierarchical char- acter of Japanese-led production networks, see Aoki (1987), Dore (1986), Gilson and Roe (1993), Sako (1989), Schonberger (1982), Sturgeon (1999), and Womack, Jones, and Roos (1990). 78 GLOBAL PRODUCTION NETWORKING IN EAST ASIA manufacturing and distribution business. After the Asian financial crisis, however, the first Japanese electronics customer began to shift its export production to its plant in Malaysia, and production at the dedicated plant declined from 12,000 to 2,000 units a month; orders ceased completely at the end of 2001. By 1998 the second customer had also withdrawn all its or- ders. The plant survived by serving a new customer, a Taiwanese (Chinese) producer of monitors. Management believes that its television business is unlikely to recover be- cause the global market is saturated, and television manufacturers are fo- cusing on high-value television sets whose large size and weight leads to a preference for assembly close to final markets. The company is trying to di- versify into broadband communications services, real estate development, corporate financial services, and food and agriculture businesses, but the future is uncertain. The injection molding unit did develop new business in beer crates and motorcycle, refrigerator, and water cooler parts, but man- agement has found that many of the Japanese firms producing these prod- ucts locally prefer to rely on their own suppliers, which have followed them to the Philippines. The recent experience of this Philippine manufacturer is far from unique. Japanese production networks throughout East Asia have been undergoing a process of rationalization. As a local supplier to both the Ford and Nissan joint ventures in Taiwan (China) noted, "Japanese firms have slowed invest- ment in new capacity, rationalized production in Southeast Asia, increased investment in China, scaled back internal capacity, and outsourced more activities, including engineering." The combination of the protracted economic downturn in Japan, the continuing weakness of other Asian economies, the lowering of tariff barriers for finished products in the region, and the saturation of markets for some products in smaller East Asian coun- tries is leading to the centralization of production at a smaller number of lo- cationsinordertoeliminateexcesscapacityandtoexploiteconomiesofscale. In many cases this has meant a new focus on production in China at the ex- pense of other East Asian locations. For local suppliers elsewhere in East Asia, therefore, the prospects for upgrading through links with Japanese lead firms do not seem as promising as in the past. In a sense, the new preference of Western lead firms to use Western sup- pliers on a global basis simply brings these production networks more in line with those led by Japanese firms. While FDI of this type can do much to create jobs and train workers in advanced technical and business processes, its contribution to the development of a vibrant locally owned supply base, and thus to the strategy of supplier-oriented industrial up- grading, is less evident. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 79 POLICY IMPLICATIONS OF THE NEW GLOBAL SUPPLY BASE FOR EAST ASIA'S FUTURE DEVELOPMENT Throughout East Asia, many manufacturing firms have moved along the in- dustrial learning curve by mastering contract manufacturing under the tute- lage of some of the world's most advanced lead firms, which have, in turn, marketed the products under their own brands. In the automobile and elec- tronics sectors on which our research has focused, East Asian managers described learning how to make products well with help from their early customers and then improving their products by selling to ever more de- manding buyers. A company's capabilities as a manufacturer are proven, in effect, by its clientele. The fact remains, however, that in this common sce- nario, the manufacturer captures only part (and often relatively little) of the final price of the good, remains at the beck and call of the buyer, and is vul- nerable to competition from lower-cost competitors that threaten to take the business away. Although this is true for many suppliers in all economies, the problem is made acute in places such as East Asia, where a great many sup- pliers are not co-located with their customers or with their final markets and so have little opportunity to move beyond this highly subordinate status. The concentration of lead firms in advanced economies and the increased demands they are imposing on their suppliers mean that many Asian firms are dealing with fewer and stronger customers than they faced in the past. This is not a situation in which the Asian suppliers are inevitably the losers. If the evidence in table 2.7 is anything to go by, some larger Asian suppliers still appear to hold their own in negotiations with their customers. Although it is difficult to conceptualize and to measure power relations in supply chains, the levels of profitability and return on capital of the suppliers compared with those of the lead firms--often their customers--show the suppliers capturing a comparable share of the rewards in the system. The suppliers compared in this table stand out, however, not only for producing at low cost but also for recruiting and nurturing well-educated managers, engineers, and technicians who can work in partnership with customers. These firms have innovative design capabilities that allow them to suggest new model designs to customers (ODM) as well as work to their clients' specifications (OEM). They have implemented major changes in their busi- ness structures and adopted advanced ICT systems that allow them to co- ordinate widely dispersed activities in different countries. Like the chip foundries and electronics contract manufacturers in Taiwan (China), some East Asian firms have aggressively exploited the possibilities of devertical- ization and so have achieved some success in capturing the business that is being spun off by formerly vertically integrated firms. 80 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Table 2.7 Profitability of Contractor and Lead Firms (percent) Profit Sales growth, Return growth, five-year on four-year Firm Economy Sector average shares average Johnson Hong Kong Micromotors 19 15 10 Electric (China) Black and United States Tools 10 13a -1 Decker Quanta Taiwan (China) Laptop maker 11b 57 73 Dell United States Laptop lead firm 9 58 45 Techtronics Hong Kong Tools 7 23 23 (China) Home Depot United States Retail 8 18 25 Hon Hai Taiwan (China) Cables 11 33 57 Intel United States Microprocessors 34 32 24 TSMC Taiwan (China) Microprocessors 0 38 28 Philips Holland Microprocessors 44 10 6 Hung Hing Hong Kong Printing 16 21 16 (China) Bertelsmann Germany Press 7 25 6 Li and Fung Hong Kong Trading 3 49 25 (China) The Limited United States Retail 8 35 6 a. Median value. b. Four-year average. Source: Boillot and Michelon (2001, p. 142). The suppliers in table 2.7 are exceptional. The resources that they have deployed for each of these advances are well beyond the reach of most small Asian suppliers. In the early days of the Asian miracle, even rather small firms might have been linked up to large international customers, who in turn provided many of the services and inputs the firms needed in order to produce to specifications and gradually upgrade their capabilities. Today, however, there are fewer such supports for the small firm. The inter- national customers are looking for suppliers that are already able to make the products--not for firms that can be brought up to the needed level of performance. For Asian manufacturers today, moving away from old-style OEM is a precondition for enhancing the value they create, for protecting their share of the market against competitors--both rivals elsewhere in the region and global suppliers based in the West--and, more broadly, for contributing to the continued economic development of their home societies. We do not THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 81 reject supplier-oriented approaches to industrial upgrading, but new actors and new requirements are posing challenges to any notion of an unprob- lematic or inevitable OEMODMOBM upgrading path. There are a variety of ways in which firms can balance or replace OEM production with new higher-value activities, and the scope for industrial upgrading in the Malaysian context is explored in chapter 3. One route is to acquire or cre- ate a brand and sell some or all of the firm's production under the brand name. We have already discussed the pitfalls of competing directly with cus- tomers in end markets, but there are approaches, such as selling products or selling in markets that customers are not interested in, that can avoid con- flict. A second approach is to continue to make products or components that will be sold under another label, but to develop internal capabilities sufficient to invest these products with design and functional features allowing them to command higher prices and to raise the barriers to the entry of competitors, the classic ODM route. A third is to specialize in process-specific technolo- gies that enable the firm to provide high-quality, low-priced manufacturing and manufacturing-related services for a number of (increasingly demand- ing) brand-name customers, some of whom may themselves have shed those functions entirely; this is essentially the OEM route with an emphasis on continuous improvement. A fourth is to move to the second tier of the sup- ply base, providing domestic content for the local affiliates of global sup- pliers. The fifth and perhaps most radical option for Asian firms is to follow the same strategy of those lead firms in the West that have shed the man- ufacturing function entirely and to tap the resources of the increasingly ubiquitous global supply base, which has in many ways commodified the manufacturing process. This will prove challenging in many Asian societies, which have for so long based their industrial upgrading and industrial poli- cies on the initial establishment of excellence in manufacturing. As we have stressed in this chapter, each of these routes to upgrading poses formidable technological and managerial challenges, but the key point is that industrial upgrading is an unending process that requires constant attention and in- vestment in time, money, and human effort. What concrete steps can East Asian governments take to strengthen the capabilities of their domestic suppliers to participate successfully in emerg- ing global production networks? The most important policies are those whose outcomes benefit all sectors of the economy: macroeconomic sta- bility; transparent and efficient capital markets; confidence in public in- stitutions; protection of individual rights under rule of law; and a strong foundation of education and training. Beyond these general measures, however, additional steps can be taken to address the specific capabilities of the supply base. The details will depend on the stage of development of the 82 GLOBAL PRODUCTION NETWORKING IN EAST ASIA economy, the strength of its public institutions, the particular capabilities of both government and industry, and the prevailing attitudes toward the role of government in the economy. But while the particular prescrip- tions will vary from one country to another, the empirical evidence on supplier-oriented industrial upgrading presented in this chapter points toward several important general lessons. First, there is more than one upgrading path. The best approach will vary from one industry to another; moreover, even in the same industry dif- ferent firms may choose different paths depending on their internal capa- bilities, the regional environment in which they operate, and the particular overseas market or markets they are serving. Government policies that do not acknowledge the range of upgrading possibilities for domestic suppli- ers, that presume that all firms will follow the same path, or, worse, that attempt to dictate such a path are likely to fail. Equally important, however, is to recognize that remaining at the same level of competence is not a viable option. This is the second key lesson for policymakers. The process of upgrading is unending; there is no threshold of adequacy. Taken on its own, the strategy of pursuing low-cost labor wher- ever it is to be found, a strategy that many Asian manufacturers have pursued, is not a viable approach in anything other than the short term. Cost and quality will remain important to customers, but performance in these di- mensions is increasingly becoming a commodity and will not be enough to yield attractive margins. Providing innovative design content, production flexibility, and the ability to dependably deliver the right product at the right time and in the right place are gaining in importance, as are the international sourcing and logistic capabilities on which these service enhancements are based. Even as the competitiveness of lead firms in Western markets is in- creasingly determined by the efficiency of their supply chains, these lead firms are electing to outsource more of the key supply chain functions, and this continued deverticalization will certainly put ongoing pressure on suppliers--from East Asia and elsewhere--to continuously upgrade their capabilities and their geographic reach. Third, the innovations that fuel new generations of products and processes grow both out of new ideas that a society develops on its own and those it finds in other countries and adapts and develops for its own pur- poses. Even an advanced economy like the United States, with its vast do- mestic research and development infrastructure, still needs mechanisms to monitor science and technology in leading centers abroad and to pull the most promising new developments into its own industries. East Asian na- tions need to enhance those institutions that enable them to tap into the most promising technologies developed abroad. In new areas of technology in which firms wish to advance, there may be much to be gained by recruit- THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 83 ing international experts with experiences in leading-edge firms. Immigra- tion and housing policies can often be effective in aiding the recruitment efforts of domestic corporations. Governments should also look for oppor- tunities to strengthen links with leading research universities overseas, in- cluding universities elsewhere in the region. The importance of information technology to the effort to upgrade sup- plier performance has been a pervasive theme of this chapter and is a fourth key lesson for policymakers. We have commented on the central role of information technology in the design, manufacture, and delivery of the service-enhanced products demanded by customers in advanced economies. For the emerging economies of East Asia, participating successfully in GPNs will require stronger domestic capabilities to exploit advanced technologies of information generation, storage, and communication. Notwithstanding the deflation of the information technology bubble in the United States and elsewhere, electronic commerce is rapidly becoming a sine qua non for small- and medium-size manufacturing enterprises throughout East Asia to compete effectively in global markets. The technologies of electronic com- merce, combined with the computerization and codification of both the de- sign and manufacturing processes, are accelerating the possibilities for effective integration in geographically and organizationally dispersed value chains. Although great strides have been made, the process of codification that supports this sort of "value chain modularity" is, in many ways, in its infancy.15 Of course, vast areas of knowledge and information remain un- codified, and some are extremely difficult to codify for technical reasons. Furthermore, the ongoing processes of technological change and innova- tion can render existing standards obsolete, restarting the clock on the process of codification (David 1995; Storper 1995). Indeed, internalizing such realms of tacit knowledge, and packaging them as services for a wide range of customers, remains one of the most promising routes for industrial upgrading. Related to this is a fifth important lesson. The cultivation of vertically in- tegrated "national champions" can no longer serve as a primary goal of in- dustrial development policies. Rather, the development of key supply chain capabilities should be the objective of these policies. The presumption that such capabilities should be combined within vertically integrated corporate structures is not consistent with the emerging pattern of GPNs and may ac- tually inhibit the acquisition of knowledge and technology on which both domestic upgrading and international competitiveness depend. 15. For more on the concept of value chain modularity, see Baldwin and Clark (2000); Dolan and Humphrey (forthcoming); Gereffi, Humphrey, and Sturgeon (forthcoming); Schilling and Steensma (2001); Sturgeon (2002, 2003). 84 GLOBAL PRODUCTION NETWORKING IN EAST ASIA REFERENCES The word processed describes informally reproduced works that may not be commonly avail- able through libraries. Abo, Tetsuo. 1994. Hybrid Factory: The Japanese Production System in the United States. New York: Oxford University Press. "Acer Plots Path to Unbeatable Service." 2001. Financial Times, March 30. Akinwande, Akintunde, Douglas Fuller, and Charles G. Sodini. 2001. "Leading, Following, or Cooked Goose: Explaining Innovation Successes and Failures in Taiwan's Elec- tronics Industry." IPC Globalization Working Paper 01-001. Massachusetts Institute of Technology, Cambridge, Mass. Processed. Amsden, Alice H. 1989. Asia's Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. Aoki, Masahiko. 1987. "The Japanese Firm in Transition." In K. Yamamura and Y. Yasuba, eds., The Political Economy of Japan. Palo Alto, Calif.: Stanford University Press. Automotive News. 1996. Available at http://www.autonews.com/ Top 50 automotive suppliers, 1995. Baldwin, Carliss Y., and Kim B. Clark. 2000. Design Rules. Cambridge, Mass.: MIT Press. Berger, Suzanne, and Richard Lester, eds. 1997. Made by Hong Kong. New York: Oxford University Press. Boillot, Jean-Joseph, and Nicolas Michelon. 2001. Chine, Hong Kong, Taiwan. Paris: Docu- mentation Française. Borrus, Michael, Deiter Ernst, and Stephan Haggard, eds. 2000. International Production Networks in Asia. London and New York: Routledge. Cooke, Phillip, and Kevin Morgan. 1993. "The Network Paradigm: New Departures in Corporate and Regional Development." Environment and Planning D: Society and Space 11(5):543­64. Crain's Detroit Business. 2001. "Largest OEM Parts Suppliers Ranked by 2001 Sales of Original-Equipment Manufacturers' Parts." Crain's List, July 8. Available at http://www.crainsdetroit.com/. Curry, James. 2000. "San Diego/Tijuana Manufacturing in the Information Age." Paper pre- pared for San Diego Dialogue: The Global Engagement of San Diego/Baja California. Available at http://www.sddialogue.org/pdfs/brpaper percent20mfg.pdf. Processed. David, Paul A. 1995. "Standardization Policies for Network Technologies: The Flux between Freedom and Order Revisited." In Richard Hawkins, Robin Mansell, and Jim Skea, eds., Standards, Innovation, and Competitiveness: The Politics and Economics of Standards in National and Technical Environments, pp. 15­35. Aldershot, U.K.: Edward Elgar. Dedrick, Jason, and Kenneth Kraemer. 1998. Asia's Computer Challenge: Threat or Opportu- nity for the United States and the World? New York: Oxford University Press. Dolan, Catherine, and John Humphrey. 2000. "The Governance of the Trade in Fresh Veg- etables: The Impact of U.K. Supermarkets on the African Horticulture Industry." Journal of Development Studies 37(2):147­76. ------. Forthcoming. "Changing Governance Patterns in the Trade in Fresh Vegetables between Africa and the United Kingdom." Environment and Planning A. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 85 Doner, Richard. 1991. Driving a Bargain: Automobile Industrialization and Japanese Firms in Southeast Asia. Berkeley: University of California Press. Dore, Ronald. 1986. Flexible Rigidities: Industrial Policy and Structural Adjustment in the Japanese Economy 1970­1980. Palo Alto, Calif.: Stanford University Press. Dunn, Darrell. 1997. "Ericsson Telecom Signs Agreement with Solectron." Electronic Busi- ness News, March 31(1051): Service and Support section. Electronic Trend Publications. 2000. "The Worldwide Contract Electronics Manufactur- ing Services Market." In A Comprehensive Study on the Trends, Issues, and Leading Com- panies in the Worldwide Market for Contract Electronics Manufacturing Services, 8th ed. San Jose, Calif. Available at http://www.electronictrendpubs.com/cems8bro.pdf. Encarnation, Dennis J. 1999. "Asia and the Global Operations of Multinational Corpora- tions." In Dennis J. Encarnation, ed., Japanese Multinationals in Asia: Regional Opera- tions in Comparative Perspective. Japan Business and Economics Series. Oxford: Oxford University Press. "Ericsson Shifts Phone Production to Contractors." 2000. Electronics Weekly, October 20. Ernst, Dieter. 1997. "From Partial to Systemic Globalization: International Production Networks in the Electronics Industry." BRIE Working Paper 98. Berkeley Roundtable on the International Economy. Evans, Peter. 1995. Embedded Autonomy: States and Industrial Transformation. Princeton, N.J.: Princeton University Press. Fine, Charles H. 1999. Clockspeed: Winning Industry Control in the Age of Temporary Advan- tage. London: Little, Brown and Company. Gereffi, Gary. 1994. "The Organization of Buyer-Driven Global Commodity Chains: How U.S. Retailers Shape Overseas Production Networks." In Gary Gereffi and Miguel Korzeniewicz, eds., Commodity Chains and Global Capitalism. Westport, Conn.: Praeger Publishers. ------. 1999. "International Trade and Industrial Upgrading in the Apparel Commodity Chain." Journal of International Economics 48(1):37­70. Gereffi, Gary, John Humphrey, and Timothy J. Sturgeon. Forthcoming. "The Governance of Global Value Chains." Review of International Political Economy. Gilson, Ronald J., and Mark J. Roe. 1993. "Understanding the Japanese Keiretsu: Overlaps between Corporate Governance and Industrial Organization." Stanford University, Center for Economic Policy Research, Palo Alto, Calif. Processed. Guillèn, Mauro. 2001. The Limits of Convergence: Globalization and Organizational Change in Argentina, South Korea, and Spain. Princeton, N.J.: Princeton University Press. Jonas, Gabrielle. 1997. "Ericsson, CEMs Sign Délas." Electronic Business News, July 14(1066): News section. Kaplinsky, Raphael. 2000. "Globalization and Unequalization: What Can Be Learned from Value Chain Analysis?" Journal of Development Studies 37(2):117­46. Keesing, Donald B., and Sanjaya Lall. 1992. "Marketing Manufactured Exports from Devel- oping Countries: Learning Sequences and Public Support." In Gerald Helleiner, ed., Trade Policy, Industrialisation, and Development, pp. 176­93. Oxford: Oxford University Press. Kenney, Martin, and Richard L. Florida. 1993. Beyond Mass Production: The Japanese System and Its Transfer to the United States. New York: Oxford University Press. 86 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Lee, Ji-Ren, and Jen-Shyang Chen. 2000. "Dynamic Synergy Creation with Multiple Busi- ness Activities: Toward a Competence-Based Growth Model for Contract Manufactur- ers." In Ron Sanchez and Aime Heene, eds., Advances in Applied Business Strategy. Vol. C: Research in Competence-Based Research, pp. 209­28. Stamford, Conn.: JAI Press. Leonard-Barton, Dorothy. 1994. "Core Capabilities and Core Rigidities." In H. Kent Bowen, Kim B. Clark, Charles A. Holloway, and Steven C. Wheelwright, eds., The Per- petual Enterprise Machine: Seven Keys to Corporate Renewal Through Successful Product and Process Development. New York: Oxford University Press. Levine, Bernard. 2000. "Sony, Solectron Set EMS Deal." Electronic News, October 18. ------. 2001. "Bear Stearns Upgrades Contractors." Electronic News, May 7. MacIntyre, Andrew, ed. 1994. Business and Government in Industrializing Asia. Ithaca, N.Y.: Cornell University Press. Mason, Mark, and Dennis Encarnation, eds. 1995. Does Ownership Matter? Japanese Multi- nationals in Europe. New York: Oxford University Press. McKendrick, David, Richard Doner, and Stephan Haggard. 2000. From Silicon Valley to Sin- gapore: Location and Competitive Advantage in the Hard Disk Drive Industry. Palo Alto, Calif.: Stanford University Press. "NEC to Sell Cell Phone Plants in Mexico." 2000. Bloomberg News [Britain], December 12. Ojo, Bolaji, and Claire Serant. 2002. "IBM Deal Points to More PC Outsourcing Oppor- tunities." Electronic Business News, January 14. Powell, W. 1990. "Neither Market Nor Hierarchy: Network Forms of Organization." Research in Organizational Behavior 12:295­336. Prahalad, C. K., and Gary Hamel 1990. "The Core Competence of the Corporation." Harvard Business Review 3(May­June):79­91. Quinn, James Brian, and Frederick G. Hilmer. 1994. "Strategic Outsourcing." Sloan Management Review 35(4):43­56. Ravenhill, John. 1995. "National Champions, Multinationals, and Networks in Asia: Why Ownership Makes a Difference in the Electronics Industry." Paper presented at the conference Does Ownership Matter? MIT Japan Program. Processed. Sako, Mari. 1989. "Competitive Cooperation: How the Japanese Manage Inter-firm Rela- tions." London School of Economics, Industrial Relations Department. Processed. Schilling, Melissa A., and Harvey K. Steensma. 2001. "Industry Determinants of the Adop- tion of Modular Organizational Forms: An Empirical Test." Academy of Management Journal 44(6):1149­69. Schonberger, Richard. 1982. Japanese Manufacturing Techniques. New York: Free Press. Scott, Allen J. 1987. "The Semiconductor Industry in Southeast Asia." Regional Studies 21(2):143­60. Serant, Claire. 2001. "Singapore No Longer a Magnet for EMS Companies." Electronics Business News, April 4. Shimokawa, Koichi. 1999. "New Trend for Component Modules in Japan." Japanese Au- tomotive News, April 1. Storper, Michael. 1995. "The Resurgence of Regional Economies, Ten Years Later." Euro- pean Urban and Regional Studies 2(3):191­221. Sturgeon, Timothy J. 1999. Turnkey Production Networks: Industrial Organization, Economic Development, and the Globalization of the Electronics Manufacturing Supply Base. Ph.D. diss. University of California, Berkeley, Department of Geography. THE NEW GLOBAL SUPPLY BASE: NEW CHALLENGES FOR LOCAL SUPPLIERS 87 ------. 2000. "Turnkey Production Networks: The Organizational Delinking of Produc- tion from Innovation." In Ulrich Juergens, ed., New Product Development and Produc- tion Networks: Global Industrial Experience. Berlin: Springer Verlag. ------. 2002. "Modular Production Networks: A New American Model of Industrial Orga- nization." Industrial and Corporate Change 11(3):451­96. ------. 2003. "What Really Goes on in Silicon Valley? Spatial Clustering Dispersal in Modular Production Networks." Journal of Economic Geography 3(April):199­225. Sturgeon, Timothy J., and Richard Florida. 1999. "Globalization and Jobs in the Automotive Industry: A Study by Carnegie Mellon University and the Massachusetts Institute of Technology." Final report to the Alfred P. Sloan Foundation. IPC Globalization Work- ing Paper 01-003. Massachusetts Institute of Technology, Cambridge, Mass. Processed. ------. 2003. "Globalization, Deverticalization, and Employment in the Motor Vehicle In- dustry." In Martin Kenney and Richard Florida, eds., Locating Global Advantage: In- dustry Dynamics in a Globalizing Economy. Palo Alto, Calif.: Stanford University Press. Sturgeon, Timothy J., and Ji-Ren Lee. 2001. "Industry Co-Evolution and the Rise of a Shared Supply Base for Electronics Manufacturing." IPC Working Paper 01-003. Massachusetts Institute of Technology, Industrial Performance Center, Cambridge, Mass. Processed. Tachiki, Dennis S. 1999. "Exploiting Asia to Beat Japan: Production Networks and the Comeback of U.S. Electronics." In Dennis J. Encarnation, ed., Japanese Multination- als in Asia. New York: Oxford University Press. Thun, Eric. 2001. "Growing up and Moving out: The Globalization of Traditional Indus- tries in Taiwan [China]." Globalization Study Working Paper 00-004. Massachusetts Institute of Technology, Industrial Performance Center, Cambridge, Mass. Processed. U.S. Bureau of Labor Statistics. Various years. "Non-Farm Payroll Statistics." In Current Employment Statistics. Washington, D.C. Venkatesan, Ravi. 1992. "Strategic Sourcing: To Make or Not to Make." Harvard Business Review 70(6):98­108. Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, N.J.: Princeton University Press. Wang, Abel S. H. 2001. "Taiwan's Hardware Industry in 2000 and Emerging Trends." MIC Special Release. Taipei, Taiwan, China: Market Intelligence Center, March 23. Avail- able at mic.iii.org.tw/. Processed. Womack, James P., Daniel T. Jones, and Daniel Roos. 1990. The Machine That Changed the World. New York: Rawson Associates. World Bank. 1993. The East Asian Miracle: Economic Growth and Public Policy. New York: Oxford University Press. CHAPTER 3 GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONICS INDUSTRY AND UPGRADING PROSPECTS IN MALAYSIA Dieter Ernst E ast Asia's catch-up in the electronics industry during the late twen- tieth century provides a fascinating example of the catalytic role that linkages with foreign firms can play in industrial development (for example, Borrus, Ernst, and Haggard 2000; Ernst 1997b; Ernst and Guerrieri 1998): an early integration into global production networks pro- vided Asian producers with access to the industry's main growth markets, helping to compensate for the initially small size of their domestic markets. Network participation also provided new opportunities, pressures, and in- centives for Asian network suppliers to upgrade their technological and managerial capabilities (Ernst and Kim 2002). As a result, East Asia emerged as the dominant base of global manufacturing for the electronics industry, especially for assembly and component manufacturing. This pattern of growth survived the 1997 financial crisis (Ernst 2001b). The "new-economy" boom in the United States provided an additional boost, increasing demand for Asian electronics exports.1 But there are lim- its to export-led growth, and recent transformations of global production networks force us to reconsider the region's prospects for industrial up- grading within these networks. The downturn in the global electronics in- dustry beginning in late 2000 brutally exposed the downside of export-led 1. The size of this investment-led boost in demand for Asian exports can be gathered from the following data. Almost half of the U.S. capital investment since 1997 has gone into information technology, up from less than 24 percent during the early 1990s, and roughly 40 percent of U.S. consumption of computers and semiconductors is imported, largely from East Asia (data are cour- tesy of U.S. Department of Commerce, December 2001). 90 GLOBAL PRODUCTION NETWORKING IN EAST ASIA industrialization: a country is more vulnerable, the higher the share of elec- tronics in its exports, the greater its integration into global production net- works, and the greater its dependence on exports to the United States. There is a broad consensus that East Asia's electronics industry is in need of upgrading (for example, Ariffin 2000 on Malaysia; Chen 2002 on Taiwan, China; Ministry of Information and Communication 2002 on the Republic of Korea; Simon 2001 on China; Toh 2002 on Penang and Malaysia). De- fined as a shift to higher-value-added products, services, and production stages through increasing specialization and efficient domestic and inter- national linkages, industrial upgrading necessitates a strong base of domes- tic knowledge (see, for example, Ernst forthcoming a). Building on existing strengths in volume manufacturing, industrial upgrading requires the de- velopment of complementary skills and capabilities in design and develop- ment (including the introduction of new products) as well as in "embedded" software, SoC (system-on-chip) design, intellectual property trade, system integration, and the management of resources, supply chains, and customer relations. Of critical importance is the capacity to bring in at short notice specialized experts from overseas who can bridge gaps in knowledge and catalyze necessary changes in organization and procedures required to de- velop these capabilities locally. In short, successful upgrading raises daunt- ing challenges; chief among them are substantial investments in long-term assets, such as specialized skills and capabilities. This chapter examines how three fundamental transformations in global production networks are providing new opportunities for upgrading in East Asia's electronics industry: (1) vertical specialization: the emergence of increasingly complex "networks of networks" that juxtapose original equip- ment manufacturers (OEMs) and global, U.S.-based contract manufactur- ers; (2) coordination and content: the increasing use of digital information sys- tems to manage these networks and to build global information service networks that complement networks centered on manufacturing; and (3) location: the emergence of China as a priority investment target for global production networks in the electronics industry. This analysis is based on the conceptual framework developed in Ernst (2002a, 2003a). It complements the discussion in chapter 2 but places greater emphasis on new opportunities that arise from vertical specializa- tion within global production networks (Ernst and Kim 2002) and from the increasing use of digital information systems to manage these networks (Ernst 2003b). I argue that the reintegration of geographically dispersed specialized production and innovation sites into multilayered global pro- duction networks and the increasing use of information technology systems to manage these networks are gradually reducing the constraints on inter- GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 91 national diffusion of knowledge. Global production networks expand inter- firm linkages across national boundaries, increasing the need for knowl- edge diffusion, while information systems enhance not only the exchange of information but also the sharing and joint creation of knowledge. In turn, this new mobility of knowledge creates new opportunities for using network participation as a catalyst for further upgrading of Asia's electron- ics industries. This chapter also documents the pressures that rising requirements for participation in these networks are putting on mid-size countries and es- pecially on local small- and medium-size Asian suppliers. These pressures result not only from the strengths but also from the limitations of the U.S.- style model of contract manufacturing that have been brutally exposed by the downturn in the global electronics industry. Based on this analysis, the chapter then explores the difficulties of devising realistic upgrading strate- gies and discusses policies and support institutions that could improve the implementation of these strategies. It is difficult in one chapter to consider the entire range of upgrading pos- sibilitiesthatfacetheeconomiesofEastAsia.Therefore,IfocusonMalaysia, a country that faces a particularly demanding challenge, due to four peculiar characteristics of its electronics industry.2 First, Malaysia exceeds most other Asian electronics producers (with the exception of Singapore) in its vulnera- bility to the vicissitudes of export-led growth: electronics constitute around 60 percent of its exports, its electronics industry is heavily exposed to global production networks, and the U.S. market absorbs 25 percent of its total ex- ports (an estimated 40 percent of electronics exports). Second, with its focus on low-end, assembly-type volume manufacturing and a weak domestic base of supply, this mid-size country is especially vulnerable to the emergence of China as a new competitor. Third, with the Penang Development Centre, with its two industrial master plans (Ministry of International Trade and In- dustry 1986, 1996), and with the Bill of Guarantees (drawn up for its Multi- media Super Corridor), Malaysia has developed one of the most aggressive sets of upgrading incentives for private companies (both foreign and domes- tic). And fourth, despite such policies, Malaysia has yet to develop a suffi- ciently diversified and deep industrial structure to induce a critical mass of corporate investment in specialized skills and capabilities. A focus on Malaysia helps to highlight three important propositions that should inform the study of upgrading prospects in East Asia's electronics industry. First, as long as peculiar characteristics of industry structure 2. For related studies on the prospects for industrial upgrading in the electronics industry in Korea and Taiwan (China), see Ernst (1994b, 2000b, 2001a). 92 GLOBAL PRODUCTION NETWORKING IN EAST ASIA constrain the incentives for firms to invest in long-term assets (such as spe- cialized skills), the prospects for upgrading will remain limited. Second, while investment incentives and infrastructure matter, the key to success is the development of specialized skills and innovative capabilities, ahead of whatthemarketwouldprovide.Ofcriticalimportanceareincentivesthaten- courage university professors, researchers, and students to interact closely with the private sector (through, for example, company internships and sabbaticals). Equally important are training institutions, jointly run by the private and public sector, like the Penang Skills Development Centre. Third, in countries where the structure of domestic industry constrains upgrading efforts, international linkages through participation in global production networks can play an important catalytic role. In such a situa- tion, it is critical to understand whether and how the current transforma- tions of global production networks can help to bypass the constraints on upgrading. This chapter argues that there is now greater scope for diversify- ing international network linkages, beyond the erstwhile exclusive linkages with OEMs, and that this could facilitate upgrading into more knowledge- intensive production and services. The chapter begins by sketching key characteristics of global production networks and introducing an operational definition of industrial upgrading. It then outlines the transformations of global production networks in East Asia's electronics industry. Finally, it asks how the upgrading prospects of Malaysian electronics firms are affected by these transformations. Specifi- cally, it explores whether these transformations provide new opportunities for relieving the constraints on domestic upgrading, outlines feasible re- sponses, and identifies options for further analysis. CONCEPTUAL FRAMEWORK: GLOBAL PRODUCTION NETWORKS AND INDUSTRIAL UPGRADING Trade economists have recently discovered the importance of changes in the organization of international production as a determinant of trade pat- terns (for example, Feenstra 1998; Jones and Kierzkowski 2000; Navaretti, Haaland, and Venables 2002).3 Their work demonstrates that (a) produc- tion is increasingly fragmented, with parts of the production process being scattered across a number of countries, hence increasing the share of trade in parts and components, and (b) countries and regions that have become part of the global production network have industrialized the fastest. And 3. For details on the characteristics of global production networks, see Borrus, Ernst, and Haggard (2000) and Ernst (1994a, 1997a, 2002b, 2003a, 2003b). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 93 leading growth economists (for example, Grossman and Helpman 2002) are basing their models on a systematic analysis of global sourcing strategies. This chapter builds on their work but uses a broader concept of global productionnetworksthatemphasizesthreecharacteristicsthatareimportant for industrial upgrading: (1) scope: global production networks encompass all stages of the value chain, not just production; (2) asymmetry: flagships dom- inate control over network resources and decisionmaking; and (3) knowledge diffusion: knowledge must be shared to keep these networks growing. Characteristics of Global Production Networks A global production network covers both intrafirm and inter-firm transac- tions and forms of coordination: it links together the flagship's own sub- sidiaries, affiliates, and joint ventures with its subcontractors, suppliers, serv- ice providers, as well as partners in strategic alliances. Although equity ownership is not essential, network governance is distinctively asymmetric. A network flagship like IBM or Intel breaks down the value chain into a variety of discrete functions and locates them wherever they can be carried out most effectively, where they improve the firm's access to resources and capabilities, and where they are needed to facilitate the penetration of im- portant growth markets. The main purpose of these networks is to provide the flagship with quick and low-cost access to resources, capabilities, and knowledge that complement its core competencies. As the flagship inte- grates geographically dispersed production, customer, and knowledge bases into global production networks, this may produce transaction cost savings. Yet the real benefits result from the dissemination, exchange, and outsourc- ing of knowledge and complementary capabilities (Ernst 2003a). Global production networks typically combine rapid geographic disper- sionwithspatialconcentrationonagrowing,butstilllimited,numberofspe- cialized clusters. Two types of clusters can be distinguished (Ernst 2003b): centers of excellence that combine unique resources, such as research and development (R&D) and precision mechanical engineering, and cost- and time-reduction centers that thrive on the timely provision of lower-cost services.4 Different clusters face different prospects for industrial upgrading, depending on their specialization and on the product composition of their globalproductionnetwork.Thedispersionofclustersdiffersacrossthevalue 4. Cost- and time-reduction centers include the usual suspects in Asia (China, Korea, Malaysia, Taiwan [China], Thailand, and now also India for software engineering and Web services) but also exist in once-peripheral locations in Europe (for example, Central and Eastern Europe, Ireland, and Russia), in Latin America (Brazil and Mexico), in some Caribbean locations (like Costa Rica), and in a few spots elsewhere in the rest of the world. 94 GLOBAL PRODUCTION NETWORKING IN EAST ASIA chain: it increases, the closer one gets to the final product, while dispersion remains concentrated especially for high-precision and design-intensive components.5 In short, agglomeration economies continue to matter, giv- ing rise to the path-dependent nature of upgrading trajectories for indi- vidual specialized clusters. Flagships. Global production networks typically consist of various hierar- chical layers, ranging from flagships that dominate such networks, because of their capacity for system integration (Pavitt 2003), down to a variety of usually smaller, local specialized network suppliers. The flagship is at the heart of a network: it provides strategic and organizational leadership be- yond the resources that, from an accounting perspective, lie directly un- der the control of management (Rugman 1997, p. 182). The strategy of the flagship company thus directly affects the growth, strategic direction, and network positionoflower-endparticipants,likespecializedsuppliersand subcontractors. The latter, in turn, "have no reciprocal influence over the flagship strategy" (Rugman and D'Cruz 2000, p. 84). The flagship derives strength from its control over critical resources and capabilities that facil- itate innovation and from its capacity to coordinate transactions and the ex- change of knowledge between the different network nodes. Flagships retain in-house activities in which they have a particular strate- gic advantage; they outsource those in which they do not. It is important to emphasize the diversity of such outsourcing patterns (Ernst 1997a). Some flagships focus on design, product development, and marketing, while out- sourcing volume manufacturing and related support services. Other flag- ships outsource a variety of high-end, knowledge-intensive support services. To move this model a bit closer to reality, we distinguish two types of global flagships: (a) original equipment manufacturers that derive their mar- 5. On one end of the spectrum is final assembly of personal computers, which is widely dispersed to major growth markets in Asia, Europe, and the United States. Dispersion is still quite extended for standard, commodity-type components, but less so than for final assembly. For instance, flag- ships can source keyboards, computer mouse devices, and power switch supplies from many dif- ferent sources, including Asia, Mexico, and the European periphery, with Taiwanese (Chinese) firms playing an important role as intermediate supply chain coordinators. The same is true for printed circuit boards. Concentration of dispersion increases the more we move toward more complex, capital-intensive precision components: memory devices and displays are sourced pri- marily from "centers of excellence" in Japan, Korea, Singapore, and Taiwan (China); and hard disk drives are sourced from a Singapore-centered triangle of locations in Southeast Asia. Finally, dis- persion becomes most concentrated for high-precision, design-intensive components that pose the most demanding requirements on the mix of capabilities that a firm and its cluster needs to mas- ter: microprocessors, for instance, are sourced from a few globally dispersed affiliates of Intel, two American suppliers, and one recent entrant from Taiwan, China (Via Technologies). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 95 ket power from selling global brands, regardless of whether design and pro- duction are done in-house or outsourced, and (b) U.S.-based global contract manufacturers that establish their own global production networks to pro- vide integrated manufacturing and global supply chain services (often in- cluding design) to the OEMs. Asian suppliers. To determine whether a local company in Asia is inte- grated into a global production network, this chapter uses a broad set of indicators that includes (a) use of dedicated parts supplied by a foreign firm, (b) contract manufacturing of parts or final products to the specifications of a foreign firm, (c) contract manufacturing of parts or final products based on its own design, or (d) the provision of knowledge support services to a foreign firm. It is necessary to open the black box of Asian suppliers. First, some of these suppliers have been around for quite a while. Since the 1960s, vari- ous groups of Asian suppliers have emerged, first in consumer electronics, then in contract chip assembly (Korea's Anam is the most prominent ex- ample), and, more recently, in contract wafer fabrication (so-called silicon foundries), in the own-design-manufacture (ODM) supply of computers and hand-held and wireless devices, and in the design of integrated circuits (Ernst forthcoming b). Second, Asian suppliers differ considerably in their capabilities, network position, and market power. Substantial differences also exist with regard to their capacity for component sourcing, design, de- velopment, and engineering, the capacity to provide global support serv- ices, and the use of digital information systems. Greatly simplifying, I distinguish two types of Asian suppliers: higher- tier and lower-tier suppliers. Higher-tier suppliers, such as the Acer in Taiwan, China (Ernst 2000c), play an intermediary role between global flagships and local suppliers. They deal directly with global flagships (both brand leaders and global U.S.-based contract manufacturers), they possess valuable proprietary assets (including technology), and they have developed their own mini global production networks (Chen 2002). Building on their strengths in volume manufacturing and the provision of detailed design services (the so-called ODM services), these higher-tier suppliers are now under pressure to develop complementary skills and capabilities in the intro- duction of new products, process reengineering, as well as embedded soft- ware, System-on-chip (SoC) design, intellectual property trade, system in- tegration, and the management of network resources, supply chains, and customer relations. With the exception of hard-core R&D and strategic marketing, which remain under the control of the global brand leader, Asian higher-tier suppliers must be able to shoulder all steps in the value 96 GLOBAL PRODUCTION NETWORKING IN EAST ASIA chain. They must even take on the coordination functions necessary to manage the global supply chain.6 Lower-tier Asian suppliers are the weakest link in global production net- works. Their main competitive advantages are low cost, speed, and flexi- bility of delivery. They are typically used as price breakers and capacity buffers and can be dropped at short notice. This second group of local sup- pliers rarely deals directly with the global flagships; they interact primarily with local higher-tier suppliers. Lower-tier suppliers normally lack propri- etary assets; their financial resources are inadequate to invest in training and R&D; and they are highly vulnerable to abrupt changes in markets and technology and to financial crises. Industrial Upgrading An appropriate long-term development strategy for Asian electronics in- dustries must focus on improvements in specialization, productivity, and linkages (as defined by Hirschman 1958, ch. 6), all of which necessitate a broad base of skills and capabilities (this section is based on Ernst forth- coming a). All four elements are essential prerequisites for improving a country's capacity to raise long-term capital for facility investment, R&D, and human resource development. The concept of industrial upgrading ties these four elements together in a cohesive framework for unlocking new sources of economic growth.7 Critical prerequisites for successful upgrading are a sufficiently large pool of specialized and retrainable skills, a strong base of domestic knowledge, forms of corporate governance that facilitate inno- vation, sophisticated information management, and strong international knowledge linkages. My definition emphasizes the importance of international linkages. I do not assume that industrial upgrading ends at the national border or that it occurs only if improved specialization generates pressures to create dense forward and backward linkages within the district or the national economy. A "closed economy" assumption is unrealistic, as globalization and infor- 6. As emphasized in chapter 2, this is a demanding agenda, but some ODM suppliers, especially Hon Hai and Mitac, have been able to mount a credible challenge to U.S.-based contract man- ufacturers. (Author's interviews with U.S. contract manufacturers in China, October 2002). 7. By focusing on knowledge and innovation as major sources of economic growth, my approach is consistent with leading-edge economic thinking, such as endogenous growth theories (Grossman and Helpman 1991; Romer 1990), Lipsey's structuralist growth theory (for example, Lipsey 2001), evolutionary economics (for example, Nelson and Winter 1982; Penrose 1995 [1959]; Richardson 1990 [1960]), and attempts to reunite economic growth and innovation theory and business history (for example, Lazonick 2000). A focus on knowledge and innovation also reflects a recent shift in policy debates within important international institutions, such as the European Commission, the Organisation for Economic Co-operation and Development, and the World Bank. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 97 mation technology have drastically increased the international mobility of trade, investment, and even knowledge (Ernst 2003b). This increases the scope for cross-border forward and backward linkages, in the manner in which improved specialization generates pressures to create dense forward and backward linkages within the economy (Ernst 2002b). Equally important, most countries are constrained by a narrow domestic base of knowledge and limited linkages. Both constraints are particularly important for small developing economies, which tend to have a narrow and incomplete set of domestic linkages (for example, Ernst, Ganiatsos, and Mytelka 1998; Lall 1997). The result is an inverted production pyramid: a growing sector of final products rests on a weak and much smaller domes- tic base of mostly inefficient support industries. Rapid growth in the final products sector necessitates considerable imports of intermediates and pro- duction equipment. In addition, highly heterogeneous economic structures constrain agglomeration economies, weak and unstable economic institu- tions obstruct learning efficiency, and a high vulnerability to volatile global currency and financial markets constrains patient capital, which is necessary for the development of a broad base of domestic knowledge. As a result of this vicious circle, very limited sharing and pooling of resources and knowl- edge occur within the country and often even within the export-oriented cluster.Thisimpliesthatthemodelofindustrialupgradingneedstointegrate international knowledge linkages. To compensate for their narrow base of domestic knowledge and limited linkages, Asian developing economies have to rely on foreign sources of knowledge to catalyze the formation of do- mestic capabilities. International linkages need to prepare the way for an upgrading of East Asia's electronics industries. Ifocusontwoaspectsofindustrial upgradingfoundin theliterature:firm- level upgrading from low-end to higher-end products and value chains and industry-level linkages with support industries, universities, and research in- stitutes. Without industry-level linkages, firm-level upgrading will soon reach its limits.8 Two additional features distinguish my concept of industrial 8. The other three forms of industrial upgrading are (a) inter-industry upgrading proceeding from low-value-added industries (for example, light industries) to higher-value-added industries (for ex- ample, heavy and higher-tech industries); (b) inter-factor upgrading proceeding from endowed as- sets (that is, natural resources and unskilled labor) to created assets (physical capital, skilled labor, social capital), and (c) upgrading of demand within a hierarchy of consumption, proceeding from necessities to conveniences to luxury goods. See Ozawa (2000) for discussion of upgrading tax- onomies. Most research has focused on a combination of the first two forms of industrial upgrading, based on a distinction between low-wage, low-skill "sunset" industries and high-wage, high-skill "sunrise" industries. Such simple dichotomies, however, have failed to produce convincing results, for two reasons. First, there are low-wage, low-skill value stages in even the most high-tech indus- try, and high-wage, high-skill activities exist even in so-called traditional industries like textiles. And, second, both the capability requirements and the boundaries of a particular "industry" keep chang- ing over time, which makes an analytical focus on the industry level even more problematic. 98 GLOBAL PRODUCTION NETWORKING IN EAST ASIA upgrading. First, firm behavior is a key dimension, allowing for the co- evolution of industry structure and firm behavior in response to actions of key participants and also to the policy environment. And second, a broad definition of innovation expands the focus beyond R&D and patenting.9 There is now a widespread consensus that "innovation efforts" should be de- fined broadly to include engineering, technology purchases, expenditures on licensing and consulting services, and technology search as well as the accu- mulation of tacit knowledge required to absorb imported technology (for example, Nelson 1990). That broader focus is necessary to capture the pro- liferation of knowledge-intensive professional services, made possible by information and communication technology. NETWORKS OF NETWORKS: OUTSOURCING BASED ON CONTRACT MANUFACTURING The new-economy boom in the United States accelerated a long-standing trend toward vertical specialization in the electronics industry: outsourcing based on contract manufacturing became the "panacea of the ´90s" (Lake- nan, Boyd, and Frey 2001, p. 3), a "new American model of industrial or- ganization" (Sturgeon 2002). Two inter-related transformations need to be distinguished: supply contracts and mergers and acquisitions (M&A). Global brand leaders, like Dell--the original equipment manufacturers-- increasingly subcontract manufacturing and related services to U.S.-based global contract manufacturers, like Flextronics. Equally important is that the very same contract manufacturers have acquired the existing facilities of OEMs, as the latter are divesting their internal manufacturing capacity, seeking to allocate capital to other activities that are expected to generate higher profit margins, such as sales, marketing, and product development. Argument SturgeonandLester(chapter2inthisvolume)emphasizethattheriseofU.S. contract manufacturers with global reach may pose a serious competitive 9. Most empirical work on industrial upgrading has explored the expansion of R&D-intensive industries. For most developing countries, that narrow focus is of very limited value. The (usu- ally) implicit notion is that potential rates of productivity growth are higher in "emergent" R&D- intensive industries (Globerman 1997, pp. 98, 99). Hence, "specializing in the "right" technolog- ical activities directly contributes to faster growth rates of real income." A related notion is that, for R&D-intensive industries, economic rents can be extracted, in part, from foreign consumers. Specializing in the "right" technological activities contributes to higher levels of national income by promoting more favorable international terms of trade. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 99 threat to Asian suppliers. Their analysis highlights rising threshold require- ments for the suppliers to participate in global production networks. Com- plementingtheiranalysis,thischapterhighlightstheothersideofthecoinand explores how Asian suppliers can exploit linkages with U.S.-based contract manufacturers, which now complement the original linkages with OEMs, for their upgrading purposes. More specifically, I ask what new opportuni- ties for upgrading may open up for Asian suppliers, as outsourcing based on contract manufacturing has created increasingly complex, multi-tier "net- works of networks" that juxtapose global ties among the two large global players (the OEMs and contract manufacturers) as well as intense regional ties with smaller firms (as argued, for instance, in Almeida and Kogut 1997). A focus on complex, multi-tier networks of networks distinguishes this analysis from Sturgeon's modular production network model (Sturgeon 2002). That model focuses on two actors only: global OEMs and contract manufacturers, most of them of American origin. OEMs and contract manufacturers are perceived to interact in a virtuous circle where each of them can only win. In that model, nothing can stop continuous outsourc- ing through contract manufacturing: "Turnkey suppliers and lead firms co- evolve in a recursive cycle of outsourcing and increasing supply-base capa- bility and scale, which makes the prospects for additional outsourcing more attractive" (Sturgeon 2002, p. 6). If that scenario were to materialize, Asian suppliers in the global electronics industry might face a considerable back- lash. Specifically, Asian suppliers might be unable to compete against the vastly superior capabilities of U.S.-based contract manufacturers in four areas: (1) component sourcing; (2) design, development, and engineering; (3) global reach, the provision of support services across multiple locations in all major macro regions; and (4) network coordination, improved network efficiencies through the use of sophisticated digital information systems. The analysis in this chapter yields a less gloomy perspective. Asian sup- pliers already play an important role as global contract manufacturers, and peculiar features of the U.S.-style model of contract manufacturers sug- gest possible limitations of that model. The U.S. model is just one possi- ble approach, and Asian electronics firms may have a role to play, based on their accumulated experience with contract manufacturing, before it was given that name. Furthermore, there are ample opportunities for the emer- gence of a variety of new specialized Asian suppliers, provided necessary changes are put in place in policies and support institutions. Three peculiar features of the U.S.-style model are important: the crit- ical role played by financial considerations, the limited share of contract manufacturing in worldwide electronics hardware production, and the lim- ited presence of American contract manufacturers in Asia relative to their 100 GLOBAL PRODUCTION NETWORKING IN EAST ASIA presence in the Americas and Europe. The downturn in the global elec- tronics industry exposed serious limitations of these arrangements, forcing both OEMs and contract manufacturers to adjust and rationalize the organization of their networks. All of this has important implications for upgrading in Malaysia's electronics industry. Drivers Outsourcing through subcontracting has a long history in the electronics industry (Boswell 1993). Yet, during the 1990s, outsourcing gained a new dimension, spreading across borders: global brand leaders (OEMs) have put up for sale a growing number of their overseas facilities and, in some cases, whole chunks of their global production networks. OEMs from North America like Compaq, Dell, Hewlett-Packard, IBM, Intel, Lucent, Motorola, and Nortel were first in pursuing such divestment strategies, but European OEMs (for example, Ericcson, Philips, Siemens) and, more re- cently, Japanese ones (for example, NEC, Fujitsu, Sony) have followed suit. The main driver is financial considerations: getting rid of low-margin man- ufacturing helps OEMs to increase shareholder returns.10 Other expected benefits include hedging against losses due to volatile markets and periodic excess capacity, scale economies (surface-mount technology requires large production runs, reflecting its growing capital and knowledge intensity), and improved capacity to combine cost reduction, product differentiation, and time-to-market. Growth and Market Share Contract manufacturers aggressively seized this opportunity through ac- quisitions and capacity expansion. Within a few years, they developed their own global production networks, which now complement the networks es- tablished by the OEMs. For instance, Flextronics has 62 plants worldwide, Solectron has factories in 70 countries, and the recently merged Sanmina/ SCI has 100 factories around the world. This expansion gave rise to an ex- tremely rapid growth of the contract manufacturing industry (figure 3.1). From 1996 to 2000, capital expenditures grew elevenfold (50 percent com- pound annual growth rate), and revenues increased almost 400 percent (81 percent compound annual growth rate). The industry's rapid growth was driven primarily by M&A (figure 3.2). 10. In response to pressures from institutional investors and financial analysts, OEM firms are eager to "slash their balance sheets by placing the low-margin operations with hungry contract manufacturers" (Lakenan, Boyd, and Frey 2001, p. 4). Figure 3.1 Growth of the Contract Manufacturing Industry, 1996­2000 5.0 4.5 4.0 3.5 3.0 (percent) 2.5 rate 2.0 1.5 Growth 1.0 0.5 0.0 1996 1997 1998 1999 2000 Capital expenditures (billions of dollars) 50% compound annual growth rate ...Revenues (billions of dollars) 81% compound annual growth rate Note: Contract manufacturing industry includes ACT Mfg., Celestica, Flextronics, Jabil Circuit, Plexus, Sanmina/SCI Systems, and Solectron. Source: Hoover's Figure 3.2 Mergers and Acquisitions in the Contract Manufacturing Industry, 1997­2000 45 40 35 snoitisiuqca 30 25 fo 20 rebm ACT 15 uN Manufacturing SCI Flextronics Jabil 10 Solectron Celestica Systems Sanmina Plexus Circuit 5 0 2000 1999 1998 1997 102 GLOBAL PRODUCTION NETWORKING IN EAST ASIA It is important, however, to emphasize the still limited share of U.S. con- tract manufacturers in worldwide production of electronics hardware. In 2001 this share was estimated to be around 13.7 percent (up from 13.0 per- cent in 2000); for 2002 it increased to 16.3 percent (e-mail from Eric Mis- coll, chief executive officer, Technology Forecasters, April 15, 2002). Late Move to East Asia The presence of American contract manufacturers in East Asia pales rela- tive to their presence in the Americas (Brazil, Canada, Mexico, United States) and Europe (including Eastern Europe and Israel). The move to East Asia came relatively late. During the 1990s American contract manufactur- ersspentmostoftheirmoneyacquiringglobalflagshipfacilitiesintheAmer- icas and Europe. During the new-economy boom in the United States, speed-to-market due to close proximity was much more important than cost considerations. With the slowdown in the electronics industry, cost reduc- tion again has become a central concern. Arguably, this may create new in- centives for contract manufacturers to expand their East Asian networks. Table 3.1 offers a few examples. Flextronics, which has its headquarters in Singapore, has the strongest presence in Asia, with 12 facilities in six na- tions: China, India, Malaysia, Singapore, Taiwan (China), and Thailand. Yet this compares with a total of 62 plants worldwide, of which 18 are in the Americas and 27 in Europe (including two in Israel).11 Solectron, the long-time industry leader, has factories in 70 economies, but only five of these (from the table, including Japan) are in Asia.12 Solec- tron began to increase its presence in Asia only after 2001. This is primarily in response to the company's projections that, by 2005, 60 percent of its turnover will come from Asia (including Japan), up from about 30 percent in 2001. Traditionally focused on Penang (since 1991), during 1996 Solectron added facilities in Johor (Malaysia) and in Suzhou (Jiangsu Province, China). During 2000 Solectron acquired two Sony factories, one in Japan and one in Taiwan (China), as well as NEC's Ibaraki production facilities for servers, workstations, and system file products. The latter acquisition provides Solectron with 500 highly skilled Japanese employees who are well trained in build-to-order manufacturing and final test and fulfillment services. The presence in Asia is even more limited for the remaining three major global contract manufacturers. The recently merged Sanmina-SCI has 11. Reflecting the growing importance of the Asia-Pacific market, Flextronics decided in December 2001 to make Malaysia the manufacturing and logistics hub for its operations in the Asia-Pacific region (excluding China, where Flextronics has five facilities). 12. During the 2001 recession Flextronics took over the leadership position. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 103 Table 3.1 Contract Manufacturing Clusters in East Asia, 2002 Company Southeast Asiaa Greater Chinab Other countries Flextronics Singapore: Taiwan (China): Taipei India: Bangalore headquarters Malaysia: Johore, China: Beijing, Changzhou- Johor Baru, Jiangsu, Doumen-Zhuhai, Melaka Shanghai Wai Gai Free Trade Zone, Xixiang-Baoan Shenzhen Thailand: Samutprakarn Solectron Singapore China: Suzhou-Jiangsu Japan: Kanagawa, Malaysia: Penang, Taiwan (China): Taipei Tokyo Johor (1991) Sanmina-SCI Singapore Taiwan (China): Taipei Malaysia: Penang, China: Shenzhen, Qingdao, Kuching-Sarawak, Kushan-Jiangsu Sana-Jaya FTZ Thailand Celestica Singapore Hong Kong (China) Malaysia: Kulim China: Dongguan- Hi-Tech Park Guangdong Thailand Indonesia Jabil Circuit Malaysia: Penang Hong Kong (China): Asia headquarters China: Guangzhou-Guangdong, Panyu-Guangdong, Shenzhen-Guangdong a. Indonesia, Malaysia, Philippines, Singapore, Thailand, Vietnam. b. China, Hong Kong (China), Taiwan (China). 100 factories around the world, of which only seven facilities are in Asia-- China,Malaysia,Singapore,Taiwan(China),andThailand.Celestica,aspin- off from IBM Canada in 1994, has acquired 36 plants around the world. Until mid-2001, it had four plants in Asia: one in Malaysia's Kulim Hi-Tech Park (part of the northern cluster), two in China, andoneinThailand.Since then, Celestica has substantially expanded its Asian presence to meet the growing demand of Japanese OEMs for outsourcing.13 13. The starting point was an $890 million acquisition of Omni, one of Singapore's leading con- tract manufacturers, which added facilities in Indonesia, Malaysia, Singapore, and Thailand. And in January 2002 Celestica acquired two optical and broadband equipment factories in Japan from NEC as part of a five-year, $2.5 billion supply agreement. 104 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Finally, Jabil Circuit, the smallest of the leading global contract manu- facturers, has 21 facilities worldwide. As with the other leaders, Jabil's pres- ence in Asia (three facilities) lags behind its presence in the Americas (11) and Europe (seven). Its involvement in Asia started in 1995, with its fac- tory in Penang. In 1998 it established its Asian regional headquarters in Hong Kong (China) and a large low-cost manufacturing plant in China's Guangdong Province, in the YiXing Industrial Estate in Panyu.14 Limitations of the U.S.-Style Model of Contract Manufacturers The downturn in the global electronics industry exposed serious limitations of the U.S. model of contract manufacturing, forcing both OEMs and con- tract manufacturers to adjust and rationalize the organization of their net- works. That model was based on the assumption of uninterrupted growth in demand. In reality, however, demand and supply only rarely match. This simple truth was all but forgotten during the heydey of the new economy. Upgrading strategies in Asian electronics industries should take note of six important limitations of the U.S.-style model.15 First, global contract manufacturing is a highly volatile industry. While powerful forces push for outsourcing, this process is by no means irreversible. Major OEMs retain substantial internal manufacturing operations; they are continuously eval- uating the merits of manufacturing products or providing services inter- nally versus the advantages of outsourcing. Second, global contract manu- facturers are now in a much weaker bargaining position than OEMs, whose number has been reduced by the current downturn and which are now much more demanding. In principle, important long-term customer con- tracts permit quarterly or other periodic adjustment to pricing based on de- creases or increases in component prices. In reality, however, contract man- ufacturers "typically bear the risk of component price increases that occur between any such re-pricings or, if such re-pricing is not permitted, during the balance of the term of the particular customer contract" (Jabil Circuit 2001, p. 49). A third important limitation represents tradeoffs between the advantages of specialization and rapid inorganic growth through M&A. In economic 14. Jabil Circuit recently expanded its long-established operations in Penang through the acqui- sition of Xircom, a wholly owned subsidiary of Intel that supplies personal computer and network cards (New Straits Times, August 23, 2001). 15. This section is based on a recent study by Booz Allen Hamilton (Lakenan, Boyd, and Frey 2001); e-mail correspondence with the study's lead author, Bill Lakenan; recent 10K reports of the leading U.S. global contract manufacturers; and author's interviews at affiliates of global con- tract manufacturers in Malaysia. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 105 theory, vertical specialization is supposed to increase efficiency, that is, to re- duce the wastage of scarce resources. It is not clear whether the recent rapid growth of contract manufacturers has produced this result. The excessive growth and diversification that we saw during the new-economy boom may well truncate the specialization and efficiency advantages of the contract manufacturer model. The leading contract manufacturers have aggressively used M&A to pursue four objectives that do not easily match: rapid growth, a broadening of the portfolio of services that they can provide, a diversifi- cation into new product markets (especially telecommunications equip- ment), as well as an expansion of their own production networks, establish- ing a global presence at record speed. Yet this forced pace of global expansion may well create an increasingly cumbersome organization that could un- dermine the primary advantage of the contract manufacturing model: a ca- pacity to scale up and down rapidly, in line with the requirements of OEMs. Fourth, rapid growth, based on the use of stock as a currency for M&A is extremely risky and contains the seed of future problems. It stretches the limited financial resources of contract manufacturers, which typically have to cope with very low margins. The downturn of the global electronics in- dustry further increased these financial pressures on leading U.S.-based contract manufacturers.16 This, of course, raises the question of whether this will lead firms to employ off-balance-sheet financing techniques to hide accumulated debt. Fifth, in contrast to the original expectation that outsourcing based on contract manufacturing would improve inventory and capacity planning, global brand leaders in the electronics industry that rely heavily on out- sourcing have experienced very serious periodic mismatches between sup- ply and demand. On the one hand, when a product unexpectedly becomes a hit, outsourcing provides OEMs with only a limited capacity for scaling up. During a recession, on the other hand, OEMs cannot abruptly reduce orders that they had previously placed with contract manufacturers.17 Lastly, there seems to be a conflict of interest between OEMs, which are looking for flexibility, and contract manufacturers, which are looking for 16. Ironically, these pressures are particularly severe for those contract manufacturers, like Solec- tron, that have aggressively diversified beyond the personal computer sector into telecommunica- tions and networking equipment, the high-growth sectors of the new-economy boom. 17. Take Cisco. During the peak of the new-economy boom, from 1999 to 2000, demand for its products grew 50 percent. Reliance on contract manufacturers initially produced severe compo- nent shortages and a massive backlog in customer orders. When demand fell abruptly, starting in the fall of 2000, Cisco was saddled with excess inventory worth $2.25 billion which it had acquired to meet expected growth in demand. In time-sensitive industries like electronics, such inventory depreciates at alarming speed. 106 GLOBAL PRODUCTION NETWORKING IN EAST ASIA predictability and scale. For instance, OEMs focus on early market pene- tration and rapid growth of market share to sustain comfortable margins. OEMs thus need flexible outsourcing arrangements that allow them to di- vert resources at short notice to a given product as it becomes a hit. This contrasts sharply with the situation of contract manufacturers: with razor- thin margins, they need to focus ruthlessly on cost cutting. Contract man- ufacturers need predictability: "They want to make commitments in advance to reap benefits like big-lot purchases and decreased overtime" (Lakenan, Boyd, and Frey 2001, p. 60). These conflicting interests complicate the coordination of contract manufacturer­based outsourcing arrangements. They also require substan- tial fundamental changes in the organization of both OEMs and contract manufacturers as well as an alignment of incentives through contract terms and agreements. Effective outsourcing requires that both flagships and con- tract manufacturers acknowledge their conflicting interests. Further, with complexity comes uncertainty. In industries with rapidly shifting technolo- gies and markets, OEMs have no way to predict with any accuracy the spec- ifications of what they will need in terms of capacity, design features, and configuration and in terms of the specific mix of performance requirements. In the electronics industry, all of these variables can change quite drastically and at short notice. Such high uncertainty has important implications for the reorganization of contract manufacturer­based outsourcing arrangements (table 3.2). Flex- ibility now becomes the key to success. Proceeding by conjecture (stochas- tically) takes over from a deterministic approach. Flagships need adjustable networks to "satisfy a range of possible demand profiles with a portfolio of customizable capacity." They "need access to--and the ability to turn off-- big chunks of production more quickly than ever contemplated in order to capture profitability" (Lakenan, Boyd, and Frey 2001, pp. 61, 62). This has important implications for East Asia's upgrading prospects. The transition to stochastic and fluid outsourcing arrangements will substantially increase the required capabilities that local network suppliers in East Asia will have to master. However, this also opens up new possibilities for upgrading, pro- vided necessary changes in policies and support institutions are put in place. COORDINATION AND CONTENTS: INFORMATION SYSTEMS AND SERVICES A second important transformation of global production networks results from the increasing use of digital information systems to manage these net- GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 107 to key that derivates is " use components with perfect which fast to scalability flexible -- load period uses and lines how starts seeking base ovide services pr native parts changes of adjustment, extended optimum hit contract decide to alter successful without and capacity ging covers a for to oduction ers with pr global that allow right extend iterative oducts the Emer e availability capacity of pr of becomes capacity the actual that in in e along eductionr eservedr manufactur arrangement of gin nover: capacity beginning go speed cent cost conjectur oduct egate tur use pr eserver befor cing per systems the you the by eserver if ning unit aggr just high-mar contract variations variations faster at as 90" up commitment OEM e for for lear outsour flexible on ect oceed eate oduction ofitability Pr Cr Scale Reserve Have capacity Requir pr Allow Allow Enable Start solution Corr Accelerate pr Combine Allow Focus Achieve Stochastic · · · Rolling · · · Flexible · · · Iterative · · · Systemic · · · · given oduction contract mance pr at with Arrangements of line optimum oach cing perfor e in chain range appr and linkages closur configuration global, and in value small Outsour e levels, of a not mix ganization or Based­ raditionalT contract over local, changes structur aspect er service, at oduct cost scalability pr network every unit achieve ements commitment mine e minimal shapes ol limited to ministic capacity Manufactur Fix equirr ecise Deter closur ozen Allow Develop strategy Contr Reduce volumes Entail Seek Deter · Pr · Fr · Strategy · · Focused · · · Contract in nance expectations planning Changes planning design gover 3.2 mance ableT oduction oduct Activity Capacity Pr Pr Network Perfor 108 GLOBAL PRODUCTION NETWORKING IN EAST ASIA works (coordination) and to build global information service networks that complement networks centered on manufacturing (contents). Digital Information Systems and Knowledge Diffusion Digital information systems--electronics systems that integrate software and hardware to enable communication and collaborative work--are in- creasingly used to manage global production networks. While still at an early stage of trial and error, these systems gradually enhance the scope for knowledge sharing among multiple network participants at distant loca- tions. Equally important, they reduce (but do not eliminate) the problems associated with the need to coordinate product design and manufacture rapidly over long distance.18 This new mobility of knowledge may provide new opportunities for Asian suppliers to upgrade their capabilities, pro- vided appropriate policies and support institutions are in place. Digital information systems provide new opportunities for improving communication routines within global production networks: new combi- nations become feasible between old and new forms of communication. The most commonly used technologies today facilitate asynchronous inter- action, such as e-mail or non-real-time database sharing. But as data trans- fer capacity (bandwidth) increases, new opportunities are created for using technologies that facilitate synchronous interaction. This involves video con- ferencing and real-time data exchange for financial control, engineering, and R&D. Digitization implies that knowledge can be delivered as a service and built around open standards. This has fostered the specialization of knowledge creation, giving rise to a process of modularization, very much like earlier modularization processes in hardware manufacturing. Under the heading of "e-business," a new generation of networking software provides a greater variety of tools for representing knowledge, including low-cost audiovisual representations. Those programs also provide flexible information systems that support not only the exchange of information among dispersed network nodes but also the sharing, use, and creation of knowledge among multiple network participants at remote locations. New forms of remote control are emerging for manufacturing processes, quality, supply chains, and customer relations. Digital information systems, and especially the open-ended structure of the Internet, substantially broaden the scope for outsourcing. They have allowed OEMs to shift from partial outsourcing, covering the nuts and 18. I am grateful to the late Keith Pavitt for this suggestion. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 109 bolts of manufacturing, to systemic outsourcing, which includes knowledge- intensive support services. This has sharpened the competition among the providers of contract manufacturing services: competition now focuses on the capacity to provide integrated manufacturing, design, and supply chain management services wherever required. In turn, this has heightened the competition among specialized clusters in the electronics industry. For lower-cost outsourcing, OEMs and contract manufacturers can now choose between alternative locations in Asia, Latin America, the former Soviet bloc, and the European periphery. For higher- end outsourcing, they can choose from among specialized clusters in Nordic countries, France and Germany, the United States, as well as in China, Hungary, Ireland, Israel, Korea, Singapore, and Taiwan (China). Global Information Service Networks Global information service networks complement the existing production networks with their primary focus on manufacturing.19 They cover a variety of knowledge support services, such as software engineering and develop- ment, development of information technology applications, business process outsourcing, maintenance and support of information systems, as well as skill transfer and training. While much of this service outsourcing involves low- cost "sweatshop" activities,20 it also provides considerable opportunities for Asian network suppliers to upgrade their capabilities. Market pull. The growth of these service networks in Asia's electronics in- dustry is due to a combination of market pull and government policies. With the drastic slowdown in major information technology markets, especially in the United States, the center of activity has shifted to Asia. During 2001 the region's information services market significantly outpaced that of other re- gions, with a growth rate double the world average and nearly three times that of North America (Gartner Dataquest data, quoted in CMPnet.Asia, December 4, 2001). The main drivers of demand are attempts by both global flagships and local suppliers to improve the efficiency and security of exist- ing global production networks. Asian lead markets are Korea, Singapore, Taiwan (China), major export platform clusters in Malaysia and elsewhere 19. This section is based on phone interviews, company websites, and the following sources: var- ious issues of CMPnet.asia and Asia Computer Weekly.com; Aberdeen Group (2001). 20. A typical recent example is a call center set up by General Electric in the city of Dalian (Liaoning Province). Staffed by Chinese people fluent in Japanese, the center handles inquiries from GE Consumer Credit's Japanese customers (Nikkei Weekly, November 12, 2001). 110 GLOBAL PRODUCTION NETWORKING IN EAST ASIA in Southeast Asia, China's electronics clusters, especially in the south and Shanghai, as well as India's software clusters. Government policies. Other important drivers of global information serv- ice networks are support policies and incentives to foster the establishment of higher-level software and service development centers, especially in China, Korea, Malaysia, Singapore, and Taiwan (China). Singapore, for in- stance, has amended its highly successful policy to develop local manufac- turing support industries (especially for the electronics industry) with a pol- icy to promote local suppliers of information services. The former policy was called the Local Industry Upgrading Programme (LIUP), and the new service-oriented program is called Infocomm Local Industry Upgrading Programme (iLIUP). A typical example of the efforts to build a support sys- tem is the iLIUP partnership agreements that link Software AG, the Ger- manvendorofXML(extensiblemarkuplanguage)software,withspecialized Singaporean solution providers21 to develop customized XML-based busi- ness solutions for foreign affiliates and Asian companies. To make this net- work operate, Software AG has brought in a global supplier of training pro- grams for XML and related technologies. That company (Genovate Solutions) is responsible for establishing and running an XML Academy in Singapore that serves the Asia-Pacific region, training enterprises and individuals on XML, SAP, Java, Oracle, Linux, Web Logic, and other enterprise software programs. Of particular interest are policies, pursued in China, to develop software and information service capabilities. A core element of these policies is the development of 10 major software bases. Three of the important ones in- clude the Qilu Software Park in Jinan (Shandong Province), Shanghai's Pudong Software Park, and the Yangtze River Software Belt. Most of these projects are quite ambitious. For instance, the Jiangsu Software Park, the center of the Yangtze River Software Belt, takes up 120,000 square meters and contains 165 software companies that focus on applications in telecom- munications, network security, and e-business. This park is expected to be- come a major export platform for software, way beyond Jiangsu Province's current software export revenues of 2.3 billion yuan (about $277 million), one-tenth of China's total software exports. The Qilu Software Park in Jinan, Shandong Province is expected to develop into China's largest software park. The park covers a vast area 21. These Singaporean solution suppliers focus on enterprise resource planning, system man- agement, network security, network knowledge management, and a variety of Internet-based application services. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 111 of 6.5 square kilometers, and sales revenues (mostly exports) are expected to reach $1.2 billion in 2005, up from sales revenues of $233 million in 2000. A very different approach has been chosen for Shanghai's Pudong Software Park. Space constraints (only 9,000 square meters are available for this park) and the high price of land have led to a very selective focus on the dominant global flagships in the information industry. The dominant flagship is IBM, which is responsible for roughly one-third ($300 million) of total current foreign investment in the Pudong park. The park's spe- cialization is shaped by two important recent developments. First, Shang- hai is rapidly developing into a regional R&D and engineering center for leading global network flagships (so far, around 40), especially in the elec- tronics industry. Second, Shanghai is about to become one of Asia's most important clusters for the design and manufacturing of semiconductors. The Pudong Microelectronic Industrial Belt aims to build around 10 chip production lines by 2005, with a projected production value of 100 billion yuan (approximately $12.05 billion). Proximity to this cluster implies that much of the Pudong Software Park's activities are geared to circuit design and related activities. Another example of joint cross-border software development is a new re- search center, established by Ericsson, the Swedish telecommunications equipment producer, in China's Southern Software Park in Zhuhai, China. This center serves as a focal point for interaction with Ericcson's local part- ners, enabling them to co-develop and test 2G, 2.5G, and 3G applications for the Chinese market. The center is jointly run with Zhongshan Univer- sity of Guangdong Province, which provides top-notch graduates to pursue joint research projects with Ericsson related to wireless Internet protocol data network compression and encryption technology, multimedia services, mobile electronics business, bluetooth technology, embedded software, and 3G systems. Skill development and training. Skill development and training are es- sential elements of these global information service networks and may open up new opportunities for industrial upgrading. Often training and service provision are closely intertwined. A first example is the development of Asian networks for wireless Java applications. This is based on the joint ini- tiative of two global flagships: Sun Microsystems, the developer of the Java language, and Nokia, the leading supplier of mobile handsets. This initia- tive brings together the individual networks, established by both flagships: the Sun Developer Network and Forum Nokia. The Asia-Pacific Sun and the Nokia wireless Java developers networks have a twofold purpose: to develop the Asian market for wireless Java applications and, at the same 112 GLOBAL PRODUCTION NETWORKING IN EAST ASIA time, to create an Asian low-cost base of developers of such systems. A key component of these networks is the Developer Training Program, which is based on tools, knowledge, and resources provided by the two network flagships. Although programs and Java tools will be available at no charge, Asian developers will have to pay for hardware, training, and technical sup- port. The objective is to "train up to 30,000 [Asian] developers to develop wireless Java applications . . . and to bring to market 1,000 Java content providers by end-2002" (AMPnet.asia, November 29, 2001). For Nokia, the objective is to create an Asian consumer market for more than 50 mil- lion mobile terminals supporting the Java platform. A second example of emerging global information service networks in- volves the transfer of skills and the outsourcing of support services for stor- age area networks (SAN), a technology where Asian markets are expected to play an important role. With the exception of Singapore, Asian economies are latecomers in the use of digital information systems. Asian companies and government agencies thus have the opportunity to base their storage infrastructure on the new SAN model, which is more flexible and cheaper than traditional direct-attached storage models. In response, all major pro- viders of infrastructure for storage area networks are rushing to establish global information service networks that, in addition to providing training for potential users, seek to develop a robust supply base for software de- velopment and SAN-related support services. An example is the emerging network of Brocade Communications Systems, which has nodes in Beijing, Hong Kong (China), Seoul, Singapore, Sydney, and Tokyo. Essential ele- ments of Brocade's network are partnerships with leading Asian universi- ties, like Beijing's Qinghua University, to establish joint technology labs and scholarship programs. 3Com Asia-Pacific, the network equipment and software supplier, pro- vides a third example of global information service networks that are cen- tered on information technology­enabled training (so-called e-training). Its 3Com University network provides on-line training and certification pro- grams in simplified Chinese as well as web-based support services to cus- tomers and suppliers in 11 major cities in China. 3Com has established sim- ilar networks in Japan and Korea and plans to expand its efforts into other Asian countries. An important motivation is the need to create new markets for 3Com's products and software. Equally important, however, 3Com needs to have access to lower-cost local supply bases for service modules. Our last example concerns Sybase, a global vendor of database technol- ogy and e-business applications. The company has strong links with China and Korea. In Korea it has partnered with Samsung to create new e-business GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 113 software designed to facilitate the management of Asian suppliers' multiple linkages with global production networks. In China it has developed strong links with leading telecommunications carriers, supplying database man- agement software to support the billing systems for these carriers.22 Sybase's entry into the e-training market is motivated by two goals: to develop the Chinese market for database management software and to reduce the grow- ing deficit of information technology skills in this country. An equally im- portant objective is to develop a robust base of low-cost human resources that Sybase can tap into at a later stage. A major component of this training network is a joint venture between Sybase and the Beijing University of Aeronautics and Astronautics to offer on-line information technology train- ing courses that are customized to the requirements of specific industries. Topics covered include database technology, e-business applications, Java applications, and mobile and wireless applications. CHINA: A SHIFT IN NETWORK LOCATION A third important transformation within East Asia is the emergence of China as a priority investment target for the leading global electronics flag- ships (whether from Europe, Japan, or the United States), their global sup- pliers from Korea and Taiwan (China), and, more recently, the leading U.S. contract manufacturers. As a result, China poses a serious challenge for mid- size countries (like Malaysia) with a focus on volume manufacturing. But the new challenge from China could also be a blessing in disguise, by catalyz- ing serious industrial upgrading efforts. Furthermore, China's huge poten- tial market for electronics products and services provides new trade and in- vestment opportunities for Asian firms. Equally important, Asian electronics firms may consider tapping into China's huge pool of low-cost engineers and scientists. China's New Role in the Electronics Industry There is no doubt that the center of gravity of FDI in the electronics industry is moving toward China, transforming the geography of global production networks within the region. China is no longer only a provider 22. Sybase's main partners include China Unicom, Heilongjiang Telecom, Shaanxi Mobile, Shandong Telecom, Sichuan Mobile, and Zhejiang Telecom (CMPnet.Asia, October 24, 2001). 114 GLOBAL PRODUCTION NETWORKING IN EAST ASIA of cheap labor. China's new attractiveness results from a combination of five developments: a booming market for information technology products and services, when the rest of the world is in recession; an unlimited supply of low-cost information technology skills; abundant land and a rapidly im- proving infrastructure; a massive rush of capital flows into China; and, catching this opportunity, support policies pursued by the central govern- ment as well as regional and local authorities to rely on FDI as an acceler- ator of industrial upgrading. The move toward China is particularly pronounced in three sectors: elec- tronic components (especially semiconductors), computers, and telecom- munications. For U.S. foreign direct investment in semiconductors, China has become the second most important recipient, after Singapore, over- taking Malaysia, which was the main recipient in 1996. Even the new incen- tives provided by Malaysia's Second Industrial Master Plan (IMP2) have not prevented this move toward China. A similar trend can be discerned for FDI by Japanese electronics firms: China has moved into first place ahead of Malaysia, which was its biggest recipient during the early 1990s (Malaysia Ministry of Finance data, quoted in Takeuchi 2001). Taiwanese (Chinese) computer companies that supply leading U.S. com- puter OEMs have played an important pioneering role in integrating China into global production networks. Since the early 1990s, they have continu- ously moved production from Taiwan (China) to China. The result is that roughly 40 percent of China's electronics exports today are shipped from Taiwanese (Chinese) factories in China (courtesy of Market Intelligence Center, Institute for Information Industry, Taipei, December 2001). Tai- wanese (Chinese) suppliers are serving a large share, around 33 percent, of their export orders from production lines in China. From the United States, OEMs like AMD, Cisco, Compaq, Hewlett Packard, Intel, Microsoft, Motorola, and Sun Microsystems have all initi- ated significant new investment projects in China. Motorola, for instance, considers China to be of critical importance. At present, Motorola has 12 affiliates in the Asia-Pacific region. While the company's involvement started out with Korea, later followed by Singapore and Taiwan (China), China has gained substantially in importance since the early 1990s. Cur- rently, six of Motorola's Asian affiliates are based in China, with two in Sin- gapore, including the regional headquarters, and one each in India, Korea, Taiwan (China), and Thailand. Motorola is counting on the continuous rapid growth of the Chinese market to reduce the negative impact of the current recession. But China is also expected to play an important role in Motorola's long- term strategy. By providing access to the world's largest pool of relatively GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 115 lower-cost information technology skills, the company expects to enhance its innovation capabilities. For instance, out of the 13,000 people that Motorola currently employs in China, 1,000--that is, almost 8 percent-- are active in R&D. Reflecting China's growing importance, Motorola held its annual board meeting for 2001 in China, its first board meeting outside the United States. China also has attracted major new investments from leading European electronics flagships, like Alcatel, Ericsson, Nokia, Philips, and Siemens. Philips, for instance, has moved its entire cell phone manufacturing oper- ations to its Shenzhen joint venture with Beijing-based China Electron- ics. And Nokia has committed itself to establish a large integrated cluster, the Xingwang (International Industrial Park) in Beijing, which incorpo- rates Nokia's own mobile phone plant and brings together 15 of its long- standing and trusted suppliers of international components. This Nokia- centered cluster involves an initial investment of about $1.2 billion and is expected to create 15,000 jobs, with a projected annual sales volume of $6 billion. In addition, all leading Japanese electronics flagships are considering major new investments in China as they struggle under a depressed do- mestic economy and high manufacturing costs. Toshiba is building a plant for computer hardware and components in Nanjing, while Matsushita, Mitsubishi Electric, and NEC all are expanding their phone production in China. Finally, a large surge in investments in China is also reported by the Korean electronics chaebol that are all gripped by a severe "China fever." Semiconductors Since 2000, the semiconductor industry has provided a telling example of the speed of China's integration into global production networks.23 In terms of policies and support institutions, China's experience in the semi- conductor industry also provides a useful reference point for a case study of Malaysia. Until 1999, investment in China's semiconductor industry lagged woe- fully behind similar investments in Korea and Taiwan (China). The turning point came in 2000. During that year, the Chinese government apparently made a strategic decision to rely on FDI to accelerate the development of this industry. The underlying expectation was that FDI can generate a 23. If not stated otherwise, the following is based on Simon (2001) and relevant industry newslet- ters, such as the Semiconductor Reporter and Interfax China. 116 GLOBAL PRODUCTION NETWORKING IN EAST ASIA critical mass, establishing new global dynamic clusters for semiconduc- tor manufacturing, primarily in Beijing and Shanghai. The hope is that, if it engages leading U.S. flagships (both semiconductor manufacturers and equipment producers), this strategy may create enough pressure within the United States to dismantle the remaining U.S. restrictions on exports of technology. This may be an unrealistic expectation, however, in light of the current resurgence of defense and security concerns in the United States. China's new pragmatic policies toward FDI have induced global flagships to announce several large investment projects worth around $7 billion. Dur- ing 2000, China's semiconductor output grew 42 percent. Exports during 2000 grew almost 35 percent in unit terms and 30 percent in value terms, reaching $2.1 billion. The role of global production networks has been crit- ical. Foreign-invested enterprises dominate China's semiconductor exports, with a share of about 94 percent, while state-owned enterprises are respon- sible for a meager 5.3 percent. After 2001 China emerged as the main growth market for semi- conductor production equipment manufacturers. Leading equipment mak- ers are all scrambling to expand their sales in the largely untapped China market.24 Table 3.3 presents information on major investment projects by global flagships, Taiwanese (Chinese) contract manufacturers (so-called sil- icon foundries), as well as domestic firms. The prospective boost in FDI has given rise to optimistic projections. In-Stat Group, a semiconductor industry consulting firm, argues that China has the potential to become the second-largest market for semi- conductors after the United States by 2005. The Chinese government cer- tainly leaves no doubt that it has very ambitious objectives. By 2005 it hopes to increase semiconductor sales revenues to $9.7 billion, or 2 per- cent of the global market. By then, domestic production is expected to meet about 30 percent of China's demand. By 2010 China expects to be where Korea is now, that is, producing roughly 5 percent of the global semiconductor market. If these projections materialize, this would obviously pose a major threat to existing semiconductor industries in Asia. This threat would be most im- mediate for Malaysia, with its heavy reliance on semiconductor assembly. However, important threats may also emerge for semiconductor manufac- 24. Applied Materials, the world's leading vendor of semiconductor production equipment, for instance, expects to increase its sales in China from $100 million in 2000 to $1 billion by 2005 (Far Eastern Economic Review, November 1, 2001). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 117 ) 4 and chipsets feet e 800,000 continued( 845 Pentium squar total: feet Capacity technology news' with e ocessors Intel used pr 120,000 (2002); squar use (billions total: to mation to dollars) infor cost- and plans U.S. (2002); in of ficiency Investment 0.302 0.108 1.9 0.010 0.200; latest technology enhance ef time-to-market , for 2000 than testing chip testing testing logic, analog to e local testing assembly of January to of and and wafer mor and and plans and Activities ce since, testing range cent ete, per Expansion assembly Assembly Integrated fabrication, and Assembly wide discr devices; outsour 50 suppliers Assembly Industry radeT ovince Pr Suzhou ee e Jiangsu Fr Park, Guangdong Semiconductor Location s' Shanghai Jiangsu aigaoqiao ovince ovince China W Zone, Suzhou, ianjinT China-Singapor Industrial Pr Dongguan, Pr in ojects filiates) Pr af cent per Investment (100 Major 3.3 flagships ola m child ableT Fir Global Intel AMD Motor Fair Philips 118 GLOBAL PRODUCTION NETWORKING IN EAST ASIA at omfr and wafers monthly 400,000 capacity to on 8-inch expansion lines; Capacity technology micr 10 easing to oduction 480,000 0.25 Capacity 6 pr incr units (billions dollars) ) U.S. of Investment 1.6 0.0028 continued( 2000 testing (Pentium USI testing cuit ds cuit January cir and with cir and Activities computer (China) foundry foundry IBM since, for Silicon Silicon Integrated assembly Personal motherboar 4) aiwanT Integrated assembly Industry e joint ventur mosa Chinese silicon Shanghai oup) silicon Shanghai Suzhou (joint For in Zhejiang Gr Semiconductor Location and in in in (possible with s' China e oup ovince China in aiwan,T Shanghai Shanghai between Plastics gr Investing foundry Investment foundry Hangzhou, Pr Shenzhen Shanghai ventur Huahang omfr ojects Pr Corp. ASE primarily Scientific of Manufacturing ers, Industry Investment oelectonics Semiconductor Corporation Company) Semiconductor Corporation) Semiconductor (Universal Micr subsidiary ecision a Pr Major manufactur Grace onics e 3.3 (Semiconductor national aiwan(T (United (Advanced m Electr ableT Fir Contract SMIC Inter Shanghai Manufacturing TSMC Manufacturing UMC ASE Engineering) USI Industrial), Siliconwar GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 119 month on per feet e micr wafers squar 0.25 40,000 50,000 8-inch, 1.2 0.200 and lines 8-inch testing testing of fabrication wafers and and fabrication wafer oduction Pr 12-inch woT fera Assembly Assembly W ", that e e Kingston vendor Harbor cent) ventur ech-T ventur per and Chinese investment (joint 80 (joint Silicon" U.S.-based aiwaneseT a, module [Asia-Pacific Kingston owns Shanghai Shenzhen Beijing involves companies plus [Chinese] fund echnologyT Development Corporation]) Pudong Shanghai with nology memory that Beijing IC Semiconductor Computer all Semiconductor W companies Hongli eat oup Xunchuang Gr Huaxia Gr Chinese Shanghai Manufacturing Sast Beijing China Shenzhen Beijing Manufacturing 120 GLOBAL PRODUCTION NETWORKING IN EAST ASIA turing and wafer fabrication in Korea, Singapore, and Taiwan (China) as well as in Japan. But it is unlikely that these countries will simply sit still and let shifting comparative advantage run its course. In fact, all major competitors in the global semiconductor industry are pursuing aggressive policies to upgrade their product composition and capabilities, giving rise to a major transformation of this industry (for example, see Macher, Mow- ery, and Simcoe 2002). In addition, it will not be easy for China to realize its ambitious upgrad- ing objectives. One reason is the relatively low level of development of China's semiconductor industry. An important indicator of this is the fun- damentalmismatchbetweenChina'sexportsandimportsofsemiconductors. In value terms, China imports more than 70 percent of all semiconductor devices that it needs. This is slightly better than the ratio found in Korea's semiconductor industry during the early 1990s (Ernst 1994b, 1998). On the one hand, most Chinese exports are lower-end devices involving fairly ma- ture and basic process and manufacturing techniques. Imports, on the other hand, are much more sophisticated. Between 1995 and 2000, China's semi- conductor imports grew at a compound annual rate of almost 92 percent, while exports over the same period rose less than 60 percent. A further fundamental constraint to a rapid upgrading of China's semi- conductorindustryarethemassiveinvestmentsrequiredinproductionequip- ment, facilities, infrastructure, R&D, and education and training. The projected expansion of international market share requires an upgrad- ing in the product mix as well as in process and design capabilities. Out of China's 25 wafer fabrication lines in early 2001, only one used 8-inch wafers, while 21 lines used outdated 5-inch wafers (six lines) and even 4-inch wafers (15 lines).25 However, by 2003 a few more 8-inch lines had been added and several 12-inch wafer plants are planned or being built--and will begin op- erating in 2004­05. Table 3.4 provides a widely used proxy, developed by the U.S. Semi- conductorIndustryAssociation,fortheinvestmentcostsofwaferfabrication. These estimates assume that sophisticated infrastructure and support indus- tries exist within the relevant clusters or at least in close proximity. This is by no means the case in China. For instance, 90 percent of the materials used to produce the existing 8-inch wafer fabrication lines must be imported. And domestic manufacturers of semiconductor production equipment meet less 25. China Electronics News, March 27, 2001. The 8-inch line is a joint venture with Japan's NEC and the Shanghai Hua Hong NEC Electronics. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 121 Table 3.4 Investment Costs of Wafer Fabrication Process technology Minimum investment requirementsa 6-inch wafers (0.5­1.2 microns) $200 million 8-inch wafers (0.35­0.5 microns) $1.2 billion 12-inch wafers (less than 0.35 microns) $2.5 billion a. Estimates, courtesy of U.S. Semiconductor Industry Association. than 10 percent of domestic requirements. In other words, China's weak semiconductor infrastructure and base of support industries clearly imply that the effective investment costs for upgrading its semiconductor industry may be substantially higher than indicated by the figures of the Semi- conductor Industry Association. Inshort,thereareclearindicationsinthesemiconductorindustryofashift in network location away from the traditional export platform sites in South- east Asia and toward China. Yet this process may not be as quick and smooth as many observers appear to believe. This may provide enough breathing space for developing appropriate upgrading strategies in those economies that are most heavily exposed to the threat from China. IMPLICATIONS FOR MALAYSIA'S UPGRADING PROSPECTS IN THE ELECTRONICS INDUSTRY In light of the fundamental transformations of global production networks in East Asia's electronics industry, what can be said about Malaysia's up- grading prospects? Do these transformations provide new opportunities for relieving domestic upgrading constraints? And what policies and sup- port institutions could help Malaysia to successfully implement upgrading strategies? This section first highlights achievements and structural weaknesses of the Malaysian electronics industry that define and constrain its upgrading prospects. I then assess current policies that try to link cluster development and global network integration, discuss adjustments in linkages with global brand leaders, and ask to what degree linkages with contract manufacturers can broaden these opportunities. I conclude by exploring new opportunities for diversifying Malaysia's international linkages that could enhance the up- grading prospects of its electronics firms, focusing on carriers of knowledge exchange that complement formal global production networks. 122 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Achievements The progressive integration into global production networks has been a primary driver of Malaysia's success in the electronics industry. This inte- gration started in the early 1970s with offshore chip assembly, primarily by U.S. semiconductor firms. The next stage, beginning in the early 1980s, centered on Japanese electronics makers that moved their export platform production for consumer electronics to Malaysia and other Southeast Asian locations. Since the late 1980s, Malaysia has been integrated into the pro- duction networks of American producers of computer-related equipment as well as those established by their Taiwanese (Chinese) subcontractors. The most recent stage involves the production of communication and net- working equipment and the acquisition of existing flagship affiliates by global contract manufacturers. The Malaysian government, through its Industrial Master Plan (IMP) (1986­95), has tried to reap as many benefits as possible from this for- tuitous tailwind of FDI (Ministry of International Trade and Industry 1986). The guiding principle has been "outward industrialization," sub- ordinated to the needs of global network flagships. The results have been impressive, in terms of production, exports, employment, and invest- ment (table 3.5). Within a relatively short period, Malaysia experienced a substantial ex- pansion in capacity and international market share for electronics products. The heavy reliance on electronics exports acted as a powerful engine of growth. Although there were periodic disruptions, like the downturn in 1985­86 and, in particular, the Asian financial crisis in 1997­98, the over- all balance was remarkably positive. During the last decade, from 1990 to 2000, Malaysia's electronics industry registered a compound average an- nual growth rate of almost 24 percent. During the same period, exports Table 3.5 Performance of Malaysia's Electronics Industry Compared with Objectives of the Industrial Master Plan (millions of ringgits) Area Goals Performance Production 11.0 30.5 Exports 10.9 32.6 Employment 5.5 21.l7 Investment value 6,298 25,985 Note: Figures for production, exports, and employment are average rates of growth between 1986 and 1992. Those for investment are the cumulative total of investment approved between 1986 and 1995. Source: Ministry of International Trade and Industry (1996). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 123 grew at an annual average of 25 percent, while employment grew almost 11 percent annually (figures courtesy of Ministry of International Trade and Industry, Kuala Lumpur, June 2001). Global electronics brand leaders--and, more recently, their contract manufacturers--played an important role. The electronics industry is the major recipient of foreign direct investment, absorbing more than one-third of total foreign direct investment in manufacturing between 1996 and 1998 (MIDA 1999). About 100 large foreign affiliates effectively dominate this industry. Their share in manufactured exports (most of it electronics) in- creased sharply, from 40 percent in 1985 to 68 percent in 1992 (Takeuchi 1997, p. 9). The 18 members of the Malaysian-American Electronic Indus- try Association accounted for more than 14 percent of Malaysia's electron- ics exports in 2001 (Business Times, Kuala Lumpur, October 3, 2001). Weaknesses Despite these achievements, a shift in strategy is now overdue. Since the summer of 2000, the downturn in the global electronics industry has bru- tally exposed six structural weaknesses of Malaysia's electronics industry that define and constrain its upgrading prospects: (1) an asymmetric industry structure; (2) a heavy dependence on imports, due to weak domestic support industries and limited Hirschman-type linkages; (3) a heavy reliance on ex- ports, especially to the U.S. market; (4) a highly concentrated composition of products, centered on low-end assembly operations; (5) a declining capacity to generate employment; and (6) a serious mismatch between the demand for and supply of skills. Although "outward industrialization" poli- cies have provided Malaysia with substantial initial advantages in terms of export and capacity growth, these policies have failed to develop sufficient sectoral breadth and depth. Asymmetric industry structure. Malaysia's integration into global pro- duction networks gave rise to the development of an asymmetric industry structure: multiple layers of electronics firms are distinguished by asym- metric control over resources and decisionmaking. At the end of 2000, roughly 900 electronics companies were registered in Malaysia, employing more than 400,000 workers. While Malaysian firms dominate in numbers, Malaysia's electronics industry continues to be shaped by strategic decisions of global flagships (both OEMs and major American contract manufac- turers). In hierarchical order, four types of firms can be distinguished: at the top of the industry pyramid are global OEMs and contract manufac- turers, followed by suppliers and contract manufacturers from Japan, 124 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Korea, Singapore, and Taiwan (China), higher-tier local suppliers, and, at the bottom, lower-tier local suppliers. In contrast to economies like Korea, Singapore, and Taiwan (China), Malaysia has failed to develop a broad and multi-tier base of support in- dustries. There are, of course, a few widely quoted successes, almost all of them in Penang, such as BCM, Eng Teknologi Holdings, Globetronics, LKT, and Unico, that have positioned themselves as higher-tier local sup- pliers for leading OEMs (Best 2001; Rasiah 1995, 2003). These companies are moving to upgrade their capabilities to cope with the new opportuni- ties and challenges that result from the transformations of global produc- tion networks. However, the majority of the local suppliers possess few proprietary ad- vantages and clearly qualify as lower-tier suppliers. They lack sufficient financial resources to invest in training (and retraining) and to invest in dig- ital information systems and leading-edge equipment. This is so despite various promotional policies, focused on the smaller suppliers, introduced by the government. Possible explanations may include the proximity to Singapore and its base of sophisticated local suppliers, which may discour- age flagships from using Malaysian suppliers. This asymmetric industry structure has given rise to a lack of efficient do- mestic linkages and an inverted production pyramid--a huge and rapidly growing final product sector that rests on a weak and much smaller domes- tic base of support industries. Import dependence. The result is a persistently high dependence on im- ports: rapid growth in the final products sector necessitates considerable imports of intermediates and production equipment. Between 1986 and 1992, imports of Malaysia's electronics industry increased at a rate of more than 24 percent, far exceeding the goal of 7.6 percent envisioned under the IMP. By the late 1980s, the Malaysian electronics industry had to import almost 43 percent of the intermediate goods that were required to produce one unit of final output, far more than Korea (37 percent) and Japan (8 per- cent; Takeuchi 1997, p. 7). This reflected the initial strategy to position Malaysia as a low-labor-cost assembly site. By the late 1980s, however, the government acknowledged that this strategy was no longer sustainable, as new lower-labor-cost locations emerged within Southeast Asia as well as in China and Mexico. Yet Malaysia's dependence on imports of electronics components, espe- cially semiconductors, kept increasing during the 1990s, both as a share of electronics imports and as a share of total merchandise imports (table 3.6). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 125 This suggests a fundamental mismatch between the country's electronics ex- ports and its imports, with negative terms-of-trade implications: while im- ports involve high-value-added core components, especially microproces- sors and other imported semiconductors, Malaysia's component exports overwhelmingly consist of low-value-added final assemblies. Export dependence. Malaysia's electronics industry remains heavily dependent on exports: in 2000 electronics manufacturing made up about 60 percent of Malaysia's total export value, of which 35­40 percent were exports from Penang. Malaysia is one of the eight economies that are most dependent on exports to the United States, six of these being from East Asia: Malaysia's exports to the United States (most of them electronics products) accounted for 24 percent of the country's gross domestic product in 1999. Malaysia's electronics industry is extremely vulnerable to a recession in major export markets, especially the United States. During 2001 when the U.S. electronics market was in free-fall, Malaysia's electronics production and exports dramatically declined, the former more than 25 percent and exports almost 19 percent. Concentrated product composition. A highly concentrated product com- position adds further to the country's vulnerability. By 1998 the share of electronics in merchandise exports had increased to almost 58 percent, up from 48 percent in 1993. And components accounted for 46 percent of all electronics exports, of which semiconductors alone accounted for more than one-third (table 3.7). Industrial electronics (including computer- related products and telecommunications equipment) accounted for slightly more than 30 percent, and consumer electronics accounted for about 15 per- cent. Recent figures from other sources show a further increase in the con- centration of product composition. By 2000 the share of electronics exports in Malaysia's total exports had reached 70 percent (Zainal Abidin 2002, slide 6). In historic terms, this constitutes an important improvement. Back in 1986, 84 percent of Malaysia's electronics exports were components (most of them assembled chips), with 14 percent consisting of consumer elec- tronics and only 2 percent consisting of industrial electronics. Unfortu- nately, this impressive change was not sufficient to reduce the country's vulnerability to abrupt changes in the world market. An important reason for this vulnerability is that semiconductor exports generate very little local value added, as Malaysia only performs assembly and testing. A heavy dependence on assembly-type operations for a handful of products can be 126 GLOBAL PRODUCTION NETWORKING IN EAST ASIA 1998 16.6 12.7 26.5 17.5 36.7 26.4 8.6 1997 12.0 ---- ---- 23.8 15.0 28.3 19.1 6.7 imports 1996 10.0 16.7 11.0 24.3 14.9 29.6 19.1 chandise 9.7 6.4 1995 16.9 11.2 26.0 16.2 28.8 18.3 Mer 9.5 6.0 9.7 1994 14.8 23.2 13.4 28.3 16.5 9.5 5.8 8.5 9.7 1993 13.7 18.6 25.3 14.2 1998 71.6 54.7 56.4 37.3 78.5 56.5 1997 60.2 43.0 ---- ---- 53.4 33.6 72.8 49.2 98­ imports 1996 55.3 36.8 65.3 43.1 55.4 33.9 76.7 49.6 1993 onics 1995 54.1 35.7 70.0 46.5 56.9 35.6 78.1 49.7 Electr Components, 1994 53.2 34.0 67.5 43.9 52.9 30.5 78.7 45.9 Imported on 1993 57.0 34.8 66.3 41.2 48.0 25.0 76.5 42.9 of Dependence and e 3.6 Rep. China cent) ea, ableT (per Economy component Kor Components Semiconductors aiwan,T Components Semiconductors Singapor Components Semiconductors Malaysia Components Semiconductors GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 127 -- -- -- -- 2.0 0.1 6.2 4.8 12.8 11.4 7.2 9.4 7.3 3.0 0.2 6.3 14.1 12.5 9.2 3.6 6.4 3.2 0.2 5.9 7.4 2.6 12.6 20.7 17.7 11.6 6.0 7.7 5.4 3.3 0.3 6.0 6.9 2.3 12.4 11.6 5.6 6.6 4.7 3.5 0.3 5.1 6.3 1.8 11.6 10.4 9.6 4.6 5.8 4.3 4.4 0.4 9.8 4.8 5.3 1.4 -- -- -- -- 2.4 33.9 40.1 19.6 53.1 22.3 2.1 58.7 29.9 34.9 27.0 33.8 40.5 20.3 55.0 21.4 2.9 60.0 30.2 54.7 46.6 37.5 39.6 20.2 49.6 17.7 3.8 59.1 28.7 43.7 30.3 46.3 39.7 20.7 44.3 14.7 4.0 55.9 26.8 40.3 28.8 48.6 38.1 18.7 43.1 12.5 4.5 55.1 26.4 37.8 28.0 46.4 38.0 18.8 40.9 11.0 Comtrade. Database China radeT UN Kong, available. ce: Not Thailand Components Semiconductors Philippines Components Semiconductors Indonesia Components Semiconductors Hong Components Semiconductors China Components Semiconductors -- Sour 128 GLOBAL PRODUCTION NETWORKING IN EAST ASIA 1998 3.91 4.1 3.46 5.34 2.71 1.9 1.97 5.98 4.20 45.29 40.86 21.77 4.64 4.1 2.1 2.10 6.42 3.59 1997 15.5 62.3 42.9 12.8 44.62 41.93 22.05 44.1 exports 1.8 2.4 1.3 6.8 4.0 1996 14.5 60.8 40.3 15.6 45.0 40.2 19.6 42.8 onics electr 1.0 2.4 1.0 8.5 4.1 in 1995 12.2 62.4 45.7 16.1 41.6 41.6 20.4 39.4 e Shar 0.8 2.7 0.6 4.8 1994 11.9 56.2 37.2 20.5 39.0 39.3 16.8 12.0 38.6 0.5 3.0 0.8 4.9 1993 14.4 50.1 30.4 22.5 39.4 37.2 13.7 12.9 40.7 98­ 0.7 1.1 2.7 3.8 1.5 0.5 5.2 1998 28.3 1993 oducts, 0.9 1.3 2.8 4.0 1.7 0.6 5.5 1997 29.2 Pr advantage Export 0.9 0.7 2.7 3.6 2.0 0.8 3.4 0.6 2.2 2.2 1.1 1.6 5.4 1996 28.8 35.8 Leading 0.8 0.4 2.8 4.1 2.0 0.8 3.0 0.5 2.1 2.0 1.2 1.6 5.1 comparative 1995 30.9 34.3 and 0.8 0.3 2.7 3.8 2.4 0.9 2.7 0.3 2.0 1.8 1.5 1.7 5.1 Revealed 1994 29.7 31.0 Advantage 0.9 0.2 2.4 3.3 2.3 0.9 2.5 0.3 1.9 1.6 1.4 1.7 4.6 1993 28.0 29.5 Comparative in in oduct onics onics pr onics exports onics exports of Revealed and data electr data electr data electr electr e 3.7 Rep. China onic of onic of onic ea, ocessing e chandise cent) ocessing e chandise cent) ocessing ableT Economy Kor Electr pr Storage Components Semiconductors Consumer elecommunicationsT Shar mer (per aiwan,T Electr pr Storage Components Semiconductors Consumer elecommunicationsT Shar mer (per Singapor Electr pr GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 129 9.81 0.74 7.42 6.8 1.1 7.42 7.6 5.74 0.73 5.11 2.8 8.7 1.1 6.3 3.1 3.9 19.3 38.9 25.0 25.3 45.1 30.1 16.5 40.9 35.8 18.3 13.1 1.3 0.3 3.3 4.1 18.8 38.0 23.9 10.4 20.5 44.6 29.3 20.1 41.4 36.5 18.8 12.0 1.6 1.1 3.2 3.7 16.0 38.5 23.8 11.6 17.1 43.9 29.8 22.6 36.8 39.9 19.5 13.3 1.9 0.2 4.0 3.9 15.7 35.4 19.4 13.7 15.6 42.6 28.3 24.2 34.0 38.8 19.3 16.4 1.9 0.0 3.8 5.2 17.6 29.4 15.7 15.9 13.8 44.6 30.5 24.2 32.2 38.4 20.8 16.7 3.7 5.0 1.7 0.6 3.0 4.1 3.9 5.2 3.5 1.5 11.4 61.4 57.5 3.7 4.8 2.3 0.7 2.9 3.7 3.9 5.3 4.1 1.7 2.5 1.6 1.7 1.7 1.2 12.4 60.6 55.8 29.6 3.6 4.5 2.9 0.8 2.3 0.2 3.8 4.9 5.0 2.0 2.4 1.6 1.6 1.5 1.2 15.3 60.7 54.7 28.4 3.4 4.2 2.9 1.1 2.0 0.8 3.6 4.7 5.1 2.0 1.9 1.5 1.4 1.4 1.0 12.8 60.7 54.9 24.9 3.4 4.0 3.2 1.2 1.8 0.1 3.7 5.2 5.0 2.4 1.8 1.5 1.6 1.6 1.1 13.4 58.8 52.5 24.0 2.7 3.2 3.1 1.1 1.4 0.0 3.7 5.6 4.2 2.1 1.4 1.4 1.7 1.3 1.2 12.9 53.0 47.6 20.8 Comtrade. in in in onics onics onics database onics exports onics exports onics exports electr data electr data electr trade electr electr electr UN of e chandise cent) onic of onic of ocessing e chandise cent) ocessing e chandise cent) ce: Storage Components Semiconductors Consumer elecommunicationsT Shar mer (per Malaysia Electr pr Storage Components Semiconductors Consumer elecommunicationsT Shar mer (per Thailand Electr pr Components Semiconductors Consumer elecommunicationsT Shar mer (per Sour 130 GLOBAL PRODUCTION NETWORKING IN EAST ASIA crippling, as those operations can be easily replicated in countries with low levels of education. The government has tried to address this issue by investing in two new silicon foundries: Silterra (in Kulim Hi-Tech Park) and 1st Silicon (in Sarawak). To justify the heavy investment, an attempt has been made to use these fabrication plants as catalysts for the development of circuit design houses. It is too early to assess the success of these investments. Bad timing has been an important constraint--these facilities became operational dur- ing the recent downturn. However, there are positive signs. As the provi- sion of silicon foundry services is becoming a commodity (see United Microelectronics 2001, on risk factors), there are new entry possibilities for low-cost foundries in China and Malaysia, while the industry leaders (IBM Microelectronics and Taiwan [China]'s TSMC and UMC) move up the ladder to combine design capabilities with advanced fabrication tech- nology (12-inch wafers; Depeyrot 2002). Declining capacity for employment generation. A capacity for job cre- ation in Asia's thriving electronics sector has been a hallmark of the region's successful export-oriented industrialization. Since the 1997 financial crisis, however, the sector's capacity to generate employment has declined. Take Seagate, the leading U.S. disk drive manufacturer. Since the mid-1980s, the company was among the largest employers in Southeast Asia, topping the list in Penang and Singapore. That golden age of employment generation has long gone. Table 3.8 documents the company's massive destruction of manufacturing jobs in Asia. In Malaysia's electronics industry, an estimated 150,000 to 165,000 jobs have been lost since the financial crisis (table 3.9). During 2001, the most recent year for which data are available, almost 19,000 workers were laid off. Malaysian labor market experts talk of a declining employment-generating capacity of the electronics industry: after earlier downturns, a substantial shareoflaid-offworkerswererehired,butthisnolongerseemstobethecase. The latest unemployment report, prepared for the Penang State gov- ernment, conveys some distressing findings (Too and Leng 2002). With job Table 3.8 Manufacturing Employment of Seagate, Peak Year and 2002 Location Peak year 2002 Job losses Malaysia 24,000 (1987) 5,500 -18,500 Thailand 40,000 (1998) 18,000 -22,000 Singapore 20,000 (1998) 9,000 -11,000 Total 84,000 32,500 -51,500 Source: Company reports and 10K forms. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 131 Table 3.9 Job Losses in Malaysia's Electronics Industry, 1998­2001 Year Number of jobs lost 1998 83,900 1999 37,400 2000 25,600 2001a 18,900 a. January­August. Source: Government figures, quoted in Hamid (2001). losses of more than 16,000 during 2001, most of them in sectors related to the electronics industry, retrenchment has been dramatic, and it has hit pri- marily low-skilled, female production workers.26 Particularly disturbing is the unusually high proportion of retrenched workers (62 percent) who could not be located, indicating a massive return of Malay females (in the 25­29 age range) to their villages. Export-led electronics manufacturing is unlikely to act again as an engine of employment growth. Mismatch between the demand for and supply of skills. Finally, an in- creasingly important weakness in Malaysia's electronics industry is the seri- ous mismatch between the demand for and supply of skills. Despite the re- cession, job vacancies reached nearly 90,000 nationwide in September 2001, withthebiggestjobopeningsinthe"managerialandprofessional"categories in the electronics industry. Data collected in Penang show that a growing deficit of specialized information technology skills is an important qualita- tive constraint on Malaysia's upgrading possibilities in the electronics indus- try (DCT Consultancy Services 2002). This is especially true for engineers with degrees in electronics, mechanical engineering, quality control, testing, and chemistry. There are also important bottlenecks for mechanics, tool and die makers, and information technology professionals, especially system analysts. All of this indicates weak incentives for firms to invest in long-term assets, such as specialized skills. This human resource bottleneck also has an important qualitative di- mension. As Too and Leng (2002) document, having the right degrees in electrical and electronics engineering, information technology, and man- agement does not guarantee entry into the labor market. They also find that 26. Production workers with limited skills account for three-quarters of the total rentrenched workers, while female workers account for almost two-thirds of total job losses. 132 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the majority of unemployed graduates have not held a job of any kind since graduation. This reflects the perception of electronics firms that local uni- versity graduates have book knowledge but are ill equipped to deal with real- world problems on the shop floor and lack basic skills in communication, negotiation, and presentation. This has led to the emergence of a bifurcated labor market, where the winners pick all the stakes: there is intense compe- tition for those engineers and managers who either graduated from over- seas universities or worked for a foreign firm. Obviously, upgrading efforts will remain truncated as long as this skills mismatch is not addressed. A Shift in Strategy (I): Clusters and Global Network Integration Two policy initiatives are important for assessing Malaysia's prospects for upgrading in the electronics industry: the Second Industrial Master Plan (IMP2) and the Multimedia Super Corridor (Ministry of International Trade and Industry 1996; Multimedia Development Corporation 2002). Both represent attempts to overcome some of the structural weaknesses of this sector. Both point in the right direction but have had only limited suc- cess. The rest of this section explores whether the transformations of global production networks provide new opportunities for relieving the constraints on domestic upgrading. The IMP2. The IMP2 document signals a fundamental change in Malaysia's industrialization strategy, away from assembly-based "outward in- dustrialization" to value chain­based manufacturing, from sector-based to cluster-based development, and from performance targets to productivity- driven growth. The strategy is defined by two key concepts: manufacturing ++ and cluster-based development. In line with Porter (1990), manufactur- ing ++ highlights activities at both ends of the value chain, that is, R&D and engineering and inbound logistics, on the one hand, and outbound logistics and sales and marketing, on the other hand. It is argued that a move into knowledge-intensivesupportserviceslikeproductdevelopment, process en- gineering, supply chain management, and select areas ofR&Dwillenhance local value added and productivity. Cluster-based development implies that, based on existing strengths, especially in components and semiconductors, developing a dense web of domestic linkages will enhance value added and deepen domestic capabilities. On paper, these two concepts represent the cutting edge in current policy debates on regional and technology policy (for example, Best 2001; OECD 1999). However, within Malaysia, there are four electronics clusters that GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 133 differ quite substantially in their upgrading objectives and capabilities:27 (1) Penang Island in the north and the Kulim Hi-Tech Park in the neigh- boring state of Kedah; (2) Selangor and Negeri Sembilan in the center; (3) the southern part around Johor, with close linkages to Singapore; and (4) the more recent Multimedia Super Corridor around Kuala Lumpur. Of these, the combined Penang-Kedah cluster has arguably been the most suc- cessful, with good chances for further upgrading. IMP2 highlights four specific objectives: (1) foster the growth of leading local companies (Malaysian brands); (2) reduce dependence on imported in- puts; (3) strengthen agglomeration economies by developing integrated manufacturing centers for global network flagships; and (4) develop cross- border clusters. Of these, the first two objectives are problematic, while the last two indicate a move in the right direction. The first objective represents an outdated concept of industrial upgrad- ing that assumes a fixed pattern of sequencing from low-end, assembly-type subcontracting to original brand name (OBM) manufacturing (for a typi- cal example, see Hobday 1995). We now know that the transition to OBM is extremely difficult--even Taiwan (China)'s Acer has had only limited suc- cess (Ernst 2000b, 2000c). The limited achievements of the Proton City cluster in automobiles also indicate that this objective may be unrealistic. As for the second objective, much depends on whether the country suc- ceeds in finding the right balance between reaping the benefits of imported foreign inputs (as described in Rodrik 1999) and developing local backward and forward linkages.28 Integrated manufacturing centers. The third objective contains some promisingelements.TakerecentdevelopmentsinthePenangcluster.Rather than just giving in to requests of foreign companies for improved incentives, the state government has been pursuing a more selective approach: incen- tives are explicitly linked with the promotion of integrated manufacturing centers. The goal is to induce global flagships to move to Penang an "entire chain of operations for a particular product" (Asia Computer Weekly Online, October 22, 2001, p. 4). It is expected that this should enable the Penang cluster to upgrade from mere assembly and testing to knowledge support services, like sales and marketing, adaptive process engineering and tooling, 27. The designation of a particular location as a "cluster" raises tricky questions that are beyond the scope of this chapter. For instance, are any of these "clusters" just a collection of firms drawn to this location purely based on incentives given to local firms? And are there rationales other than policy for the observed level of agglomeration in the electronics industry in Malaysia? 28. Hirschman (1958) emphasizes the need to combine both effects. 134 GLOBAL PRODUCTION NETWORKING IN EAST ASIA financial planning, and, eventually, parts of R&D like design and develop- ment.29 Table 3.10 provides examples. Equally important is the effort by the Penang State government to de- velop a photonics industry cluster. Photonics is the technology of harness- ing light for digital applications, covering CD-ROMs, fiber-optic com- munications, lasers, sensing and measuring devices, and liquid crystal displays. Penang has developed volume-manufacturing capabilities in all of these areas. Major global players in this sector, such as Agilent, Finisar, Osram, and Solectron, are already active in Penang. In August 2001 the Penang Photonics Consortium was established. Its main objective is to bun- dle existing activities into a dynamic cluster that could provide a broad range of contract manufacturing services in optical components. One of the main local players is Chahaya Optronics, a company that has received funding from the U.S.-based hard disk manufacturer, Komag, and leading venture capital firms. The recent downturn in the global telecommunications industry drasti- cally slowed the pace of these efforts. Firms are struggling to survive. But, in principle,theideaissound,andthechallengeistobeprepared,oncedemand Table 3.10 The Integrated Manufacturing Center in Penang, 1999­2001 Year Company and product Remarks 1999 Komag: world largest supplier of Relocates entire U.S. operation to Penang, thin-film disks except R&D and sales and marketing 2000 Dell Computer: personal Establishes Penang as built-to-order hub for computers, servers, and storage the Asia-Pacific region products Reverses the decision one year later, when much of these activities are moved to Xiamen, China 2001 Quantum: hard disk drives Plans to move to Penang entire manufacturing line for digital linear tape 2001 Intel: computer chips For embedded 8-bit processors, Penang covers all value stages, including design of chip and motherboard Plans to locate two new design centers in Penang 2001 Motorola: software Motorola's software centers in Penang and MSC receive ACI level-five certification (highest level of software certification) 29. Beginning in the late 1980s, Japanese flagships, like Matsushita, had relocated design and development activities to places like Malaysia and Taiwan, China (for example, Ernst 2000a; Ernst and Ravenhill 1999). GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 135 for optical components begins to grow. Penang's photonics cluster develop- ment is characterized by attempts to leverage multiple and diverse sources of knowledge and capital, both foreign and domestic, to create a critical mass for local clustering. For instance, the Penang Development Corporation has organizedtwoworkingmeetingswithengineersandmanagersfromTaiwan's Photonics Industry and Technology Development Association, with a twofold objective: to learn from Taiwan's experience and to develop joint projects. Equally important are linkages, developed by the Penang Pho- tonics Consortium, with U.S. and Taiwanese (Chinese) venture capital firms. Finally,aconcertedeffortisunderway,incooperationwiththePenangSkills Development Centre and several local universities, emphasizing photonics in their core curriculum. Upgrading Linkages with OEMs The fourth objective of the IMP2 is of greatest interest. Extending exist- ing clusters beyond national borders originally was driven by two concerns: to ease the severe shortage of information technology skills by establishing joint growth triangles with neighboring countries that would attract low- cost engineers from throughout Asia. Yet competition among Asian economies for scarce information technology skills has drastically intensi- fied, frustrating progress along these lines. It is time now to redefine the meaning of cross-border clusters and to ask how Malaysia's existing elec- tronics clusters could reap greater benefits from participation in global pro- duction networks. As part of the concept of integrated manufacturing centers, Malaysia was able to induce some OEMs to contribute to the development of specialized clusters. Such policies, which build on earlier successful policies in Singa- pore (Wong 2000) and Taiwan, China (Ernst 2000b), can play an important catalytic role and need to continue. Yet, to a large degree, the outcome of these policies depends on sector- specific developments that are beyond the control of a mid-size country like Malaysia. Accumulated agglomeration economies matter, of course, in terms of human resources, infrastructure, and support industries (Best 2001; Ernst and others 2001). They also need to be continuously improved. Beyond that, the availability of incentives may tip the balance in favor of a particu- lar location, but only if all the other conditions are in place. Otherwise, in- centives will not have the desirable effect. A brief comparison of the cases mentioned in table 3.10 illustrates this point. The decisions by Komag and Quantum reflect the relentless pressure within the hard disk drive industry to move volume manufacturing and sup- 136 GLOBAL PRODUCTION NETWORKING IN EAST ASIA port services to locations in close proximity to Singapore, the dominant global cluster for these activities (Ernst 1997a). Footloose nature of foreign direct investment. An especially serious con- cern is that much of Malaysia's inward FDI remains highly "footloose" and prone to sudden decisions to relocate to lower-cost locations. Equally im- portant is that global flagships forced to downsize to retain shareholder value in a recession are inclined to cut employment in export platform lo- cations first, reflecting their flexible labor market regulations.30 Table 3.11 provides examples of both cases. Of particular interest is the recent decision by Dell to relocate its desk- top production for the Japanese market from Penang to Xiamen, China Table 3.11 Footloose Foreign Direct Investment in Malaysia, 1998­2001 Year Company and product Remarks 1998 Read-Rite: hard disk drives Penang facility closed and relocated to the Philippines and Thailand; 4,000 jobs lost 2000 Seagate: hard disk drives Shareholder-driven downsizing; facility in Ipoh closed; 2,000 jobs lost 2001 Seagate: hard disk drives One plant in Prai closed; 4,000 jobs lost 2001 Motorola: software 10 percent of 4,000 employees laid off in (April) plant in Sungei Way, Selangor; 400 jobs lost 2001 Intel: computer chips Worldwide workforce reduced 5 percent; expected job loss in Malaysia: 500; massive expansion in China 2001 AMD: computer chips 1,300 jobs lost in Penang plant (52 percent of worldwide job cuts); massive expansion in China 2001 Dell: personal computers, Desktop production for Japan market (August) servers, and storage relocated from Penang to Xiamen, China; products Xiamen becomes exclusive supply base in China for Dell's complete product line; with exception of desktops, Penang remains hub for the rest of Asia-Pacific; main reason: limited flight connections between Malaysia and Japan 2001 Lucent: telecommunications Regional technology center closed in equipment Malaysia; 150 jobs lost; workforce at manufacturing plant cut 50 percent 30. Thus far, contract manufacturers have been reporting only limited retrenchments. It remains to be seen to what degree the new limitations of contract manufacturer­based outsourcing may force contract manufacturers to lay off workers and close factories in Asia. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 137 and to assign Xiamen to be the exclusive supply base in China for Dell's complete product line. Although Dell's two plants in Penang remain the build-to-order shipment hub for the rest of Asia-Pacific (with the exception of desktops), this constitutes a major blow for Malaysia. While immediate job losses are only 60 (out of a total of 2,000), more such redeployments may be in the offing. Dell gives three reasons for its redeployment to China: good and low-cost Chinese engineers, cheap land, and a limited number of flight connections between Japan and Malaysia. This indicates how unpredictable and fragile Malaysia's upgrading prospects are. Upgrading opportunities: embedded software and RosettaNet. Linkages to OEMs also provide important opportunities for upgrading, including embedded software and RosettaNet. For Malaysia, important upgrading opportunities reside in embedded software, a no-frills program used in a broad array of electronics systems that does only the specific task it is meant to perform. The program takes very specific inputs from its usage environment, processes these, and pro- duces very specific outputs. Typical applications are car electronics, avion- ics, intelligent consumer products, communication and tracking devices, industrial automation, and medical equipment. In Malaysia, examples in- clude the joint software development projects of Intel and Motorola. These projects are in line with a general industry trend. There is a world- wide shortage of the specific skills required to develop embedded software. Embedded software requires very distinctive skill sets that are closer to hardware design than to mainstream software development. Essential pre- requisites are experience in manufacturing and hardware design and state- of-the-art equipment and quality control. Places, like Penang, with accu- mulated experience in manufacturing and product design (even if it involves only product customization) apparently have some advantages over tradi- tional locations for software outsourcing, such as in India, with less expo- sure to hardware design. Another interesting example is the 5 million ringgit (RM) grant allocated in the 2002 budget to promote the adoption of RosettaNet e-business stan- dards (interview with participants of Penang Skills Development Centre's seminar Jump Start Your e-Business with RosettaNet/XML, July 11, 2002). RosettaNet is a global consortium of more than 400 of the world's leading OEMs and contract manufacturers for electronic components, semiconduc- tors, computers, and telecommunications equipment, working to create, im- plement, and promote open process standards for e-business. Malaysia is the fifth economy in Asia to join RosettaNet, after Japan, Korea, Singapore, and Taiwan (China). 138 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Two tools are available to implement the RosettaNet initiative: incentives and participation in standard definition. Out of the RM 5 million grant, roughly 10 percent has been spent to set up the local operations of Rosetta- Net, with the Penang Skills Development Centre responsible for providing the backbone infrastructure. The remaining RM 4.5 million will be given out to eligible companies with no more than RM 100,000 per company. The grant will be administered by the Small and Medium Industry Development Corporation. The idea is to involve major global network flagships that are already on the RosettaNet, such as Agilent, AMD, Cisco, Dell, Hitachi, Intel, Motorola, Quantum, Siemens, and Solectron. These flagships could then be used to pressure and cajole their local suppliers to upgrade their infor- mation technology infrastructure so that the local suppliers become eligible for the grants. Another criterion for receipt of a grant is financial strength, that is, the eligible company must finance out of its own funds another 50 percent of the project cost. It is an open question, however, whether smaller lower-tier suppliers can overcome the substantial constraints on their adoption of the RosettaNet standards. Participation in defining the RosettaNet standards is probably the more immediately relevant tool. Six Malaysian electronics engineers, on loan to RosettaNet for two years, will work for six months at the California-based RosettaNet headquarters alongside American engineers to define XML- based specifications for the global electronics industry. The companies that providetheseMalaysianengineersincludeglobalflagships(Intel,Microsoft), leading local suppliers (BCM Electronics, Globetronics Multimedia Tech- nology), and two employees of the Malaysian Institute of Microelectronics Systems, a Web developer and a key public infrastructure developer. Obvi- ously, these six Malaysian engineers will play an important role as multipli- ers and upgrading catalysts, once they return from their U.S. mission. They will also act as gatekeepers for these more knowledge-intensive linkages with global flagships. Developing Multiple Linkages with Contract Manufacturers To what degree can linkages with contract manufacturers broaden Malaysia's opportunities in information technology? Within Asia, two regions have ex- perienced the greatest concentration of clusters of contract manufacturers: first Malaysia and Singapore (with a few additional sites in Thailand) and then, during the latter part of the 1990s, China (table 3.5). As for Malaysia and its neighboring countries, four important developments affect the opportunities for information technology: the arrival of major U.S. contract manufacturers, the acquisition of second-tier Asian contract manufacturers GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 139 by major U.S. contract manufacturers, the mutation of component suppli- ers from Japan and Taiwan (China) into contract manufacturers, and the up- grading efforts of Malaysian higher-tier suppliers. Arrival of major U.S. contract manufacturers. All the main U.S. con- tract manufacturers are now present in the Northern Penang­Kulim Hi- Tech cluster or in the Southern Johor­Singapore cluster. Solectron is pres- ent in Johor, Penang, and Singapore; Flextronics, in Johor and Singapore; Sanmina/SCI, in Penang and Singapore; Celestica, in Kedah's Kulim Hi- Tech Park; and Jabil Circuit, in Penang. There are also a few important in- vestmentselsewhereintheregion,suchasinIndonesia(Celestica),Malaysia's Kuching­Sarawak (Sanmina-SCI), and Thailand (Celestica, Flextronics, and Sanmina-SCI). The arrival of leading global contract manufacturers thus far has created only limited upgrading opportunities for countries like Malaysia. Against initial expectations, a website search for the five leading global contract manufacturers, which was conducted in December 2001, did not provide evidence that operations in Asia had moved significantly beyond manufac- turing. Compared to a few years ago, the main progress was an increasing sophistication in assembly technologies, especially multi-tier surface-mount technology, used for printed circuit board assembly. Most of these sites rou- tinely provide support services related to manufacturing, with the exception of assets and logistics management. Typically, this also includes electrical and mechanical design services, global test services, printed circuit board layout services, and detailed process engineering (known as advanced man- ufacturing technology research). These services provide manufacturing so- lutions that enable a quick ramping up of volume manufacturing. But more recent interviews in late 2002, primarily with leading Taiwanese ODM companies, showed an increase in more knowledge-intensive activities, in- cluding electronic design, indicating a growing similarity in the distribution of capabilities across the Americas, Europe, and Asia (table 3.12). A few locations, primarily in Penang and Singapore, are also involved in the introduction of new products. These two locations are witnessing the development of ODM capabilities, but on a very limited scale. Overwhelm- ingly, leading global contract manufacturers concentrate design (and espe- ciallycircuit,advancedoptical,andsystemsdesign)inEuropeandtheUnited States. One would, of course, expect such a disparity in design and product development,duetotheirhighknowledgeintensity.This,however,ischang- ing, as Taiwanese (Chinese) contract manufacturers are providing such ODM services (Wu 2002). Industry observers expect that leading Taiwanese (Chinese) design firms will soon provide ODM services from their overseas network sites in China as well as in Penang and Singapore. 140 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Table 3.12 Geographic Dispersion of Capabilities in Contract Manufacturing, 2001 Capability Americas Europe Asia Manufacturing and distribution Supply base and logistics management · · · Printed circuit board assembly · · · Complex systems assembly · · · Build-to-stock systems assembly · · · Build-to-order systems assembly · · · Configure-to-order systems assembly · · · Channel assembly · · Penang (2002) Systems integration and reconfiguration testing · · a Environment stress screening · · a Custom packaging · · a Logistics and distribution management · · Support services Repair systems and printed circuit boards · · · Product refurbishment and remanufacturing · · · Assets and logistics management · · · Product upgrades · · · Sustaining engineering · · · End-of-life manufacturing · · · Warranty processing · · · Technology Interconnection and packaging consulting · · · Process development · · · Reliability and failure analysis · · · Manufacturing technology roadmap · · · Test technology roadmap · · · Design Design of application-specific integrated circuits · · b Circuit design · · b Radio-frequency and wireless design · · b Mechanical design · · b Systems design · · b Test process design · · b Design validation · · b Product development New product introduction management · · · Component engineering · · · Design-for-manufacturability · · · Design-for-testability · · · Printed circuit board layout · · · Test development · · · Quick-turn prototyping · · · Quick-turn testing · · · a. Projected to change: focus on China. b. Projected to change: Taiwanese ODM companies as trend-setters. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 141 A comparison with locations in China, Hong Kong (China), and Taiwan (China) also demonstrates an important weakness of contract manufacturer locations in Southeast Asia. Very few final or system assembly activities are locatedintheMalaysia-Singaporeclusters.Overwhelmingly,theyarelocated in Europe and the United States, in close proximity to the traditionally dom- inant markets. But over the last few years, and especially in response to the recession in Europe and the United States, leading contract manufacturers havestartedtoestablishfinalassemblylocationsandbuild-to-ordershipment hubs in China, Hong Kong (China), and Taiwan (China). The obvious mo- tivation is to be as close as possible to the potentially huge Chinese market. In short, the inflow of substantial investments from contract manufac- turers thus far has produced some opportunities for Malaysia to move be- yond its traditional focus on volume manufacturing. But these opportu- nities have not yet reached a level sufficient for a major push into more knowledge-intensive activities. The acquisition of second-tier Asian contract manufacturers. A sec- ond important development in Malaysia's cluster is that leading U.S. con- tract manufacturers have recently rushed to acquire second-tier Asian con- tract manufacturers, primarily in Singapore, but also in Malaysia and elsewhere in the region. These acquisitions reflect the rapid concentration intheglobalcontractmanufacturerindustry,whichhasbeendrivenbyM&A (see figure 3.2). The recession has further accelerated these trends toward concentration. Important recent examples include Solectron, which acquired the Singa- porean contract manufacturers Natsteel Electronics and Singapore Shinei Sangyo (the latter an affiliate of a Japanese component supplier); Flextron- ics, which acquired second-tier Singaporean contract manufacturer JIT Holdings and Li Xin Industries; and Celestica, which, through its acquisi- tion of Singapore's contract manufacturer Omni (October 2001), acquired facilities in Indonesia, Malaysia, Singapore, and Thailand, with almost 9,000 employees. To the degree that these acquisitions will result in plant closures and lay- offs, they may constrain opportunities for industrial upgrading. However, they could have positive effects as well, if they insert new capital, customers, and management approaches. Component suppliers from Japan and Taiwan (China). A third impor- tant development predates the arrival of American contract manufacturers, which has absorbed most public attention. Suppliers of parts and compo- nents from Japan and Taiwan (China), whose arrival in Malaysia goes back 142 GLOBAL PRODUCTION NETWORKING IN EAST ASIA to the second part of the 1980s, catalyzed the development of Malaysia's local support industries (for example, Ernst 1997a; Takeuchi 1993). Japanese component manufacturers are concentrated primarily in the consumer elec- tronics sector. Some of them, however, have also branched out into the com- puter sector. An interesting example is Kobe Precision (Malaysia), a com- pany that, in October 2000, was acquired by one of the leading Malaysian contract manufacturers, Eng Technologi Holdings (Business Times, Kuala Lumpur, October 27, 2000). Taiwanese (Chinese) firms have played an important role in Malaysia's computer industry. Prominent examples are Acer Peripherals and Iventech. Both companies became involved in Malaysia during the late 1980s. Over time, the Malaysian affiliates of these and other Taiwanese (Chinese) firms upgraded from simple volume manufacturing, according to designs owned by the global flagships, to sophisticated contract manufacturing for leading computer network flagships. In addition to manufacturing, the Malaysian sites of these Taiwanese (Chinese) affiliates now provide product and com- ponent design, supply chain management services, and other knowledge- intensive support services. A handful of large Taiwanese (Chinese) contract manufacturers, led by firms like Acer, Delta Networks, Iventech, Kinpo Electronics, and USI, have pioneered the use of ODM capabilities in Asia (report by Technology Forecasters, Alameda, Calif., quoted in EMS Insight, supplement on Circuits Assembly, September 2000). This apparently has forcedmajorU.S.-basedcontractmanufacturerstofollowsuit.ForMalaysian firms that interact with affiliates of these Taiwanese (Chinese) contract man- ufacturers, the move to ODM capabilities may provide opportunities for industrial upgrading. This example illustrates that FDI policies and incen- tives should not target only the world industry leaders. Wherever possible, they should try to attract second-tier actors that are willing to bring along more knowledge-intensive activities. Upgrading efforts of Malaysian higher-tier suppliers. There might be significant opportunities for higher-tier Asian suppliers to compete as low-cost niche contract manufacturers. Recent interviews in Penang (July 2002) indicate that leading higher-tier local suppliers all understand that they need to move up within the hierarchy of contract manufacturing arrangements, from low-end box-build and consignment arrangements to the provision of ODM and total solutions. However, they face major prob- lems in sustaining and expanding their upgrading efforts. They all face the demanding challenge of pursuing simultaneously the following upgrading strategies, each of which requires major investments: establish a credible position as a low-cost niche contract manufacturer; develop a global pres- GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 143 ence, through overseas investment; achieve diversification and market seg- mentation; develop knowledge-intensive support services; and invest in design and R&D. LC1, one of the most successful local companies, is building on existing strengths in contract manufacturing and the provision of ODM services in an effort to become a lower-cost "total solution provider" for carefully cho- sen niche markets. To do this with low overhead requires strong capabilities in six highly inter-dependent functions: manufacturing, quality, materials, procurement, engineering, and human resources. It is important to empha- size the systemic nature of the required capabilities. The implementation of these upgrading options requires the develop- ment of a broad and diverse set of capabilities. Take manufacturing services. The move from printed circuit board assembly and box-build to testing necessitates the development not only of testing capabilities (which are scarce) but also of system engineering and maintenance capabilities. Fur- thermore, developing design and engineering capabilities requires substan- tial funds for R&D. Or take after-ship services. A seemingly mundane activ- ity like repair requires technicians trained in failure analysis, while end-of- life program management requires capable supply chain managers. As for the upgrading of procurement and outbound logistics, substantial funds are required to gradually upgrade the necessary information systems. In short, the ability to upgrade and become an Asian niche market con- tract manufacturer requires substantial investments in training, equipment, facilities, and, most important, R&D. LC1 identifies the following seven challenges that result from this strategy. Substantial improvements are re- quired for supply chain management and efficient use of the company's as- sets. These are the most fundamental requirements for staying in this busi- ness. Yet their implementation requires substantial resources and attention from management. Challenges three to five constitute the medium-term challenge: the company needs to develop a strong portfolio of designs (so- called intellectual properties), it needs to capture new global niche market opportunities, and it needs to develop a global presence. Finally, the last two challenges highlight critical changes in industry organization, that is, the move toward flexible domestic supplier networks that can complement LC1's own capabilities and the overriding importance of human resource development as a constant process of acquiring new skills and knowledge. A Shift in Strategy (II): International Knowledge Sourcing Despite impressive achievements, Malaysia's knowledge base in the elec- tronics industry remains too weak to sustain industrial upgrading into more 144 GLOBAL PRODUCTION NETWORKING IN EAST ASIA knowledge-intensive activities. There is a heavy reliance on technological capabilities developed within affiliates of global flagships and their eventual spillover into local firms. This traditional pattern of transfer does not seem to work any longer. Searching for new sources of productivity growth. In Penang, for in- stance, a disturbing slowdown in total factor productivity growth has been observed since 1995 (State Government of Penang 2001). Between 1995 and 1997, total factor productivity declined 0.5 percent for all manufacturing compared to an increase of 8.9 percent between 1990 and 1995. In the elec- tronics industry, total factor productivity growth fell to 2 percent (from 14 percent during the earlier period), hardly sufficient for an industry that is supposed to be the engine of upgrading. For all of Malaysia, most estimates put total factor productivity growth at about 1­2 percent annually (until 2000). This is way below the 3.2 per- cent minimum total factor productivity growth (for the period 2001­10) that is necessary if Malaysia wants to achieve the projected growth rate of 7.5 percent. Compared with historical patterns of productivity growth in industrial countries, this slowdown in productivity growth comes much too early. For Malaysia, such a massive slowdown in total factor productivity growth is certainly premature, in light of the limited progress made in its specialization by product and production stage. In short, while FDI by major global electronics OEMs used to play a catalytic role in boosting Malaysia's productivity growth before 1995, it may no longer play that role. Except for China, and possibly India, OEMs are unlikely to increase their inward FDI in Asia. These changes are structural rather than cyclical, and they are here to stay. Hence, Malaysia needs to develop a set of alternative international linkages that could play a complementary role as external sources of productivity growth. The multimedia super corridor. A widely known attempt to address this issue is the government's initiative to establish a $40 billion Multimedia Super Corridor that was supposed to leapfrog the country into the status of a fully developed nation by 2020 (Multimedia Development Corporation 2002). In 1996 the government hired McKinsey, the global consulting firm, to draft a blueprint for a 15-kilometer-by-50-kilometer strip intended to be Malaysia's answer to Silicon Valley. An unprecedented set of incentives, en- shrined in the Bill of Guarantees, was offered to companies involved in the creation, distribution, integration, or application of multimedia products and services within the Multimedia Super Corridor. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 145 According to the Multimedia Development Corporation (2002), these incentives include the following commitments: · Provide a world-class physical and information infrastructure · Allow unrestricted employment of local and foreign knowledge workers · Ensure freedom of ownership by exempting companies with Multimedia Super Corridor status from local ownership requirements · Give the freedom to source capital globally for Multimedia Super Cor- ridor infrastructure and the right to borrow funds globally · Provide competitive financial incentives, including pioneer status (100 percent tax exemption) for up to 10 years or an investment tax allowance for up to five years, and remove duties on the importation of multimedia equipment · Become a regional leader in intellectual property protection and cyber laws · Ensure no censorship on the Internet · Provide globally competitive telecommunications tariffs · Tender key infrastructure contracts to leading companies willing to use the Multimedia Super Corridor as their regional hub · Provide a high-powered implementation agency to act as an effective one-stop super shop. It would be hard to find a more aggressive list of incentives. By 2000, $3.7 billion had been spent, but the results were disappointing (however, there was some increase in investor interest in 2003). A leaked confidential report by the very same company that designed the project (McKinsey) con- cluded in February 2001 that the Multimedia Super Corridor "had not at- tracted much interest from global investors nor made an impact on the do- mestic economy" (Prystay 2001).31 In the meantime, the debate has moved on to explore what one can learn from the experience to date (author's in- terviews with members of the National Information Technology Council, July 4, 2002, in Kuala Lumpur). Three conclusions emerge. First, initiating the Multimedia Super Corridor was a step in the right direction. How- ever, it is now time to expand its geographic coverage and to extend Mul- timedia Super Corridor status to other electronics clusters in Malaysia. The Penang State government, in particular, has been lobbying for such an ex- tension, arguing that this is necessary to attract specialized skills from over- seas on a contract basis to overcome critical shortages. 31. Some 500 companies, including 45 international ones, are located in the corridor. But the McKinsey report describes the level of investment of those international electronics firms as "not very significant" (Prystay 2001). 146 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Second, lavish tax incentives and massive investment in infrastructure are insufficient to bring about the development of dynamic clusters. Especially for the information technology sector, infrastructure is a highly perishable "public good": the infrastructure for the Multimedia Super Corridor was perhaps state-of-the-art when it was established a few years ago, but it be- came obsolete very rapidly. Third, perhaps the most important ingredient of successful cluster formation is missing: specialized skills and innovation capabilities. As emphasized throughout this chapter, the keys to success are incessant efforts on a massive scale to continuously upgrade existing skills and capabilities. The lack of depth and horizontal mobility in Malaysia's labor market increases the risk of individual investment in specialized skills. Therefore, the importing of scarce skills should be given greater emphasis. While this may seem obvious, this simple fact is frequently forgotten. In the case of Malaysia, the gap between the supply of and the demand for special- izedinformationtechnologypersonnelhascontinuouslyincreased,especially for engineers.32 In short, as long as this critical human resources bottleneck is not overcome, there is little hope that the Multimedia Super Corridor will act as a "breeding ground for technological innovations, new businesses, and companies through the cooperation between the industrial and academic circles" (Takeuchi 1997, quoting from official background documents). The following major priority areas for reducing the skills mismatch in the Malaysian electronics industry were identified during interviews with gov- ernment agencies and leading companies (June-July 2002):33 (1) a massive reskilling and retraining requirements of production workers, (2) graduates, especially for electrical and electronics engineering, information technol- ogy, and circuit design, who are able to combine hardware, software, and ap- plication knowledge, (3) experienced managers, especially for strategic mar- keting, upgrading management, and management of international linkages, (4) entrepreneurs who combine streetwise commercial and financial in- stincts with analytic capacity for strategic decisionmaking, (5) experienced and industry-savvy administrators who are willing to stick out their necks 32. On this indicator, Malaysia is continuously ranked at the bottom in the annual World Com- petitiveness Yearbook (Institute for Management Development various years), lagging substantially behind India, Korea, Singapore, and Taiwan (China). 33. In Kuala Lumpur, I am especially indebted to discussions with Dr. Zawai Ismail, director, Commerce Asset Ventures, who set up brainstorming sessions with relevant government agen- cies and venture capital firms. In Penang, I am especially indebted to discussions with Dr. Koh Tsu Koon, chief minister of Penang, Dr. Toh Kin Woon, Penang State executive councilor, Mr. Boonler Somchit, executive director of the Penang Skills Development Centre, Dr. Ganesh Rasagam, chief executive officer, DCT Consultancy Services, and Dr. Anna Ong, senior analyst, Socio-Economic and Environmental Research Institute. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 147 and to do more than just follow the rules (this, of course, requires some in- centive alignment), (6) incentive alignments for university professors and academics that encourage close interaction with the private sector (com- pany internships and sabbaticals), (7) dense interactions with expatriate na- tionals who are based in Australia, Europe, the United States, or elsewhere in Asia, and (8) a capacity to bring in at short notice specialized experts from overseas who can bridge existing knowledge gaps and catalyze necessary changes in organization and procedures required to develop these capabil- ities locally. Diversifying international linkages. In light of the transformations of global production networks documented in this chapter, Malaysia should exploit new opportunities for diversifying international linkages. Three complementary international linkages deserve particular attention: (1) links with foreign universities and research institutes; (2) links with information service and consulting firms, especially smaller, second-tier firms; and (3) links with informal global peer group networks that are playing an increasingly important role as carriers of knowledge. Foreign universities and research institutes. Current transformations in the organization of global production networks, accelerated by the use of digi- tal information systems, have substantially increased the mobility of knowl- edge. Digitalization implies that knowledge can be delivered as a service and built around open standards. This has fostered the specialization of knowl- edge creation, giving rise to a process of modularization, very much like ear- lier modularization processes in hardware manufacturing. These developments may well create new opportunities for more ag- gressive forms of industrial upgrading that no longer need to avoid the cre- ation of original knowledge. At this stage, this is largely uncharted territory, as these developments are very new and there is practically no research. Nevertheless, it is time for a country like Malaysia to strengthen linkages with overseas universities that can help to upgrade research, development, and design capabilities in Malaysian universities and public labs. The start- ing point is to correct the current policies. The focus thus far has been on a handful of global elite institutions that bring in their standard, routine in- formation technology and business courses at very high cost. Instead, col- laboration should focus on specific niche areas, in line with Malaysia's needs. Possible examples include certain areas of chip design, packaging technol- ogy, and photonics. Realistically, the search should move beyond the exclu- sive ranks of the Ivy League universities: there is ample choice of smaller, less well-known universities and research institutes that are more willing 148 GLOBAL PRODUCTION NETWORKING IN EAST ASIA to develop innovative courses customized to the specific needs and capa- bilities of Malaysia's electronics clusters. These may include institutions in economies like India (the well-regarded Indian Institutes of Technology and the Indian Institutes of Information Technology), Korea (the Electronics and Telecommunications Research Institute, the Korea Advanced Institute of Science and Technology, the Information and Communications Univer- sity of Korea, and other specialized research institutes), and Taiwan, China (the Industrial Technology Research Institute, the Electronics Research and Service Organization, and the Institute for Information Industry). International consulting firms. Equally important is a reconsideration of link- ages with consulting firms. For information technology, these firms now play a critical role in diffusing both codified and tacit knowledge. The prob- lem is that this market is overwhelmingly dominated by a handful of gi- ant corporations like IBM and consultancies like Accenture that grew out of global accounting firms. These firms thrive on the economies of scale in knowledge sharing (which information economists call network economies). As flagships of global information service networks, these firms provide a standard product wherever they go. Customization is possible only within the limits of a package of standard solutions. This approach to customization is extremely costly: customers are charged for the time required to adjust the package of standard information technology and to achieve effective imple- mentation, and these costs are inflated by massive delays. The result is that new systems often come in late, over budget, and unable to solve problems they were meant to address. This has created a demand for smaller, special- ized niche players that do not start from a package of standard solutions and that offer clients fixed-price projects. A wide choice of smaller, less well- known but proven information service and consulting firms is now available. This may include firms from some of the Asian economies that already have some experience with knowledge-intensive information services. Informal peer group networks. Malaysia also needs to tap into an increasingly important carrier of international knowledge diffusion: transnational tech- nical communities (Saxenian 2002) of technically skilled immigrants with business experience and connections in Europe, Japan, and the United States that play an important and complementary role to network flag- ships in global production networks. Such informal peer group global networks have created new opportunities for industrial upgrading in for- merly peripheral economies around the world. By linking their home coun- try with the world's centers of information and communication technology (encompassing Silicon Valley and other centers of excellence in less well- GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 149 known places like Grenoble, Helsinki, Kista [Stockholm], Munich, Tel Aviv, and Tsukuba), these informal social networks transform what used to beaone-waybraindrainintoatwo-wayprocessof"braincirculation."These networks generate invaluable knowledge on trends in global markets and technology in a way that addresses the needs of domestic firms much bet- ter than linkages with global flagships or, for that matter, with global con- sulting firms. They also provide entrepreneurs and venture capitalists who can function in both worlds. This has created alternative and robust mechanisms for exchanging knowledge across geographic borders and firms. Examples include China, India, Israel, Korea, Taiwan (China), as well as Brazil and Mexico. In Malaysia, the Penang cluster has obviously benefited from students who have studied engineering and management overseas, whether in Aus- tralia, Japan, the United Kingdom, or the United States, and who have re- turned with business experience and connections. Predominantly, these connections have been with global flagships like Intel and Motorola in semi- conductors or with Matsushita and other Japanese flagships in consumer electronics. Overwhelmingly, the technology, skills, and knowledge gener- ated by these immigrant engineers have focused on manufacturing-related activities. It is time for Malaysia to adjust this brain circulation to encom- pass new areas like knowledge-intensive support services, circuit design, and chip packaging. CONCLUSIONS AND POLICY SUGGESTIONS Based on operational definitions of the key characteristics of global pro- duction networks and industrial upgrading, this chapter has explored how East Asia, and especially Malaysia, could build on recent transformations in the structure, coordination, content, and location of global production net- works to promote a continuous upgrading of the electronics industry. We have shown that the reintegration of geographically dispersed, specialized production and innovation sites into multilayered global production net- works and the increasing use of information technology­based information systems to manage these networks are gradually reducing the constraints on the international diffusion of knowledge. Global production networks ex- pand inter-firm linkages across national boundaries, increasing the need for knowledge diffusion, while information systems enhance not only the ex- change of information but also the sharing and joint creation of knowledge. This new mobility of knowledge has created greater opportunities for us- ing network participation as a catalyst for upgrading Asia's electronics in- 150 GLOBAL PRODUCTION NETWORKING IN EAST ASIA dustries further. But realizing this potential has become more difficult for mid-size countries like Malaysia. While Malaysia has developed some of the most ambitious sets of policies and incentives, the country has a long way to go to exploit these new opportunities. The best choice arguably is to move forward in incremental steps and to build on existing strengths in assembly and volume manufacturing by adding knowledge-intensive support services. While public policy documents frequently talk about "catching up" and "forging ahead" through technology leadership strategies (ISIS Malaysia 2002), the de facto strategies appear to focus on technology diversification, that is, the recombination of (mostly) known technologies. Defined as "the expansion of a company's or a product's technology base into a broader range of technology areas" (Granstrand 1998, p. 472), such strategies focus on products and services that draw "on several . . . crucial technologies which do not have to be new to the world or difficult to acquire" (Granstrand and Sjölander 1990, p. 37). The transformations in global production networks that this chapter has explored in detail may have created opportunities for Asia's leading electronics-exporting countries to engage in technology diversification that did not exist before. Asian electronics firms may also have important latecomer advantages, building on their accumulated capa- bilities to implement, assimilate, and improve foreign technologies, as tech- nology diversification often involves the exchange of knowledge with for- eign parties (Ernst 2003c). Of critical importance is the absorptive capacity of local suppliers, that is, their resources, capabilities, and motivations. The absorptive capacity is shaped by pressures exerted by network flagships and by existing incentives. This chapter documents that, to stay on the global production networks, local suppliers must constantly upgrade their absorptive capacity by invest- ing in their skills and knowledge base. Equally important are attempts to strengthen the country's innovative capabilities through selective sourcing of international knowledge. Trans- formationsinglobalproductionnetworksaregradually reducing the barriers to effective knowledge diffusion. As an immediate policy instrument, it is ad- visabletoimportcriticalskillsfromoverseas.Thiscouldhelptocatalyzenec- essary reforms in the domestic innovation system. The timing may be good, as massive retrenchments in the U.S. and European electronics industries and a more hostile attitude toward foreign researchers, especially from the Middle East but also from Asia, may induce foreign researchers to work in Malaysia. Adequate incentives are required to generate sufficient investments in the development of skills and capabilities, as illustrated, for instance, by the Nordic countries in Europe and by Singapore and Taiwan (China). Poli- GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 151 cies toward both OEMs and contract manufacturers need to move beyond "incentive tournaments." Infrastructure development is critical but needs to move beyond a widespread bias toward hardware. As illustrated in this chapter, successful upgrading within global production networks requires policies that support local firms through the development of local suppli- ers, (cofunded) development of skills, setting of standards, and provision of investment and innovation finance through a variety of sources, including venture capital and initial public offerings. Of particular importance for East Asia are new opportunities to tap into international flows of human capital and knowledge through informal peer group networks of technically skilled immigrants with business experience and connections in Europe, Japan, and the United States. These international social networks can play an important and complementary role as carriers of knowledge and capital to OEMs, contract manufacturers, and global consulting firms. To reap the benefits of integration into global production networks re- quires a very active involvement of the state (that is, local, regional, and cen- tral government agencies as well as a variety of intermediate institutions). But this involvement is taking on a very different form from earlier top- down, "command economy" industrial policies. It also differs from the new- economy liberalization doctrine. Traditional Asian developmental policies are no longer feasible. With their top-down approach, controlled investment finance, and reliance on state-owned enterprises or chaebol-type conglomerates, these policies are too rigid to cope with the complex challenges and opportunities of the global network economy that have been explored in this chapter. These policies cannot cope with the conflicting needs of multiple and increasingly vocal domestic actors. In addition, traditional developmental policies are unable to cope with the high uncertainty and rapid changes in technology and mar- kets that are typical of the electronics industry. Finally, in light of their pro- tectionist focus, these policies are unlikely to generate and benefit from international knowledge linkages that are of critical importance for upgrad- ing Asian electronics industries. Neither can the new-economy liberalization doctrine cope with the new opportunities and challenges for Asian electronics industries that are exam- ined in this chapter. This doctrine claims that, except for education, infra- structure, and a few general incentives (for training and R&D), the state should get out of the way and let transnational technical and venture capi- tal communities make the necessary investments in innovative capabilities (for example, Bresnahan, Gambardella, and Saxenian 2001). While such a selective approach is an important condition for breaking up the stalemate of the traditional Asian developmental policies, it is insufficient to cope with 152 GLOBAL PRODUCTION NETWORKING IN EAST ASIA the new opportunities and challenges raised by the transformation of global production networks. It is ironic that the U.S. government does not do what the proponents of the new-economy doctrine claim it does. Driven by na- tional security concerns and by fears of losing its technological leadership, the federal government spends $2 billion a year on information technology research, coordinated through a variety of agencies including the Defense Advanced Research Projects Agency, which has played a prominent role in the development of the U.S. computer and Internet industry (for example, Flamm 1988). In short, for Malaysia's upgrading efforts in the electronics industry to succeed, new policy approaches are required (Ernst forthcoming c): · Strengthen the state's steering and coordination capacity · Provide public goods and create assets (infrastructure, bottleneck skills, training, and education) · Facilitate access to and diffusion of knowledge; balance this with the need to protect intellectual property rights · Encourage innovations in the financial sector · Generate dialogue at various levels among multiple participants (local and foreign) in production and innovation networks · Foster interactive learning and innovation · Provide social protection and retraining options for the losers from innovation · Facilitate international knowledge sourcing through corporate networks, institutional collaboration, and diverse social networks (global knowledge communities and expatriates). REFERENCES The word processed describes informally reproduced works that may not be commonly avail- able through libraries. Aberdeen Group. 2001. "Offshore Software Development: Localization, Globalization, and Best Practices in an Evolving Industry." Boston, Mass. Processed. Almeida, Paul, and Bruce Kogut. 1997. "The Exploration of Technological Diversity and the Geographic Localization of Innovation." Small Business Economics 9(1):21­31. Ariffin, Norlela. 2000. "The Internationalisation of Innovative Capabilities: The Malaysian Electronics Industry." Ph.D. diss., University of Sussex, Science and Technology Policy Research Unit. Best, Michael H. 2001. The New Competitive Advantage: The Renewal of American Industry. Oxford: Oxford University Press. Borrus, Michael, Dieter Ernst, and Stephan Haggard, eds. 2000. International Production Networks in Asia: Rivalry or Riches? London and New York: Routledge. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 153 Boswell, David. 1993. Subcontracting Electronics: A Management and Technical Guide for Pur- chasers and Suppliers. London: McGraw-Hill. Bresnahan, Timothy, Alfonso Gambardella, and Anna-Lee Saxenian. 2001. " `Old Econ- omy' Inputs for `New Economy' Outcomes: Cluster Formation in the New Silicon Valleys." Industrial and Corporate Change 10(4):835­60. Chen, Shin-Horng. 2002. "Global Production Networks and Information Technology: The Case of Taiwan (China)." Industry and Innovation 9(2). Special issue on global pro- duction networks. DCT Consultancy Services. 2002. "Annual Survey of the Manufacturing Industries in Penang Development Corporation Industrial Areas." Report submitted to the Penang Development Corporation, Penang, Malaysia. Processed. Depeyrot, Michel. 2002. "Taiwan Foundries Acquiring Design Capabilities." Electronic Engineering Times, September 25. Ernst, Dieter. 1994a. "Network Transactions, Market Structure, and Technological Diffu- sion: Implications for South-South Cooperation." In Lynn K. Mytelka, ed., South-South Cooperation in a Global Perspective. Development Centre Documents. Paris: Organisation for Economic Co-operation and Development. ------. 1994b. What Are the Limits to the Korean Model? The Korean Electronics Industry under Pressure. BRIE Research Monograph. University of California, Berkeley Roundtable on the International Economy. ------. 1997a. "From Partial to Systemic Globalization: International Production Networks in the Electronics Industry." Data Storage Industry Globalization Project Report 97-02. University of California at San Diego, Graduate School of International Relations and Pacific Studies for the Sloan Foundation. Processed. ------. 1997b. "Partners in the China Circle? The Asian Production Networks of Japan- ese Electronics Firms." In Barry Naughton, ed., The China Circle. Washington, D.C.: Brookings Institution. ------. 1998. "Catching Up: Crisis and Industrial Upgrading, Evolutionary Aspects of Technological Learning in Korea's Electronics Industry." Asia Pacific Journal of Man- agement 15(2):247­83. ------. 2000a. "Evolutionary Aspects: The Asian Production Networks of Japanese Elec- tronics Firms." In Michael Borrus, Dieter Ernst, and Stephan Haggard, eds., International Production Networks in Asia: Rivalry or Riches? London: Routledge. ------. 2000b. "Inter-Organizational Knowledge Outsourcing. What Permits Small Taiwan Firms to Compete in the Computer Industry?" Asia Pacific Journal of Management 17(2, August):223­55. Special issue on knowledge management in Asia. ------. 2000c. "Placing the Networks on the Internet: Challenges and Opportunities for Managing in Developing Asia." Paper presented at the Second Asia Academy of Man- agement Conference, Singapore, December 15­18. Processed. Also forthcoming in Bengt-Âke Lundvall and Keith Smith, eds., Knowledge Creation in the Learning Economy. Cheltenham: Edward Elgar. ------. 2001a. "Catching up and Post-Crisis Industrial Upgrading: Searching for New Sources of Growth in Korea's Electronics Industry." In Frederick C. Deyo, Richard Doner, and Eric Hershberg, eds., Economic Governance and the Challenge of Flexibility in East Asia. Lanham, Md.: Rowman and Littlefield. ------. 2001b. "Moving beyond the Commodity Trap? Trade Adjustment and Industrial Upgrading in East Asia's Electronics Industry." In Richard S. Newfarmer and Christina 154 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Wood, eds., East Asia: From Recovery to Sustainable Development. Washington, D.C.: World Bank. ------. 2002a. "The Economics of Electronics Industry: Competitive Dynamics and Indus- trial Organization." In William Lazonick, ed., The International Encyclopedia of Business and Management (IEBM), Handbook of Economics. London: International Thomson Busi- ness Press. ------. 2002b. "Global Production Networks and the Changing Geography of Innovation Systems: Implications for Developing Countries." Journal of the Economics of Innovation and New Technologies 11(6):497­523. ------. 2003a. "Digital Information Systems and Global Flagship Networks: How Mobile Is Knowledge in the Global Network Economy?" In Jens F. Christensen and Peter Maskell, eds., The Industrial Dynamics of the New Digital Economy. Cheltenham, U.K.: Edward Elgar. ------. 2003b. "The New Mobility of Knowledge: Digital Information Systems and Global Flagship Networks." In Robert Latham and Saskia Sassen, eds., Digital Formations in a Connected World. Princeton, N.J.: Princeton University Press for the U.S. Social Science Research Council. ------. 2003c. "Pathways to Innovation in Asia's Leading Electronics Exporting Countries: Drivers and Policy Implications." Paper prepared for the conference the Common Future of the Twenty-first Century Pacific, Democratic Pacific Assembly, Taipei, Taiwan, September 18­20. Available as East-West Center Economics Working Paper 62. East- West Center, Honolulu, October. Processed. Forthcoming in International Journal of Innovation and Technology Management (IJITM). ------. Forthcoming a. "Global Production Networks and Industrial Upgrading: A Knowledge-Centered Approach." In Gary Gereffi and Eric Hershberg, eds., Who Gets Ahead in the Global Economy? Industrial Upgrading, Theory and Practice. Baltimore, Md.: Johns Hopkins University Press. ------. Forthcoming b. "Internationalisation of Innovation: Why Is Chip Design Moving to Asia?" International Journal of Innovation Management, special issue in honour of Keith Pavitt (Peter Augsdoerfer, Jonathan Sapsed and James Utterback, guest editors). Also published as East-West Center Economics Working Paper 62. ------. Forthcoming c. "Pathways to Innovation in the Global Network Economy: Asian Perspectives." Oxford Development Studies. Special issue in honor of Linsu Kim. Ernst, Dieter, Tom Ganiatsos, and Lynn Mytelka, eds. 1998. Technological Capabilities and Export Success: Lessons from East Asia. London: Routledge. Ernst, Dieter, and Paolo Guerrieri. 1998. "International Production Networks and Chang- ing Trade Patterns in East Asia. The Case of the Electronics Industry." Oxford Develop- ment Studies 26(2): 191­212. Ernst, Dieter, Paolo Guerrieri, Simona Iammarino, and Carlo Pietrobelli. 2001. "New Challenges for Industrial Districts: Global Production Networks and Knowledge Dif- fusion." In Paolo Guerrieri, Simona Iammarino, and Carlo Pietrobelli, eds., The Global Challenge to Industrial Districts: Small and Medium-Sized Enterprises in Italy and Taiwan. Aldershot, U.K.: Edward Elgar. Ernst, Dieter, and Linsu Kim. 2002. "Global Production Networks, Knowledge Diffusion, and Local Capability Formation." In Research Policy 31(8­9):1417­29. Ernst, Dieter, and John Ravenhill. 1999. "Globalization, Convergence, and the Transfor- mation of International Production Networks in Electronics in East Asia." Business and Politics 1(1, April): 53­62. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 155 Feenstra, Robert C. 1998. "Integration of Trade and Disintegration of Production in the Global Economy." Journal of Economic Perspectives 12(4):31­50. Flamm, Kenneth. 1988. Creating the Computer: Government, Industry, and High Technology. Washington, D.C.: Brookings Institution. Globerman, Stephen. 1997. "Transnational Corporations and International Technological Specialization." Transnational Corporations 6(2, August): 95­114. Granstrand, Ove. 1998. "Toward a Theory of the Technology-Based Firm." Research Policy 27(5, September):465­89. Granstrand, Ove, and Sören Sjölander. 1990. "Managing Innovation in Multi-Technology Corporations." Research Policy 19(1, February):35­60. Grossman, Gene M., and Elhanan Helpman. 1991. Innovation and Growth in the Global Econ- omy. Cambridge, Mass.: MIT Press. ------. 2002. "Integration versus Outsourcing in Industry Equilibrium." Quarterly Journal of Economics 117(1, February):85­120. Hamid, Hamisah. 2001. "Lay-Offs Unlikely to Be as Severe as Three Years Ago." Business Times, October 3. Hirschman, Albert O. 1958. Strategy of Economic Development. New Haven, Conn.: Yale University Press. Hobday, Michael. 1995. Innovation in East Asia: The Challenge to Japan. Aldershot, U.K.: Edward Elgar. Institute for Management Development. Various years. World Competitiveness Yearbook. Lausanne, Switzerland. ISIS (Institute of Strategic and International Studies) Malaysia. 2002. Knowledge-Based Econ- omy Master Plan. Kuala Lumpur, September. Jabil Circuit. 2001. Jabil 10K Report to the United States Securities and Exchange Commission. New York, September 30. Jones, Ronald, and Henryk Kierzkowski. 2000. "A Framework for Fragmentation." In Sven Arndt and Henryk Kierzkowski, eds., Fragmentation and International Trade. Oxford: Oxford University Press. Lakenan, Bill, Daren Boyd, and Ed Frey. 2001. "Why Cisco Fell: Outsourcing and Its Per- ils." Strategy + Business 24(3rd quarter):54­65. Lall, Sanjaya. 1997. "Technological Change and Industrialization in the Asian NIEs: Achievements and Challenges." Paper presented at the international symposium on "Innovation and Competitiveness in Newly Industrializing Economies," Science and Technology Policy Institute, Seoul, Korea, May 26­27. Processed. Lazonick, William. 2000. "Understanding Innovative Enterprise: Toward the Integration of Economic Theory and Business History." Unpublished mss., University of Massa- chusetts, Lowell and European Institute of Business Administration (INSEAD), Fontainebleau, May. Processed. Lipsey, Richard G. 2001. "Understanding Technological Change." East-West Center Work- ing Paper, Economics Series 13. East-West Center, Honolulu, February. Processed. Macher, Jeffrey T., David C. Mowery, and Timothy S. Simcoe. 2002. "eBusiness and the Semiconductor Industry Value Chain: Implications for Vertical Specialization and Integrated Semiconductor Manufacturers." In Industry and Innovation 9(3, Decem- ber):155­81. Special issue on global production networks. MIDA (Malaysian Industrial Development Authority). 1999. Malaysia's Manufacturing Sector: Into an Era of High Technology. Kuala Lumpur, Malaysia. 156 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Ministry of Information and Communication. 2002. IT Korea 2002. Seoul, Korea. Ministry of International Trade and Industry. 1986. Industrial Master Plan 1986­95. Kuala Lumpur, Malaysia. ------. 1996. Second Industrial Master Plan. Kuala Lumpur, Malaysia. Multimedia Development Corporation. 2002. "Unlocking the Full Potential of the Infor- mation Age." Available at www.mdc.com.my/package/. Processed. Navaretti, Giorgio B., Jan I. Haaland, and Anthony Venables. 2002. Multinational Corpora- tions and Global Production Networks: The Implications for Trade Policy. Brussels: European Commission, Directorate General for Trade, March. Nelson, Richard R. 1990. "Acquiring Technological Capabilities." In Hadi Soesastro and Mari Pangestu, eds., Technological Change in the Asia-Pacific Economy. Sydney: Allen and Unwin. Nelson, Richard R., and Sidney G. Winter. 1982. An Evolutionary Theory of Economic Change. Cambridge, Mass.: Bellknap Press. OECD (Organisation for Economic Co-operation and Development). 1999. Boosting Inno- vation: The Cluster Approach. Paris. Ozawa, Terutomo. 2000. "The `Flying-Geese' Paradigm: Toward a Co-evolutionary Theory of MNC-Assisted Growth." In Khosrow Fatemi, ed., The New World Order: Internation- alism, Regionalism, and the Multinational Corporations. New York: Pergamon. Pavitt, Keith. 2003. "What Are Advances in Knowledge Doing to the Large Industrial Firm in the `New Economy'?" In Jens F. Christensen and Peter Maskell, eds., The Industrial Dynamics of the New Digital Economy. Cheltenham: Edward Elgar. Penrose, Edith T. 1995. The Theory of the Growth of the Firm. Oxford: Oxford University Press. Porter, Michael. 1990. The Competitive Advantage of Nations. London: Macmillan. Prystay, Chris. 2001. "Malaysia Seeks to Lift Super Corridor." Asian Wall Street Journal, August 24. Rasiah, Rajah. 1995. Foreign Capital and Industrialization in Malaysia. London: St. Martin's Press. ------. 2003. "Export Experience and Technological Capabilities: Evidence from East and Southeast Asian Firms." Unpublished ms. United Nations University, Institute for New Technologies, Maastricht, The Netherlands. Processed. Richardson, G. B. 1990. Information and Investment: A Study on the Working of the Competi- tive Economy. Oxford: Oxford University Press. Rodrik, Dani. 1999. The New Global Economy and Developing Countries: Making Openness Work. ODC Policy Essay 24. Baltimore, Md.: Johns Hopkins University Press for the Overseas Development Council. Romer, Paul M. 1990. "Endogenous Technological Change." Journal of Political Economy 98(5, pt. 2):S71­S102. Rugman, Alan M. 1997. "Canada." In John H. Dunning, ed., Governments, Globalization, and International Business, ch. 6. London: Oxford University Press. Rugman, Alan M., and Joseph R. D'Cruz. 2000. Multinationals as Flagship Firms: Regional Business Networks. New York: Oxford University Press. Saxenian, Anna-Lee. 2002. "The Silicon Valley Connection: Transnational Networks and Regional Development in Taiwan, China, and India." In Industry and Innovation 9(2, August). Special issue on global production networks, information technology, and local capabilities. GLOBAL PRODUCTION NETWORKS IN EAST ASIA'S ELECTRONIC INDUSTRY 157 Simon, Denis. 2001. "The Microelectronics Industry Crosses a Critical Threshold." China Business Review (November­December):42. State Government of Penang. 2001. The Second Penang Strategic Development Plan, 2001­2010. Penang, Malaysia. Sturgeon, Timothy. 2002. "Modular Production Networks: A New American Model of Industrial Organization." Industrial and Corporate Change 11(3):451­96. Takeuchi, Junko. 1993. "Foreign Direct Investment in ASEAN by Small- and Medium-Sized Japanese Companies and Its Effects on Local Supporting Industries." RIM Pacific Busi- ness and Industries 4(22). ------. 1997. "The New Industrialization Strategy of Malaysia as Envisioned in the Sec- ond Industrial Master Plan." RIM Pacific Business and Industries 3(37). ------. 2001. "Comparison of Asian Business by Japanese and American Companies in the Electronics Sector." RIM Pacific Business and Industries 1(3). Toh, Kin Woon. 2002. "The Political Economy of Industrialization in Penang." Unpub- lished ms. Processed. Too, T., and T. P. Leng. 2002. "Unemployment Situation in Penang." Economic Briefing to the Penang State Government 4(4). United Microelectronics. 2001. "Form 20-F Report." Submitted to the U.S. Securities and Exchange Commission, December 31. Processed. Wong, Poh Kam. 2000. "Riding the Waves: Technological Change, Competing U.S.- Japanese Production Networks, and the Growth of Singapore's Electronics Industry." In Michael Borrus, Dieter Ernst, and Stephan Haggard, eds., International Production Networks in Asia: Rivalry or Riches? London: Routledge. Wu, Nancy. 2002. "Design Trends and the EDA Market in Taiwan: A White Paper." Re- port prepared for the Electronic Engineering Times, Design and Engineering Group, Gartner Dataquest, February. Processed. Zainal Abidin, Mahani. 2002. "Issues on Malaysia's Competitiveness." Slide presentation at the seminar on Asian competitiveness, Penang State Government, September 30. Processed. CHAPTER 4 PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY Richard F. Doner, Gregory W. Noble, and John Ravenhill E ast Asia, with the exception of Japan, is a far less significant player in the auto parts industry than in other manufacturing sectors, most notably electronics. Nonetheless, the rapid evolution of Asia's auto industry has created opportunities for parts producers to expand production both for local assembly and for export. Recent investments in the region by foreign assemblers aim to produce components and vehicles for export throughout regional and global networks as well as for domestic sale. Indeed, assemblers face an imperative to export because of significant local overcapacity created by excessive investment driven by protectionist barriers and, more recently, by excessively optimistic expectations of rapid growth in the region's auto markets. Intense cost pressures and opportuni- ties for profits in nonmanufacturing segments of the auto industry are caus- ing assemblers worldwide to outsource an increasing proportion of parts and components. They are seeking to increase local parts production to avoid foreign exchange fluctuations, meet short delivery times, adapt to lo- cal demand, and make use of cheaper local inputs. In sum, domestic over- capacity, combined with the efforts of assemblers to balance the benefits of localization with the need for centralization and economies of scale, has created significant opportunities for the export of locally produced auto- motive components. Translating these opportunities into efficient businesses is not always smooth, however, because barriers to entry for locally owned parts produc- ers are rising significantly. Although trade protection in East Asia economies in attenuated and uneven form, reliance on profitable, protected domestic markets is less of an option for domestic producers than in the past. Gov- ernments are reducing tariff levels and abandoning trade-related investment 160 GLOBAL PRODUCTION NETWORKING IN EAST ASIA measures (TRIMs) under pressure from requirements of the World Trade Organization (WTO) and, for Southeast Asian states, from obligations of the Association of South East Asian Nations (ASEAN) Free Trade Area. Faced with new competitive challenges, assemblers are demanding that parts suppliers match the levels of quality control, productivity, and cost- effectiveness prevailing in industrial economies. At the same time, they are demanding that first-tier suppliers take responsibility for the design as well as the manufacture of modules and not just individual components. Locally owned firms that lack financial, technological, and managerial resources will not survive these demands; indeed, many have already begun to disappear or to merge with newly arrived foreign component producers. The role of foreign capital in the region's production of auto parts has grown and will probably continue to do so. Although the new environment offers signifi- cant opportunities for local production, policymakers will have to meet the challenge of dealing with an industry that is increasingly denationalized. The central message of this chapter is that the location-specific assets that the economies examined here--China, the Republic of Korea (here- after Korea), Malaysia, Taiwan (China), and Thailand--can offer to foreign automakers and to first-tier global suppliers vary significantly, and thus so do their strategic options. Small domestic markets, weak indigenous tech- nical capacities, and modest automotive-related human resources mean that Malaysia, Taiwan (China), and Thailand can only optimize within the regional and global frameworks created by the strategies of global assem- blers. However, even within such "constrained optimization," these three countries have potentially attractive options in product niches in both orig- inal equipment manufacture (OEM) and replacement equipment manu- facture (REM), particularly in the development of regional markets. Larger domestic markets in China and Korea--and, in the Korean case, established local assemblers and significant experience with auto parts exports--provide these two countries with greater opportunities to mold the strategies of global assemblers to national priorities. This might translate into more as- sembler acceptance of de facto protection and more willingness to transfer technology and, therefore, not only more extensive local production of parts but also greater opportunities for indigenous producers, possibly in joint ventures with the dominant multinational first-tier suppliers. But potential resources do not translate automatically into leverage. The capacity to convert strategic options into reality depends on policies, insti- tutions, and politics. Effective policies in areas such as technical training, infrastructure, marketing, trade policy, finance, and ownership can greatly magnify the attractiveness of location-specific assets and thus a country's leverage vis-à-vis multinational assemblers and parts producers. Public sec- PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 161 tor expertise and cohesion, public-private sector linkages, and coordination within the private sector are all important influences on a country's capac- ity to formulate and implement effective policy. And politics, especially the distributional consequences of automotive policies, influences both the strength of relevant institutions and the effectiveness of implementation. This chapter develops these themes as follows. First, it provides a brief overview of the region's auto industry, highlights its weakness through a brief comparison with the region's booming electronics industry, and iden- tifies the auto industry "drivers" that account for the differences. Then it identifies the new drivers that are creating opportunities and challenges for regional auto parts production and presents concise overviews of the indus- try in China, Korea, Malaysia, Taiwan (China), and Thailand. A final sec- tion identifies the location-specific assets of these five economies, their strategic options, and the policies, institutions, and politics influencing their ability to pursue these options. Finally, we offer a caution: East Asian automotive industries are in an ac- tive process of transformation. Whether it is Japanese and U.S. firms shift- ing all Asian-based production of pickup trucks to Thailand, substantial new investments by foreign companies in China, Malaysia's transfer of management control of Perodua to Daihatsu, General Motors' purchase of Daewoo, or the resurgence of Samsung Motor under Renault's control, the geographic and corporate contexts of East Asian parts production are be- coming both more extensive and more interdependent than anyone would have predicted a decade ago. The objective of this chapter is not to make predictions but rather to identify potential areas of growth and the factors determining whether such potential will be realized. AUTOMOTIVE PARTS PRODUCTION IN EAST ASIA In contrast to their stellar performance in the electronics sectors, East Asian economies remain relatively small players in the global auto parts industry. In this section of the chapter, we first review the record of East Asian economies in auto parts production and export and then contrast it with the integration of East Asian economies into global production networks in electronics. (Uneven) Weaknesses With the exception of Japan, East Asia remains a relatively small player in the global production of automotive parts. Exports in 1999 from the five 162 GLOBAL PRODUCTION NETWORKING IN EAST ASIA economies in this study accounted for only 4.3 percent of total world ex- ports of auto parts (calculated from the U.N. Commodity Trade Statistics Database, known as U.N. Comtrade).1 This modest global position reflects the relatively small role of parts production within each national economy and, with some exceptions, the lack of international competitiveness of this industrial sector in East Asia. In none of the economies do auto parts ac- count for more than 3 percent of total export earnings, a figure lower than the share of this commodity in overall global exports (figure 4.1), even though the value of parts exports has risen over the past decade (figure 4.2). Only Korea exports significant volumes of automotive products to Canada and the United States, the most open and competitive of the large markets, but Korea's market share in the two countries (1 and 3 percent, respec- tively) is dwarfed by that of Mexico (6 and 17 percent). Korean exports to Japan are inconsequential, and while Taiwan (China) and Thailand both export to Japan, each accounting for 2 percent of Japanese automotive im- ports, their individual shares of the Japanese market are less than that of Mexico (3 percent; WTO 2002, p. 63). Further evidence of the region's weaknesses in auto parts production is found in generally negative automotive trade balances (table 4.1), the con- tinued high levels of protection that governments perceive are required by the industry (table 4.2), and the low-value-added character of exported auto parts. Exports are dominated by natural resource­based products, such as Figure 4.1 Share of Auto Parts in Total Exports, by Economy, 1999 6 5 4 3 Percent 2 1 0 of China Rep. World (China) Indonesia Malaysia Thailand Philippines Korea, Taiwan 1. Available from the United Nations Statistics Division at http://unstats.un.org/unsd/comtrade/. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 163 Figure 4.2 Exports of Auto Parts, by Economy, 1988­99 (millions of U.S. dollars) Millions 4,000 3,500 Korea, Rep. of 3,000 Taiwan (China) 2,500 China Thailand 2,000 Indonesia 1,500 Malaysia 1,000 500 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 tires, and labor-intensive goods often produced by subsidiaries of Japanese or Western component firms, such as batteries and wire harnesses (table 4.3). There are, to be sure, some notable bright spots in this picture. Korea and Taiwan (China) both run trade surpluses in auto parts. Korea benefits both fromasizabledomesticmarketaswellasfromasignificantreplacementmar- ket based on Korean-made vehicles sold abroad and has developed substan- tial expertise in some components, such as climate control systems. Taiwan (China) has begun to exploit the replacement market, with its products now accounting for 80 percent of global production of aftermarket sheet-metal bodyparts(Liang2002).Overall,however,theindustrycontinuestobechar- acterized by low levels of technology and low-skill production. Table 4.1 Balance of Trade in Auto Parts, by Economy, 1999 (thousands of U.S. dollars) Economy Exports Imports Trade balance China 2,757,721 2,885,638 -127,917 Indonesia 710,734 721,305 -10,571 Korea, Rep. of 4,008,995 2,160,772 1,848,223 Malaysia 485,116 635,464 -150,348 Philippinesa 818,023 417,344 400,679 Singapore 1,037,477 1,468,775 -431,298 Taiwan, China 2,874,750 1,716,435 1,158,315 Thailand 1,220,796 1,420,653 -199,857 a. Data for 1998. Source: UN trade database Comtrade. 164 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Table 4.2 Tariff Protection of Auto Assembly and Parts Industries, by Economy, 2002 (percent) Economy Assembled cars Auto parts China 70­80 9­50 Indonesia 75­200 25 Korea, Rep. of 8 8 Malaysia 50­140 5­30 Philippines 30 3­10 Taiwan, China 60 7.5­25 Thailand 30­80 20­33 Source: APEC tariff base (http://www.apectariff.org); for Thailand, see "Car Tax" (2002). Contrast with Electronics The weakness in auto parts exports stands in marked contrast to the re- gion's experience in electronics. From 1995 to 1999, East Asia accounted for more than 37 percent of global electronics exports, compared with less than 3 percent of automotive vehicles and parts. The region's electronics industry, especially industrial electronics and components, has taken the form of networks composed of cost-efficient, specialized, interdependent production nodes linked to, and in many cases created or nurtured by, U.S. firms interested in exporting products back to developed-country markets (Borrus, Ernst, and Haggard 2000; McKendrick, Doner, and Haggard 2000; Scott 1987). Until the mid- to late 1990s, in contrast, East Asian auto production took the form of "hub and spoke" arrangements composed of inefficient, largely self-contained, and thus duplicative facilities produc- ing for protected, national markets and communicating only with the head- quarters of dominant Japanese assemblers, not with other production sites in the region (Hatch and Yamamura 1996). In electronics, local production facilities were sources of globally competitive components for use in final products exported to industrial markets. In the auto industry, local assem- bly facilities relied on a combination of major functional components im- ported principally from Japan and locally produced or assembled compo- nents whose quality and price typically failed to meet global standards.2 These differences reflected different drivers in the auto industry. High transport costs, extensive protection, divergence in the preferences of local assemblers, limited interest of Western firms in small Asian markets, and relatively slow technological change combined to produce a regionally fragmented structure of production. 2. See the discussion of parts quality in chapter 5 of this volume. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 165 Table 4.3 Principal Auto Parts Exports, by Economy, 1999 (percent) Economy and product Share of auto parts exports China Tires 27 Nickel-cadmium accumulators 8 Brake parts 8 Wiring sets 7 Filament lamps 6 Wheels 5 Korea, Rep. of Tires 37 Lead acid accumulators 5 Body parts 4 Engines 2 Malaysia Nickel-cadmium accumulators 24 Tires 16 Lamp parts 6 Wheels 5 Lead acid accumulators 3 Engine parts 3 Oil filters 3 Taiwan, China Tires 14 Engine parts 10 Lamp parts 9 Wheels 9 Bumpers 3 Accumulators 3 Thailand Tires 24 Wire harnesses 22 Diesel engines 5 Engine parts 5 Wheels 5 Radiators 4 Source: Calculated from UN trade database Comtrade. This structure provided local auto parts producers in most countries with neither extensive economies of scale nor incentives for technological improvement. Only Korea, whose protection of the auto industry was coupled with an emphasis on exports and whose parts exports followed its vehicle exports, became a partial exception to this pattern of domestically oriented and inefficient parts sectors. In electronics, low transport costs, 166 GLOBAL PRODUCTION NETWORKING IN EAST ASIA low rates of protection, the irrelevance of local consumer preferences, and Western firms' active interest in exporting to developed-country markets all combined with rapid technological change to encourage the growth of efficient, locally based component production. CHANGING AUTO INDUSTRY DRIVERS AND MULTINATIONAL STRATEGIC RESPONSES IN EAST ASIA New auto industry drivers--the entrance of Western assemblers and in- tensified competition for the region's growing markets, rapid technologi- cal changes, and shorter product cycles--have prompted new assembler strategies and the evolution of more complex production networks. The resulting picture involves good and bad news for East Asian parts produc- tion: increased opportunities exist for local parts manufacture, but they are associated with steeper entry barriers for local producers. Industry Drivers Global assemblers must now contend with an increasingly challenging set of competitive pressures. The first involves sources of new market growth. The market for new cars in the Triad ( Japan, the United States, and Western Europe) has been relatively flat for the past 10 years, with average demand growing less than 1 percent a year. Countries outside the industrial world have become the source of significant growth: East Asian automobile mar- kets averaged around 15 percent annual growth rates before 1997, and sev- eral have recovered to pre-crisis levels. The Asia-Pacific region (excluding Australia and Japan) is projected to account for some 45 percent of incre- mental global industry volume from 1998 to 2006 "and will remain a cata- lyst for vehicle manufacture and supplier growth for years to come" (Bur- well and Ferris 2000). The largest markets, China and Korea, are predicted to double in the next decade and together to reach Japanese volumes (table 4.4). And these projections may underestimate the potential for growth in China. Although ASEAN markets are smaller, their projected annual growth rates are impressive--between 10 and 20 percent (Veloso 2000, p. 24). Global automakers, much like the producers of silicon wafers and elec- tronic components and products, also have to contend with market frag- mentation and overcapacity (current excess capacity is roughly 24 million units, the equivalent of 96 assembly plants). Capacity utilization rates worldwide have fallen since the early 1990s from 80 to 69 percent (PWC 2000, p. 2). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 167 Table 4.4 Current and Prospective Vehicle Sales of Asian Economies, 1999, 2005, and 2010 Economy 1999 2005 2010 Japan 5,861 6,816 7,000 Korea, Rep. of 1,275 1,948 2,650 China 1,925 3,238 4,673 Thailand 218 687 1,253 Taiwan, China 423 560 638 Malaysia 289 504 747 Indonesia 94 439 696 Philippines 74 244 533 India 830 1,432 2,209 Pakistan 67 99 114 Australia 787 898 1,051 Others 102 148 186 Total 11,944 17,014 21,750 Source: Standard & Poor's DRI, cited in Veloso (2000, p. 24). Within the region, relative scale economies are achieved mainly in Korea, where in the past the state limited the number of assemblers and the finan- cial crisis brought consolidation of the parts industry (although significant overcapacity remains, necessitating a heavy dependence on export markets). China has more than 120 vehicle assemblers, many producing fewer than 15,000 units annually (Veloso 2000, p. 29). In ASEAN, protective measures favoring assembly over parts manufacture have resulted in a proliferation of assembly plants, most of which produce well under the 200,000 units tradi- tionally assumed necessary to achieve efficient scale economies in a single plant (Abrenica 2000, p. 4; Legewie 2000, p. 220). Some scale economies are obtained in Thailand, where tax policies and local demand have generated a large market for one-ton pickups, and in Malaysia, where heavy protection has allowed the "national car"--Proton--to hold on to about half of the lo- cal market with sales of 155,000 in 2003. But overinvestment has produced excess capacity in all countries: utilization rates in East Asia were estimated to lie between 65 and 75 percent as of 2000 (Veloso 2000). Assembler Strategies Whereas auto assemblers tended to concentrate on dominating their home region through the 1980s, in the 1990s they made significant investments throughout the world, especially in emerging markets, to tap into the most rapidly growing sources of demand (Fine, LaFrance, and Hillebrand 1996, p. 24; Veloso 2000, p. 7). This trend is clear in East Asia, where the region's 168 GLOBAL PRODUCTION NETWORKING IN EAST ASIA potential has intensified the competition among a larger number of auto manufacturers (Noble 2001). Japanese assemblers, long dominant in the ASEAN countries, expanded investments there in the 1990s, especially in Thailand. Most significantly, Western firms--assemblers, first-, second-, and third-tier suppliers, and service providers--have recently entered or reentered East Asian markets. Western firms have invested both on their own and through tie-ups with Asian OEMs. These alliances are designed to help Western firms to broaden their brand and product portfolios and to draw on Asian firms' expertise in the manufacture of attractive and af- fordable small vehicles. The principal focus of Western initiatives has been China and Thailand, although Korea has also attracted significant invest- ments in the post-crisis environment. Consolidation is an important part of assembler strategies as these firms search for scale economies in design, development, and manufacture. Auto manufacturers are merging and are developing strategic alliances to gain access to volume-generating markets, to expanded skill sets and compe- tencies, and to innovation. PricewaterhouseCoopers anticipates that only six assemblers will account for some 80 percent of total vehicle output in the next decade, and one Japanese source predicts that only those assem- blers producing 4 million units a year will survive (Matsushima 1999 and PWC 2000). A similar consolidation is occurring in East Asian operations.3 An emphasis on exports and support for trade liberalization are two ad- ditional components of emerging assembler strategies in East Asia. Exports were already a priority of the newly entered U.S. OEMs, but the 1997 cri- sis led them to increase exports to compensate for the drop in local vehi- cle sales (de Jonckheere 1998, p. 6). Recently, Japanese auto assemblers have begun to follow the trend of electronics producers in using offshore facilities for production for export back to the Japanese market. These ex- port efforts have been undertaken through regional and global sourcing networks that allow OEMs to contend with, and even take advantage of, currency shifts (de Jonckheere 1998; Takayasu and Mori, chapter 5 of this volume). Complementing these export efforts is intensified support by assemblers for regional trade liberalization, which is perceived as essential for the cre- ation of efficient production networks. While East Asia's traditionally high 3. Of all the Japanese producers, only Honda and Toyota remain independent, although they have signed agreements with General Motors to share parts and technology. Two Korean auto companies--Daewoo and Samsung Auto--were acquired by General Motors and Renault, respectively. Hyundai has absorbed Kia. A process of consolidation and the disappearance of independent Asian companies is thus accompanying the entry of new competitors. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 169 levels of automotive protection probably will not disappear completely,4 all countries are reducing the level of tariff support given to the auto industry. ASEAN countries are now engaged in two sets of liberalization efforts. One is the ASEAN Industrial Cooperation (AICO) program--a bilateral com- plementation scheme launched in 1996 and aimed primarily at the auto in- dustry. Under this scheme, firms pay only 0­5 percent tariffs if 40 percent of the product's value originates in another, participating ASEAN country. The second is the ASEAN Free Trade Area (AFTA), have cut automotive tariffs to 0­5 percent by 2003 (but Malaysia has successfully sought an ex- emption until 2005). Global automakers operating in East Asia have ex- pressed strong support for both of these initiatives.5 Indeed, foreign as- semblers and suppliers, led by the Japanese, have established some 75 bilateral exchange programs under AICO. Honda, Mitsubishi, Nissan, Toyota, and the giant parts supplier Denso have been leaders in develop- ing multicountry complementation efforts in Southeast Asia (Legewie 2000, p. 230). Even as they attempt to increase scale economies, automakers must sat- isfy changing region- and country-specific tastes and conditions. "Plat- form" strategies are a key response to this challenge, and their adoption has implications for parts suppliers. These strategies, in which automakers use a small number of underbody platforms as the basis for a greater number of vehicle models, are detailed elsewhere in this volume (chapters 2 and 5). Here we emphasize three aspects of most platform strategies: consolida- tion, specialization, and localization. To cut the costs of platform develop- ment and to encourage component sharing among models, automakers, es- pecially U.S. firms, have begun to reduce their number of platforms even as they strive to maintain flexibility by broadening the definition of a plat- form (for example, PWC 2000).6 Corporate consolidation is facilitating the reduction of platforms. For example, platform sharing between Chrysler and Mitsubishi will allow Mitsubishi to slice its number of light-vehicle platforms from 12 to six or seven (Treece and Sherefkin 2001, p. 53). This 4. As a recent analysis of the Chinese auto market has noted, "Few people think Beijing will allow a flood of imports to swamp its domestic carmakers, any more than have South Korea, Malaysia, Indonesia--or for that matter, Japan" (Brooke 2002, sec. W, p. 1). Governments are likely to turn increasingly to less opaque nontariff barriers as a means of providing support to local industries. 5. There are thus some differences among the assemblers with regard to trade liberalization. U.S. firms tend to place more emphasis on the success of AFTA, whereas the Japanese have moved quickly to make use of AICO (for example, Yap 2001). 6. R&D expenditures in the OECD auto industries rose from 2.4 to 3.2 percent of production costs between 1972 and 1992 (Humphrey 1998, p. 3). 170 GLOBAL PRODUCTION NETWORKING IN EAST ASIA process of platform development and consolidation is clearly in evidence in East Asia, where, for example, Honda will produce the Odyssey based on the same platform as the Accord.7 Similarly, Renault (including Renault Samsung) and Nissan will share platforms in the future. National specialization is an important component of platform strate- gies. The Japanese have begun to promote specialization through the com- plementation efforts noted above. Equally notable is the emergence of Thailand as a hub for pickup truck production by almost all assemblers. In fact, Isuzu, Nissan, and Toyota have stated their intention to make Thailand the production site for all one-ton pickups produced outside of North America (Chatrudee 2002). Finally, wherever possible, assemblers are attempting to localize the manufacture of components. Localization can help assemblers to facilitate just-in-time supply of local assembly facilities, to reduce foreign exchange risks, to meet local tastes and requirements, and, when local conditions permit, to reduce costs through local procurement (chapter 5 of this vol- ume). Moreover, this localization is occurring through assembler "deverti- calization." As they focus on capturing more of the downstream value chain (between assembler and final customers), assemblers are outsourcing the manufacture and, in some cases, the design of platform components. Implications for Parts Suppliers: The Promises and Pitfalls of Globalization The increased transaction costs generated by such highly demanding out- sourcing have had important consequences for assembler-supplier rela- tions. U.S. assemblers have moved away from relations with a large num- ber of suppliers toward closer involvement with a smaller number of larger, richer, more technologically competent producers. Whereas Japanese OEMs historically maintained very close ties with a small number of sup- pliers, faced by new competitive pressures, they have begun to impose 7. In China General Motors is building a small car--the Buick Sail--off an Opel Corsa platform that was first modified for production in Brazil and then modified again for China. Volkswagen is producing the Audi A6 and the Passat in two different locations in China, appealing to two dif- ferent price segments and customers. The two vehicles are built on the same platform. In Thailand Ford and Mazda produce two different brands of pickup--virtually identical except for cosmetics--off the same platform. Also in Thailand, General Motors and Isuzu are planning to produce two pickup trucks based on the same S-10 platform. The platform was jointly developed (Dunne 2001). Ford is assembling a new common sedan in four ASEAN plants and develop- ing a regional component supply network based on the combined volume of the four facilities (de Jonckheere 1998, p. 4). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 171 tougher requirements on existing suppliers and to source from new, some- times Western, producers of parts. In both instances, heightened respon- sibility for design development by suppliers has meant that consolidation in the parts sector is proceeding in parallel with consolidation of the assemblers. A small number of independent global "mega suppliers" is emerging.8 These new strategies have raised the barriers facing developing-country firms attempting to supply parts and components to OEMs. In the past, as- semblers designed and manufactured core components while outsourcing easy-to-produce, detail-controlled parts. In some cases assembler sub- sidiaries even had some autonomy to introduce changes in design. This pro- vided local producers with opportunities to enter, to acquire competencies, and gradually to ascend automotive supply chains (Humphrey 1998, p. 10). Now, however, assemblers are relying on suppliers to produce and even de- sign standardized, core components based on common platforms. Given the technological weaknesses of most developing-country suppliers, which pre- clude them from meeting the assemblers' requirements, assemblers have be- gun turning to what Humphrey terms the "follow source"--the large firm already producing the part elsewhere in the world. In developing-country operations, first-tier firms such as Bosch, Delphi, Denso, TRW, Visteon, or Yazaki are being asked to supply the local subsidiary of the assembler and will "be responsible for ensuring that the rest of the supply chain meets the assembler's standards" (Humphrey 1998, p. 5).9 Intensified competition among assemblers for East Asia's growing markets, combined with technological and managerial innovations and with increased currency instability, is encouraging greater loca- lization of parts production within expanding regional and global production networks. The extent and nature of local parts produc- tion will vary according to two OEM firm-specific strategies and such industry- and product-specific factors as parts' value-to-weight ratios, position in assembly sequence, and degree of standardization. 8. In some systems, such as brakes and seats, the global industry is dominated by three or four global producers. PricewaterhouseCoopers predicts, "The 1,500 Tier 1 suppliers in 1998 will whittle down to about 150 large system integrators and 450 direct suppliers who will compete largely on price" (PWC 1999). By 2000 the five largest suppliers of automotive safety parts accounted for roughly 80 percent of global market share, up from around 55 percent in 1995 (McMaster 2001, p. 11). 9. First-tier suppliers sell directly to OEMs. They are typically capable of producing components, modules, or entire systems (for example, Lear produces complete interior systems; Delco pro- duces electronics systems; Dana produces drive trains; Bosch produces ABS systems). For lists of U.S. suppliers, see Brunnermeier and Martin (1999, ch. 2). 172 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Other things being equal, standardized and easily shipped parts such as printed circuits and switches will be sourced globally; parts such as alter- nators, heat exchangers, wiring harnesses, or fuel tanks that are large or more difficult to transport will be sourced regionally; and in-line parts supplied to the vehicle manufacturer in the sequence that vehicles are assembled will be supplied locally, often by firms in close proximity to the assembly site (Brooker Group 1997, p. II-8).10 How host countries fit into and influence these calculations depends on location-specific assets, such as market size and demand characteristics, OEM network presence, preexisting supplier base, human resources, and government policies. The possibility for developing countries is that the localization compo- nent of assembler strategies will translate into denationalization and into intensified competition for becoming a regional hub in the new production networks. For instance, as of late 1997 Brazil--a large domestic market with a sizable base of local suppliers--had only one locally owned firm among the 13 largest component producers (Humphrey 1998, p. 10). In Thailand, almost all of the 250 first-tier suppliers are foreign subsidiaries or affiliates (Brooker Group 2002, p. 208). Even in Korea, the post-crisis opening to FDI has facilitated significant acquisitions of larger Korean parts companies by Western first-tier suppliers. With assemblers at- tempting to squeeze first-tier suppliers, and they, in turn, pressuring their component manufacturers to reduce costs, profit margins are almost cer- tain to be smaller at the lower rungs of the value chain. And with markets for first-tier suppliers being dominated by foreign firms, the possibilities for technological learning and upgrading for local companies may be re- duced unless appropriate public policies and incentive structures are put in place. CASE STUDIES This section of the chapter provides a more detailed examination of the policy framework for auto parts in our five economies and reviews their recent performance in domestic production and in exports. 10. Another hybrid scenario, this one proposed by a major producer of safety components, antic- ipates local and national sourcing of unique parts: local assembly of some parts in the firm's com- ponent assembly plant (10 percent of total parts), regional sourcing of region-specific parts (30 percent), and global sourcing of standard parts (60 percent). See McMaster (2001). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 173 China The lure of China's potentially huge market, large foreign investments induced by the government's protective policies, and the accumulation of experience and infrastructure in export-oriented manufacturing have cre- ated a formidable base for the growth of the Chinese auto industry. None- theless, the time required and the political and social turmoil involved in reaching global competitiveness should not be underestimated: China's market for passenger cars is still modest, and even amidst overall growth, pervasive excess capacity will ensure that many small parts producers fail and that large numbers of workers lose their jobs. Demand for and production of autos. Volume production, mostly of trucks, began in the 1950s at the First Auto Works (FAW) in Changchun; FAW and a host of vertically integrated suppliers located in the same com- plex provided most components. In the 1960s fears of conflict with the Soviet Union and Maoist doctrines of local self-reliance spurred the creation ofahostofnewassemblersandpartssuppliersininlandprovinces.Economic opening in the late 1970s led to joint ventures with foreign auto companies, notably Volkswagen. In the 1990s General Motors and Toyota garnered as- sembly licenses by pledging to establish local design capacities and culti- vate a network of local suppliers. As the government pressed to achieve economies of scale, the production of passenger cars rose steeply over the 1990s and reached 2.07 million units in 2003. High tariffs and other pro- tective measures largely discouraged imports. Nevertheless, excess ca- pacity and padded work forces remain pervasive, and the capabilities of firms vary drastically by location, size, and ownership. Parts production. Production of parts grew quickly in the 1990s, doubling in value between 1994 and 1999. Parts production is geographically dis- persed. The leading production network is centered in Shanghai (including Jiangsu and Zhejiang); other significant production areas are the northeast, Hubei, and Sichuan-Chongqing. Beijing-Tianjin and Guangzhou host major assembly sites and boast strong industrial bases, so they could emerge as centers for parts production in the future. Many producers are linked to state, provincial, or municipal conglomerates. Design capacities of local firms vary but are generally low; quality and delivery times, though greatly improved, are still weak, particularly at lower tiers of the supply pyramid. Even joint ventures are often plagued with excess labor. Of approximately 1,540 firms supplying auto and motorcycle parts, about one-fourth have attracted foreign capital (Ministry of Machinery 174 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Industry 2000, pp. 10, 314). Whereas foreign firms can own at most 50 per- cent of assembly operations, foreign participation in parts manufacture ranges from technology transfer through joint ventures to wholly owned subsidiaries. Firms with foreign participation are especially important in the production of more demanding parts, such as piston rings (two of 21 firms account for 65 percent of total production of this component), fuel injectors (one of six firms, but with 70 percent of production), and air con- ditioning (six of 10 firms, with 73 percent of production; Fourin 1999). They are also a major force in the production of standardized items such as tires and batteries. Trade. From 1994 to 1999, imports of parts doubled, while exports nearly tripled. Before long, China is likely to generate a trade surplus in parts. Major exports include tires, wheels, wire harnesses, glass, and other heavy or labor-intensive items, but complex parts such as transmissions and axles are also exported in small numbers. Japan is the major source of imported autos and auto parts, followed by Germany, the United States, and Taiwan (China); the United States is the leading export market, followed by Japan, Germany, Hong Kong (China), and a mixture of European and developing countries (Ministry of Machinery Industry 2000, pp. 179­88). With China's entry into the WTO, local content requirements were eliminated and tariffs began to drop, but the industry will be cushioned by the slow pace of liberalization and remaining promotional and protection- ist policies, including taxes, controls on FDI, and local protectionism. Lo- cal production of mainline vehicles and their parts will continue to grow as manufacturers adapt their vehicles to the Chinese market and take advan- tage of inexpensive labor to produce for export markets. Production networks. China contains a number of imperfectly integrated component production networks. Joint ventures and wholly owned opera- tions from Taiwan, China (officially 166 as of 2002) tend to be efficient and produce with acceptable quality, often for export markets, but they rarely extend beyond low-medium levels of sophistication (tires, batteries, radios, steering wheels, lights, piston rings, brakes, crank shafts). Until recently, Japanese parts producers entered China only reluctantly, to support local assembly, producing such parts as cylinder heads and blocks, cam shafts, universal joints, steering systems, alternators, starters, and brake master cylinders. Industry leader Denso, for example, has only five plants in China (three in Tianjin). With the exception of Denso and other suppliers linked to Toyota, Japanese firms have been criticized for supplying only dated technology, but they do pass along quality control and on-time production PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 175 techniques to their subcontractors. Western supply firms have concen- trated in the Shanghai area. Nine of Delphi's 15 Chinese plants, for exam- ple, are located in or around Shanghai, producing wire harnesses, car audio equipment, shock absorbers, drum brakes, generators, fuel pumps, and other low- to mid-technical products as well as the more sophisticated fuel injec- tion and engine management systems (EMS) at which the company excels. All plants export, with a group average of one-third. Other major global component firms with significant investments include Borg-Warner (trans- missions), Bosch (spark plugs, electronic fuel injection systems, oil pumps), and T&N (cylinder gaskets, piston rings). Summary. China is likely to continue experiencing rapid, but uneven and painful, growth. Older, state-owned firms located in inland areas and pro- ducing only for the domestic market face a dark future. Prospects are bright- est for Shanghai-area firms supplying parts for small cars and light com- mercial vehicles, with capital and technical assistance from foreign owners or partners. Korea Korea has been more successful than any other industrializing country in developing a national automobile industry. The Automotive Industry Pro- motion Law, promulgated in 1962, part of the First Five-Year Develop- ment Plan, marked the birth of the modern automobile industry. It was subsequently nurtured by the full panoply of trade and industrial policy in- struments that the Korean state had at its disposal. Imports of cars were prohibited--a ban that was not lifted for a quarter of a century. The gov- ernment provided subsidized loans and tax incentives for investments as well as export subsidies, including export promotion loans that enabled Korean cars to be sold in foreign markets at less than half the domestic market price. Imported components were exempted from tariffs, but the state encouraged assemblers to increase local content by working with local suppliers. Immediately before the financial crisis, car production in Korea (all from domestically owned companies) was the fourth largest in the world, and half of this output was exported. But repeated state attempts to consolidate the industry failed, and the consequent overcapacity left the industry vulnerable to the collapse in domestic demand that followed the onset of financial crisis, with dire consequences for the parts industry.11 11. For further consideration of the accomplishments and failures of the Korean automobile strat- egy, see Ravenhill (2002). 176 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Parts production. Of the economies included in this study, Korea has by far the largest parts industry in terms of overall turnover and exports. By 1999 the turnover of component companies was estimated to have grown to close to $10 billion (Economist Intelligence Unit 2001, p. 78). The do- mestic aftermarket accounted for about one-fifth of this total, a market then dominated by the assemblers.12 Subsequently, the component manu- facturing arms of Korea's largest assemblers were spun off into independ- ent companies: Hyundai-Mobis and Daewoo Precision (reconstituted in mid-2002 after General Motors' acquisition of Daewoo). Parts producers in Korea benefit from one of the region's largest do- mestic markets for assembled cars, with annual sales approaching 1.5 mil- lion vehicles, of which slightly more than 1 million are passenger cars. Even at the end of the 1990s, more than a decade after the government lifted its ban on imports, Korean companies enjoyed the luxury of a de facto sanc- tuary in the domestic market.13 Korea has one of the most liberal trade regimes for auto imports in Asia, with assembled vehicles and most auto parts subject to an 8 percent tariff (table 4.2). The most significant barriers to auto imports today are the strong nationalist sentiment against pur- chasing foreign cars (with the limited exception of the luxury end of the market), the price competitiveness of small cars produced by domestic as- semblers, and these companies' well-established distribution networks. Besides the large domestic market, the auto parts industry also benefits from Korea's strong national innovation system, engineering and metal- working skills, and backward linkages with the steel industry. At the onset of the financial crisis in 1997, small- and medium-size enterprises domi- nated the industry, accounting for 95 percent of a total of approximately 1,400 parts manufacturers. These firms typically had limited capital and rudimentary technology and produced solely for the domestic market. Indigenous technological development was confined primarily to the 50 largest parts suppliers. In many instances, local development rested on improvement on technologies previously produced under license. Research and development (R&D) expenditures remain low and heavily concen- trated in a few companies: only 1.4 percent of the industry's overall turn- over was invested in R&D in 1999. Korean exports of car parts were de- rived almost entirely from these larger firms and from the assemblers 12. Only at the beginning of 2000 were component makers permitted to sell their products directly to maintenance shops or consumers rather than indirectly through the assemblers. 13. In 2000 imports of all motor vehicles totaled 11,168, and motor vehicle registrations totaled 1,441,628 (data on the Korean Automobile Manufacturers Association's web page: http:// www.kama.or.kr). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 177 themselves. Daewoo, Hyundai, and Kia provided about one-third of total exports, while another 10 companies contributed a further quarter of export earnings. Trade. As Korean auto exports grew, so did exports of auto parts: the value of auto parts exports tripled between 1990 and 2000 (when they constituted 2.8 percent of total exports, up from 2 percent in 1988). Exported parts were divided almost equally into three categories: those used in assembly overseas, those used for the aftermarket for Korean cars, and those used for the aftermarket for cars produced by other (foreign) makers. By 1999 Korea enjoyed a large surplus in its trade balance in auto parts. More than one-third of all exports went to the North American market, the largest single export market for Korean cars. Asia accounted for 25 percent and Europe for 21 percent of the balance of parts exports. Most auto parts exports remain at the low to medium end of the spec- trum of technological complexity. Tires are by far the single most valuable export, constituting close to one-third of the total value. Other significant exports include air compressors, batteries, air conditioners and parts, bear- ings, wheels, clutch covers, window regulators, and steering parts. Korean assemblers still rely heavily on imports for the more technologically de- manding components, including transmissions and engines (with the excep- tion of Hyundai, which has developed its own engines for some models), safety systems, and electronic equipment. Production networks: post-crisis consolidation. The economic crisis in the second half of the 1990s exacerbated the pressures that global consoli- dation was exerting on the domestic auto parts industry. The bankruptcies of four of the five domestic assemblers--Kia and Ssangyong (these two be- fore the onset of the financial crisis), Samsung, and Daewoo--had flow-on effects for many of their suppliers, which were unable to withstand the nonpayment of invoices. Between 300 and 500 small- and medium-size enterprises exited the auto parts industry following the bankruptcies of the assemblers. The financial crisis dramatically changed the ownership struc- ture of many of the larger Korean parts companies. Bankruptcies among some of the largest conglomerates forced divestment of some auto com- ponent subsidiaries. Even more significant, it opened the way for foreign investors to acquire significant shareholdings in Korean components producers--often at fire-sale prices (the advantage of low stock market pricesbeingcompoundedbyafavorableexchangerate).Morethan100parts suppliers have transferred 50 percent or more of their equity to foreign in- vestors since the crisis began. This has facilitated the entry or the expansion 178 GLOBAL PRODUCTION NETWORKING IN EAST ASIA of Korean operations of the leading global first-tier suppliers, including Autoliv, Britax, Delphi, FAG, Valeo, and Visteon. The financial crisis hastened the consolidation of the Korean compo- nents industry and led to further differentiation between larger firms, on the one hand, increasingly with foreign equity as well as technology licens- ing agreements, and small- and medium-size enterprises, on the other. Re- structuring and the opening to foreign investment, in turn, have facilitated the negotiation of new technological partnerships. Prospects. The size of Korea's domestic auto market (even though domes- tic growth is expected to be modest), coupled with its continued dominance by local assemblers, provides Korean auto parts makers with a springboard for penetrating global markets, as does the increasing aftermarket for parts for Korean cars on foreign roads. The future of the Korean components industry inevitably rests in large part on the success of the assemblers. And here the prospects appear much brighter than many believed at the time of the crisis: Hyundai and Kia have dramatically increased their exports, es- pecially to North American markets, going a long way toward overcoming the negative brand image that previously afflicted Korean-built cars. Gen- eral Motors intends to use its new Daewoo subsidiary to supply small vehi- cles not only to the region but also to Europe and the United States. Renault has committed itself to investments that will substantially raise the output of Renault Samsung. Nevertheless, investment in foreign assembly plants by Hyundai and Kia (with plants planned for China, Europe, and the United States) may reduce both the total exports of assembled autos from Korea and the opportuni- ties for auto parts exports. Hyundai and Kia are encouraging a number of their larger suppliers to establish subsidiaries close to their overseas oper- ations: more than 20 are planning to establish plants adjacent to Hyundai's facility in Alabama. China has also attracted more than 40 Korean auto parts companies, seeking to utilize local lower-cost labor to maintain their presence in the production of low-end components. As domestic assemblers seek to increase their presence in global mar- kets, their insistence on modularization and on Korean suppliers reaching international standards will force the domestic auto parts industry to be- come more competitive if it is to survive. Korean-owned companies face a potential pincer movement. At the low-tech end, they are not price com- petitive with products from China and Taiwan (China). At the high-tech end, the emphasis on modularization provides advantages to the trans- national first-tier suppliers. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 179 A possible strategy for the larger Korean parts producers, which takes advantageofthesignificanceofthedomesticmarket,isoutlinedbyHyundai- Mobis president and chief executive, Park Jeong-in: "Our weakness is we lack the technology, and we don't have time to learn it. We aim to select top- ranked makers of automotive components and offer joint venture produc- tion bases in Korea in exchange for the chance to supply parts to Hyundai and Kia" (quoted in "Hyundai-Mobis" 2001). Hyundai-Mobis is the largest supplier to Hyundai and Kia and is focused entirely on the production of modules, referring to itself as a "tier-0.5" company. Further consolidation will be required if the Korean parts industry is to become competitive internationally. Too many Korean companies produce on too small a scale to be competitive--and their comparative inefficien- cies are handicapping Korea's domestic assemblers. The pressure for mod- ularization will inevitably hasten industry consolidation: Hyundai-Kia, which now has a joint procurement operation, plans to reduce its suppliers from the current figure of 750 to fewer than 200. Many more of the small- and medium-size enterprises are likely to disappear. Some current first-tier suppliers will drop to the second tier. Hyundai Motors' research unit esti- mates that a further 200,000 jobs will be lost through restructuring in the auto industry. With a large portion of Korean auto parts exports now coming from foreign-invested companies, a key question that remains unanswered is what role the new joint ventures will play within the overall production networks of the largest players in the global industry. Will they be used primarily for local production to supply the Korean market, or will they be integrated into sourcing for global markets? And what technology will be transferred to the Korean subsidiaries? The history of joint ventures in the Korean automotive industry has not been a particularly happy one for local firms, which have been denied access to leading-edge technologies and prevented by their joint venture partners from exporting. Whether this will change in Korea's more positive climate for foreign investment (with a significant shift to wholly or majority-owned subsidiaries) and for protection of intel- lectual property rights remains to be seen. Malaysia Production networks in auto parts in Malaysia have been shaped by gov- ernment policies aimed at promoting national champions in the assembly industry and at increasing the participation of ethnic Malays in commerce, objectives that have sometimes proved less than completely compatible. 180 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Government promotion and ongoing high levels of protection of auto as- semblers have enabled the two national car companies, Proton and Pero- dua, to capture more than 90 percent of the domestic car market. Although Malaysia has the largest domestic car market in ASEAN, with annual sales of close to 350,000 vehicles, and their domination of the domestic market gives Proton and Perodua top position in all car sales in ASEAN (together they shared 28 percent of the regional market in 2000 compared with Toyota's figure of 21 percent), overall production volumes remain too small for most local parts makers to be internationally competitive.14 Although some parts firms have entered into joint ventures with foreign partners and into technology licensing agreements, government restrictions have provided a rather inhospitable climate for foreign investors--in marked contrast to Malaysia's generous treatment of investments in the electron- ics sector.15 In the early 1980s government disappointment with the slow progress of local auto firms, coupled with its desire to promote heavy industry to re- duce dependence on commodity exports, led to a decision to develop a na- tionally owned auto industry. The first national car project, Proton, a joint venture with Mitsubishi, was launched in 1983. With heavy protection, sales of Proton vehicles soared: within three years of its first production Proton had secured 73 percent of the domestic market. Exports began in 1986. Volumes remain tiny, and sales have been heavily concentrated on the U.K. market. Proton cars have a negative brand image outside Southeast Asia, often being described as depending heavily on previous-generation Mitsubishi technologies. In 1993 the government launched a second national car project, Perodua, to produce small vehicles. Perodua was established as a joint venture between Daihatsu and its trading company partner--Mitsui of Japan--and three local companies. Besides the objectives of increasing local content and achieving techno- logical spillovers, the government also hoped to use the national car proj- ects to increase the participation of ethnic Malays in an industry that had previouslybeendominatedbyChineseentrepreneurs.Protonwasinstructed to purchase components from ethnic Malay­owned companies; it was not permitted to employ experienced labor laid off by other (Chinese- 14. Proton had 18 percent of the ASEAN market; Perodua had 10 percent. In Malaysia in 2000, Proton had 63 percent of the domestic car market, and Perodua had 30 percent. 15. The Malaysian government has permitted few foreign companies in the auto industry to es- tablish majority-owned local subsidiaries. Volvo, with 100 percent ownership of its assembly plant in Selangor, is an exception. Ford, Honda, and Toyota have all been confined to a 49 percent stake in joint ventures. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 181 dominated) assemblers and component suppliers after the establishment of the national car producer (Rasiah 1996).16 Trade policies. The national car companies receive the highest levels of protection in the region. Tariffs on motor vehicles were increased in 1998 in response to the economic crisis. The current rates on assembled pas- senger car imports range from 140 to 300 percent, depending on engine size. National car firms also are protected by various nontariff barriers, in- cluding reductions in excise taxes and import licensing. Malaysia has also maintained extensive local content requirements, including a list of com- ponents that must be purchased locally by all car manufacturers or assem- blers and a percentage local content requirement (abolished on January 1, 2002, as part of an agreement with the WTO). Malaysia successfully sought an extension until 2005 of the time period under AFTA for lowering its tariffs on auto imports from other ASEAN members (40 percent local content required) to the 0­5 percent range. Some reports have suggested that the government will continue to protect the local assemblers after this date through various discriminatory tax measures. Some relief from Malaysian duties for parts producers in other ASEAN countries is provided through the AICO scheme (of the 63 joint ventures in the auto industry currently approved under the scheme, 29 in- volve companies or their subsidiaries in Malaysia). The local car parts industry receives direct assistance through govern- ment measures aimed at promoting the participation of ethnic Malay en- trepreneurs in the economy. The Vendor Development Programme guarantees local producers a market with the national car companies. By 1998 the Vendor Development Programme for Proton had 188 suppliers responsible for 4,319 parts, including metal-stamped and pressed parts, plastic injection molded parts, wire harnesses, wipers, lamps, radio­ cassette players, and air conditioners. By guaranteeing a market, the Ven- dor Development Programme removes any pressure for suppliers to be- come internationally competitive. Exports. One of the strongest indicators of the lack of international com- petitiveness of the auto parts industry in Malaysia is its small volume of ex- ports. By 1999 the auto parts industry contributed only half of 1 percent of total Malaysian exports, constituting a smaller share in overall export earn- ings than in the other economies included in this study and, indeed, than 16. In 1984 ethnic Malay equity in the auto assembly industry was 30 percent compared to the 43 percent held by Chinese Malaysians. In 1988, 94 percent of Proton employees were ethnic Malays, the vast majority of whom had no previous experience in the auto industry ( Jomo 1994, pp. 266, 285). 182 GLOBAL PRODUCTION NETWORKING IN EAST ASIA in other ASEAN countries such as Indonesia and the Philippines that have much smaller assembly industries than does Malaysia.17 Tires traditionally constituted the single most valuable auto parts export for Malaysia. In 1999, however, their position was supplanted by nickel cadmium accumulators. Other major exports are similarly low tech. Although the export market for Malaysian auto parts is diversified, ASEAN is by far the single most im- portant outlet, accounting for more than a quarter of total exports, a sub- stantial portion of which can be assumed to be spare parts for Proton and Perodua cars sold in other ASEAN countries. Europe is the second most important market, again linked to Proton sales in the United Kingdom. The Malaysian auto parts industry. Despite or, probably more accu- rately, because of the very high levels of unconditional protection afforded by the government, the auto parts industry in Malaysia remains small scale, inefficient, and lacking in technological competencies. The gov- ernment estimates that 350 companies are involved in the manufacture of automotive components. Most parts manufacturers operate on a very small scale. Few have more than 200 employees. In total, they employ around 17,000 workers, only slightly more than 1 percent of all employ- ment in manufacturing in Malaysia. The larger firms typically are joint ventures with foreign partners.18 Despite the government's objective of increasing the value added by the domestic auto industry through localizing the production of parts, the trend in value added per unit of output in the parts and accessories industry as well as the manufacturing and assembly industry has been downward, largely because of the high import content of locally produced parts (Tyndall n.d.). The principal problems faced by the auto parts industry in Malaysia can be summarized as (a) a lack of scale economies, (b) relatively high labor costs, (c) shortages of skilled labor, (d) lack of research and development, and (e) ongoing dependence on imported components. Government poli- cies have either contributed to many of these problems or failed to address them effectively. The decision to use the national car project to promote ethnic balancing exacerbated the industry's problems of small scale and lack of skills. Meanwhile, the government has done little to alleviate the prob- lems within the national system of innovation. Locally trained engineers are in short supply and are likely to find a more attractive career path in the electronics industry than in auto parts. 17. The definition of auto parts used in this study does not include car radios or compact disk play- ers, which are a substantial export for Malaysia. 18. Tyndall provides a list of parts manufacturers and information about their products and num- ber of employees at http://www.asean-auto.org/mal/sheet0017_1.htm. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 183 The government has failed to provide institutions to assist the domestic auto industry to develop technology. No specific research institution exists for the automotive industry; existing institutions merely provide facilities for testing components. The government thus has not compensated for the lack of research and design capabilities of the local industry, a weakness deriving in part from the small size and limited budgets of local firms. The industry itself is poorly organized in comparison with its counterparts in other economies in the region. The principal business association, the Malaysian Automotive Component Parts Manufacturers Association, has only two full- time employees and plays a negligible role in facilitating technological upgrading. Malaysian auto components companies generally appear poorly equipped to cope with the increasing use of electronic commerce in the industry. Despite the government's requirements for high levels of local content and its claims that these requirements are being met, the reality is less encouraging. Leutert and Sudhoff (1999, p. 258) estimate that the actual ratio of local content may be as low as 35­40 percent. Moreover, Proton suppliers have not kept up with the trend in the industry toward providing complete systems: 60 percent of Proton suppliers merely provide single parts. The parts industry consequently does not reflect to any degree the trend elsewhere toward suppliers being organized into various tiers. And because only a minority of its suppliers are capable of operating on a just- in-time basis, Proton is compelled to carry higher levels of inventories than most of its international competitors. The after-sales market is dominated by components supplied by Mitsubishi. Prospects. Current regulations in Malaysia, though intended to promote the auto parts industry, often appear to have undermined its competitive- ness. Because it has been insulated from market forces, Malaysia's auto parts industry has not participated in the consolidation that has occurred else- where in the region following the financial crises. Protection of the indus- try through import licensing and the provision of guaranteed markets for parts makers with the national car companies have encouraged a climate in which rent seeking thrives. Although the Malaysian government can be ex- pected to continue to drag its feet on liberalization of trade in both assem- bled vehicles and auto parts, there being no significant domestic political constituency currently arguing in favor of such liberalization, it will even- tually have to comply with AFTA and the WTO commitments on TRIMs. To do so will place the parts industry in some jeopardy: it is caught in a pin- cer movement because it competes on the basis neither of low wages nor of advanced technology. 184 GLOBAL PRODUCTION NETWORKING IN EAST ASIA The future of the parts industry inevitably will be closely tied to that of the national car producers. Whether Proton can survive in a globalizing in- dustry without a strong foreign partner is questionable. Former Prime Minister Mahathir indicated his willingness to sell a 30 percent stake in Proton to a foreign manufacturer. Proton is reported to be actively seek- ing a foreign partner, but the investment climate in the automobile sector in Malaysia remains uncertain. There has also been discussion in the press of the possibility that Proton's plant will be made available to assemble for- eign cars. Some observers consider the decision at the end of 2001 to allow Daihatsu to take management control of Perodua's production facilities as an attempt by the government to set up a test case to see what will happen if a foreign producer is allowed to take over a national car company. If deemed successful, it may provide a model for the future of Proton ("New Era" 2001).19 The future of the national car makers themselves, however, will be determined in part by the competitiveness of their parts suppliers. Current government attempts to use the auto parts industry to pursue social and political objectives handicap the national car makers. If forced to choose between maintaining an assembly industry (probably) in alliance with foreign car manufacturers or attempting to prop up an inefficient locally owned auto parts industry, the government is likely to opt for the former, as strongly suggested by its decision to allow Daihatsu to assume control of Perodua.20 Taiwan (China) Taiwan (China) is a medium-size, medium-tech auto market with extensive ties to Japan and emerging links to China and (to a lesser extent) Southeast Asia. Local firms export a large volume of replacement parts, especially to the United States, and a small but growing amount of original equip- ment, particularly to Southeast Asia and Japan. WTO entry will shrink, but not eliminate, the original equipment market. 19. Perodua plans to reach a production capacity of 300,000 units a year by 2005. Under the new arrangement with Daihatsu, it will also manufacture Daihatsu brand cars (Daihatsu lacks a pro- duction plant in Southeast Asia). As the current producer of the lowest-priced car in the region and, in partnership with Daihatsu, the only producer of cars with engines smaller than 1 liter in Southeast Asia, Perodua may carve out a niche for itself. 20. Malaysian sources suggest that Daihatsu gave a commitment to Perodua that it will provide ex- isting Malaysian suppliers with an opportunity to meet its cost and quality requirements and supply assistance to facilitate technological upgrading. It also made it clear, however, that if local suppliers are unable to match its requirements, it will source components elsewhere ("New Era" 2001). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 185 Demand for and production of autos. In the late 1990s congestion, slow- ing economic growth, the belated maturing of urban mass transit systems, and the popularity of motorcycles depressed auto sales from their peak of 575,000 units to little more than 420,000. Passenger cars have established a dominant position, but small utility vehicles remain popular. Local assem- blers have increased their share of the market to more than 80 percent and have developed modest independent design capabilities. Production is frag- mented among about a dozen assemblers, led by affiliates of four Japanese companies (Honda, Mitsubishi, Nissan, and Toyota) and Ford. Automobile policy. After decades of vacillation, in the mid-l980s the gov- ernment of Taiwan (China) embarked on a slow process of liberalization. By 2000 tariffs on passenger cars had fallen to 30 percent (although strict quo- tas limited imports from Japan and Korea). Tariffs on parts averaged about 15 percent, while local content requirements were reduced to 40 percent. The government provided some promotional help with research and gave locally designed components a reduction of three percentage points off the stiff commodities tax. The government also strongly promoted "central- satellite relations" between assemblers and their suppliers. Parts production. Taiwan (China) has about 2,200 parts firms, of which about 300--almost all linked to Japan by capital investment or technology licenses--produce original equipment. By some estimates, perhaps 100 of the OEMs would be competitive even without protection. In 2000 parts production reached about $4 billion. Areas of strength include tires, steel and aluminum wheels, batteries, bumpers and other body parts, lights, and some engine parts. One of the weakest areas is transmissions, most of which are imported. Quality levels and design skills are considerably more ad- vanced than in China or Southeast Asia but lag behind those in Japan. Qual- ity standard 9000, ISO 9002, and ISO 14000 certification is common among large firms, but not small ones. Research and development spending in- creased rapidly over the 1990s, reaching 1.3 percent of sales by 1999. Despite the pervasive influence of Japan, production networks are quite different from those in Japan or Korea. Low volumes and fragmented assembly have precluded the development of distinctive Japanese-style keiretsu: parts firms supply all assemblers evenly. Nevertheless, the cluster- ing of assemblers and parts suppliers in the industrial region just south of Taipei has facilitated the development of a sophisticated division of labor. Trade and WTO accession. Taiwan's (China's) auto industry is still pro- tected by significant tariffs, quotas, and nontariff barriers, although those have loosened in recent years, especially for parts. Parts makers made up for 186 GLOBAL PRODUCTION NETWORKING IN EAST ASIA a modest loss of domestic original equipment orders with increased exports. As a result, both imports and exports grew relative to production, and local firms began to run a surplus in auto parts by 1996. Production is diverse, but replacement parts and accessories such as body parts, tires, batteries, light- ing, and bumpers lead the way. The ratio of exports to imports increased even in technically demanding areas such as engines and transmissions. In 1999 and 2000 surpluses averaged about 15 percent of production value. North America is far and away the largest export market, but China and other Asian destinations are catching up. The government has actively pro- moted exports to Japan and (especially) its overseas production bases by allocating licenses to import some Japanese cars to firms that increase their "reverse" exports to Japan. After steady progress, the sister firms China Motors and Yulong (with large minority investments by Mitsubishi and Nis- san, respectively) have nearly balanced their trade with the rest of the Japan- ese production networks. The elimination of local content requirements following accession to the WTO should reduce local production of some items. However, the slow pace of liberalization for local assemblers and the increased skill and export prowess of the parts exporters should cushion the blow considerably. Outward foreign direct investment. Taiwan's (China's) auto parts firms are unusual in combining a high degree of reliance on transnational com- panies, overwhelmingly from Japan, with a high degree of outward invest- ment. Both on their own and in conjunction with investments in China and Southeast Asia by the leading assemblers Yulong and China Motors, local parts firms have aggressively invested abroad, both to cut the costs of land and labor and to increase economies of scale. As of early 1999, Taiwanese (Chinese) auto firms had 106 ventures in China, largely divided between Fujian (home to the CMC and Yulong assembly operations and just across the Taiwan Strait) and China's economic heart: the Shanghai-Jiangsu-Zhe- jiang area. For a time Japanese assemblers largely ignored Taiwan (China). More recently, as Japanese firms have been forced to rationalize operations, they seem to have concluded that local firms provide reasonable quality at a reasonable price and that they can effectively complement operations in Southeast Asia and help to manage investments in mainland China. The auto industry, like Taiwan (China) more generally, appears to be finding its niche between China and Japan. Thailand Bythemid-1990stheThaiautoindustryhadbecomethestrongestinSouth- east Asia in volume of production (559,000 units in 1996), number of parts PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 187 producers (roughly 1,200 firms), and range of automotive parts pro- duced.21 This growth reflected roughly two decades of investments by Japanese OEMs, producing mostly for the protected Thai domestic market. In the early 1990s this supplier base and accompanying infra- structure, along with partial trade and investment liberalization and a growing domestic economy, drew export-oriented Western assemblers and, eventually, their first-tier suppliers. Despite being hit hard by the 1997 crisis, the industry is characterized by growing clusters of assem- blers and suppliers generating significant growth of autos and auto parts exports. The major downside of this picture is the continuing general weakness of Thai-owned producers. Demand for and production of autos. Thailand's automotive market grew rapidly from 1981 until the Asian economic crisis. The volume of do- mestic sales rose from 78,000 in 1986 to 304,000 in 1990, to 572,000 in 1995, and to 589,000 in 1996. After falling sharply to 144,000 in 1998, sales rose to 255,000 units in 2000, while production hit 459,000 units.22 During this same period, the volume of domestic production rose from 74,000 in 1985 to 304,000 in 1990, to 525,000 in 1995, and to 559,999 in 1996. The difference between domestic sales and production reflects the growth of Thai auto exports, which increased from 68,000 units in 1998, to 126,000 in 1999, to 153,000 in 2000, and to 175,000 in 2001.23 Auto parts exports grew as well. The structure of vehicle demand and production also merits note. The majority (58 percent) of vehicles produced in the last two decades were one-ton pickup trucks, which also dominated export sales (Brooker Group 2002, p. 106), with pickups accounting for more than 70 percent of all vehicles produced in Thailand in 2000 (see chapter 5 of this volume). The country's large domestic market for pickup trucks (second-largest in the world after the United States) has also been important in enabling Thai sup- pliers to achieve scale economies. 21. Production volumes are from the Automotive Industry Club of the Federation of Thai Indus- tries, cited in Brooker Group (2002, table 3.14). Numbers of firms are from Brimble and Sherman (1999, p. 21). Brooker Group (2002, p. 208) cites a figure of 1,600 parts firms. Takayasu and Mori (chapter 5 of this volume) also note that Thailand has the most developed raw materials industry in Southeast Asia. 22. Sales volume from the Auto Industry Club of the Federation of Thai Industries are cited in Brooker Group (2001, pp. 2­13). Production volumes from the Auto Industry Club of the Fed- eration of Thai Industries are cited in Brooker Group (2002, table 3.14). 23. Vehicle figures from the Thai Automotive Industry Association and MMC Sittipol are cited in Brooker Group (2002, p. 103, diagram 3.17). Vehicle exports rose from less than 5 percent of total production in 1995 to almost 40 percent in 2001 (Brooker Group 2002, p. 103, diagram 3.18). 188 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Japanese brands have dominated the Thai market, accounting for more than 90 percent of sales throughout the 1990s, declining only slightly to 88 percent in 2001 (Toyota Motor Thailand, cited in Brooker Group 2002, p. 108). And although some 15 assemblers operate in Thailand, the market is dominated by Toyota (28 percent in 2001), Isuzu (24 percent), Honda (13 percent), Nissan (12 percent), and Mitsubishi (8 percent). Western firms entering the market in the mid-1990s, especially BMW, Daimler- Chrysler, Ford, and General Motors, operate largely through Japanese partners and Japan-designed vehicles.24 Automotive policy. Thai auto policy has been a balancing act between pro- tection and product-specific incentives, on the one hand, and gradual liber- alization, on the other. The auto industry expanded initially as a result of import substitution industrialization efforts beginning in the late 1960s. Key policy components included tariff protection, local content require- ments (54 percent for passenger cars, 60­70 percent for one-ton pickups), and limits on new assemblers, facilities, and models. High protection en- couraged the growth of local parts producers, which numbered more than 1,000 firms. Also aiding local suppliers was a special, low tax rate on pickup trucks that, until 1991, resulted in pickup truck prices that were about half the price of mid-size passenger cars. More suitable to Thailand's rural areas, pickups offered parts firms higher scale economies in lower-technology parts (such as brake drums). Pickups were, in a sense, Thailand's national car, albeit one not limited to one brand, as was the case for Malaysia's national car programs. Beginning in around 1990, as Thailand's economy experienced rapid growth, the government began to loosen constraints on the industry. Liberalization included tariff reductions in 1991 and 1992; permission for existing assemblers to increase capacity and models; permission for investments in new assembly plants and parts production in 1994; pro- vision of tax incentives to encourage automotive exports in 1994; lifting of local content requirements in January 2000 to comply with WTO regulations; and ending of special decentralization incentives to encour- age clustering.25 24. Both Ford models produced in Thailand were designed by (Ford-owned) Mazda in Japan. General Motors operates through Isuzu. 25. Other measures included liberalization of taxi registration, which raised demand for 1,600 cu- bic centimeter passenger cars and a requirement that all producers install catalytic converters, which compelled assemblers to reduce their stock of autos prior to the deadline (see Nipon 1999, pp. 11­12). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 189 In addition, the government expanded infrastructure and investment in- centives for industrial estates in the eastern seaboard area and port facili- ties for container vessels at Laem Chabang (Unger 1998, ch. 6). Building on the overall growth of the Thai economy, these measures encouraged overall market expansion and attracted Western assemblers as well as first- tier suppliers such as Arvins, Bosch, Dana Spicer, Delphi, Johnson Con- trols, Lear, TRW, and Valeo. Among other consequences of these investments, two merit special note at this point. First, they provided further impetus for the development of automotive clusters, already initiated by Japanese assemblers and suppliers in the early 1990s.26 Second, they expanded capacity. By 1999 the auto in- dustry had a total capacity of roughly 920,000 units (Ministry of Industry, cited in Nipon 1999, table 5). Because the domestic market was relatively small, this expanded capacity added to the pressure for export expansion. There are limits to this liberalization, however. To help local parts pro- ducers prepare for the end of local content requirements, in 1997 the government initiated a number of tariff adjustments to encourage local assembly and sourcing.27 In addition, the Thai parts industry continued to be plagued by a biased tax structure for finished products and raw materials that raised the cost of locally made products relative to equivalent im- ported goods.28 Parts production. Local production of parts expanded with the industry's overall growth. By 2001 parts "made" in Thailand accounted for roughly 26. These include the Laem Chabang Cluster (Mitsubishi along with suppliers such as Thai Summit and Siam Michelin); Gateway Cluster (Toyota, with only a few suppliers); Eastern Seaboard Industrial Estate (Ford-Mazda and General Motors), and several mini clusters and stand-alone facilities (Thai Automotive Institute presentation and website). Takayasu and Mori (chapter 5 of this volume) characterize the Thai cluster as the largest automobile indus- try cluster in ASEAN. 27. Adjusted tariff rates increased the gap between tariff rates applying to assembled cars and to completely knocked down passenger car kits but cut the differentials for commercial vehicles. In 1999, in a further move to help parts producers, the government raised the tariff rate for pas- senger cars and pickup trucks from 20 to 33 percent, making imported parts some 5 percent more expensive than locally produced ones (Brooker Group 2001, p. 2.4; Brooker Group 2002, pp. 209­10; Nipon and Chayanit 2000, p. 6). More recently, Thailand is preparing to re- structure its tariffs to protect local parts firms prior to tariff reductions required by AFTA. After tariffs on finished vehicles are cut to 0­5 percent from 30­80 percent to comply with AFTA, du- ties on completely knocked down kits will also be cut to 0­5 percent from 20­33 percent to en- courage local assembly. Further, tariff cuts on completely knocked down vehicles will be se- lective (Watcharapong 2002). 28. The problem is that tariffs for many raw materials are higher than those levied on finished parts. See, for example, Nipon (1999, p. 21). 190 GLOBAL PRODUCTION NETWORKING IN EAST ASIA 60 percent of the components used in vehicles produced in Thailand (Brooker Group 2002, p. 205). The variety of locally produced products is the broadest in Southeast Asia, ranging from engines, to wiring harnesses, to radiators, to body parts, to tires. Auto parts exports expanded signifi- cantly, with the value of parts exports rising more than tenfold during the 1990s, from $78 million in 1991 to $871 million in 1995 and to $1.4 bil- lion in 1999 (Brooker Group 2002, p. 214). From 1998 to 2001, exports were dominated by wire harnesses, engines, ignition systems, and tires.29 Much of this export growth seems to have come from non-Thai produc- ers. According to one Japanese executive, half of all parts exports came from just two Japanese electronics firms: Denso and Thai Arrow (cited in Nipon and Chayanit 2000, p. 7). Thus, in 1999, only 30 percent of Thai exports came from locally owned firms; the other 70 percent came from foreign and Thai-foreign joint ventures.30 The minimal Thai role in exports reflects weaknesses apparent even be- fore the 1997 crisis. A 1995 study found that Thai-owned firms' level of production technology was a C+, below the international OEM require- ment of a B grade (Nipon 1999, table 9; see also chapter 5 of this volume). This rating reflects generally low productivity in assembly (11 cars per worker compared with the global benchmark of 45), weaknesses in preci- sion tool making, as well as a lack of personnel trained in general design and engineering. These problems are, in turn, a function of broader gaps in vocational training and persistently low enrollments in technical areas such as computing, engineering, mathematics, and science, especially com- pared with Korea and Taiwan (China).31 With regard to tertiary technical enrollment, Thailand's percentage of population enrolled in 1997 was 0.19 percent, almost the same as in 1985 (0.16 percent), whereas enroll- ment in Korea and Taiwan (China) doubled from levels that in 1985 were already three to four times higher than Thailand's 1995 figures (from 0.86 to 1.65 percent in Korea and from 0.59 to 1.06 percent in Taiwan, China; Bell 2003, p. 14). Indeed, a 2003 report concluded that "the scale of re- sources and capabilities for exploiting technology and generating innova- tion lags far behind what might be expected on the basis of international 29. Data on market destinations of these exports are available only for "general auto parts," which do not include the most important export products: for example, engines, tires, and electrical igni- tions (Brooker Group 2002, p. 215). 30. According to Brooker Group (2002, p. 214), in 1996 exports accounted for less than 2 percent of sales of one of the largest Japanese component producers operating in Thailand. By 2000 the per- centage was 24 percent. 31. For a recent overview of the mismatch between work force skills and company needs, see Ketels (2003). On vocational weaknesses, see Ritchie (2001). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 191 comparisons" with regional competitors.32 The industry's trade balance and real level of local content reflect the weaknesses of local firms. In 1996 Thailand's total trade deficit was $14.6 billion, of which $3.8 billion, or 28 percent, was due to the vehicle and parts section (Brooker Group 1997, p. IV-2). And actual local content--that is, real value added of vehicles as- sembled in Thailand--was much lower than the official 54­70 percent, be- ing closer to 20 percent according to one estimate (Veloso, Soto Romero, and Amsden 1998, p. 17). The Thai parts sector has experienced sharp consolidation since 1997. Because most local firms were severely undercapitalized before 1997, the crisis hit them hard. After the government lifted the 49 percent ceiling on foreign shareholding in 1997, many local firms were taken over by foreign partners that also provided access to new export markets. The result was a shakeout in which all of roughly 230 first-tier suppliers survived, albeit under foreign ownership, while some 600 smaller, largely Thai-owned, producers went out of business. By 2001­02 the roughly 1,200 Thai parts producers operating prior to the crisis had been reduced to some 500­600 firms.33 Prospects. Given the small size of its domestic market, Thailand must op- timize within the regional and global frameworks created by Western and Japanese assemblers. But the country also has important assets. Owing to its import substitution policies, product-specific incentives, local parts producers, infrastructure development, and recent, gradual liberalization, Thailand has become the site of several foreign-dominated automotive clus- ters. Given East Asia's expected market growth, these production sites are well poised to play important roles within the region. Relatively low-technology, high-volume vehicles offer significant op- portunities for local parts production. Since several firms--Ford, General Motors, Isuzu, Mitsubishi, Toyota--are committed to using Thailand's do- mestic market as a springboard for regional and global exports of pickup trucks, parts for pickup trucks, including engine components, constitute a major opportunity for local parts production. Recently, for example, Ford, which exports pickups from Thailand to 130 countries around the world, 32. Not only do Thailand's current capacities lag, but its commitment to developing these capaci- ties also "lags behind the efforts of [Korea, Taiwan (China), and Singapore] around 10­20 years ago when they had levels of economic development similar to those of Thailand now" (Bell 2003, p. 5). 33. The total of Thai firms is 600 according to Brimble and Sherman (1999). The Thai Auto- motive Industry Association lists roughly 525 producers as of 2001 (cited in Brooker Group 2002, p. 206). The number of first-tier survivors is from Brooker Group (2002, p. 208). 192 GLOBAL PRODUCTION NETWORKING IN EAST ASIA announced plans to expand its Thai facilities for regional sales, since the company expects Asia to account for a substantial portion of the growth in total global auto sales in the coming decade (Pichaya 2002). Another source of demand for local parts is the production, largely by Japanese firms, of entry-level, global platform­type vehicles for the Asian market. These in- clude low-priced vehicles, such as Toyota's multipurpose vehicle, suitable for developing countries, as well as low-cost passenger vehicles built on global platforms (so-called Asia cars) produced by Honda and Toyota. The Japanese have already begun to integrate these projects into regional com- ponent exchanges through AICO. Successful liberalization under AFTA will strengthen Thailand's position as a base of regional production. With regard to particular types of parts, expanding the range of elec- tronics components may be possible given Thailand's extensive consumer and industrial electronics sectors and the fact that electrical parts are already significant export items for the country's auto industry. This is advisable since Thai wage rates are no longer highly competitive relative to those of regional competitors (for example, China), and producers of labor- intensive products such as wire harnesses may opt to go elsewhere. Build- ing on its long experience in assembly, the country also has strengths in press parts and related supporting industries such as dies, molds, and jigs. Unlike more technologically sophisticated components, these products of- fer opportunities for entry and technological upgrading based on the par- ticular needs of developing-country markets. It is worth note that a Thai company has emerged as one of the world's best suppliers of low-volume tooling.34 Finally, Thai parts firms unable to break into OEM markets may pursue a "Taiwan (China) model" by exploiting aftermarket and REM opportunities.35 The preceding has several important consequences. First, as shown in chapter 5 of this volume, the growth opportunities for local parts manufac- turers are shrinking. The best OEM options for local firms lie, at least for the medium term, in becoming second- or third-tier suppliers or providers of tooling (jigs, molds, dies) and of intermediates and raw materials, such as plastics, leather goods, glass, paint, rubber goods, and petrochemical prod- ucts (Ki 2002). The replacement market offers real opportunities, but even here, Thailand is something of a "middle child": lacking broad technologi- 34. The firm, Aapico, recently won the contract to design, supply, and test jigs for the worldwide assembly of Mercedes's new luxury E-Class. See Crispin (2002); Deyo and Doner (2001). 35. The Thai Ministry of Industry recently announced a policy to reduce dependence on OEM production and follow "Taiwan's successful model" of REM exports to the U.S. market ("Auto Parts Exports" 2002). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 193 cal competencies, but no longer able to build on inexpensive labor (Brooker Group 1997, p. V-13). The degree of indigenous participation in Thailand's growing role as a regional production base thus depends on broader public policies and institutions devoted to human resources, small- and medium- size enterprise financing, acquisition of technology, information technol- ogy, logistics, and general value chain management. Efforts in most of these areas have already begun.36 Even absent indigenous participation, Thailand's role as a base of regional productionislikelytogrowasaresultofitspreexistingassets,theweaknesses of other Southeast Asian industries, and the pressures on global producers to localize supplier and support industries. But, given China's market, labor costs, and growing technological assets, Thailand cannot afford to rely on existing assets and investments alone. Sustained growth will require con- siderable attention to the same areas required for the development of lo- cal firms: human resources, information technology, logistics, and other types of infrastructure development. ASSETS, MARKET STRATEGIES, AND POLITICS The strategic options open to Asian governments in adjusting to this new environment of great opportunities but increasing barriers to entry depend in good measure on the location-specific assets that they bring to nego- tiations. But the translation of assets into desired outcomes depends on policies, and the effectiveness of those policies is, in turn, a function of institutions and political factors. Assets First and most important are the size of the local market and its prospects for future growth. Autos are complex, heavy products whose appeal is deeply influenced by local consumer preferences and government policies, from gas taxes and road quality to safety and environmental standards. Even absent tariffs and other trade barriers, automakers usually prefer to produce locally when possible, but the profitability of local production is greatly affected by the size of the market. For assemblers or large compo- nent suppliers considering new investments, market growth relative to ex- isting capacity is also crucial. Of the five economies discussed in this chap- ter, China offers the most attractive domestic market, while Korea, whose 36. An interesting example is the establishment of a joint effort linking Thai parts producers and Singaporean precision engineering firms looking for lower-cost production sites (Nareerat 2002b). 194 GLOBAL PRODUCTION NETWORKING IN EAST ASIA domestic market is largely saturated, has considerable potential as a re- gional hub for exports; these economies are best placed to attract new in- vestors. They are followed by Thailand and then Malaysia, while wealthy but saturated Taiwan (China) is the least promising. To the degree that ASEAN free trade is realized, Malaysia and Thailand will become more attractive sites for foreign investors, since similarities in geography and cli- mate make it possible to sell identical or similar cars from facilities in these countries to the rest of Southeast Asia. A second major consideration is the size and strength of the pool of ex- isting assemblers and suppliers and the degree to which they are located in economically efficient geographic clusters. The countries vary with regard to the presence of OEMs, with China's pool of assemblers growing, Korea's and Taiwan's (China's) fairly strong, and Thailand's and Malaysia's weak. With regard to clusters, the gap across our five economy cases is not great. China, with its continental reach, is somewhat more complicated than the others, although the location of most regional clusters near the coast ame- liorates the situation somewhat. All of the others have significant clusters of suppliers within easy reach of sea transport, usually concentrated in one part of the economy (for example, for Taiwan, China, the area south of Taipei, and for Malaysia, the area near Kuala Lumpur) or in two (for ex- ample, Busan and Seoul). Thailand's four clusters, especially the agglom- eration of assemblers and suppliers in the eastern seaboard, constitute an especially attractive asset. Cross-economy differences with regard to the numbers and manufac- turing capacities of local suppliers are somewhat greater. Parts firms in Ko- rea and Taiwan (China) are at the upper end of the regional scale with re- gard to technological competence. Thai firms, although numerous, are generally weak in the kinds of manufacturing capacities required by as- semblers. The situation is generally similar for local Chinese suppliers. Malaysian parts producers are probably the weakest in the region as a re- sult of high levels of protection and distortions generated by ethnic pref- erences. Despite some strengths, in none of the economies covered here, with the partial exception of Korea, do local suppliers constitute a major asset, except at the lower tiers of the supply pyramid. The strength of local suppliers is also affected by the strength, stability, and flexibility of each economy's financial system. None of the East Asian financial systems lacks problems, as the financial crisis of 1997 and the more chronic prob- lems in Taiwan (China) and Japan have demonstrated. China, the largest of the five, is most severely beset with nonperforming loans and has the fewest alternatives to the traditional banks, while Malaysia, the smallest of the five, probably has the soundest banking sector. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 195 Finally, physical and human infrastructure clearly differentiates the richer from the poorer economies. Korea and Taiwan (China) have the best trans- portation systems, the highest penetration rates for the Internet and broad- band telecommunications, and the most sophisticated engineers. Malaysia and Thailand are weaker (but improving) in areas such as transport and com- munications, but they present idiosyncratic problems: a very weak edu- cational system in the case of Thailand and continuing ethnic tensions (and educational and economic gaps) in the case of Malaysia. The most interest- ing case--and the most difficult for foreign investors to read--is China. On the one hand, it is the poorest and the most unevenly developed of the five, unemployment and other potentially explosive social problems abound, and local protectionism still remains a barrier to the integration of the domestic auto industry. On the other hand, at least along the coastal provinces, the im- provement in physical infrastructure, including the creation of an extensive highway system, is remarkable. And since the first stirrings of the industrial revolution over two centuries ago, the capacities of local manufacturers have probably never increased more rapidly than they have in China over the last two decades, a tribute not only to the determination of China's policymak- ersandthediligenceandflexibilityofChinesemanagersandworkersbutalso to the increasingly streamlined global system of technology transfer. Market Strategies In principle, these location-specific assets suggest somewhat different lever- age and options. Large and growing markets in China and Korea, and sig- nificant experience with auto parts export in Korea, mean that these two are in the best position to help local firms jockey for a better position in the global division of labor. Conversely, smaller domestic markets mean that Malaysia, Taiwan (China), and Thailand will find it more difficult to shape comparative advantage, with Malaysia and Thailand burdened further by weak indigenous technical capacities.37 But in spite of the constraints they face, even these economies have options. We can identify four categories of options for national auto suppliers, three of which involve market orientation: domestic or export market, original or replacement equipment manufacturing, and product-specific 37. For example, owing to delays in implementing AFTA's multilateral tariff reductions, auto- makers that selected Thailand as their primary regional hub might opt to divert some investment to other countries in the region to take advantage of lower (bilateral) tariffs under the AICO scheme. Such a development would result in a network of smaller plants, a possible increase in investment to China, and a further stimulus to localization, since AICO exchanges require 40 per- cent ASEAN content (Nareerat 2002a; Srisamorn 2002). 196 GLOBAL PRODUCTION NETWORKING IN EAST ASIA niches. The fourth has to do with the degree of local versus foreign pres- ence in locally based parts production. China and Korea have the op- portunity to expand exports from companies benefiting from economies of scale based on large domestic markets. In China's case, the growth of the domestic market provides significant opportunities for production and export of parts for lower-end cars and light commercial vehicles, although, as noted in the case study, the value added of these exports might be strat- ified among local operations linked to Taiwanese (Chinese), Japanese, and Western parts producers (see, for example, Murphy and Lague 2002). In Korea there are opportunities both for domestic OEM and exports of REM parts for Korean cars sold overseas. New linkages with global assemblers may provide new opportunities for marketing Korean-built cars and parts in regional and extraregional markets (for example, Latin America). The challenge for local parts firms is to increase their level of technological competency to enable them to tap into these new openings. As Taiwanese (Chinese) firms shift production of many OEM parts to China, their best options for domestic production seem to be in replacement parts such as bumpers and metal stampings for export markets. This is an area in which locally owned firms have managed to hold their own despite gradual liber- alization. Whereas continued growth in their domestic market is likely to provide Thai producers with some OEM sales, the greatest opportuni- ties are for OEM exports to regional markets in components for pickup trucks. Except for lower-tier components, this process will be dominated by foreign producers. Malaysia is the weakest player among the five. At best, there are opportunities for lower-value-added parts for national cars sold in the domestic market and, potentially, for regional sales of Proton, a small-volume passenger car that might be marketable elsewhere in the region. Whether this occurs, of course, will depend on trade, ownership, and other policies to which we now turn. Policies In this section of the chapter we review how policies in specific areas--trade, ownership, finance, infrastructure, and industrial policy more generally-- have exerted an impact on the auto parts industries in our five economies. Trade. Regional and global trends toward freer trade are power- ful. Everywhere from ASEAN to the Republic of Korea tariffs, quotas, and nontariff barriers are declining; with the accession of China and Taiwan (China) in late 2001, virtually all of the Asian economies have joined the WTO. Still, within the general trend of liberalization, some room for government maneuver remains. Taiwan (China), for PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 197 example, continues its path of consistent but slow liberalization, while Thailand, feeling pressure from the rush of foreign investment to China and benefiting from a strong assembler-supplier complex, is moving more quickly. The hand of government weighs heavily in China and Malaysia, but (with the partial exception of some leftovers from the financial crisis) more lightly in Korea. Ownership. A crucial issue concerns policies and informal attitudes toward ownership and autonomy. Local ownership can provide greater opportuni- ties for skill acquisition and technology transfer as well as for more inde- pendent corporate strategies. The automotive history of East Asia strongly suggests that the best-performing local firms have limited foreign ownership and diversified sources of technology rather than depending on foreign par- ent companies for capital or technology (Toyota versus Nissan in Japan, Hyundai versus Daewoo in Korea, Yulong­China Motors versus Liu Ho in Taiwan; see, for example, Amsden 1989; Cusumano 1985). At the same time, independence is no guarantee of success, and with increasingly liberal trade and a powerful trend toward global consolidation, it is unrealistic to expect that Proton, for example, could repeat the success of Toyota. Even local-for- eign joint ventures, which can be difficult to manage, are increasingly giving way to full foreign control. The overall trend is clearly toward liberalization of ownership requirements: to compete effectively in regional and global markets, assemblers and major parts suppliers are demanding and getting a freer hand. In assembly, this is illustrated by General Motors' takeover of DaewooinKoreaandtheacquisitionofeffectivecontroloverMalaysia'sPer- odua by Toyota's subsidiary Daihatsu as well as the increasing control by Western firms of their Japanese affiliates where such linkages are proving to be commercially successful. Still, some room for maneuver remains for the countries with the best location-specific assets, especially China and Korea.38 38. The Korean government, for example, ignored antitrust considerations in permitting the acqui- sition of Kia by Hyundai, thus greatly strengthening the position of Hyundai as de facto national champion. Another telling case is the September 2002 joint venture between Nissan and China's sec- ond-largest assembler, Dongfeng. In return for an initial investment of more than $1 billion and de- tailed promises to continue investing in China, Nissan received the right to assemble a full range of vehicles, something China had not granted to foreign producers in the past. Nissan did not, how- ever, gain majority control, and press reports suggest that the speed with which it will reduce Dongfeng's bloated work force will be substantially slower as a result (AutoAsia, September 19, 2002; Business Week, October 7, 2002). Although the motivation may be largely domestic--sustaining em- ployment among the urban working class, which has traditionally been the bedrock of support for the Communist Party--the effect of such agreements may well be to send "costly signals" that Sino- foreign joint ventures will inevitably be committed to sustaining and expanding production. In a re- gion troubled by domestic overcapacity, such signals could deter investments in Korea, Taiwan (China), and Southeast Asia. 198 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Systematic information on policies regarding the ownership of parts pro- ducers is less readily available, but Taiwan (China) and Thailand seem to be movingtowardcompleteliberalizationofownershiprequirements.Yetsome diversity in ownership patterns evidently persists, as reflected in the overseas operations of Denso, Japan's largest producer of auto parts.39 China and Ko- rea appear to be the two partial exceptions to the trend toward complete for- eign control, although the number of cases is too small to allow definitive conclusions. Finance. A third important policy area is finance, particularly banking. The vast majority of auto parts producers remain dependent on banks for financ- ing; only the largest suppliers have the skill and scale to finance their opera- tions primarily through offerings of stocks or bonds. The quality, stability, and accessibility of local banks thus are crucial to the viability of the auto industry. Although weaknesses plague all Asian banking systems, cross- economyvariationsareemerging.Chinaisespeciallytroubled,andthespeed at which new banks, including foreign banks, enter the market may become an important issue for local companies. Thailand's banking system also re- mains weak. Korea, in contrast, is turning around what had been one of the least effective banking systems in the region. An additional source of financ- ing is support from Japan--both from the parent companies of local sub- sidiariesandfromtradingcompanies--andfromJapaneseforeignaid,agood chunk of which has found its way toward local subsidiaries of Japanese firms, especially in the aftermath of the financial crisis. Local parts producers not closely affiliated with the Japanese may thus find themselves at a disadvan- tage if financial problems create another credit crunch. Infrastructure. With the slowdown of growth in Southeast Asia and mas- sive spending on roads, railways, and ports in China, transportation con- gestion has generally eased, at least for a few years. But the generation and transmission of electricity have sometimes failed to keep pace, even in rela- tively advanced areas such as Taiwan (China). To some degree, the private sector can itself deal with energy issues. With advances in technology, it is increasingly feasible for large firms to generate electricity on site, perhaps selling excess capacity to the national power grid. But this may create dilem- 39. Denso has adopted full or almost full ownership in most markets, particularly for its largest plants, but 50-50 joint ventures or even minority shareholdings are not uncommon, particularly in Asia. As of 2001, Denso's shares in three Korean subsidiaries were 40, 51, and 100 percent, re- spectively; in Thailand they were 51.3 and 100 percent; and in China they were 30, 40, 51, 85.9, and 94.2 percent (Denso 2001, pp. 8­9). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 199 mas for small producers not clustered near larger firms. Telecommunica- tions facilities may also manifest some clustering tendencies, as suppliers find it economical to extend fiber optic cables to large and tightly packed industrial users near urban areas. Smaller and richer areas--Hong Kong (China), Korea, Singapore, Taiwan (China)--are pulling away from the de- veloping countries, despite rapid installation of broadband access through- out the region. Again, China is in between--uneven and behind, but rap- idly improving. And again, gaps are also opening up by firm size, as virtually all large suppliers can offer products on the Internet, but few smaller companies can do so. In electronics, Dell and other foreign assem- blers are requiring that all procurement be handled through e-commerce; as similar trends spread to autos, economies where the entire supply base can work effectively through the Internet will find themselves with a major competitive advantage. Full implementation of e-commerce is dependent not solely or even pri- marily on physical infrastructure but also on skills and organizational in- frastructure. Indeed, skills may become the great differentiator between winners and (relative) losers in the auto business. Technical training and upgrading of local skills is largely a "win-win" situation for foreign assem- blers and their local hosts and suppliers (see, for example, the agreement between General Motors and Thailand; Noble 2001) but implementation remains difficult, particularly in countries with weak or uneven educational systems. Takayasu and Mori, chapter 5 of this volume, also refer to the weaknesses of Thai auto component suppliers with respect to information technology. Japanese assemblers have proven more willing to provide tech- nical assistance to their suppliers (even in North America, suppliers re- port that the Japanese are significantly more helpful), but they also face a greater language and cultural gap, since few locals know Japanese, while Japanese managers and technicians often have weak English skills. Less no- ticed, but important for the long run if firms and workers are to be con- vinced to invest heavily in specific skills that could become vulnerable to market vicissitudes at the firm or industry level, is the provision of re- training opportunities and employment protection for auto parts workers (for example, see Estevez-Abe, Iversen, and Soskice 2001). Implementation of training programs and many other collective policies, such as format standardization, electronic commerce, and interactions with other industries, often are greatly facilitated by effective industry associa- tions. Here again, the Japanese role has been particularly important. Japan- ese advisers have taken the lead in establishing many industry associations, Japanese agencies have often insisted on working through local associa- tions, and Japanese-affiliated firms are usually the leading participants in 200 GLOBAL PRODUCTION NETWORKING IN EAST ASIA both national and regional auto associations (Doner 1997). To improve in- dustry associations requires financial support, capable staff, and the willing- ness of governments to communicate with and delegate many tasks to the associations. Korea and Taiwan (China) are well ahead of Southeast Asia, while China is hampered by its size and diversity. Industrial policy. Improving physical, human, and organizational infra- structure is a difficult but essential task for East Asian economies. The de- gree to which regional governments should focus on particular products, technologies, or firms--that is, articulate an explicit or implicit industrial policy--remains a matter of considerable controversy. The usual argu- ments in favor of industrial policy--promoting economies of scale, scope, learning, and agglomeration as well as overcoming coordination dilem- mas, market externalities, and weaknesses in local financial markets--all apply to autos, and first-mover advantages, once less true of autos than of electronics, may be growing. Increasing oligopolization of global assembly and component production also makes host economies reluctant to trust to markets that are less than perfectly competitive. Given these considerations, some version of industrial policy is likely to persist, despite the challenges and dangers involved, such as daunting in- formation requirements, difficulty in keeping up with rapid market devel- opments, infant industries that fail to grow up, politicization and rent seek- ing, and a world trading system that is much less willing to acquiesce in government policies of protection and promotion than it was when Japan and Korea developed their auto industries. Such persistence may be par- ticularly the case for China and Korea, which export significant quantities of auto parts to world markets. Such industrial policies will probably take a soft form that does not at- tempt to target highly specific products if such targeting is inconsistent with the strategies of subsidiaries of global firms. Instead, policies will be designed to promote and expand comparative advantage, even if such ad- vantages are the result of highly targeted policies from the past. In Thailand, for example, tax policy has created an overwhelming preference for one- ton pickups that has allowed assemblers to achieve economies of scale in pickup production. Along with assembly, production of pressed parts and jigs for pickups is also concentrating in Thailand. Building on this base, the Thai government has initiated policies to make auto parts production, in- cluding for the replacement market, one of the country's target growth in- dustries ("Competitiveness" 2002). To the extent that economies attempt to build on (somewhat fortuitous) outcomes of past policies as opposed to imposing sectorally neutral policies, PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 201 they must address two further issues. The first is whether governments should encourage local parts producers to concentrate on OEM or the pro- duction of replacement parts. At first glance, production of aftermarket goods appears more appropriate for developing economies, since these tend to be price sensitive and to impose only modest technical requirements. For many years, Taiwan (China) exported more auto parts than Korea, de- spite the much larger size of the Korean auto industry, precisely by concen- trating on replacement parts. The size of the no-brand replacement market is limited, however, and pressure by Western consumers to prevent insur- ance companies from paying for only the lowest-quality replacement parts may shrink it further. In recent years, the expansion of exports of Korean cars has propelled the export of both OEM and replacement parts, so that Korean auto parts exports now easily outpace those of Taiwan (China). Production of OEM parts, however, requires tighter coordination with global suppliers and first-tier component suppliers, and, as noted, require- ments for quality and interoperability are higher. Local firms hoping to break into OEM markets, and even foreign firms looking to expand local production to cut costs, may well petition governments for help on a vari- ety of fronts, from technology to marketing. At the same time, Taiwan's (China's) experience, in which public-private sector linkages were critical to certification and "national reputation," suggests that local firms succeeding in developed-country REM markets also require help in areas such as qual- ity control and delivery. A second issue is whether governments should orient policies toward smaller parts makers at the bottom of the production pyramid or toward medium and larger firms in the first or second tier. Opportunities for rapid productivity growth and high-paying jobs are greater near the top of the pyramid, but establishing a major position in the regional and global production networks of global assemblers, including increasing responsi- bility for design, will be difficult for most local suppliers, which, once again, are likely to petition governments for support. At the same time, promotion of small- and medium-size enterprises itself poses important challenges in areas such as information diffusion, quality certification, technical training, and financing. Thus, while the general trend of trade and industrial policies in the re- gion is clearly toward more liberalization and openness, the incentives for companies to continue to seek and for governments to supply some degree of protection and, most critically, promotion are not likely to disappear. The policy challenge will be to do so intelligently and with realistic goals well suited to factor endowments. Successfully meeting this challenge is a function not of some abstract "political will," but rather of the distribu- 202 GLOBAL PRODUCTION NETWORKING IN EAST ASIA tional and institutional factors influencing the formulation and implemen- tation of policy. Politics: Interests and Institutions Local auto parts producers tend to press for protection against imports and market downturns: production is generally labor intensive and, at the higherend,involveslargeandoftenhighlyasset-specificinvestments.Never- theless, corporate reorganization has proceeded at a rapid pace in recent years, and Asian governments have generally acquiesced. Korea and Thai- land are the clearest cases. In Taiwan (China), General Motors, Honda, and Peugeot have all dumped their long-time joint venture partners and put more emphasis on imports, while many parts firms have moved much or all of their production to China. Even though urban workers have tradition- ally constituted the core support of the Chinese Communist Party, the Chi- nese government has allowed and even pushed state-owned enterprises to shed labor and explicitly favors the top three assemblers over all others.40 Even Malaysia may be the exception that proves the rule: the government has relinquished majority ownership in Perodua, and former Prime Minister Mahathir, the godfather of the domestic industry, signaled his in- tention before his retirement to ease the ethnic component of industrial policy; in any event, a fundamental review of auto policy is likely now that Mahathir has retired. Resistance to liberalization has not evaporated, but it is relatively muted: organized labor is a significant force only in Ko- rea, the reluctance of Taiwan (China) to allow unimpeded trade and in- vestment with China has had only a minor restraining impact on autos, and so forth. But distributional issues are far from resolved. For if particularistic pressures in the auto industry are proving weaker than expected in the face of regional and global trends toward liberalization and consolidation, it is less clear that East Asian governments have developed the coalitional bases to develop competent bureaucracies with effective and relatively transparent ties with the private sector. Thailand, for example, has resis- ted protectionism but has proved woefully weak at provision of crucial public goods and, despite considerable prodding from Japan, has been in- consistent in its attempts to build institutional links to a well-coordinated private sector. Korea has a stronger bureaucracy but is still struggling to 40. Toyota's joint venture to build engines in Tianjin, for example, has used a variety of methods to cull employment by 20 percent, and management is determined to cut another one-third in the next couple of years (Noble interview, Tianjin, April 2002). PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 203 build an open and effective relationship with the giant chaebol business groups, a task now complicated by the presence of major transnational partners. Democratization in a context of personalized and factionalized politics has yet to dampen the dramatic policy swings associated with a one-term presidential cycle. While both Thailand and Malaysia have tended to promote one partic- ular product niche, they have differed in a number of important ways. In assembly, Thailand's emphasis has been on small commercial vehicles-- one-ton pickups--whereas Malaysia's has been on passenger vehicles. Also in assembly, Thailand has pursued a private sector­led strategy with few limits on foreign ownership, whereas Malaysian efforts have been state led and fairly restrictive with regard to foreign ownership. In part owing to its more open policies regarding FDI and in part due to its lack of discrim- ination against ethnic Chinese­owned parts producers, Thailand's indige- nous parts sector is larger and more dynamic than Malaysia's. The result has been a greater Thai willingness to pursue liberalization, both regional and global. But if the two countries differ with regard to the size and strength of their parts sectors, their support for local parts firms, their trade policies, and their approaches to foreign investment, they resemble each other in their indigenous technological weaknesses, their relatively small domestic markets, and the fragmentation of their policies and insti- tutions devoted to technological improvement. Institutional weaknesses are likely to become all the more damaging as the region's economies attempt to implement the kinds of policies, espe- cially in areas such as infrastructure and human resource development, necessary to meet the growing demands of assemblers and their first-tier suppliers. Several features make these issue areas especially challenging. First, effective policies in these areas require information that is techni- cal and dynamic. Second, the required number of participants is relatively large, and the breadth of participation is relatively wide. For example, the increasing importance of first- and second-tier components firms from Japan and the West suggests that governments will need to enter com- plex negotiations with a wider variety of foreign firms and will no longer be able to depend on relations with a handful of global assemblers. Third, areas such as technical training tend to be plagued by free rider and asso- ciated collective action problems. Fourth, distributional tensions may worsen as liberalization, rationalization, and consolidation result in the disappearance of local firms and the rise of unemployment. Finally, the kinds of efficiency-related measures required to benefit from today's auto- motive production networks exhibit long time-to-payoff. They do not lend themselves to quick fixes. Investments in infrastructure, human re- 204 GLOBAL PRODUCTION NETWORKING IN EAST ASIA sources, and technology diffusion often take time to implement and even longer to yield tangible benefits. All of these considerations mean that effective policies must be backed up by institutions that can collect, exchange, and diffuse information, co- ordinate large numbers of interests, provide selective benefits and punish- ments, and operate on the basis of long time horizons. Absent such capaci- ties, it is unrealistic to expect significant local participation in and, especially, benefits from regional and global auto value chains. CONCLUDING OBSERVATIONS Great opportunities beckon the auto parts industry in East Asia, but ex- ploiting those opportunities is becoming more difficult. Strong growth in demand for autos and parts in the region combines with global trends to- ward liberalization, consolidation, and delegation of manufacturing to lower parts of the production chain to open up lucrative opportunities to local firms with the skills and savvy to assume vital roles in regional and produc- tion networks. Liberalization and consolidation have created losers as well as win- ners, particularly among smaller local firms and those without foreign capital or technological links, but so far, protectionist impulses have re- mained surprisingly weak. Governments have generally accepted the ver- dict of the market and have made only limited efforts to moderate the pace of liberalization. Less progress has been made in the positive task of creating capacities in the public and private sectors to boost local skills and technology to take advantage of the opportunities created by growth and liberalization. In much of the region, design and management skills remain weak, and e-commerce is far from universal. In recent years the Northeast Asian economies, while still modest actors in the global industry, have developed some impressive location-specific assets. Korea, in particular, combines a large market with solid engineering capacities. China's uneven develop- ment is more than offset by rapid growth in demand and skills, the lure of its vast population, and the commitment of its leadership to industrial up- grading. The picture in Southeast Asia, where many locally owned firms have failed and others are likely to succumb to regional and global liberal- ization, is less sanguine. Even Thailand, an emerging center of regional as- sembly of cars and pickups, lacks the organizational capacity and human skills necessary to meet global standards for quality, design, speed, and re- liability of delivery. PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 205 Nevertheless, assemblers continue to prefer local assembly and pro- curement of parts whenever production volumes allow. With combined efforts from local governments, industry associations, aid donors, and inter- national economic organizations, Southeast Asian producers still have ample opportunity to develop new capacities to supply local, regional, and global markets, both independently and in conjunction with the leading global assemblers and particularly the two or three dozen first-tier suppli- ers that are increasingly responsible for the manufacture of automobiles and their parts. Whether such efforts occur and succeed will depend in large measure on political dynamics: partisan competition, coalition for- mation, and their impact on bureaucratic development. REFERENCES The word processed describes informally reproduced works that may not be commonly avail- able through libraries. Abrenica, Joy V. 2000. "Liberalizing the ASEAN Automotive Market: Impact Assessment." Pacific Economic Cooperation Council project on the ASEAN auto industry. Processed. Amsden, Alice H. 1989. Asia's Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. "Auto Parts Exports: Strategy Shift for Local Firms." 2002. Nation, June 25. Bell, Martin. 2003. "Knowledge Resources, Innovation Capabilities, and Sustained Com- petitiveness in Thailand. Final Report." University of Sussex, Science and Technology Policy Research Unit, January. Geneva Processed. Borrus, Michael, Dieter Ernst, and Stephan Haggard, eds. 2000. International Production Networks in Asia: Rivalry or Riches? London: Routledge. Brimble, Peter, and James Sherman. 1999. "Mergers and Acquisitions in Thailand: The Changing Face of Foreign Direct Investment." Paper prepared for the UN Conference on Trade and Development. Geneva. Processed. Brooke, James 2002. "Japan Carves out Major Role in China's Auto Future." New York Times, July 11, sec. W, p. 1. Brooker Group. 1997. "Automotive Industry Export Promotion Project: Thailand Indus- try Overview. Final Report." Report prepared for the Thai Ministry of Industry, Bangkok. Processed. ------. 2001. "Thailand." Paper prepared for a project on the international competitive- ness of Asian economies: a cross-country study. Bangkok, October. Processed. ------. 2002. "Thailand's Automotive Industry." Bangkok. Processed. Brunnermeier, Smita, and Sheila A. Martin. 1999. "Interoperability Cost Analysis of the U.S. Automotive Supply Chain: Final Report." Research Triangle Institute, Research Triangle Park, N.C. Processed. Burwell, Michael, and Lisa Ferris. 2000. "Road to Riches." Daily Deal (New York), August 2. Available at LexisNexis: www.TheDeal.com. Processed. 206 GLOBAL PRODUCTION NETWORKING IN EAST ASIA "Car Tax: Major Overhaul to Precede AFTA." 2002. Nation, May 20. Chatrudee, Theparat. 2002. "Toyota Targets 100 Percent Local Pickup Content in Four Years Awaiting Final Word on Shift to Thailand." Bangkok Post, August 28. "Competitiveness: PM Unveils Global Drive." 2002. Nation, May 21. Crispin, Shawn. 2002. "Fast Lane to Success." Far Eastern Economic Review, September 12. Cusumano, Michael A. 1985. The Japanese Automobile Industry: Technology and Management at Nissan and Toyota. Cambridge, Mass.: Harvard University, Council on East Asian Studies. de Jonckheere, Terry. 1998. "Asian Crisis: Challenges and Opportunities." Presentation to Asian Autos after the Crash: Where to from Here? Ford Motor Company, Dearborn, Mich.; University of Michigan, Ann Arbor, Office of Study of Automotive Trans- portation. Processed. Denso. 2001. "Corporate Profile 2001." Kariya, Japan. Processed. Deyo, Fredrick, and Richard F. Doner. 2001. "The Enclave Problem: Challenge of Flexi- ble Production in a Weakly Coordinated Market Economy." In Fredrick Deyo, Richard Doner, and Eric Hershberg, eds., Economic Governance and the Challenge of Flexibility in East Asia. Boulder, Colo.: Rowman and Littlefield. Doner, Richard F. 1997. "Japan in East Asia: Institutions and Regional Leadership." In Peter J. Katzenstein and Takashi Shiraishi, eds., Between Two Worlds: Japan in Asia. Ithaca, N.Y.: Cornell University Press. Dunne, Michael. 2001. "When the Barriers Come Down." Presentation to Automotive News International Asia-Pacific Congress, March 28­29. Processed. Economist Intelligence Unit. 2001. "Automotive Forecast Asia." London. Processed. Estevez-Abe, Margarita, Torben Iversen, and David W. Soskice. 2001. "Social Protection and the Formation of Skills: A Reinterpretation of the Welfare State." In Peter A. Hall and David W. Soskice, eds., Varieties of Capitalism: The Institutional Foundations of Com- parative Advantage. Oxford: Oxford University Press. Fine, Charles H., John C. LaFrance, and Don Hillebrand. 1996. The U.S. Automobile Man- ufacturing Industry. Washington, D.C.: U.S. Department of Commerce, Office of Technology Policy. Fourin. 1999. Kaigai Jido-sha Cho-sa Geppo-. (Monthly Survey of Global Car Manufacturing) 172. Nagoya, December. Hatch, Walter, and Kozo Yamamura. 1996. Asia in Japan's Embrace: Building a Regional Pro- duction Alliance. Cambridge, U.K.: Cambridge University Press. Humphrey, John. 1998. "Assembler-Supplier Relations in the Auto Industry: Globalization and National Development." Unpublished mss. Institute of Development Studies, University of Sussex, Brighton. Processed. "Hyundai Mobis, Bosch Ink Brake Technology Deal." 2001. AutoAsia Online, November 16. Jomo, K. S. 1994. "The Proton Saga: Malaysian Car, Mitsubishi Gain." In K. S. Jomo, ed., Japan and Malaysian Development: In the Shadow of the Rising Sun, pp. 263­90. London: Routledge. Ketels, Christian H. M. 2003. "Thailand's Competitiveness: Key Issues in Five Clusters." Unpublished mss., Harvard Business School, Cambridge, Mass. Processed. Ki, Woo. 2002. "Auto Sector Rises Again." Nation, October 11. Legewie, Jochen. 2000. "Driving Regional Integration: Japanese Firms and the Develop- ment of the ASEAN Automobile Industry." In Verena Blechinger and Jochen Legewie, PRODUCTION NETWORKS IN EAST ASIA'S AUTOMOBILE PARTS INDUSTRY 207 eds., Facing Asia: Japan's Role in the Political and Economic Dynamism of Regional Cooper- ation. Munich: Iudicium. Leutert, Hans-Georg, and Ralf Sudhoff. 1999. "Technology Capacity Building in the Malaysian Automotive Industry." In K. S. Jomo, Greg Felker, and Rajah Rasiah, eds., Industrial Technology Development in Malaysia: Industry and Firm Studies, pp. 247­73. London: Routledge. Liang, Quincy. 2002. "Taiwan Leads the World in Sheet-Metal Auto Body Parts." Avail- able at http://www.taiwanfurtniture.net/linerpt/20000302063.html. Processed. Matsushima, Noriyuki. 1999. "Japan's Auto Industry Plunged into an Era of Great Realign- ments." Interview with the director of Equity Research, Nikko Salomon Smith Barney, June 15. Available at http://www.fpcj.jp/e/gyouji/br/1999/990615.html. Processed. McKendrick, David, Richard F. Doner, and Stephan Haggard. 2000. From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry. Palo Alto, Calif.: Stanford University Press. McMaster, T. D. 2001. "Autoliv." Presentation to Automotive News International Asia- Pacific Congress, March 28­29. Processed. Ministry of Machinery Industry. 2000. Zhongguo Qiche Gongye Nianjian 2000 [China Auto- motive Statistical Yearbook 2000]. Beijing. Murphy, David, and David Lague. 2002. "Time for a Tune-up." Far Eastern Economic Review, July 4, pp. 24­27. Nareerat, Wiriyapong. 2002a. "ASEAN Free Trade: AFTA Delay Stifles Auto Plans." Nation, July 17. ------. 2002b. "Joint Vehicle-Parts Council: S'pore Agrees to Form Auto Club." Nation, October 31. "New Era for Perodua." 2001. Autoworld.com.my, December 6. Available at http://www. autotrade.com.my/EMZine/Review/article_t2.asp?awReviewID=846&aw Parent ID =846. Processed. Nipon, Paopongsakorn. 1999. "Thailand Automotive Industry (Part I)." Thailand Devel- opment Research Institute, Bangkok. Processed. Nipon, Paopongsakorn, and Wangdee Chayanit. 2000. "The Impact of Technological Change and Corporate Reorganization in the ASEAN Automotive Industry." Pacific Economic Cooperation Council, project on the ASEAN auto industry. Processed. Noble, Gregory W. 2001. "Congestion Ahead: Japanese Automakers in Southeast Asia." Business and Politics 3(2, August):157­84. Pichaya, Changsorn. 2002. "Ford Mulls $100m Boost." Nation, November 8. PWC (PricewaterhouseCoopers). 1999. "Major Auto Mergers Drive Sweeping Changes in the Parts Industry, According to PricewaterhouseCoopers Survey." Available at www.pwc.com. Processed. ------. 2000. "The Second Automotive Century Executive Summary." Processed. Rasiah, Rajah. 1996. "Rent Management in Proton." Working Paper 2. University of Kebangsaan, Institut Kajian Malaysia dan Antarabangsa (IKMAS), Sarawak, Malaysia. Processed. Ravenhill, John. 2002. "From National Champions to Global Partners: Crisis, Globaliza- tion, and the Korean Auto Industry." In William W. Keller and Richard Samuels, eds., Crisis and Innovation: Asian Technology after the Millennium, pp. 108­36. Cambridge, U.K.: Cambridge University Press. 208 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Ritchie, Bryan. 2001. "Innovation Systems, Collective Dilemmas, and the Formation of Technical Intellectual Capital in Malaysia, Singapore, and Thailand." International Journal of Business and Society 2(2):21­48. Scott, Alan J. 1987. "The Semiconductor Industry in Southeast Asia." Regional Studies 21(2):143­60. Srisamorn, Phoosuphanusorn. 2002. "Brooker Confident over Local Industry's Outlook." Bangkok Post, July 17. Treece, James B., and Robert Sherefkin. 2001. "Platform Sharing Key to Profits." Auto- motive News 75(5919):53. Tyndall, Paramjit Singh. n.d. "The Malaysian Automotive Industry." Pacific Economic Co- operation Council project on the ASEAN auto industry. Available at http://www. asean-auto.org/Malaysia.pdf. Processed. Unger, Daniel. 1998. Building Social Capital in Thailand: Fibers, Finance, and Infrastructure. New York: Cambridge University Press. Veloso, Francisco. 2000. "The Automotive Supply Chain: Global Trends and Asian Per- spectives." Background paper for a project on international competitiveness of Asian economies: a cross-country study. Massachusetts Institute of Technology, Cambridge, Mass. Processed. Veloso, Francisco, Jorge Soto Romero, and Alice Amsden. 1998. "A Comparative Assess- ment of the Development of the Auto Parts Industry in Taiwan and Mexico: Policy Implications for Thailand." Massachusetts Institute of Technology, Cambridge, Mass. Processed. Watcharapong, Thongrung. 2002. "Major Overhaul to Precede AFTA." Nation, May 20. WTO (World Trade Organization). 2002. International Trade Statistics 2001. Available at http://www.wto.org/english/res_e/statis_e/its2001_e/chp_4_e.pdf. Yap, Chips. 2001. "Will Malaysia Ever Open Up?" Automotive Resources Asia, October 15. Available at www.auto-resources-asia.com/experts. Processed. CHAPTER 5 THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS AND THE IMPLICATIONS FOR THE THAI AUTOMOBILE INDUSTRY Ken'ichi Takayasu and Minako Mori E ast Asian automobile industries have experienced dramatic changes since the late 1990s. These far-reaching changes are being driven by powerful forces that include trade and investment liberalization, strategic partnerships among vehicle assemblers in the advanced in- dustrial economies, and the rising standard of Asian automobile industries. In this chapter, we analyze the major changes that are occurring in East Asian automobile industries from four interrelated perspectives. First, the East Asian and global strategies of assemblers based in advanced economies are having an even greater influence on automobile industries in the East Asian economies than previously. The aim of those strategies is to enhance competitiveness by taking advantage of trade and investment liberalization to improve the efficiency of production networks in East Asia. Second, vehicle assemblers recognize the location-specific assets of East Asian economies, including markets, technologies, human resources, and industry clusters (see chapter 4 of this volume). Faced with escalating international competition, they are giving priority in their investment allocations to the economies with the best assets. Their aim is to build com- petitive production networks. When the governments of target economies imposed various restrictions on investment, assemblers merely needed to manufacture whatever vehicles they could under these limitations at prices that allowed them to earn a profit in local markets. In the environment cre- ated by trade and investment liberalization, assemblers cannot survive 210 GLOBAL PRODUCTION NETWORKING IN EAST ASIA against competition unless they build structures that allow them to provide markets with products that meet international standards of quality and price. Third, an extremely important policy issue from the viewpoint of East Asian economies is the positioning of their economies in the international production networks of vehicle assemblers and primary parts manufactur- ers based in advanced industrial economies. This is true not only of mem- bers of the Association of South East Asian Nations (ASEAN) but also of the Republic of Korea, which has its own internationally active vehicle assemblers, and China, which has a vast market and extensive industrial in- frastructure. Even in these economies, many domestic vehicle manufactur- ers have formed strategic partnerships with manufacturers from advanced industrial economies. Fourth, the present situation offers both opportunities and challenges to automobile industries in East Asian economies. The expansion of foreign assemblers' activities in these economies will bring increased business op- portunities for local parts manufacturers. To turn these potential busi- ness opportunities into reality, however, parts manufacturers will need to meet international requirements in terms of such factors as quality, price, and delivery. Domestic manufacturers that are unable to reach these stan- dards will either fall by the wayside or be forced to reposition themselves as secondary or tertiary parts suppliers for foreign parts manufacturers. From a national economic perspective, increased investment by foreign vehicle as- semblers and parts manufacturers will lead to the expansion of production and exports and to higher added value and export earnings. Yet if this growth is driven by foreign manufacturers, it will not necessarily raise the standards of domestic manufacturers. For these reasons, East Asian governments need to develop policies that make their assets more attractive, attract in- vestment by foreign manufacturers, and enhance the competitiveness of do- mestic manufacturers. In Thailand, for example, the focus of efforts to develop the automobile industry is shifting from the protection of domestic industry to the devel- opment of human resources and infrastructure needed to link the global strategies of foreign assemblers to the advancement of the domestic auto- mobile industry. This chapter seeks to clarify the ways in which the regional and global strategies of vehicle assemblers have influenced automobile industries in East Asian economies by focusing on the dramatic expansion of investment in Thailand by Japanese vehicle assemblers. Japanese assemblers have been involved in vehicle manufacturing in Thailand since the 1960s, and they have made a major contribution to the accumulation of production tech- THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 211 nology. In recent years the choice of which type of vehicle to manufacture has become a highly significant question in relation to the expansion of investment in Thailand. Ultimately, the regional and global strategies of assemblers are what determine decisions about which vehicles will be pro- duced in which economies and at what prices and quality standards. An assembler will decide to start production of a new model in Thailand af- ter making decisions on a range of other factors, including (a) the choice of vehicle type, based on analyses of consumer preferences and the struc- ture of vehicle markets, (b) the choice of manufacturing method, including the selection of a platform and technology to be transferred from Japan, and (c) the role of Thailand in East Asian and global production and sales networks in determining where the vehicles should be produced and sold. There are two reasons for our decision to focus on Thailand. First, Thailand has experienced the growth and development of its domestic automobile industry as a consequence of policies that actively welcomed investment and participation by foreign assemblers and parts manufac- turers. Second, Thailand has become a major base for vehicle assemblers from advanced industrial economies as they implement East Asian and global strategies. In fact, there is a clear trend among vehicle assemblers toward the posi- tioning of Thailand as a production and export base in the ASEAN region (see chapter 4). Since the 1990s, not only Japanese assemblers but also American and European assemblers have regarded Thailand as an appro- priate target for massive investment. We have sought to identify the ways in which the strategies of Japanese assemblers have influenced the Thai automobile industry by dividing assemblers' vehicle production operations into individual processes so that we could determine, through interviews with representatives of Japanese assemblersandpartsmanufacturers,whichprocessesarecarriedoutinJapan and which in Thailand and what division of roles is planned for the future. (The interviews were conducted in July 2001.) By analyzing the allocation of production functions between parent companies and overseas production sites,itispossibletoascertainanassembler'sspecificstrategyregardingtech- nology transfer. Manufacturers take profitability into account when trans- ferring the technologies that are needed to implement their strategies. Vehicle assemblers do not channel their investment into economies that lack the technology required to realize their strategies at a reasonable cost. The first half of this chapter consists of an overview of automobile mar- kets and manufacturing infrastructure in East Asia, together with an analysis of the issues confronting Japanese vehicle assemblers, especially from a pro- duction perspective, as they implement their new strategies in the ASEAN 212 GLOBAL PRODUCTION NETWORKING IN EAST ASIA region. This is followed in the second half by an examination of issues in Thailand, including the allocation of production processes between Japan and Thailand now and in the future, priorities for the acceleration of tech- nology transfers from Japan, and policy priorities for the government. OVERVIEW OF THE EAST ASIAN AUTOMOBILE INDUSTRY Most of the world's automobile industries form a four-part structure con- sisting of Asia, Europe, Japan, and North America.1 However, the automo- bile industry is substantially smaller in Asia than in the other three regions, in terms of market scale and volume of production. Moreover, Asia has long been divided into separate national markets, each with many production facilities, a factor that has limited the potential for economies of scale. In 2000 the combined automobile market of the seven automobile- producing economies in East Asia, not including Japan, amounted to 4.9 million units (table 5.1). China is the biggest market, followed by Korea, Taiwan (China), Malaysia, Indonesia, Thailand, and the Philippines. The total market for the 12 East Asian economies, including those that produce no automobiles, amounts to 5.0 million units. This is substantially smaller than the figures for other regions, which include 18.0 million units for the American market, 16.5 million units for the European Union market, and 6.0 million units for the Japanese market. The biggest single market in the ASEAN region is Malaysia (343,000 units). This is smaller than the to- tal for Brazil (1.3 million units) and Mexico (871,000 units), which are the two biggest automobile-producing economies in Latin America. In 2000 the seven automobile-producing economies in East Asia had a combined output of 6.5 million units. The fact that this total exceeds the size of the region's market is attributable to the large contribution made by ex- ports in the case of Korea. Korea is the region's biggest producer, followed by China, Taiwan (China), Thailand, Malaysia, and Indonesia. The United States produces 12.8 million units, the European Union produces 17.1 mil- lion, and Japan produces 10.1 million. Notwithstanding its small size, the East Asian market remains important because of its growth potential. While the market for new cars in Japan, the United States, and Western Europe has been stagnant for the past 10 years, 1. Japanese vehicle assemblers divide the world into Asia, Europe, Japan, and North America. From the viewpoint of American and European assemblers, the world consists of three regions, since Japan is part of Asia. a 7 b otalT 340 -- 348 42 362 293 285 326 795 317 2,702 2,964 1,935 2,076 1,671 3,115 2,069 20,275 17,142 17,699 12,800 16,643 10,144 58,296 vehicles a 2 b 25 97 36 26 26 cial 324 656 430 101 323 -- 513 190 287 96 2,693 2,236 9,327 7,258 1,413 4,136 1,782 1,464 16,996 oduction Commer Pr a b vehicles 5 16 372 542 239 2,378 1,551 1,279 1,646 1,348 -- 324 605 265 257 259 136 507 221 8,363 2,602 17,582 14,907 12,508 41,299 Passenger 84 871 307 776 873 373 301 343 262 111 640 otalT 2,435 1,586 2,038 1,300 5,963 1,431 2,087 2,603 17,030 16,518 20,450 17,993 10,955 57,800 vehicles 82 47 55 29 305 737 268 439 233 277 253 373 110 254 179 299 234 cial 2,282 2,206 9,136 4,225 1,703 1,475 10,141 18,456 Sales Commer 2000 vehicles 47 29 83 82 849 603 225 499 617 613 263 297 406 2,130 8,857 1,570 1,067 6,732 4,260 1,058 2,302 14,747 14,312 10,309 39,312 Automobiles, of Passenger (2002). ope oduction Pr Eur JAMA and omfr units) ope Central of of Sales economy Union Eur or n and China 5.1 n States East Rep. Africa. A America Asia total Compiled available. ope ester opean Asia of gentina ea, ce: ableT (thousands Region Eur W Eur Easter NAFT United Canada Mexico South Ar Brazil Middle Oceania East Japan Kor China aiwan,T Indonesia Malaysia Philippines Thailand Others orld Australia. South Rest Africa W --Not a. b. Sour 214 GLOBAL PRODUCTION NETWORKING IN EAST ASIA East Asian markets averaged almost 15 percent annual growth rates before 1997. The Asia-Pacific region (excluding Australia and Japan) is projected to account for slightly less than half of incremental global volume between 1998 and 2006, with the largest markets, China and Korea, doubling over the next decade. ASEAN markets are projected to grow at rates between 10 and 20 percent a year. The automobile industry is currently affected by global overcapacity. East Asia is no exception to this pattern. In the early 1990s predictions of sustained market growth triggered substantial expansion of capacity in ASEAN and Korea. However, the pace of market recovery in the wake of the currency crisis has been slow. Despite this situation, factories planned before the currency crisis are now going into production in Thailand, while foreign assemblers are increasing their capacity in China in parallel with suppliers of parts. In addition, assemblers in ASEAN are undertaking new investment in preparation for the introduction of new models. Total production value and production value per worker of the East Asian automobile industries are substantially smaller than the correspon- ding statistics for Japanese industry. As shown in table 5.2, the output of the Japanese automobile industry in 1998 was worth $307.8 billion, which is far higher than China's $50.9 billion, Korea's $46.4 billion, and Thailand's $18.4 billion.2 Productivity is also low in East Asia. In 1998 production per worker was worth $209,397 in Korea and $171,589 in Thailand, compared with $409,277 in Japan. Until the first half of the 1980s, there was little need for governments or automobile assemblers to be concerned about international competitive- ness. The industry was in the early stages of mass production, and govern- ments used a variety of policy tools, such as tariffs, domestic taxes, import restrictions, and local content regulations, to encourage import substitution production. Governments saw local content ratios as indicators of the de- velopment of automobile industries, and their greatest priority for foreign manufacturers was to fulfill local content requirements while coping with a variety of limiting factors. Between the second half of the 1980s and the middle of the 1990s, gov- ernments and automobile assemblers began to envision a future for the automobile industry based on a view of ASEAN as a single economic unit. This led to the introduction of brand-to-brand complementation schemes, whereby parts were supplied reciprocally within the ASEAN region, as a way of maximizing the benefits of mass production at the regional level. In 2. A billion is 1,000 million. THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 215 per dollars) 15,074 96,513 oduction employee 209,397 151,599 114,427 126,829 171,589 409,277 (U.S. Pr b age (U.S. -- 965 W 6,508 3,638 4,776 rate dollars) 21,094 17,053 43,007 m fir 72 25 98 72 verage 498 166 120 122 A number per all of in cent 8.5 6.8 2.9 0.9 2.9 2.7 4.4 7.6 Per Employees employees manufacturing 72 38 38 30 222 107 752 persons) 3,375 Number (thousands of of all of 7.2 4.1 1.0 3.4 6.1 of oduction cent 10.2 13.2 12.6 1998 pr Per value alueV manufacturing Asia, in U.S. 4,394 4,832 2,934 automobile Millions of dollars 46,444 50,876 10,940 18,398 307,776 all . of Manufacturing ers ms cent fir 3.3 4.6 1.8 1.1 1.4 1.7 4.6 2.8 industry Per manufacturing manufacturing. . all Automobile to (1999). manufactur equipment ms worker Automobile fir 232 317 249 of 3,083 6,779 2,838 1,095 per Number of 10,436 elativer UNIDO e ar costs omfr earY 1997 1998 1996 1998 1997 1997 1996 1998 transportation Overview the annual centages 5.2 of Compiled a available. Per ea, ce: ableT Economy Kor Rep. China iwan,aT China Indicates Indicates Indonesia Malaysia Philippines Thailand Japan --Not Note: a. b. Sour 216 GLOBAL PRODUCTION NETWORKING IN EAST ASIA 1996 agreement was reached on introduction of the ASEAN Industrial Cooperation (AICO) scheme, which contributed to the expansion of intra- regional trade in automobile parts. As of April 2002, a total of 61 AICO approvals had been granted to automobile manufacturers. The ASEAN automobile industry has experienced dramatic changes since the second half of the 1990s. Assemblers no longer have self-contained ASEAN strategies and are now focusing on the effective use of their pro- duction facilities in ASEAN within the framework of their global networks. This shift began with the exporting of commercial vehicles from Thailand to markets throughout the world. This reflects two environmental changes affecting the Asian region and the automobile industry. First, there has been progress toward trade liber- alization within the region. This process will accelerate further in the future. Under the framework of the ASEAN Free Trade Area (AFTA), all of the original ASEAN signatories (Brunei, Indonesia, Malaysia, the Philippines, Singapore, and Thailand) except Malaysia will reduce their tariffs on built- up vehicles imported from within the region to between 0 and 5 percent. (Malaysia moved its deadline for this change back to 2005.) In addition to these moves to lower tariffs within the ASEAN region, there is also intense interest in a free trade agreement with China. At a sum- mit meeting in November 2000, ASEAN and Chinese leaders agreed to work toward the signing of a free trade agreement by 2010. Since then, con- sultation has proceeded faster than expected. There is also the possibility of progress toward the negotiation of free trade agreements between Japan and ASEAN and between Korea and ASEAN. Drastic reductions in tariff rates for built-up vehicles and automobile parts would have a major influence on the Asia strategies of automobile manufacturers and on trade within the re- gion. ASEAN members currently impose tariffs of between 30 and 70 per- cent on built-up vehicles imported from outside the region. A significant re- duction in those rates would have a particularly important impact. Second, leading automobile assemblers began to build strategic part- nerships based on capital relationships in 1998 and 1999. Of particular sig- nificance were the relationships that formed between Japanese automobile manufacturers and their counterparts in Europe and the United States. These developments would radically change the profiles of automobile manufacturers in Asia. The formation of strategic alliances between Japanese assemblers, which have numerous facilities in East Asia, especially in ASEAN, and their European and American counterparts, which have production facilities in other regions, such as Eastern Europe, Europe, Latin America, and North America, is starting to have a profound impact on the situation. American THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 217 and European automobile assemblers have benefited from the powerful brands, procurement networks, and technology of Japanese manufacturers. They have also been able to establish production and sales structures in East Asia almost overnight. Now that their strategies are linked to those of American and European assemblers, Japanese assemblers will need to restructure their business operations from a global perspective. Each group has production facilities in all or most of the seven economies listed. However, only three groups had formed strategic alliances with Korean manufacturers as of the end of 2001. Those groups are Daimler- Chrysler, which has acquired a shareholding in Hyundai, Renault, which has acquired Samsung Motor, and General Motors, which now owns Daewoo. Vehicle assemblers have been able to pursue these increasingly complex global strategies thanks to the existence of information and communica- tions technology (ICT) networks. These networks are being used not only to improve the efficiency of internal information flows but also to exchange information with suppliers, sales outlets, and consumers. This strengthening of links with global networks is not limited to ASEAN. In Korea, the only mass exporter of passenger vehicles in East Asia, local assemblers are starting to link their networks with the global networks of the American and European manufacturers that have become their share- holders. Many automobile assemblers based in the advanced economies are starting to establish facilities in China. In addition, China will have to phase down its tariff rates for imported automobiles as a requirement for mem- bership in the World Trade Organization (WTO). As illustrated by these trends, the rapid implementation of global strategies by automobile assem- blers will necessitate a change in the structure of automobile industries in East Asia. PRODUCTION-RELATED PRIORITIES FOR ASSEMBLERS SEEKING TO ENHANCE THEIR COMPETITIVENESS IN EAST ASIA Vehicle assemblers have started to clarify the roles of the production sites that they have established in East Asia and to use them in the context of global production networks. This process will require assemblers to con- front two issues. First, because they have started to introduce the same models in East Asia as in other regions, assemblers will need to reallocate functions by deciding which operations should remain with the parent com- pany in Japan and which should be transferred to production sites in East Asia. Second, because products are required to meet international standards in terms of quality and price, assemblers need to manufacture products that 218 GLOBAL PRODUCTION NETWORKING IN EAST ASIA can compete internationally on price and quality under conditions that are less favorable than in their home country. Selecting Models to Produce and Allocating Functions For automobile assemblers, the goal of global management is to maximize profitsthroughtheefficientuseofmanagementresourcesscatteredthrough- out the world. To achieve this, companies must decide where to carry out basic research, design, and development, which models to produce at which locations and at what prices and quality standards, and where to sell those models. Multinationals have long faced the task of balancing the benefits of centralization, which allows economies of scale, against localization, which allows products to be tailored to local needs (figure 5.1).3 By launching the same model on multiple markets, assemblers can reduce research and de- velopment (R&D) costs per vehicle. This issue can be paraphrased in the following two-part question: Which models should be sold on East Asian markets, and which parts of the processes leading up to mass production should be carried out in the country concerned? Model strategies for ASEAN illustrate these changes in corporate strate- gies. The models sold in ASEAN can be divided into three categories. First, there are vehicles for ASEAN. Second, there are vehicles for emerging markets for sale in developing regions, such Latin America. Third, there are models for sale in both ASEAN markets and the markets of advanced economies. There has been a shift in the models that vehicle assemblers have introduced or are planning to introduce in ASEAN, away from prod- ucts designed specifically for a single market (so-called "Asia car" models) and toward models that share common features with vehicles designed for other regions, which fall into the second and third categories. One reason for this change is the fact that ASEAN consumers are now demanding newer models. A second reason is that it is becoming possible for vehicle assemblers to concentrate production of specific products and parts within a single country. A characteristic development in relation to passenger cars was the announcement early in 2001 that production of Asia car models would be terminated. First launched onto the market in 1996, these low- priced vehicles were targeted toward the urban middle class, especially in Thailand. To minimize costs, design modifications included the removal 3. Japanese automobile assemblers are pursuing localization strategies. For example, they have launched localized versions of sport utility vehicles not only in East Asia but also in North America. THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 219 offer merits that Functions centralization Use Asia Japan. in and large Asia and localization: Japan Japan. in available in functions) with Asia products. those used clusters Strategies between other produced to in from and needs) complaints methods supplier networks Global functions R&D centralization Vehicles reflected the with of to transferred of regional be different consumer can are from and merits information Sharing and countries Reconciling of in Relation national available differently in centralization changes only meet localization characteristics Centralization Localization to minor through Asia offer materials produced applied is to market that and are with volumes. scale Localization of products Asian parts parts export strategy If If If Dealing and for Functions (1) (2) (3) (4) This production economies (supplying Japan in (achieving Centralization produced 5.1 Vehicles Figure of and global goal sell efficiently this price we Institute using as produce, by we what strategy deployment should Goal achieve at profit resources to and Research should and where Global used and Japan models where, maximize possible To management as Methods (Development technology) What when, quality, them? Source: 220 GLOBAL PRODUCTION NETWORKING IN EAST ASIA of features not essential for ASEAN economies. Now some groups plan to produce vehicles in specific ASEAN economies for export to neighboring economies or to introduce strategic vehicles for emerging markets. Many assemblers are also planning to introduce "world car" models. These are mostly compact vehicles intended for use in multiple markets throughout the world, including markets in the advanced economies and in the ASEAN region. Thailand has also started to produce models, such as one-ton pickup trucks, that can be exported to many countries. These changes in the types of vehicles introduced have been accompa- nied by increased efforts of vehicle assemblers to localize production as part of the sharing of functions between Asia and Japan. The localization strate- gies of assemblers depend on design and development capabilities in the re- gion for two reasons. First, assemblers must produce a wide range of models in East Asia to suit market needs. Assemblers need to be able to make rapid design changes to reflect local characteristics in East Asian economies, es- pecially when introducing common models in Asia and other regions. To do this, they need to build systems that will allow their plants in ASEAN to turn market information into product plans and to advance those plans in East Asia to mass production without delay. In the case of commercial models that are not produced in the advanced economies, it is quite possible that minor changes that have no impact on the basic performance, frames, dy- namic performance, or safety of vehicles will be carried out at plants in East Asia. Second, the most important requirement in terms of bringing prices in line with international levels is to increase local procurement. To do this, manufacturers will need to establish product engineering capabilities at the East Asian plants and to develop production methods that support the op- timized use of locally available materials. Both of these aspects are discussed in detail in the case study of Thailand. Resolving Two Issues Affecting the Smooth Implementation of Model Strategies Assemblers will need to resolve two key issues as they implement their model strategies in East Asia: meeting consumer needs and meeting inter- national standards of price and quality. Meeting consumer needs. The first issue is the need to create a structure capable of supplying a wide range of models to meet consumer needs. East Asian markets have significant numbers of consumers who purchase vehi- cles configured for local markets (Fujimoto and Sugiyama 2000, p. 406) THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 221 and manufactured using platforms and core components, such as engines, that are no longer current in the country of origin. At the same time, ad- vances in information technology mean that consumers now have easy ac- cess to the latest information from the advanced economies, with the result that it is becoming more difficult to sell outdated models, especially in the passenger car category. The slowdown in sales of Asia car models is attrib- utable in part to this change in consumer behavior. To satisfy the demands of East Asian consumers, manufacturers will increasingly need to shorten the time lag between the launch of new models in the markets of advanced economies and their introduction in East Asia. Achieving international standards of price and quality. The second is- sue is that manufacturers are under pressure to measure up to international standards in terms of both price and quality. The major assembler groups are introducing a succession of new models onto Asian markets. As a result, manufacturers must now enhance their price and quality competitiveness. International competitiveness is also crucial for models that are exported to the advanced economies, such as one-ton pickup trucks. The removal of local content regulations and the reduction of intrare- gional tariffs under AFTA will force suppliers to compete with imported parts. Moreover, as assemblers switch in earnest to global procurement systems, international prices will become the only prices for parts and ma- terials. Assemblers will first select suppliers on the basis of price infor- mation and then consider whether each supplier is able to meet their re- quirements in terms of quality, delivery dates, and other criteria. Obviously, cost reduction is an ongoing priority in the automobile in- dustry. It has become a major issue in East Asia over the past few years be- cause the automobile industry, which was previously protected by govern- ment policy, has been forced to reduce its costs to international levels over a very short time, despite the inadequacy of its infrastructure. According to an interview survey conducted in Thailand in July 2001, assemblers are tar- geting cost reductions of around 30 percent over the next three years.4 In the past, target prices for parts manufactured in Thailand were calculated by adding transportation costs and tariffs to procurement prices in Japan. Recently, however, prices are being based directly on Japanese or inter- national price levels. Moreover, manufacturers previously placed a higher 4. Assemblers are applying uniform cost-reduction targets, not only in Japan but also in their overseas facilities, including those in East Asia. The three-year target of 30 percent is the same as the target for facilities in Japan. 222 GLOBAL PRODUCTION NETWORKING IN EAST ASIA priority on quality than on price in the case of key components because of the emphasis on automobile safety. Today, however, suppliers must meet increasingly high standards in terms of both price and quality. Three Approaches to the Improvement of Price Competitiveness Assemblers and suppliers must now achieve substantial improvements in their total competitiveness, based on both price and quality, over a very short time. In particular, manufacturers are using a variety of measures based on regional characteristics in Asia to reduce their costs at the inter- national level. There are three approaches to the improvement of price competitiveness. Increasing local content ratios. The most important way to enhance price competitiveness is to increase local content. In the case of Japanese assemblers with production facilities in Thailand, parts and raw materi- als account for about 82 percent of automobile manufacturing costs (cost in 1998 = 100). Depreciation makes up 8 percent, and other manufactur- ing costs make up 6 percent. Labor accounts for a mere 4 percent of to- tal manufacturing costs ( Japan Economic Foundation 1999, p. 18). Sim- ilarly, the regional content ratio (the percentage of parts and materials procured in local currencies) in ASEAN was 52 percent in 1998. On a net basis, excluding parts and materials imported for use in the production of parts purchased locally by assemblers, the regional content ratio was around 24 percent. The major assemblers still regard local procurement as a key issue, de- spite the abolition of local content requirements and the trend toward glob- ally optimized procurement of parts and materials. This is because they can reduce their exposure to exchange rate risk if they can source local parts that measure up to international standards for price and quality. To raise local content ratios, it is also necessary to achieve economies of scale. When the number of vehicles produced by each assembler is small, as is the case in East Asia, it is essential to build business relationships that allow suppliers to serve multiple assemblers across corporate groups and to boost production to the required scale through exporting or other means. Adopting a modular approach. Another way to reduce production costs is to produce parts under modular systems, which increase the unit for production processes to the subassembly level. Modular parts are being in- troduced throughout the world as part of efforts to cut production costs. In ASEAN there is considerable scope for reducing processes and costs by THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 223 using similar production methods for parts manufactured using the modu- lar approach in the advanced economies. To achieve these cost savings, mul- tiple suppliers need to cooperate with first-tier suppliers.5 Achieving the optimal mix of labor and machinery. The design of ef- ficient production lines is also crucial to attaining the required breakeven point. By deploying workers efficiently, manufacturers can lower the de- preciation cost of machinery, thereby achieving cost performance on a par with highly automated production lines in the advanced economies. The percentage of processes automated on assemblers' production lines in ASEAN amounts to a fraction of that in advanced economies. Many processes that are carried out mechanically in the advanced economies are handled by human workers in East Asia. Capital investment and automation tend to be limited to areas in which consistent quality is required. Even then, manufacturers seek to minimize investment costs and do not necessarily in- stall the most up-to-date equipment. CHANGES IN PRODUCTION STRUCTURE AND TECHNOLOGY TRANSFERS IN THAILAND How can the East Asian economies take advantage of the global strategies of automobile assemblers to develop their own automobile industries? In this and the following two sections, we use Thailand as the case study for our analysis of this question. We examine the business strategies of multinational corporations, especially Japanese companies, in Thailand and the develop- ment of the Thai automobile industry. This analysis focuses on the ways in which manufacturers have positioned their operations in Thailand and on the future role of those operations. We also attempt to predict the future shape of the industry and identify policy priorities for its development. We focus on Thailand because Thailand is a pertinent example of an economy that has achieved industrial development through FDI and tech- nology transfers from MNCs, especially Japanese companies. A compar- ison of automobile industries in East Asia (excluding Japan) shows that 5. For reasons of profitability, plants in ASEAN are not used to develop new modular parts that have never been produced at plants in advanced economies. As in the United States, the wage gap between assemblers and suppliers is not large, so only minimal cost benefits would be achieved by allowing suppliers to develop new modular parts. Even where suppliers are able to develop and produce modular parts profitably, there are other issues. For example, the automation of pro- duction lines by assemblers is still limited, and neither plants nor production lines are configured for modular parts. 224 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Thailand is the region's third-biggest producer of automobiles after Korea and China. Its industry has grown to the extent that it now occupies an important position in the Asian strategies of automobile assemblers. Overview of the Thai Automobile Industry The Thai automobile industry has changed radically since the late 1990s. Until the mid-1990s, the industry produced automobiles only for the do- mestic market under a policy designed to protect domestic industry. Since 1997, however, there has been a dramatic shift toward a role for Thailand as part of the global strategies of automobile assemblers. The shift reflects en- vironmentalchanges,includingthecurrencycrisisandthesubsequentslump in domestic sales, global moves toward trade and investment liberalization, and the formation of strategic alliances among automobile assemblers. Environmental changes. Table 5.3 provides an overview of automobile production in Thailand. One-ton pickup trucks accounted for 63 percent of all vehicles produced in Thailand in 2001. This type of vehicle makes up a large share of the domestic market because of taxation advantages (also noted in chapter 4). Thailand's exports of one-ton pickup trucks also have expanded in recent years. These trends are reflected in the continuing growth in the share of this type of vehicle in production. Table 5.3 Production, Sales, and Exports of the Automobile Industry in Thailand One-ton Commercial vehicles pickups as a Indicator Passenger One-ton Total percentage of and year vehicles Total pickups vehicles total vehicles Production 1991 76,938 206,177 169,940 283,115 60.0 1996 138,579 420,849 351,920 559,428 62.9 2001 156,066 303,352 289,349 472,351 63.0 Domestic sales 1991 66,779 201,781 155,366 268,560 57.9 1996 172,730 416,396 327,663 589,126 55.6 2001 104,502 192,550 168,639 297,052 56.8 Export 1991 -- -- -- -- -- 1996 -- -- -- 14,260 -- 2001 -- -- -- 175,299 -- --Not available. Source: Thai Automotive Industry Association. THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 225 There are 13 automobile assemblers in Thailand, mainly Japanese. Japanese companies first moved into Thailand in the 1960s, and they have adapted to Thailand's local content policies by building production capacity there. They accounted for almost all vehicles produced in Thailand until Ford established a joint venture in Thailand with Mazda in 1998, followed by General Motors in 2000. As shown in table 5.4, Japanese brands still pre- dominate, accounting for 78 percent of total production in Thailand (esti- mated as the sum of domestic sales and exports in 2001). In addition, many Japanese parts manufacturers have followed Japanese assemblers into Thailand, as noted in chapter 4. Global move toward trade liberalization. Since the late 1980s, economic growth and the resulting rise in incomes have been paralleled by a steep up- ward trend in the number of vehicles sold in Thailand (figure 5.2). In the early 1990s, market growth was spurred by falling prices and the intro- duction of new models. These changes were driven by a gradual easing of policy pertaining to the automobile industry, including the reduction in tariff rates and the removal of restrictions on participation in the industry, and by the escalation of competition as assemblers vied to secure their share of an expanding market. By 1995 manufacturers had increased their pro- duction capacity on the strength of forecasts that the domestic market would reach 800,000 units by 2000. It was in this period that Ford and General Motors decided to move into Thailand in order to strengthen their Asia strategies. Investment by assemblers encouraged parts manufacturers to ex- pand into Thailand, leading to formation of the biggest automobile indus- try cluster in ASEAN. The onset of the currency crisis in 1997 was followed by a sharp decline in the domestic market. This caused the Thai automobile industry to shift its focus from domestic demand to exports. In 2001 Thailand exported Table 5.4 Share of Each Nation's Brands in Thailand, 2001 Number of units Share of all brands Domestic Domestic Brand sales Export Total sales Export Total Japanese 261,939 105,918 367,857 88.2 60.4 77.9 American 20,196 69,381 89,577 6.8 39.6 19.0 European 11,405 0 11,405 3.8 0.0 2.4 Korean 1,936 0 1,936 0.7 0.0 0.4 Others 1,576 0 1,576 0.5 0.0 0.3 Total 297,052 175,299 472,351 100.0 100.0 100.0 Source: Thai Automotive Industry Association and others. 226 GLOBAL PRODUCTION NETWORKING IN EAST ASIA Figure 5.2 Automobile Sales and Production in Thailand (Thousand units) 600 500 400 300 200 100 0 61 71 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 sales production Source: The Thai Automotive Industry Association. 175,299 vehicles, an increase of 15 percent over the previous year. This total was equivalent to 38 percent of total automobile production. Exports thus helped to supplement domestic demand, allowing the industry to main- tain production levels during the slow recovery in the domestic market. In addition to Mitsubishi Motors, which has used Thailand as a production base for one-ton pickup trucks since the mid-1990s, Thailand has also be- come an export base for fully assembled vehicles manufactured by Auto Alliance Thailand, which is a Ford-Mazda joint venture, and General Motors. The presence of these three companies has played a major role in the sustained growth of exports (table 5.5). Together they accounted for 86 percent of total automobile exports in 2001. Formation of strategic alliances. Thailand needs to improve its interna- tional competitiveness. The Thai automobile industry, which produced for the domestic market under protectionist policies, must now make rapid im- provements in its ability to compete as an exporter of fully assembled ve- hicles. In the space of a few years, the benchmarks for the Thai automobile industry reached international levels in terms of price and quality. Automobile manufacturers are beginning to use their operations in Thailand from a global perspective. This change reflects trade and invest- ment liberalization under the WTO and AFTA and the formation of al- liances among American, European, and Japanese companies. For example, some manufacturers are making Thailand their main center for Asian pro- duction and export of specific types of vehicles. Alliances between compa- THE GLOBAL STRATEGIES OF JAPANESE VEHICLE ASSEMBLERS 227 ease incr 5.5- of 679.7 15.8- 25.0- 11.6 35.3- 73.7- 27.5- 14.7 Rate total of 6.9 3.9 2.1 0.7 0.2 2001 e 34.2 27.9 24.0 100.0 Shar units of 392 6,900 3,683 1,206 60,027 48,987 42,077 12,027 175,299 Number 541 2000 6,283 6,184 5,689 4,590 63,540 49,977 16,031 152,835 519 665