24971 September 2002 ~ F CAN EAST ASIA COMPETE? INNOVATION FOR GLOBAL MARKETS '/, /SHAHID YUSUF AN D H THE WORLD BANK; - SIMON J. EVENETT CAN EAST ASIA COMPETE? Innovation for Global Markets SHAHID YUSUF AND SIMON J. EVENETT Ekwj A copublication of THE WORLD BANK the World Bank and Washington, D.C. Oxford University Press ©D 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NTXV Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbanLkorg All rights reserved. First printing Septemiber 2002 1 2 3 4 05 04 03 02 A copublication of the World Bank and Oxford University Press. Oxford University Press 198 Madison Avenue New York, NY 10016 The findings, interpretations, and conclusions expressed here 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 govemments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other infornmation shown on any map in this work do not imply on the part of the World Bank any judgment of the legal starus of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any forn or by any means, electronic or mechanical, uicluding phorocopyinig, recording, or inclusion in any infornma- tion storage and retrieval system, without the prior writteni pemission of the World Bank. The Wlorld Bank encourages dissemination of its work and will nonnally grant pemuission promptly. For permission to photocopy or reprinit, please send a request with complete infomsation to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephonie 978-750-8400, fax 978- 750-4470, www.copyright.com. All other queries on rights and licenises, including suibsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, fax 202-522-2422, e-niiail pubrights @worldbank.org. ISBN 0-8213-4998-8 Cover design by Naylor Design Inc., Washington, DC Liln-ay of Congress Caealoging-in-Publication Data Yusuf, Shahid, 1949- Can East Asia compete?: innovation for global markets / Shahid Yusuf and SimonJ. Evenett. p. cm. Includes bibliographical references and index. ISBN 0-8213-4998-8 1. Technological innovations-Economic aspects-East Asia. 2. East Asia-Economic policy. 3. East Asia-Economic conditions. 4. Finance-East Asia. 5. Competition, Intemational. I. Evenett, Siimon J. H. Title. HC460.5.Z9 T422 2002 338'.064 095-dc2 i 2002028868 CONTENTS Foreword vii Preface ix Abbreviations and Acronyms xiii East Asia Enters Round Two 1 A New Dynamic: Why and How 1 East Asian Development: Round Two 14 Notes 1 5 2 The National and Regional Development Context 17 Industrial Restructuring and the New Face of Policy 17 Context for Reforn in East Asia 2 1 Notes 3 1 3 The Warp and Weft of Innovation 3 3 Role of the Business Environment 36 Clusters: The Spatial Dimension to Innovation 47 Local-Global Networking 55 Global Circulation of Knowledge and of Knowledge Workers 56 Research and Innovation by Firms: The Growing FDI Dimension 59 Logistics and Supply Chains 62 Conclusion 66 Notes 69 4 Financial and Business Services 73 The Banking Sector: Crisis Avoidance, Macroeconomic Stability, and Growth 74 Salience and Competitiveness of Business Services 89 Future Role of Business Services 94 Notes 96 5 ICT: Beyond the Hype 99 What's at Stake? 99 Making the Most of the Internet: The State's Role 117 From Fixed-Line to Wireless Internet Access 124 Wireless Infrastructure, Competition, and Policy 134 Internet-Based Growth in East Asia 137 Notes 138 6 Managing the Global Interface 143 Reaping the Benefits of Open Trade and Investment Regimes 144 Conclusion 166 Notes 167 7 Innovating to Win 171 Why Innovation Matters 172 Can East Asia Retain Its Lead? 177 What Should East Asia Do? 180 Notes 182 References 183 Index 205 Box 5.1 Impacts of the Internet 116 Figures 1.1 Number of Patents with a Foreign Coinventor, by Sector, OECD Countries, 1980-84 and 1992-96 8 1.2 Share of Patents with a Foreign Coinventor, Major OECD Economies, 1985-87 and 1993-95 9 1.3 U.S. Patent Grants, Selected East Asian Economies, 1995 and 2000 12 iv 2.1 Trade Barriers, Selected Country Groups and China, 1984-87 and 1994-97 22 2.2 Inequality, Selected East Asian Nations, Selected Years 27 2.3 Japan's Declining Influence in East Asia, Selected Years 29 2.4 Allocation of Japanese Bank Lending to East Asia, 1990 and 2000 30 3.1 Opinions as to Whether Intellectual Property Is Well Protected, Selected Economies 39 3.2 Likelihood that the Well Educated Will Remain at Home, Selected Economies 59 4.1 Electronic versus Paper Payments, Selected Years 91 4.2 Connections to Leading U.S. Law Firms, Selected Asian Cities, 1997 93 5.1 Use of the Internet by Fortine Global 500 Companies 107 5.2 Comparison of Number of Cellular Subscriptions with Fixed- Line Telephones, Indonesia, 1996-2001 132 5.3 Investment in Telecommunications as a Share of GDI1, Selected East Asian Economies, 1990-99 135 6.1 Projected FDI, Selected East Asian Economies, 2000 and 2001 145 6.2 Shift of FDI to Northeast Asia, Selected Years 146 6.3 Growth of China's Low- and Medium-Tech Exports, 1990-2000 149 6.4 Degree of Protection in the Telecommunications Sect-or, Selectecd Economies 163 6.5 Effect on Net Interest Margins of Trade and FDI Barriers to Banking Services, Selected Economies 164 Tables 1.1 Research and Development Expenditures, Selected Economies, 1997 7 1.2 Cell Phone Subscriptions, Selected East Asian Countries, 1996-2000 11 3.1 Estimates of TFP Growth, Selected East Asian Economies, Selected Years 34 3.2 Business Environment Rankings, Selected Economies, 2001-02 36 3.3 Enrollment Ratios, Selected East Asian Economies, 1996 40 3.4 Science and Mathematics Achievement Scores at Age 14, Selected Economies 42 3.5 Indicators of R&D Effort and Outcomes, Selected East Asian Economies, Selected Years 43 V 3.6 Perceived Role of Clusters, Selected East Asian Economies 54 3.7 Patents Filed, Selected Economies, 1968-97 and 1992-97 61 4.1 Legal Protection for External Financiers of Firms, Selected East Asian Economies 85 4.2 Financial Restructuring, Selected East Asian Economies 86 4.3 Trade in Services, Selected Economies, 1987 and 1997 90 5.1 Contribution of Computer Hardware to U.S. Output Growth, Selected Years 101 5.2 Decomposition of Productivity Acceleration, United States, 1995-2000 Compared with 1973-95 102 5.3 Contribution of ICT Capital and TFP in ICT Production to Growth in GDP and Labor Productivity, Selected Countries and Years 105 5.4 Contributions of Computer Hardware to GDP Growth in the Republic of Korea and Singapore, Selected Years 106 5.5 U.S. Public Attitudes toward Government Regulation of the Internet 118 6.1 Indexes of the Restrictive Effect of Policies toward FDI, Selected East Asian Economies and the United States 163 vi FOREWORD The recent history of economic development argues against compla- cency-rapid and stable growth is something of a high-wire act that is not easily sustained. Yet a group of East Asian economies did man- age to defy expectations for nearly three decades. The 1997-98 eco- nomic crisis interrupted this remarkable sprint, and even though some of the countries appear to have gotten their second wind, their average growth rates are slower. All economies in the region are in the grip of uncertainty and are casting around for a mix of policies and a political economy that will restore their earlier and effortlessly magical performance. As is becoming increasingly apparent, an investment-led, state-di- rected approach that derives much of the demand pull from exports of manufacturers may not yield the impetus it did in the past. The mid- dle- and upper-income economies of East Asia in particular are past the stage when such a recipe would be sufficient to generate annual gross domestic product growth of 7 percent or more. For most of them the answer to their economic ambitions is likely to lie in more complex and variegated approaches that blend significant institutional deepening with efforts to combine resource mobilization and much greater levels of innovativeness and productivity. How they might achieve this is the subject of this study, which is part of a project on East Asia's future development prospects conducted jointly by the World Bank and the government of Japan. By drawing upon some of the latest research findings, the authors show how the East Asian model responsible for the "miracle" of past decades can be modified vii and augmented so as to return those East Asian economies that perse- vere with reforms to a high growth trajectory. Against the backdrop provided by the 1997-98 crisis and the tenta- tive recovery of East Asian economies thereafter, this study presents the context in which new policies will be deployed, and by analyzing the options the authors indicate how a combination of institutions and polices would buttress the innovativeness that will be the basis of East Asia's competitiveness. The great merit of this book, and one that should recommend it to a broad audience, is that it lucidly summarizes the issues currently be- ing debated and suggests how the discourse can be moved to the next stage and provide East Asian policymakers with the guidance they are urgently seeking. Nicholas Stern Jemal-ud-din Kassum Chief Economist and Vice President Senior Vice President East Asia and Pacific Development Economics The World Bank The World Bank viii PREFACE In late 1999 the government of Japan suggested that the World Bank examine the future directions of economic change in East Asia, with the emphasis on how the approaches to development followed in the region might evolve in the early 21st century. The Bank was highly receptive to the idea of analyzing the longer-term implications of the 1997-98 crisis and assembled a team that worked closely with senior Japanese policymakers and scholars. This team defined a broad-rang- ing study to ascertain the directions of change and to suggest policies for East Asian economies that would sustain their past growth mo- mentum within a changing and more competitive globalizing envi- ronment. This volume is the first of a series of publications that we expect to emerge from the study, and as such seeks to map the terrain that forthcoming books and papers will scrutinize much more closely. East Asia's industrial competitiveness was the stuff of legend in the 1980s and 1990s. Can the region sustain this competitiveness in man- ufactures and extend it to new products and services? This is the cen- tral question this book addresses. Experience suggests that most East Asian economies can rise to the challenge, but success is by no means a foregone conclusion. The 1997-98 crisis revealed a number of ob- stacles that countries will have to surmount. Moreover, their future performance is likely to be inextricably bound with their capacity to implement institutional changes and to build innovation capability, which is increasingly the arbiter of competitive strength. How East Asian economies, especially those in the middle- and upper-income ix groups, might fully realize their potential to innovate will require policy, organizational, and institutional initiatives at many levels. By drawing on a wide spectrum of research, this book delineates the key issues and suggests how East Asian economies might grasp the policy nettles that lie ahead. We hope that for the general reader interested in development, this book will provide an accessible window on fu- ture possibilities and a sense of the complexity of the development ef- fort that encourages deeper reading. We also hope that the specialist will find this a handy and stimulating progress report of events along a broad front together with a look beyond the frontier that con- tributes to the debate on development. The Development Economic Research Group at the World Bank provided us with the ideal environment for conducting the research for this volume, and we are deeply grateful to Paul Collier, the direc- tor of the group, for his support, for providing us with a spacious in- tellectual environment, and for encouraging us to exploit the oppor- tunities offered to the full. Equally vital was the financial backing of the government of Japan through its Policy and Human Resources Development Fund and the access we enjoyed to senior public officials, who gave generously of their time and provided valuable feedback on earlier drafts of this book. We are deeply indebted to Haruhiko Kuroda, Masahiro Kawai, Kiyoshi Kodera, Rintaro Tamaki, Junichi Maruyama, and Takatoshi Ito. The study has been greatly enriched through our many detailed conversations with Eisuke Sakakibara, who has taken a deep interest in the research since its inception. Members of the international steering committee of the study re- viewed the initial draft of this work. Our special thanks go to Masahiko Aoki, Jemal-ud-din Kassum, Homi Kharas, Eisuke Sakak- ibara, II Sakong, Sven Sandstrom, Tharman Shanmugaratnam, Nicholas Stern, Joseph Stiglitz, and Xiaochuan Zhou for their advice and for sharing with us their deep knowledge of the East Asian re- gion. The staff of the World Bank's Tokyo office organized the semi- nars at which drafts of the study were discussed, and we very much appreciate the assistance provided by Shuzo Nakamura, Mika Iwasaki, Tomoko Hirai, and Hitomi Sasaki. Our research efforts were greatly enhanced by Soumya Chattopad- hyay, Farhan Hameed, and Marc Shotten. We thank them for adding x value in many ways. Mir Anjurn Altaf, Kaoru Nabeshima, Shekhar Shah, and three anonymous referees gave us detailed comments on later drafts and contributed to the quality of the final product. Rebec- ca Sugui expertly facilitated the production of the manuscript. The leap from finished manuscript to published book was efficiently or- chestrated on the tightest of schedules by our editors, Nicola Marrian and Janet H. Sasser of the World Bank Office of the Publisher. They good-humoredly introduced much-needed discipline and made sure that what started as a glint in our eyes achieved publication. September 7, 2002 xi ABBREVIATIONS AND ACRONYMS AFTA ASEAN Free Trade Area AMC Asset management corporation APEC Asia-Pacific Economic Cooperation ASEAN Association of Southeast Asian Nations B2B Business-to-business CDMA Code division multiple access DSL Digital subscriber line EU European Union FDI Foreign direct investment G Generation GDP Gross domestic product HDD Hard disk drive HTML Hypertext markup language ICT Information and communication technologies IPR Intellectual property regime ISP Internet service provider MNC Multinational corporation NPL Nonperforming loan OECD Organisation for Economic Co-operation and Development R&D Research and development TFP Total factor productivity TVE Township and village enterprise (China) WAP Wireless Application Protocol WTO World Trade Organization XML Extensible markup language xiii CHAPTER 1 EAST ASIA ENTERS ROUND TWO In 2000 the East Asian economies collectively grew by nearly 6 per- cent, compared with just 1.6 percent in 1998. This performance would seem to provide convincing evidence of the speed and strength of the region's recovery from the 1997-98 crisis. It con- founded some predictions that the severe financial shock was likely to result in a gradual U-shaped revival of business activity, as occurred in Latin America in the 1980s. This growth also appeared to reaffirm beliefs in the robustness of the region's key economies and confirmed the findings of research on 160 crisis episodes, which suggested that countries are likely to return to their precrisis growth rates within two to three years (Park and Lee 2001). By 2001, however, growth rates again faltered as the economic slowdown in the United States and Europe spread to East Asia. Even though growth revived in 2002, the questions remain as to whether East Asian economies have ade- quately remedied the systemic weaknesses that precipitated the crisis, can sustain their competitiveness, and are likely to return to their ear- lier growth trajectory. A NEW DYNAMIC: WHY AND HOW Despite Japan's Herculean fiscal efforts to offset weak consumer spending and revive its economy from a decade-long stagnation, its economic outlook remains uncertain. 'Thus the region's largest econ- omy is providing less stimulus to the rest of East Asia through import demand, foreign direct investment (FDI), and bank credit than in the recent past. Moreover, even if Japan could resolve the problems 2 CAN EAST ASIA COMPETE? caused by banks' nonperforming assets and improve the investment climate, the economy is unlikely to expand by much more than 1.5 to 2.5 percent over the coming decades. The Japanese approach to organizing and managing the economy lost much of its luster during the 1990s. This and the problems the crisis revealed in other economies has undermined faith in the East Asian model, which emphasized growth through resource mobiliza- tion and government-coordination of heavy investment in export- oriented industries. The recent history of East Asia suggests that growth should accel- erate as the tempo of economic activity in the countries of the Organ- isation for Economic Co-operation and Development (OECD) picks up. However, whether East Asia can maintain rapid growth over the medium and long term will depend on how the acquired strengths of individual countries react to institutional reforms and efforts to de- velop knowledge-based sectors. These would sustain a dynamic con- ducive to long-run competitiveness and growth, and would encour- age leading, as well as some lagging, sectors to push toward the frontier of technological possibilities. Much depends, though, on how these measures link with policies to enhance the economies' openness to trade, factor flows, and ideas that contributed much to periods of rapid growth in the second half of the 20th century (Mad- dison 2001). Innovating to Grow The economies of East Asia can be divided into two groups. In the lower- and lower-middle-income category are countries such as Cambodia, Indonesia, the Lao People's Democratic Republic, and Vietnam, whose comparative advantage is in the resource-based and low-tech manufacturing industries. The middle- and higher-income economies range from the Philippines and Thailand at one end of the spectrum to Hong Kong (China), Japan, and Singapore at the other. While some of these middle- and higher-income economies still have substantial low-tech industries in the tradable sectors, they are under increasing pressure to move into middle- and high-tech activities. China straddles these two groups and is emerging as a formidable competitor for economies in both categories as it acquires high-tech EAST ASIA ENTERS ROUND TWO 3 capabilities, while at the same time consolidating its position in low- tech products, forcing the low-income countries to match or exceed China's industrial productivity in order to remain competitive in world markets. In the longer term China's domestic market is likely to offer significant opportunities to its neighbors. Over the near term this market will remain modest for the kinds of products most East Asian economies export, although demand for resource-intensive products from Southeast Asia will increase. China is also a market that is difficult to penetrate for a variety of reasons that have less to do with tariff barriers than with constraints imposed by severe com- petition from local producers, marketing arrangements, logistics, and local regulations. The closer integration of East Asian economies with others in the region and with the global economy-an integration that is likely to increase if trade and capital account liberalization continue-opens the door to fresh opportunities for growth. By the same token East Asian economies, especially the smaller and more open ones, will also be exposed to sharper competition and the risk of shocks. As we see it, the middle-income economies have little to gain by turning back the clock in an attempt to find security in delaying or reversing eco- nomic integration. Such a step would almost certainly slow develop- ment, and while it might have short-run political support from some groups, this would be unlikely to endure as the costs became more apparent. The drawback of temporarily halting reform and liberaliza- tion is that precious momentum is lost, and countries that do so risk being left behind if capital and technology go to their competitors. The middle- and upper-income economies of East Asia are past the stage where increasing resource mobilization and investment is a recipe to embrace for the next decade and beyond. In practice such a strategy, which resembles the one still being followed, could entail a substantial waste of resources. Levels of investment are already high in much of East Asia, and ratcheting up accumulation, even where this is possible, will yield meager, if any, returns. However, given the wide gaps in industrial efficiency and the possibilities for catching up, if not leapfrogging, technology provides the surest avenue to growth. Except for the low-income economies, innovation will be the engine of growth for much of East Asia now that the initial resource-inten- sive phase of industrialization is ending. Innovation in a broad range 4 CAN EAST ASIA COMPETE? of areas, from products to services and business organization, will be the principal source of increases in productivity and in export com- petitiveness. In this study we argue that innovation in three intersecting areas will raise the likelihood that East Asia's middle- and higher-income economies will resume sustainable growth. The first part of our argu- ment is that an environment that stimulates innovation derives from spending on research and development (R&D) by public and private entities in a competitive milieu that, in turn, maximizes the incentives to innovate and commercialize findings by existing firms as well as by new entrants. Networking among firms can promote innovation and efficient resource use and is both more likely to occur and more fruit- ful when many producers agglomerate in urban clusters. The most ef- fective clusters are ones where a few core activities, such as electronics, biotechnology, apparel, and software, are powerfully reinforced by suppliers of inputs and business services, such as financiers, lawyers, marketing specialists, and accountants. Moreover, successful clusters rely on policies supporting openness, which encourage many-stranded links with other clusters worldwide. In other words, product and process innovation is a function not just of investment in R&D, but also of the clustering of networked firms in an open and competitive policy environment. The second and related set of arguments presented in this study focuses on the financial sector and business services. This is because a significant share of future growth will derive from the expansion of services, both traditional ones as well as newly created services; gains in the efficiency levels of the services sector, because these currently lag far behind those of the industrial countries; and closer interaction between services and manufacturing. A strong services sector facili- tates industrial development, and some of the commercially most successful innovation occurs when consulting, information manage- ment, and marketing services, for example, are combined with manu- facturing in the biotechnology sector, automotive, engineering, and apparel industries as well as many others. Efficiency gains and inno- vation depend on competitive pressures and openness, but through much of East Asia, the dynamism of services is predicated on imple- menting policy reforms that stimulate competition, revisiting the role of regulation, and building a larger skill base. EAST ASIA ENTERS ROUND TWO 5 A third area closely interconnected with the other two is informa- tion and communication technologies (ICT). This is important in three respects. First, ICT is a field where a huge amount of innova- tion is occurring and is likely to continue. In the 1990s growth in the higher-income countries was increasingly influenced by [CT-related investment and the productivity gains triggered by advances in ICT. A second reason why ICT is so intrinsic to the innovation process is that by greatly facilitating communication, collaboration, and compe- tition, it has enhanced the productivity of research globally, while at the same time multiplying the pathways for the diffusion of technolo- gy. A third reason for seeing ICT as integral to innovation-led growth derives from the fusing of ICT with key business services. Whether considering finance, marketing, or logistics, the future effi- ciency and competitiveness of these subsectors depends largely on leveraging ICT. The cost, convenience, and accessibility of telecom- munications services will in large part determine the extent to which ICT contributes to innovation. Hence an analysis of these aspects of ICT and the regulation of both telecommunications and the Internet becomes a central part of the ICT story in the context of growth. An innovative economy draws much of its energy from openness and competition. In the East Asian context these are synonymous with trade and capital flows. For close to three decades trade and cap- ital flows, together with resource mobilization, have served as the drivers of growth in the region. Now, as development shifts to a high- er plane, growth fed more by innovation will continue to derive its strength from closer regional and global integration by way of in- creased trade, FDI, and knowledge flows. Research suggests that such integration of the middle- and high-income economies can not only provide the spur of wider markets, greater competition, and faster dissemination of technology, but that through policy coordination it can also enhance macroeconomic stability, further strengthening the incentives for accumulating knowledge with a long-run payoff. Growth through innovation thus depends on the investment in R&D and the environment in which findings are put to effective use; a strong and efficient business services sector; the promotion of ICT as an end in itself and as a means of bolstering both manufacturing and other business services; and the coordination of these with steps to increase the regional integration of East Asia and its ties to world markets. 6 CAN EAST ASIA COMPETE? Technological Capability The contribution of technological capability to competitiveness and to growth is inducing both governments and firms to build the R&D capital essential for developing and assimilating technology. R&D capital entails investment in skills, research facilities, technology ac- quisition, and institutions to promote the deepening of knowledge and to protect intellectual property rights. R&D outlays, which cap- ture some of the accumulation of R&D capital, are rising, reflecting governments' actions to raise spending both directly by using public funds and indirectly by offering inducements to private entities. This increase also demonstrates firms' greater readiness to secure or en- hance their own market positions by embarking on research, whether in-house or through contracts with institutes and universities. In terms of research spending, the middle- and high-income East Asian economies have drawn abreast of OECD front-runners. Japan and the Republic of Korea, for example, both spend close to 3 per- cent of their gross domestic product (GDP) on R&D, with business spending providing most of the impetus (table 1.1). China has been raising its outlay on R&D since the early 1990s, reaching 0.65 per- cent of GDP in 1997, and Malaysia has also increased its spending.2 Thus technological advances are unlikely to be hampered by the overall volume of resources available for this purpose. For the mid- dle- and lower-income countries the constraints will arise from their limited capacity to use available funds productively if research talent is scarce, institutional supports and incentives are weak, and firms are not geared toward commercializing their research findings effective- ly. Hence the significance of other factors described in the following paragraphs. Joint Research Research partnerships have grown as firms in the OECD countries, as well as in East Asia, have multiplied their links with each other and more fully developed and leveraged the potential of universities and research institutes to produce marketable innovations. This is reflect- ed in part by the rise of cross-border ownership of patents and cross- country strategic technology alliances (figures 1.1 and 1.2). Innova- tive activity is further promoted by an upsurge in entrepreneurship, EAST ASIA ENTERS ROUND TWO 7 Table 1.1 Research and Development Expenditures, Selected Economies, 1997 Economy Percentage of GDP Sweden 3.85 Japan 2.92 Korea, Rep. of 2.89 Finland 2.78 Switzerland, 2.74 United States 2.60 Germany 2.31 Israel 2.30 France 2.23 Netherlands, 2.09 Denmark 2.03 Taiwan (China) 1.92 United Kingdom 1.87 Australia, 1.68 Norway 1.68 Canada 1.60 Singapore 1.52 Italy 1.08 Russian Federation 0.95 China 0.65 a. 1996. Source: National Science Foundation (2000). and with it the entry of firms that are important vehicles for introduc- ing new goods and services, production techniques, and mechanisms for building more commercially productive relationships with vendors and customers. However, by intensifying competition, a heightening of innovation can also accelerate the exit of firms whose business mod- els fail the market test. The surge of new entrants in Korea is reflected in the number of workers employed by new firms-approximately 9 percent. The numbers are also rising in Japan and Singapore (Entre- preneurial Fresh Air 2001). In China the collective and private sectors, which account for three-quarters of industrial output, are expanding rapidly. These signs of initiative reinforce accumulated manufacturing capabilities. Industrial Clusters Although the direction of causality between clustering and industrial growth is debated, clustering is associated with specific localization 8 CAN EAST ASIA COMPETE? Figure 1.1 Number of Patents with a Foreign Coinventor, by Sector, OECD Countries, 1980-84 and 1992-96 1,000 900 800 700 600 -o E -_0 Total Biotechnology Information New materials technology O 1980-84 * 1992-96 Note: OECD = Organisation for Economic Co-operation and Development. Source: OECD (2000). economies resulting from the contiguous presence of many firms drawn from a single industry. Clustering spurs innovation and allows firms to benefit from generalized urbanization economies associated with the availability of diverse skills and infrastructure, from the cross- fertilization of ideas and learning, and from face-to-face contacts (OECD 2001a). Numerous clusters already flourish in East Asia and produce everything from buttons to electronics and cane furniture and are a source of continuous, incremental innovation; however, few high-tech clusters exist outside Japan. A prominent example of a high- ly networked and innovative cluster is Hsinchu Park, outside Taipei in Taiwan (China), and similar clusters are developing in China in Bei- jing and Shanghai or are being seeded in the Multimedia Develop- ment Corridor near Kuala Lumpur in Malaysia. An understanding of the spatial dimensions of development, the potential for creativity and learning in cities (Glaeser 1997; Hall 2000), and the factors nurturing urban clusters will be central to development in the 21st century. EAST ASIA ENTERS ROUND TWO 9 Impact of Large Urban Clusters Cities can strongly influence production and productivity by serving as centers for the clustering of industries and services serving domes- tic and global markets. They are also crucibles for stimulating con- sumption, developing tastes, spreading new consumer habits, and en- gendering different lifestyles. As East Asian economies are rapidly discovering, high savings and investment can launch a country's in- dustrialization, and for a time exports can be an important source of demand. Ultimately, however, domestic consumer spending, espe- cially by urban consumers, is vital to sustain stable growth. Moreover, sophisticated and cosmopolitan urban consumers can push local pro- ducers into refining their offerings and also provide the market for testing and launching new products. As populations age, the mix of products people demand will change, and services that provide med- ical care and living assistance, for example, will absorb increasing shares of spending in cities. These trends foreshadow the emergence Figure 1.2 Share of Patents with a Foreign Coinventor, Major OECD Economies, 1985-87 and 1993-95 8 6 p 4 a) tL 2 0 Japan European Union United States E1 1985-87 * 1993-95 Note: OECD = Organisation for Economic Co-operation and Development. Source: OECD (2000). 10 CAN EAST ASIA COMPETE? of economies dominated by services, among which high value added businesses in areas such as entertainment, media, education, and medical care will be the major determinants of growth and employ- ment, as they already are in the United States and Europe. Opportunities for Service Sectors Recent studies by Baily and Zitzewitz (1998) and McKinsey and Company (2001) have underscored the inefficiency of the services sectors of middle- and high-income East Asian economies. In Japan and Korea productivity in such services as construction and retailing is between 40 to 60 percent of the U.S. level. Closing this gap is one of the surest ways to accelerate growth. This lends urgency to the steps being taken by, for example, Japan, Korea, and Malaysia, to consolidate their banking sectors, clean up their loan portfolios, re- capitalize their banks, improve their accounting practices, and strengthen their legal frameworks for enforcing commercial and fi- nancial rules. Deregulating other services and opening sectors to for- eign competition are badly needed. The case for deregulation of the services sector is reinforced by the greater role that producer services will play in firm competitiveness, first, by bringing into existence market-monitoring institutions, such as rating agencies and account- ing and legal services that reliably scrutinize company performance with respect to clearly defined indicators; and second, by facilitating entry through venture capital funds. ICT and Technological Change The growth in and diffusion of ICT permits the rapid and efficient transfer and use of vast quantities of information, impinging on a range of factors significant for future growth. As computerization and con- nectivity expand in East Asia, the role of ICT will widen even more, es- pecially for purposes of e-commerce. In 2000 Korea had the highest proportion of people with mobile phones of any country worldwide (54 percent), while China had the largest absolute number of people using cell phones-85 million (table 1.2). Moreover, as of 2001 individuals whose first language was not English accounted for more than half of all Internet users, with the majority of these being in East Asia. EAST ASIA ENTERS ROUND TWO 11 Table 1.2 Cell Phone Subscriptions, Selected East Asian Countries, 1996-2000 (millions) Country 1996 1997 1998 1999 2000 China 6.85 13.23 23.86 13.30 85.26 Indonesia 0.56 0.92 1.07 2.22 3.67 Japan 26.91 38.25 47.31 56.85 66.78 Korea, Rep. of . 3.18 6.88 14.02 23.44 26.82 Malaysia 1.52 2.00 2.20 2.99 4.96 Philippines 0.96 1.34 1.73 2.85 6.45 Singapore 0.43 0.85 1.09 1.63 2.75 Thailand 1.84 2.20 1.98 2.34 3.06 Source: International Telecommunications Union data. By dramatically accelerating the sharing and diffusion of scientific knowledge, advances in telecommunications and connectivity will do more than increase the rate of catching up by countries able to build a strong base of technical skills. Their active participation in interna- tional scientific discourse is likely to feed creativity and to enlarge the contribution of East Asian economies in pushing the frontiers of knowledge outward. Japan's-and to a lesser extent Korea's and Tai- wan's (China)-success in generating innovations is well established, as evidenced by patent data (figure 1.3). In the high-tech areas Japan remains a leader in consumer electronics and game consoles, for ex- ample.3 It has achieved early successes in wireless telephony intro- duced by NTT DoCoMo, which in a matter of years has built up a subscriber base of more than 40 million customers. Anecdotal evi- dence indicates significant gains by China in a number of fields, in- cluding lasers, biochemistry, electronics, and materials, aided by glob- alizing processes that are drawing Chinese researchers into the first tier of scientific endeavors. These processes include the widening use of the Internet; the researchers' greater regional and trans-Pacific mo- bility; and the establishment of research facilities in China by multina- tional corporations (MNCs), such as Motorola, Philips, and IBM. FDI and Production Networking FDI combined with advances in supply chain management and logis- tics services is resulting in a networked geography of technology transfer and production that more fully harnesses the comparative 12 CAN EAST ASIA COMPETE? Figure 1.3 U.S. Patent Grants, Selected East Asian Economies, 1995 and 2000 35,000 30,000 25,000 20,000 E z 15,000 10,000 5,000 Japan Taiwan (China) Korea, Rep. of El 1995 * 2000 Note: Shows utility (invention) patents. Source: U.S. Patent and Trademark Office database. advantages of individual economies, or rather, of key urban areas within national economies.4 A slicing of the value chain, the dispersal of production to maximize cost competitiveness, and business-to- business (B2B) dealings have been aided by the scaling down of trade barriers and by the use of ICT, which is dramatically lowering trans- action costs associated with paperwork, ordering, fulfillment, track- ing, payment, and even clearing customs. Countries such as Singa- pore pride themselves on the paperless efficiency of their ports and EAST ASIA ENTERS ROUND TWO 13 on the speed with which goods can be processed at points of entry. In addition, production networking and the ever finer subdivision of the production chain are also being propelled by the falling costs of air shipping time-sensitive goods (Hummels 2001),5 as well as by the growing expertise of courier services such as UPS, and of intermedi- aries such as Li and Fung that can design, order, assemble, and deliv- er goods to a final buyer (Magretta 1998).6 East Asia's manufacturing capability, which is concentrated in key urban regions, has a hand in this, but production networking as it is now evolving would be incon- ceivable without the great leap in telecommunications, Internet serv- ices, and air transport capacity.7 Unless reversed by a widespread re- turn to protectionist policies, the stage is set for a tightening of regional interdependence. This, in turn, could enhance the growth- inducing benefits of further regional trade reform, financial market linkages, informal-and eventually formal-institutional harmoniza- tion, and exchange rate policy cooperation. Thus the resumption and persistence of rapid growth in East Asia rests on a new model that relies on the spread of facilitative institu- tions, the industrial and research initiatives that promote innovation, the effective deployment of ICT to build stronger knowledge economies that foster urban clusters and tap the full potential of in- ternational production networks, and a host of trade and regulatory reforms that can create a more integrated regional economy. Suc- cessful economies will be those that are both open and that have cul- tivated the innovative bent, institutional resilience, and growth mo- mentum needed to deflect or absorb shocks (Easterly, Islam, and Stiglitz 2000).' The manifold acquired strengths of the East Asian economies pro- vide the springboard for future development. These include, most notably, their high rates of saving, relative abundance of human capi- tal resources, manufacturing capability, and worldwide industrial rep- utation. In conjunction with strong export demand, these strengths facilitated the V-shaped rebound seen in 1999-2000. They can be the foundation for a new growth trajectory, and they provide some insur- ance that the changes promoting longer-term prosperity-if widely shared-will be implemented, but they offer no guarantees, and bad policies could easily plunge more countries in East Asia into a pro- longed spell of low growth. Indeed, international experience suggests 14 CAN EAST ASIA COMPETE? that consistent rapid growth is the exception. The dominant tenden- cy is for rapidly growing countries to lapse back to slower rates of ex- pansion (Easterly and others 1993; Temple 1999). EAST ASIAN DEVELOPMENT: ROUND TWO Just a decade ago the model of development that enjoyed the widest currency in East Asia was an amalgam of the Japanese and Korean models (see, for instance, Woo-Cumings 2001). The recent financial crisis and Japan's stagnation throughout the 1990s called into ques- tion the relevance of this recipe as a prescription for future economic growth in the region. The rest of this study examines how East Asia can capitalize on its achievements to date and the momentum of the past several decades to take full advantage of the opportunities pre- sented by the growing services- and knowledge-based economy. Chapter 2 provides a backdrop to the future growth outlook in the region by sketching how the role of the state needs to shift from a rel- atively dirigiste approach toward industrialization to one that strengthens market institutions and promotes activities that generate knowledge spillovers. The chapter also shows how the pattern of re- gional trade and capital flows could reinforce governments' efforts to enlarge the contribution of technology. The chapter then highlights some of the constraints to future development, namely, macroeco- nomic instability; rising inequality associated with technological change, globalization, and inadequate safety nets; and tensions relat- ed to Japan's economic weakness and China's growing economic and strategic presence in the region. Chapter 3 examines innovation from three angles: the macroeco- nomic, the firm level, and the transnational. The chapter starts by looking at macroeconomic and sectoral policy that influences invest- ment, R&D spending, and the training of technical personnel with the requisite skill levels. It then concentrates on business manage- ment, firm turnover, and the clustering of businesses in urban net- works. Finally, it explores how the national business environment is modulated by international links among clusters; circulation of knowledge workers among countries; FDI; and innovation in supply chains, which tie together producers and their buyers. EAST ASIA ENTERS ROUND TWO 15 As we argued earlier, a dynamic and innovative economy needs supporting institutions and efficient service providers. Subsequent re- search into the 1997-98 crisis revealed that most East Asian economies need to strengthen their market and financial institutions and to raise the volume and efficiency of core business services. Chapter 4 assesses the institutional gaps, particularly in the financial sector, and the reforms of institutions and regulations that are needed for the growth of innovative services activities. The first part of chapter 5 looks at the contribution of ICT to productivity growth, to innovative activity, and to business develop- ment. The chapter also discusses the main hurdles to the widening use of the Internet for business and research purposes. The chapter then describes how advances in telecommunications are supporting ICT and examines the importance of sound regulation for healthy development. Chapter 6 turns to the external dimension of an innovative econo- my and focuses on the shifting pattern of FDI, the evolution of pro- duction networking, and the likely pace of economic integration. By affecting the degree of openness and market access for East Asian firms, these developments will determine the international market environment that for most East Asian countries has an equal bearing on the course of innovation-led growth. Chapter 7 summarizes the case for the centrality of innovation for future growth in East Asia and weaves together the main policy strands defined in the earlier chapters. NOTES 1. Regarding fiscal policy, Bayoumi (2001) contends that the fiscal stimuli lacked sufficient "real water" and were blunted by the weakness of the banking sector. The volume of spending channeled into construction may also have reduced the multiplier cffect of government outlay on national income. 2. A related issue is how R&D spending is classified and how it is used. Where gov- ernments offer tax incentives, firms will classify some outlays that are only mar- ginally related to research as R&D expenditures, and a significant share of what is classified as R&D is probably more akin to product testing or development. 3. The electronics games industry had estimated sales of US$50 billion in 2001, with Sony and Nintendo dominating the market, and the market is expected to grow to US$86 billion by 2006 (Red HIerring, "The Game of War," October 15, 2001). 16 CAN EAST ASIA COMPETE? 4. See Scott (2001) for recent thinking on global city regions and Hanson and Feenstra (2001) on the relationship between globalization and the development of urban areas as nodes in production networks. 5. Retailers can replenish their stocks of time-sensitive goods more frequently and order in smaller lots to lessen the accumulation of inventories. 6. UPS, for example, now maintains parts inventories at its airline hubs to conduct repairs under warranty for companies such as Dell. This enables PC manufac- turers to better meet the needs of their clients (Kirby 2001). 7. The surge in airport building throughout East Asia is feeding the process. Air- ports in Bangkok, Hanoi, Singapore, Tokyo, and other cities are being expand- ed; new airports have been built at Kansai, Kuala Lumpur, Macau, Pudong, and Pusan; and many more are under construction, for example, China is planning to build 118 new airports over the next 15 years (Far Eastern Economic Review, "Flying High," May 10, 2001; Flight Inte7natio77al, "Traffic Wars," February 20-26, 2001; Oxford Analytica, "Southeast Asia, Airport Rivalry," April 3, 2002). 8. The incidence of shocks has certainly risen since the 1980s, although they seem unchanged in severity and the likelihood of contagion appears no greater (Bordo and Murshid 2000; Bordo and others 2001). CHAPTER 2 THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT T he shift from an accumulation-based model of growth to one that relies more on technological change will be shaped by the broader policy environment-national as well as interna- tional-and by a number of long-term trends. This chapter discusses the state's role in initiating and implementing economic re- forms and creating a climate that is friendly to innovation. It also notes the implications of regional integration, globalization, and the changing salience of major economies in the region. INDUSTRIAL RESTRUCTURING AND THE NEW FACE OF POLICY If innovation and efficiency are to be the drivers of growth, then the state's role in framing industrial strategy in East Asia needs to recon- figured. The role of industrial policy in this region needs to be re-ex- amined for three reasons. First, for most East Asian economies, par- ticularly the middle- and higher-income ones, mobilizing additional resources is no longer a top priority, because the availability of capital has ceased to be the binding constraint it once was. Second, industrialization by way of large, state-sponsored, but pri- vately-owned corporations or by public entities themselves has ex- hausted its potential. Markets are now mature enough to coordinate industrial decisions at least as well as, if not better than, the state can. Aggressively pushing exports of manufactures arouses resistance from importing countries, and the policy may be approaching the stage of 18 CAN EAST ASIA COMPETE? diminishing returns for those East Asian economies where gross ex- ports equal half or more of GDP. The kind of dirigisme practiced by several East Asian states had a hand in accelerating industrialization in the 1960s and 1970s and exports later, but after some early successes, the increasing complexity of the economy blunted its effectiveness (Yusuf 2001). Now government direction is associated with the misal- location of resources; the sheltering of sunset industries; the worsening of corruption; and an industrial structure favoring large, often highly leveraged, conglomerates (Porter, Takeuchi, and Sakakibara 2000). Third, the state-guided approach to development failed to insulate governments from capture by powerful interest groups. In several economies public bureaucracies initially sustained a measure of "em- bedded autonomy" (Evans 1995), but this eroded in the 1980s and 1990s with the comingling of government and business interests that led to the spread of corrupt practices and, in a few countries, to ram- pant crony capitalism. As a consequence, resources were misallocated and banks were systematically looted by owners and debtors, accumu- lating liabilities that now burden taxpayers (Hutchcroft 1999). Re- search by Kaufman, Kraay, and Zoido-Lobaton (1999); Mauro (1995); and Wei (2000) reveals the costs of corruption and of rent-seeking ac- tivities in the form of reduced productivity and growth (corruption slows growth by raising transaction costs and uncertainty and by dis- couraging FDI). State guidance of industrial development and efforts at promoting knowledge spillovers did yield some successes. They widened Korea's industrial base in the 1970s, and in the 1980s in Korea and Taiwan (China) they were responsible for introducing high-tech activities, such as the semiconductor and computer parts industries (Mathews and Cho 2000; Perkins 2001). However, similar efforts to build auto- mobile and other industries in Indonesia (Aswicahyono, Basri, and Hill 2000) wasted large sums and created sectors with political voices that continue to lay claim to public resources, and after 30 years of protection Indonesia's automobile sector has a negative rate of pro- ductivity growth (Okamoto and Sjoholm 1999). Although industrial policy as practiced in the recent past is rightly on the wane, the economies of a technologically dynamic East Asia will continue to demand highly competent and capable bureaucracies that can cope with complex regulation issues and that enable rapid THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 19 technological change by promoting activities that generate spillovers and providing the framework for a competitive business environment. Those East Asian countries such as Japan, Korea, and Singapore, which have meritocratic systems in place for recruiting able individuals, pay them competitive salaries, and offer attractive career paths, need to ensure that they retain these policies and adapt to changing governance arrangements. Other economies need to build bureaucratic capability, not necessarily following a single model, but as in Singapore and Tai- wan (China), going beyond the formal government structure to create task forces or statutory boards to achieve specific ends (Evans 1998). The developmental energies of the state could also keep accelerat- ing the ongoing, piecemeal restructuring of industry throughout the region. Such restructuring is helping to create leaner, more competi- tive, and financially resilient industrial systems, and this change is af- fecting both the public and private sectors. China, Indonesia, and Korea, for example, are privatizing publicly owned firms, and these countries are now prepared to consider majority foreign ownership in key sectors. They are also closing some loss-making public entities that cannot be salvaged. Moreover, the governments' willingness to curtail financial support for private firms and to reshape the public sector have also led to a substantial and potentially far-reaching shakeout of private industrial organizations. Korean chaebol are under pressure from the government, public opinion, the market, and for- eign investors with rising stakes in local businesses to focus on their core businesses or risk being liquidated by their creditors. The de- mise of Daewoo, despite the prolonged rearguard action by the com- pany, its unions, and its banks to save it, testifies to the strength of market forces. The magnitude of Daewoo's debt (US$74 billion) and the management practices leading to this debt buildup serve as a warning of how much damage can be wrought when market and reg- ulatory vigilance is relaxed.' Elsewhere firms and banks are seeking alliances; are being taken over by foreign companies, for example, Natsteel Electronics, one of Singapore's leading producers of semiconductors, has been acquired by Flextronics; or, as in Korea, Malaysia, Singapore, and Thailand, are being required to merge by the authorities to form, apparently, more optimally sized units. Even in Japan, the decade-long recession has triggered a change in industrial policy and organization that is 20 CAN EAST ASIA COMPETE? likely to prove irreversible. The network of relationships among the government, politicians, banks, businesses, and workers has been sub- jected to public criticism and is being partially unstitched. The im- mediate effect has been to weaken keiretsu relationships and lifetime employment obligations,2 to increase foreign shareholdings, and to weed out some of the weaker firms. However, the longer-term conse- quences are likely to reach deep into Japan's economy and society as the manifest weaknesses of Japan, Inc. are forcing bureaucrats, firms, and most important the public to scrutinize the key interrelation- ships, to acknowledge their drawbacks, and to consider alternative ways of doing business. Restructuring of the economy is proceeding along several different axes. Samsung Electronics, for example, is 40 percent foreign owned, and shareholders' views now count. Even Malaysia is ready to sell part of the state-owned PROTON automobile company in the inter- ests of gaining technology, capital, and global reach. Similar trends are apparent in the transport, power generation, and utility subsec- tors, all of which have been opened to foreign investment. In China thousands of small and medium state-owned enterprises have been divested, and the government is prepared to sell its stake in some of the larger state-owned enterprises if buyers are forthcoming. The shift away from industrial policy and public ownership and to- ward foreign acquisition of firms and banks is a welcome one, and this shift is prompting a number of associated changes in the areas of banking, accounting, corporate governance, and company law, all of which will strengthen market institutions and redefine the role of the state. The pull of inertia remains strong, however, and a marked re- luctance to let market forces determine outcomes persists, especially when large companies such as Hynex (formerly part of the Hyundai group) in Korea and major banks, as in Japan, are faced with painful choices to merge with foreign partners, to restructure their opera- tions, or to face liquidation as their credit runs out. The 1997-98 crisis has driven home the need to create banking sectors that are independent, efficient, profitable, and innovative. In time this could revolutionize the relationship between governments and banks, several of which (in Indonesia, Korea, and Thailand) were taken over by their governments after the crisis. The crisis pushed governments to eschew directed lending, a prominent feature of the THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 21 East Asian model, which is responsible for a sizable share of the non- performing loans (NPLs) weighing down bank portfolios and limit- ing the supply of credit to potential entrants. The crisis also clearly underscored the need for strict adherence to capital adequacy rules, which had been largely implemented throughout the region by 2001, and begin the inevitably time-consuming processes of privatizing state-owned banks; improving corporate governance, accounting standards, rules of disclosure, and legal remedies; and monitoring bank performance using adequately trained public supervisors and private rating agencies. The entire ongoing effort to recapitalize banks, clean up their loan portfolios, and dispose of their underpinning corporate assets has served to broaden the reform effort to include the business sector. As a result, governments are starting to address competition policy, includ- ing rules for mergers and takeovers, and relationships between banks and their corporate customers. Ultimately, it will be the depth and durability of these institutional changes, which in several instances ex- ist only on paper, that will decide whether East Asian development ac- quires a new dynamic appropriate for the early 21st century.3 CONTEXT FOR REFORM IN EAST ASIA In the past the economic change in East Asia was associated with the growth of trade; FDI; and more recently multilateral and bilateral agreements governing trade barriers, corporate law, and intellectual property rights. Collectively these are increasing East Asia's global integration and are continually raising the extent of intellectual agreement about the gains that can result from openness and about ways to improve market functioning. By participating more exten- sively in international trade and seeking to enlarge the share of FDI, East Asian countries have adopted market-based rules and lowered tariff and nontariff barriers since the mid-1980s (figure 2.1). Trade, Foreign Direct Investment, and Regional Networks The desire to attract FDI has also contributed to industrial restructur- ing by increasing foreign penetration.4 This has accelerated since the 22 CAN EAST ASIA COMPETE? Figure 2.1 Trade Barriers, Selected Country Groups and China, 1984-87 and 1994-97 Tariffs Nontariff barriers 0)40 30 tD3- 20 1984-87 1994-97 Post-Uruguay 1984-87 1994-97 Round * Newly industrialized economies 0 China [ Southeast Asia Note: As China did not join the World Trade Organization until 2001, no post-Uruguay Round tariff rate applies to its imports. Source: Dowling and Ray (2000). 1997-98 crisis. Although countries such as Malaysia seek to defer low- ering tariffs on certain imports, such as automobile parts, the region as a whole remains committed to freer trade (A New Front 2002; South- east Asia 2002); however, alongside multilateral reform, economies in the region have moved toward bilateral treaties and regional agree- ments that could divert trade. The bilateral treaties being negotiated, even byJapan, are a response to regional trade agreements springing up in the Americas and Europe and to the temporary stalling of multi- lateral negotiations post-Seattle. While the bilateral agreements could become building blocks to freer multilateral trade if they adhere to World Trade Organization (WTO) rules and eliminate any trade bar- riers included in bilateral agreements for all trading partners in a stip- ulated period, engender regional support for particular issues, and serve as models for broader agreements, they could also divert East Asia away from arguably more fruitful multilateral initiatives (East Asia 2001). For the past three decades trade, FDI, portfolio flows, and tech- nology flows contributed to East Asia's prosperity by reinforcing the push provided by factor inputs, and their importance could increase in the next phase of development. Pressure from competing imports, THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 23 for example, could raise the productivity of manufacturing industries (Lawrence and Weinstein 2001 showed how import competition has had a stronger effect on total factor productivity [TFP] in Japan and Korea than export growth). Furthermore, as countries build up their skills base, the technological diffusion from FDI and productivity gains could be magnified. Similarly, high-tech products and services will require expanding export markets to reap productivity gains and attract FDI (Choudhri and Hakura 2000; Parker and Lee 2001). It is this conviction, further buttressed by the advances in ICT, that is likely to sustain reform. Six economies-China, Indonesia, Malaysia, the Philippines, Sin- gapore, and Thailand-absorbed almost the entire flow of FDI to the region (90 percent), but what is notable is the scale of intraregional FDI-60 percent-most of it originating from Japan and the four newly industrializing economies (note that much of the US$64 bil- lion in FDI in Hong Kong [China] flowed out to China, tightening the links between the two economies). This reflects the increasing dependence of Japanese, Korean, and Taiwanese firms on their East Asian affiliates for intermediate products. Thus forJapanese firms the share of East Asia in affiliates' total procurement and total sales rose from 51 and 29 percent, respectively, in 1986 to 55 and 32 percent in 1995. Other East Asian economies have begun procuring more from the region as well as exporting more to Japan and to each other (Parker and Lee 2001; Urata 2001). Aside from intraregional FDI, global production networking also attracts a large volume of FDI from outside the region. With prosperity so firmly rooted in the external sector and with in- dustrial development, as well as technological advances, paced by in- flows from abroad, FDI inevitably leaves a strong imprint on the econ- omy. In a matter of years foreign investors have acquired more than 40 percent of all shares on the Korea Stock Exchange. As a consequence policy, institutions, and attitudes are more exposed to demonstration effects emanating from trading partners, with Japan and the United States in the forefront. As long as Japan offered an alternative econom- ic model of development and the international environment was hos- pitable to authoritarian regimes, institutional and policy convergence toward a Western model could proceed at a leisurely pace. East Asia 24 CAN EAST ASIA COMPETE? could pursue its own brand of development policies and business prac- tices and cultivate distinctive "Asian" value systems. These aspects of the East Asian model are now losing their appeal, and not just because the relative attractiveness of the U.S. model for the business communi- ty has risen (Rozman 2002). With globalization, pressure to institu- tionalize the tightening of international links through bilateral and multilateral agreements has increased. These agreements inexorably nudge countries to implement reforms that more firmly entrench the institutions of a market-based system and that strengthen competitive- ness through adherence to formal rules (though smoldering resistance remains, most recently visible in Korea and Thailand, and the threat of a backlash against globalization cannot be dismissed). Observers differ in their opinions about the degree of institutional convergence occurring or likely to occur. Some such as Ohmae (1990) perceive the coming of a borderless world manifesting increas- ing homogeneity. Others see a process of hybridization at work as countries selectively borrow and adapt institutions while maintaining their distinct identities. Still others feel that East Asia's successful business models, labor market institutions, and social relationships are resilient and will ride out the shocks of globalization with only minimal changes (Berger and Dore 1996). We believe that cultural borrowing and hybridization, plus some evolution, accounts for past successes in Japan and the United States, and that East Asia would benefit from approaching reform in this spirit. The pace of change in East Asia, parts of South Asia, and Latin America depends in no small part on the institutional infrastructure supporting globalization and regional integration. This infrastructure comprises rules governing interaction among countries in many spheres, most notably political and economic. It is sustained by such bodies as the WVTO, the World Bank, the International Monetary Fund, the International Labour Organisation, the Bank for Interna- tional Settlements, the United Nations Development Programme, and others operating at the global level, together with bodies such as the Asian Development Bank and the Association of Southeast Asian Nations (ASEAN) that operate on a regional scale. Participation is "owned" by governments, which give the organizations their legiti- macy and determine what authority they will have. THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 25 This infrastructure is. instrumental in shaping thinking on policy and underpinning global integration through trade and capital flows. Although many barriers to trade in East Asia remain, they are much sparser than they were 15 years ago. Multilateral agreements and cer- tain bilateral trade pacts are lowering barriers, which affects a host of other policies. Particular groups find that demanding special treat- ment is becoming harder and governments find that granting special treatment is more difficult when formal commitments curtail specific types of interventions. By participating in international agreements, governments can more easily deflect pressure from domestic interest groups. Policy reversal also becomes much less likely. While global integration has progressed furthest in the sphere of trade, and the V/TO/General Agreement on Tariffs and Trade rounds have probably contributed more to globalization than any other institutional developments, by altering the content of key poli- cies, trade liberalization has influenced the tenor of other institutions and induced change toward increasing openness, enhancing market competition, and protecting intellectual property rights. An entire cohort of policymakers is now in place that views competition, open- ness, and global integration as intrinsic to development, and while they face resistance from often quite powerful groups, the dissolution of regional and multilateral agreements and the dismantling of inter- national institutions seems unlikely. Countries in East Asia and other regions will change at their own pace, but in every instance the trends point in the same direction. The change sweeping through East Asia over the past decade is likely to persist for another important reason: the international insti- tutional infrastructure is only the visible part of a much more elabo- rate, albeit less formal, mesh of relationships that brings together policymakers, business people, members of professional groups such as lawyers and accountants, journalists, and opinion makers of all stripes. As a result, a continuous exchange of opinions and views takes place, feeding policies and institutions, both national and interna- tional. East Asia's dense and proliferating trading links mirror an equally rich and complex system of relationships binding transnation- al communities drawn from state bureaucracies, academic institu- tions, and the large pools of knowledge workers and business people. 26 CAN EAST ASIA COMPETE? Networks of nongovernmental organizations are also becoming an active part of these communities. Personal contacts vitally supple- ment electronic communication, and this mix of commentary, obser- vations, and exchanges is then amplified and diffused by the media. East Asia is an ethnically and culturally heterogeneous region, but following the lead of the Chinese diaspora, bureaucrats, business peo- ple, and academics have been remarkably quick to coalesce into infor- mal groups that facilitate the circulation of information, ideas, and opinions on the central issues confronting the region. Role of Macro Policies, Social Safety Nets, and Key Economies Three counterdevelopments could slow or reverse the tendencies sketched out here. We will note their importance without tracing their likely effects, as to do so is well beyond the scope of this study. These are the region's macroeconomic policies; the sharing of the benefits as well as the burdens of the changing economic order; and the future role of the region's two major economies, China and Japan. Given the large literature that exists on each of these issues, a few ob- servations must suffice (for a more detailed discussion of issues relat- ed to inequality and safety nets, governance and corporate restructur- ing see World Bank forthcoming). Macroeconomic stability and competitiveness are ultimately a function of governance, and countries such as China, Indonesia, and the Philippines face many difficult choices in this area. These diffi- culties are compounded in open economies, where exchange rate fluctuations can have adverse, economywide effects. The debate on whether to fix or float exchange rates and on the use of inflation tar- geting under a regime of flexible rates, while instructive, has failed to define clearly the choices individual countries face. On this subject policymakers will have to learn by doing, using early warning indica- tors and international coordination mechanisms to minimize risks. A widening of inequality is a distinct possibility, although as yet the evidence is equivocal (figure 2.2). Globalization is one factor affecting income distribution;5 others are technological advances that benefit knowledge workers and the changing demographic profile of some East Asian economies.6 These will compound existing regional in- come divergences that are already a source of tension in several East THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 27 Figure 2.2 Inequality, Selected East Asian Nations, Selected Years 50 *~40 0 EF 30 20 1989 1995 1992 1998 1993 1998 1988 1997 1993 1996 Malaysia Thailand Vietnam Philippines Indonesia Note: For comparison, the Gini coefficient for China took a value of 40.3 in 1998. Source: World Bank, World Development Indicators, various years. Asian economies. The winners are likely to be geographically mobile skilled workers, who will enjoy access to better paying job opportuni- ties. However, migration from the interior to coastal cities will also benefit the unskilled in countries such as China and Vietnam. Unlike the OECD countries, with the exception of Japan and Sin- gapore East Asian economies have allocated few resources for in- traregional or intragroup transfers. Poorer regions are mainly left to fend for themselves, and individual households rely to a considerable degree on self-insurance and their extended families. Under more democratic political regimes, rising incomes and openness will proba- bly entail either a significant upheaval of public finances to cover the costs of government-mediated transfers or public regulations to sup- port the private provision of insurance. If East Asia follows OECD trends, governments may need to raise and channel a large volume of resources into transfers, especially if tackling intracountry inequali- ties becomes a political imperative in countries such as China and In- donesia. This has implications not just for tax and expenditure poli- cies, but also for the development of financial markets to absorb social insurance funds efficiently. Rising transfer payments have dis- incentive effects and impose deadweight losses that detract from growth. Deeper pools of capital and more effective financial markets, however, will facilitate the entry of firms and innovation, as discussed in chapters 3 and 4. 28 CAN EAST ASIA COMPETE? Aging populations in China, Japan, and Korea will generate demand for larger transfers for social security and expenditure on medical bene- fits. Recent forecasts prepared by the OECD for its members convey a sense of how these costs could rise, and such costs range from 15 per- cent of GDP in the United Kingdom in 2050 to 35 percent of GDP in Denmark, with most of the countries having to devote 25 percent of their GDP to this purpose. Moreover, in the East Asian economies people in older age groups are acquiring the political weight to demand a shift in the composition of public expenditures, first, to ameliorate the risks arising from greater economic openness and from the indus- trial restructuring now reducing the size of the public sector, and sec- ond, to supplement private provision for pensions and health care (Garrett 1998; Iverson and Cusack 2000; Rodrik 1998). The age profile of Asian countries raises other issues. Bloom and Williamson (1999) showed that up to a quarter of East Asia's prior growth could be explained by the rapid increase in the work force, which acted as the source of the region's dynamism, entrepreneur- ship, and higher savings. Such expansion has ended in Japan, and will begin to taper off first in Northeast Asia, and later in Southeast Asia. Thus starting with Japan, the region will gradually lose its demo- graphic growth bonus. So long as the labor force continues to in- crease, however, the work of Beaudry and Green (2001) suggests that countries could experience more rapid absorption of technology and higher rates of employment, but slower rates of wage increase. Two decades from now, demographic effects that result in a slowed in- crease in the size of the labor force, rising average age of workers, de- clining rates of saving, and higher social security payments could erase the current growth advantages from this source. Japan will have to cope with this situation by the end of the decade (Adams and Gangnes 2001; Mason and Ogawa 2001). Nevertheless, a trend re- duction in growth is by no means inevitable, and could be offset by further technological change and by equipping older people with fresh skills and extending their working lives,7 along with rapid moves to set up fully financed pension schemes. Employing an aging work force productively through lifetime learning and suitable incentives will be a major test for Northeast Asia, starting with Japan. During the latter part of the 20th century two countries exercised the greatest political, economic, and social influence on the development THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 29 of East Asia: Japan and the United States. Overall, the United States has been, and for the foreseeable future is likely to remain, the more influential of the two, but Japan's economic links with the rest of Asia, supported by economic diplomacy and official assistance, have deci- sively affected industrialization, not least through FDI and associated trade flows, infrastructure development, and technology transfer. Fur- thermore, other nations have widely imitated aspects of the Japanese economic model, based on their view that it was more suited to East Asian societies. However, as figures 2.3 and 2.4 show, Japan's declining economic prowess has markedly affected its economic relationships with other East Asian economies via trade FDI and credit flows. At the same time Japan's ability to project an alternative model of growth has also waned. In the second half of the 1990s, East Asian economies and the United States were contending with the emergence of China as a Figure 2.3 Japan's Declining Influence in East Asia, Selected Years Share of exports from East Asia, 1980 Share of exports from East Asia, 2000 - 20%/ 12% J1 To Japan * To the rest 80% of the world 88% Share of FDI in East Asia, 1990 Share of FDI in East Asia, 2000 40% 7% El From Japan , From the rest 60% of the world 93% Note: FDI = foreign direct investment. FDI excludes Hong Kong (China). Source: JETRO (1996, 2001); International Monetary Fund data. 30 CAN EAST ASIA COMPETE? Figure 2.4 Allocation of Japanese Bank Lending to East Asia, 1990 and 2000 1990 2000 59% 33% El[ East Asia * The rest 41% of the world 67% Source: Callen and McKibbin (2001). fierce competitor in the sphere of trade, a magnet for FDI, and a re- gional power whose concerns-and claims in the South China Sea- must be factored in by all other parties. China's economic weight and political presence will surpass that of Japan in 15 to 20 years and will pose a serious challenge to that of the United States. This transition could be an exceedingly delicate one, and will involve major adjust- ments by all countries in the region, including China. Given its size and gathering industrial competitiveness, China's neighbors will need to scramble to retain an edge. Many industries in Japan will also be- come susceptible to competitive pressure from China. China's ex- panding military power will further complicate the geopolitical situa- tion in the region and could heighten perceptions of a strategic threat among certain circles in the United States. One possible scenario is a pooling of economic interests that leads to a mutually advantageous integration of countries in the region, with China displacingJapan as the main engine of growth (Enough for Everyone 2002). We explore the possibilities for such integration in chapter 6. Other less positive scenarios cannot be ruled out, however. How East Asia integrates and how countries work together to co- ordinate their policies and institutions so as to sustain growth with stability will become clearer over the coming decade. Throughout this period Japan will remain, at least in size, East Asia's dominant economy, and its economic performance and leadership will affect re- gional integration and growth. Meanwhile, now that China has ac- ceded to the WTO and is beginning to implement the agreed on THE NATIONAL AND REGIONAL DEVELOPMENT CONTEXT 31 terms, its trading and other links will intensify. A rapidly growing and more open Chinese economy that enlarges market opportunities in East Asia could be advantageous for the other countries in the region (Lardy 2002; lanchovichina, Martin, and Fukase 2000), but this de- pends on how competition and coordination evolve and on how East Asian countries diversify their comparative advantage through meas- ures to stimulate productivity and innovation in leading sectors (A Panda Breaks 2001). Since the crisis, East Asian countries have made some difficult choices, but even harder ones remain to be made if the region is to embrace a new dynamic and loosen the grip of political and institu- tional rigidities. Japan's attempt to restructure its banking industry during the 1990s and China's 20-year effort to reform its state sector and banks are but a few examples of institutional inertia and the ca- pacity of large, inefficient industries to mobilize political support to resist change for long periods, even in the face of severe market pres- sures. Those East Asian economies that defer institutional reforms and are slow to restructure their industry will see their dynamism ebb. Foreign capital and the advanced export industries will move elsewhere. Southeast Asia had a warning of this in 1999-2001. Coun- tries willing to change policy direction can use their capacity to mobi- lize domestic resources, their elastic supply of human capital, their reputation in export markets, and their participation in international production networks to enter a new phase of rapid growth. The vir- tuous spirals of tomorrow will have a different sectoral focus than those of yesterday and will be powered by innovation and increasing allocative efficiencies throughout the economy. These are precisely the kinds of opportunities that East Asia, because of its history of eco- nomic achievement, is well positioned to grasp. NOTES 1. A billion is 1,000 million. 2. On balance, Japanese firms have avoided major layoffs that would have disrupt- ed established labor practices. Instead they have absorbed weakening demand by hiring fewer young workers and by transferring employees to subsidiaries (Kato 2001). Nevertheless, the number of employees on regular contracts providing for full-time employment over an indefinite period fell fromn 77 percent in 1996 32 CAN EAST ASIA COMPETE? to 72 percent in 1999, and nonstandard work arrangements are becoming in- creasingly common (Oxford Analytica, "Japan: Labor Market," March 1, 2001). 3. The United States took close to five decades, starting in the 1880s and under the prodding of British investors in railway bonds, to establish institutions re- quiring the disclosure of financial information, the auditing of accounts, the strengthening of minority shareholders' rights, and the monitoring of industrial concentration (Bordo, Eichengreen, and Irwin 1999). Even today, U.S. regula- tory agencies and courts struggle to cope with the volume and complexity of the workload generated by financial regulation and antitrust policies, not to men- tion the problems posed by the esoteric nature of new business activities associ- ated with the ICT revolution. The U.S. savings and loan crisis during the early 1980s demonstrates how difficult it is for regulators to deflect overtures from their clients and to limit forbearance. Thus the East Asian economies face a con- siderable challenge in meeting the political and administrative costs of trans- forming corporate governance, legal systems, and business cultures to comport more closely with global norms when benefits will obviously be slow to materi- alize. All the countries are struggling to raise their institutional standards, if only to remain competitive in a globalizing world. 4. The developing countries received US$240 billion in FDI in 2000. Of this total US$141 billion went to developing East Asia, of which $41 billion went to Chi- na (UNCTAD 2001). 5. Although global inequality has risen over the past two centuries, Lindert and Williamson (2001) showed that globalization as such may not be the culprit, and that other factors account for the increasing divergence. 6. See, for instance, Behrman, Birdsall, and Szekely (2001), who show that reforms in Latin America are affecting inequality more through technology than through trade. 7. Boersch-Supan (2001) analyzed the German experience and underlines the sig- nificance of further education in offsetting the effects of aging through in- creased productivity. CHAPTER 3 THE WARP AND WEFT OF INNOVATION T hroughout the 1970s and 1980s, a consensus existed regard- ing the primacy of factor inputs, especially capital, as the principal determinants of growth, with total factor productiv- ity (TFP) providing a small, but steadily expanding, share of the impetus. More recently the contribution of TFP to past East Asian growth has been extensively debated. According to some estimates, the growth of TFP in some of the most rapidly growing economies was negligible (table 3.1). Although others have contested this ex- treme position, without minimizing the ingenuity of firms throughout East Asia in assimilating and adapting new technologies, most esti- mates show that the proportional share of TFP in East Asia (excluding Japan) is still well below that observed in the industrial countries.' The results of such analyses, while they leave some methodological and econometric questions unanswered, clearly indicate that technological assimilation, innovation, and gains in productive efficiency based on an accumulation of skills and tacit knowledge are among the principal vehicles for potential future growth (Easterly and Levine 2001; Robertson 2000; Temple 1999). The growth potential inherent in technological change scarcely needs reaffirming. It is central to growth theory, and the empirical documentation is both voluminous and convincing. What we lack is a reliable means of translating a widely shared aspiration into a self- sustaining process. In other words, economies or firms seeking to harness their future growth to technological advances must still rely on trial and error. However, research is reducing the element of chance, while at the same time revealing the complexity of successful innovation systems. In this chapter we present what are arguably the 34 CAN EAST ASIA COMPETE? Table 3.1 Estimates of TFP Growth, Selected East Asian Economies, Selected Years (percent per year) Adjusted Young Collins and Young estimates Bosworth Sarel (1994, 1995), (1994, 1995), (1996), (1997), Economy 1966-90 1966-90 1960-94 1978-96 China 4.6b Hong Kong (China) 2.3 2.4' Indonesia 1.2d 0.8 1.2 Korea 1.7 1.3 1.5 Malaysia l.1d 0.9 2.0 Philippines -0.4 -0.8 Singapore 0.2 1.0 1.5 2.2 Taiwan (China) 2.6 1.9 2.0 Thailand 1.5d 1.8 2.0 Note: TFP = total factor productivity.The column heads show the publications and the years they covered. a. Adjusted Young estimates use the revised factor share weights with capital assumed to have a weight of 0.35. b. 1984-94. c. 1966-91. d. 1970-85. Source: Crafts (1998). necessary elements of an innovation system appropriate for East Asian economies. These range from national-level policies to impart macroeconomic stability, protect intellectual property, increase re- search expenditures, enhance the supply of skills, and reduce the en- try barriers for firms to the emergence of clusters that support local and global networks with global links, further strengthened by ICT and the building of international supply chains. Innovation will be spearheaded mainly by firms making decisions shaped by the overall business environment. The universe of innova- tion at the beginning of the 21st century looks quite different from what it was a decade ago. This is not because the tempo of innovation has accelerated substantially, although product cycles have shortened, the speed of commercialization has increased, and the volume of patents granted for software is much larger. More significant is the shift in the composition of innovation: where it is done and how it is done. Of the four main types of innovation-product, process, serv- ice, and transactional-ICT has enlarged the role and penetration of THE WARP AND WEFT OF INNOVATION 35 the latter two types. Innovation increasingly stems from research in universities and firms concentrated in clusters that feed off knowl- edge spillovers from the diverse activities and intellectual energies generated by networks. Its cost and complexity demand that innova- tion be done cooperatively, both locally and internationally, while ICT has reduced the cost of doing so.' Research on industrial and emerging economies reveals no single and clearly superior means to achieving greater innovativeness. Many different factors appear to contribute, and varying mixes of these fac- tors can give good results. Two findings deserve emphasis. First, in- novation depends on private and public measures across a broad front and not just on increased R&D or on skill development, although both of these are necessary conditions. Second, innovativeness does not automatically guarantee rapid growth. Japan remained capable of great innovativeness in certain manufacturing subsectors during the 1990s-although the growth of this capacity slowed relative to the 1980s-without this being reflected in either its productivity statistics or its growth (Branstetter and Nakamura 2002; Posen 2001).3 This chapter examines the range of interrelated factors driving innovation by focusing on eight proximate determinants of innovation, namely: * The building of R&D capital * The business environment, including ease of entry by firms, level of competition, and protection of intellectual property * The effectiveness of the education system in producing an ade- quate supply of skilled and technical workers * The links among businesses, universities, and public and private re- search institutes that stimulate innovation and its commercialization * The interaction among firms and agglomeration economies in in- dustrial clusters * The extent of technology generation and absorption by firms through their own R&D, licensing, FDI, assistance from lynchpin buyers in a production network, new equipment purchases, and support from equipment or component suppliers * The degree of access to an international pool of professionals and to centers of excellence in East Asia and the West * The development status of production networking, supply chain management, and logistics. 36 CAN EAST ASIA COMPETE? For a more in-depth discussion of these determinants, see World Bank (forthcoming). ROLE OF THE BUSINESS ENVIRONMENT East Asia's past economic performance relative to that of other devel- oping regions highlights the significance of the business environment as a necessary condition for rapid, outward-oriented industrialization. The business environment is a broad term, but for our purposes it refers to macroeconomic stability, openness, effective public policy management, and a reliance on market institutions to achieve devel- opment objectives. Combined with judicious regulation, these factors induce businesses to make long-lived investments and to innovate and strive after competitiveness in domestic and foreign markets. Several East Asian economies-with Hong Kong (China), Korea, Singapore, and Taiwan (China) being among the foremost-have cre- ated attractive business environments (table 3.2), and most main- tained the fiscal and balance of payments discipline essential for con- taining macroeconomic imbalances through much of the 1990s. Although this discipline weakened in Korea and Thailand prior to the crisis, macroeconomic stress indicators have been restored to normal.4 Most East Asian economies have welcomed FDI, while also Table 3.2 Business Environment Rankings, Selected Economies, 2001-02 Rank Economy Rank Economy 1 Finland 14 Denmark 2 United States 15 Switzerland 3 Canada 16 Iceland 4 Singapore 17 Germany 5 Australia 18 Austria 6 Norway 19 Belgium 7 Taiwan (China) 20 France 8 Netherlands 21 Japan 9 Sweden 22 Spain 10 New Zealand 23 Korea, Rep. of 11 Ireland 24 Israel 12 United Kingdom 25 Portugal 13 Hong Kong (China) Source: World Economic Forum (2001). THE WARP AND WEFT OF INNOVATION 37 promoting exports. In recent years they have been reducing nominal tariffs on imports, a step that can stimulate industrial productivity, because import competition forces out inefficient firms and induces those that remain to raise their productivity (Lawrence and Wein- stein 2001). Furthermore, increased trade with technologically ad- vanced countries can promote productivity through the importation of embodied and disembodied technology, although large countries are the main beneficiaries (Coe, Hoffmaister, and Helpman 1997; de la Potterie and Lichtenberg 2001).5 Building R&D Capital and Intellectual Property In chapter 2 we noted the rise in R&D expenditures in several East Asian economies and the pace of innovation as displayed by patent sta- tistics. While spending on research does not ensure innovation or rapid growth rates, such spending is a necessary condition. In the absence of an adequate volume of R&D outlay a country will most likely not as- cend the technology ladder. The experience of the OECD countries, especially that of the United States, suggests that private and social re- turns from R&D can be substantial. Private returns have been estimat- ed in the range of 10 to 20 percent, with social returns ranging from 20 to 60 percent (Griliches 1992). If 25 percent is the median social rate, then a permanent increase in R&D spending equivalent to 2 percent of GDP could raise growth by 0.5 percent. A more conservative estimate by the OECD points to a 0.05 to 0.15 increase in output for every 1 percent increase in the stock of R&D. The evidence also shows that more of the R&D in the upper-income countries is used to push the technology frontier outward, whereas, predictably, in the middle-in- come countries it assists in technology assimilation (OECD 2001a). Furthermore, decentralized research activities seeking to derive codifi- able findings from an existing body of knowledge yielded more fruitful outcomes in terms of usable results in the middle-income countries. Advancing basic science, methodology, or tacit knowledge generally calls for a strong accumulated base of knowledge and a critical mass of skills that favors the advanced countries with long track r ecords in re- search (Feldman and Lichtenberg 1997). Thus the empirical work to date indicates the high potential value of investing in R&D. When the conditions are right it can lead to a 38 CAN EAST ASIA COMPETE? flow of innovation that promotes growth, but as the experience of Japan in the 1990s again suggests, such innovation must be broadly based and supported by a mix of demand and institutions, such as an intellectual property regime (IPR), that permit innovation to feed into growth. With the exception of Japan, Korea, and Singapore, the other economies have yet to move beyond a rudimentary and weakly enforced IPR, which, as in the case of Taiwan (China), can affect FDI inflows into the software industry and domestic initiatives to develop packaged software. This influences domestic innovation in the longer term and the transfer of technology from abroad, especially in the ICT sector.6 Once the industrializing East Asian economies begin marketing their own innovations, IPRs will become increasingly im- portant as a safeguard against piracy in foreign markets. IPRs have many variants, and given its stage of development, each country must design a regime that keeps its own interests in the fore- front. Now that several East Asian economies are signatories of the WTO's Trade-Related Agreement on Intellectual Property Rights, which sets a minimum threshold, the process could accelerate. For most the task is large and urgent. It involves translating a formal in- tention into viable domestic institutions that send clear signals, pro- vide the appropriate incentives, and build as well as effectively har- ness the legal infrastructure to protect ownership rights. The increasing importance of innovation in the ICT sector and of trade in ICT products has drawn more attention to the issues pertaining to intellectual property. It has also revealed the gaps in regulatory capa- bility and legal infrastructure more starkly, not just in East Asia, but also in the United States and European countries (see figure 3.1 for a subjective ranking of these institutions). East Asian countries such as Japan, Malaysia, and Singapore are tackling both competition and in- tellectual property concerns on the legislative plane, as well as at the practical level of enforcement, but overall, progress toward more ro- bust IPR regimes is slow. The Education System as a Fulcrum of Research The supply and quality of skills available influence the search for new technology; the monitoring of information about scientific discover- ies around the world; and the assimilation of relevant local technology, THE WARP AND WEFT OF INNOVATION 39 Figure 3.1 Opinions as to Whether Intellectual Property Is Well Protected, Selected Economies (7 = strongly agree, 1 strongly disagree) 7 6 Note: The score reported is the mean assessment of a survey of an economy's business people. Source: World Economic Forum (2000). whether from the spillover effects of FDI, from supplying original equipment manufacturers, from imports of capital equipment, or from licensing. The key to absorbing technology, a task that general- ly rests on a country's indigenous R&D effort, is the availability of human capital and the readiness to fully master new products and means of production (Forbes and Wield 2000; Goto and Odagiri 1997; Hobday 1995; L. Kim 1997; Lan 1996; Lau and Higuchi 1999). The endogenous growth literature views human capital as provid- ing the push for longer-term advances through new ideas arising from formal R&D and from other informal processes of innovation (Romer 2000; Temple 2001). This is in line with the findings of labor economists, who have estimated private returns of between 5 and 15 percent for a year of extra schooling. Initially attempts at confirming an equivalent level of social return from education were unsuccessful, but more recent work, which takes full account of externalities, sug- gests that the macroeconomic effects are equal to the private benefits and account for one-fifth or more of the growth in gross output (Temple 2001). 40 CAN EAST ASIA COMPETE? Table 3.3 Enrollment Ratios, Selected East Asian Economies, 1996 (percent) Economy Secondary enrollment ratio Tertiary enrollment ratio China 6 Hong Kong (China) 71 Indonesia 42 11 Japan 98 41 Korea 97 68 Malaysia 12 Philippines 60 29 Singapore 74 39 Thailand 56 22 Sources: UNESCO (1999, table 8; 2000, table 6); World Bank (1999); Wodd Bank World Development Indicators database. Most East Asian economies have already achieved high literacy rates, and the supply of human capital has been associated with rapid growth in the past (see table 3.3 for data on secondary and tertiary enrollments). However, investigators have frequently questioned the association between levels of schooling and growth (Bils and Klenow 2000; Easterly 2001; Pritchett 2001). Research now leans toward the quality of the education imparted rather than its mere volume, the scale of tertiary education, and the degree to which pedagogical tech- niques inculcate creativity among students. Hanushek and Kimko (2000) have strongly argued the case for quality, using cross-country analysis to trace its effect on economic growth. The importance as- cribed to quality and depth of schooling is growing steadily for an- other reason as well. As the volume of knowledge in each field in- creases, the time needed to acquire the expertise to make creative or innovative contributions expands, to as much as 10,000 to 20,000 hours in some areas of science and mathematics.7 Solving Fermat's Problem, for example, required command of more than half a dozen subfields of number theory, each highly evolved and arcane. As the frontiers of knowledge advance, the children who might contribute to tomorrow's advances must acquire mastery over many more facts and techniques than in the past, while also developing a capacity for creative thinking. A greater appreciation for educational quality than in the past echoes throughout East Asia. Experiments in moving from rote THE WARP AND WEFT OF INNOVATION 41 learning to discovery-oriented approaches are taking place in Japan, Korea, and Singapore, alongside the extensive introduction of very young children to computer equipment and the Internet. Korea proposes setting up special schools, similar to China's key schools, to nurture the scientific talent of gifted youngsters (Groom- ing a Science 2001). China, meanwhile, encourages competition be- tween schools by giving legal recognition to the private schools that currently train 7 million primary and high school students (China Moves to Encourage 2001). Research in the United States suggests that increased expenditures will make little difference to performance outcomes, although smaller classes can influence results by permitting teachers to give individual children more attention. In Japan and Ko- rea older teachers are reluctant to acquire computer and Internet skills, and are therefore unable to help students derive the maximum advantage from new equipment and technology. Stronger incentives and retraining could yield superior outcomes, and Korea is attempting to refresh its pool of teachers by lowering the retirement age. Coun- tries could also augment the supply of teachers by persuading those who recently acquired doctorates in science and mathematics to spend some time teaching (How to Produce 2000). For East Asian economies with effective lower and middle school systems that deliver good test results, the next step is to produce more scholars with high-level skills and to move schools up several notches in quality to heighten pupils' creativity and computer literacy without sacrificing the schools' current good attributes in the areas of science and mathematics skills (table 3.4).8 Much more ground must be covered at the high school and tertiary level, because at these levels education systems in most of East Asia are mediocre and uncompetitive at best.9 Only a few elite institutions- such as the Korea Advanced Institute of Science and Technology, set up in 1975, and the Hong Kong University of Science and Technology -consistently produce high-caliber graduates. The initiatives East Asian and Western universities are taking to build alliances and estab- lish joint programs are promising, because they import educational techniques and adapt them to local environments, something that the overseas training of thousands of East Asians has failed to bring about. Many of the leading schools in Australia, France, the United Kingdom, and the United States now have joint programs or other arrangements 42 CAN EAST ASIA COMPETE? Table 3.4 Science and Mathematics Achievement Scores at Age 14, Selected Economies Mathematics Science average average Economy achievement Economy achievement Singapore 604 Taiwan (China) 569 Korea, Rep. of 587 Singapore 568 Taiwan (China) 585 Hungary 552 Japan 579 Japan 550 Netherlands 540 Korea, Rep. of 549 Canada 531 Netherlands 545 Finland 520 United Kingdom 538 United States 502 Finland 535 United Kingdom 496 Canada 533 Thailand 467 United States 515 Israel 466 Thailand 482 Indonesia 403 Israel 468 Philippines 345 Philippines 345 International average 487 International average 488 Source: Science, 'Third International Mathematics and Science Study,' December 8, 2000. with universities in Japan, Korea, Malaysia, and Singapore. This is an important step in globalizing knowledge, especially tacit knowledge, and in developing institutions to govern and encourage the production of new knowledge. East Asia is at the early stages of this task, however, and just as resistance impedes industrial and financial reform, educa- tion systems will also change slowly. Few universities in East Asia can impart effective research skills, and those that can are located in the principal cities. This has a localiz- ing effect on knowledge generation and promotes technology-based clusters in these cities. A local supply of higher-level skills in conjunc- tion with market competition induces firms to invest in research. Rec- ognizing that research is the key to future advances in important in- dustries, governments in such countries as China, Japan, Korea, and Malaysia are investing in tertiary-level education, encouraging R&D, and funding government research institutes. Table 3.5 shows how these investments are increasing the supply of professional skills, patents granted (by the U.S. Patent and Trademark Office), and high- tech exports, particularly in Korea and Singapore. '0 THE WARP AND WEFT OF INNOVATION 43 Table 3.5 Indicators of R&D Effort and Outcomes, Selected East Asian Economies, Selected Years Scientists and Number of High-tech exports engineers in R&D, patents as a percentage of per million population, granted, manufacturing exports, Economy 1985-95 1996 1997 China 350 46 21 Hong Kong (China) 98 88 29 Indonesia 1 20 Japan 6,309 23,052 38 Korea, Rep. of 2,636 1,493 39 Malaysia 87 12 67 Philippines 1,299 4 12 Singapore 2,728 88 71 Taiwan (China) 1,897 Thailand 119 11 43 United States 3,732 61,107 44 Note: R&D = research and development. a. By the U.S. Patent and Trademark Office (utility patents). Sources: World Bank (1999); Taiwan Statistical Data Book 1999; Council for Economic Planning and Development, China; www.uspto.gov. Taking their cue from trends in the United States and Europe, more recently East Asian universities have also become more actively involved in commercial research. Starting with consulting activities, they are expanding into contract research, a path followed by the Massachusetts Institute of Technology starting in the late 19th centu- ry under the leadership of a faculty committed to a practical engi- neering education and with close ties to local industry (Lecuyer 1998)." Some universities are now venturing into self-financed re- search, which the institutions hope to capitalize on by way of patents or through new start-ups. Licensing income at U.S. universities amounted to US$700 million in 1998, compared with US$12.7 mil- lion in Japan Japan Acts to Speed 1999). Faculty from the University of California have founded fully one-third of the world's biotechnolo- gy companies (Is the University-Industrial Complex 2001).12 In Japan new government legislation similar to the technology transfer provi- sions of the Bayh-Dole Act in the United States is beginning to en- courage this tendency, but many hurdles remain, for example, limits 44 CAN EAST ASIA COMPETE? on the amounts universities can receive from private donors, how the money can be disbursed, and what it can be used for (University In- dustry Cooperation 2000). Many of the genome research institutes in China have been spun off by academic and government research institutes, forcing them to be more financially independent by capitalizing on their research findings. Alongside this, investment by private companies in universi- ty-based research, particularly by industrial country pharmaceutical companies with large research budgets, has increased, mostly in the West, but now also in East Asia. This trend worries some faculty members, who fear that it will divert universities from teaching and basic research and dilute their autonomy. Industrial Organization and Management In the current economic milieu macroeconomic regulation and skill development alone cannot sustain a virtuous spiral of productivity, in- novation, and growth. As East Asian companies have begun to enter high-tech areas and to compete not just on the basis of price, but also of technology, service, and speed of response, management quality has emerged as a major determinant of further success. Most firms in East Asian economies are family-owned and closely managed by fam- ily members."3 This has its advantages, especially for small firms. Fi- nancing, changes of direction, and decisionmaking are all made easi- er, and staff loyalty is more likely to be assured. The achievements of family enterprises in Hong Kong (China) and Taiwan (China) testify to the initiative and nimbleness of family firms, especially in manu- facturing fields in which technology is stable. Several large family- dominated companies in Korea and Taiwan (China) have performed remarkably well, and many have honed their management skills through formal training at some of the most prestigious Western business schools. Moreover, many of the larger family-run firms and their affiliated banks can mobilize "patient money" to finance signifi- cant innovations that can take years to achieve commercial success. The Corning Company in the United States is an example of a firm that throughout its history has emphasized innovation and backed it with patient investments (Graham and Shuldiner 2001). THE WARP AND WEFT OF INNOVATION 45 However, the managerial skills and organizational structures ade- quate for competing in standardized product markets are often not suitable for companies that must compete on the basis of innovative capacity. When business conditions are fluid and product cycles are short, a company environment that favors conformity and the readi- ness to follow orders rather than excellence and initiative can be a se- rious hindrance. Family-owned businesses frequently possess hierar- chical structures, and these are likely to place strong curbs on managerial activism and excellence. They may offer no strong finan- cial incentives for employees to excel and no mechanisms for weeding out poor performers. Cautious managers eager to please owners by meeting or exceeding performance targets are unlikely to suffuse the organization with a culture that nurtures innovation, even if goal ori- entation and energy levels are high. Moreover, such management will not be constantly alert for new ideas from within the company that may disrupt the smooth running of the business, and may also show limited initiative in seeking continuous feedback from other market participants (such as venture capitalists) on emerging possibilities to be tapped or in entering potentially fruitful strategic partnerships with other companies that might require determined, coordinated ef- forts to share knowledge, both tacit and formal, and to locate such functions within the organization (Dyer, Kale, and Singh 2001). The new generation of owners and managers that is taking over as business founders begin to retire could change the tenor of manageri- al practice; however, the motivation to change is more likely to come from tougher competition from foreign firms and new entrants, which can become a conduit for better managerial practices (Asian Capitalism 2000; Pulling Away 2000). Firm Turnover The entry of new firms and the maturing of those able to keep up the tempo of commercially viable innovations and to emerge as effective competitors by acquiring new management and organizational skills will be critical for East Asian competitiveness. Ease of entry and growth of firms to medium or large size is not commonplace in devel- oping countries because of regulatory barriers (in China, for example, 46 CAN EAST ASIA COMPETE? interprovincial, and even intercounty, barriers still pose a significant handicap); weak entrepreneurship; labor market constraints; limited organizational skills; unwillingness to dilute family ownership; poor infrastructure; and insufficient capital, especially risk capital. In China exit barriers sustained by supervising bureaus with the help of banks keep loss-making state enterprises alive, which dampens competition and diverts capital that more efficient producers could have used. Banks guided by signals from governments have also enabled troubled private firms to limp along in Indonesia, Korea, and Malaysia. Re- search on Taiwan (China) shows that the turnover of firms-entry and exit-has contributed to the changing mix of industry, technological gains, and advances in productivity because of the differential between those entering and those leaving (Aw, Chen, and Roberts 1997). The leading East Asian economies are notable for their wide dis- persion in relation to the scale of entry barriers to domestic firms. These include the time and resources expended on meeting registra- tion requirements as well as labor, environmental, and other regula- tions. Barriers are lowest in Hong Kong (China) and Singapore and highest in Indonesia and Vietnam (Djankov and others 2000). At least for small and medium firms, exit barriers are low. Large private and state-owned firms, however, use their leverage with governments and banks to defer breakup or liquidation, which may help explain the pro- ductivity differentials between the United States and China, Japan, and Korea in several major manufacturing and services subsectors (Baily and Zitzewitz 1998; McKinsey and Company 2001). Korea and Singapore have sought to lower the transaction costs of setting up en- terprises by minimizing regulatory and licensing requirements, espe- cially for technology-intensive firms and those whose products have export potential. Efforts to keep down entry barriers in these coun- tries extend to providing serviced land, industrial extension, and re- search grants; reducing labor market rigidities; and augmenting the supply of credit to smaller enterprises. Other East Asian economies, such as China, Malaysia, and the Philippines, have not attained the same level of regulatory efficiency or labor market flexibility, but have shown initiative by creating special economic zones with minimal en- try restrictions, high-quality infrastructure, and on-site services to at- tract foreign investors. Penang and the Multimedia Development THE WARP AND WEFT OF INNOVATION 47 Corridor in Malaysia and Subic Bay in the Philippines are examples of such zones. In China, township and village enterprises (TX/Es) compose a vast and more lightly regulated enclave within the larger economy. Even though the TVE sector has undergone significant consolidation since 1998 with a concomitant decline in employment from a peak of about 130 million, TVEs still employ close to 90 million workers and are re- sponsible for 40 percent of China's exports (Lin and Yao 2001). The growth of TVEs was almost inadvertent, and occurred because ad- ministrative and fiscal decentralization gave lower-level governments the latitude and incentives to create a favorable microeconomic envi- ronment for business within a relatively rigid planned economy. Beck- er (2000, p. 68) quotes Deng Xiaoping as saying to Yugoslav visitors: Generally speaking, our rural reforms have proceeded very fast, and farm- ers have been enthusiastic. What took us completely by surprise was the development of township and rural industries. All sorts of small enterpris- es boomed in the countryside, as if a strange army had appeared suddenly from nowhere. This is not the achievement of our central government. Every year, township and village enterprises achieve 20 percent growth. This was not something I had thought about. Nor had the other com- rades. It surprised us. As yet, most East Asian economies have only taken the first steps toward institutionalizing a competitive business environment by defining and implementing specific laws. While several have formally embraced competition policy, the elaboration of rules and their en- forcement by a court system equipped to handle cases involving mergers, predatory pricing, collusion, and monopolistic behavior is in its infancy. In addition, little attempt has been made to date to har- monize such policies among countries. CLUSTERS: THE SPATIAL DIMENSION TO INNOVATION While the previous section dwelt on the importance of the macroeco- nomic environment and the enabling institutional architecture, an equally significant driver of innovation is the nature of the municipal environment and how it contributes to the emergence of directly 48 CAN EAST ASIA COMPETE? networked clusters of firms, the transmission of knowledge, and ag- glomeration economies. Porter (1998, p. 5) defines a cluster as geographic concentrations of interconnected companies and institutions in a particular field.... They include suppliers of specialized inputs, com- ponents machinery and services. Clusters also often extend downstream to channels and customers and laterally to manufacturers of complementary products.... Finally many clusters include governmental and other insti- tutions. The proximity of firms to one another generates the productivity boost associated with clusters. The traditional benefit to firms of proximity to others-namely, reduced transportation costs between and among suppliers and purchasers-has become less important over time, principally because of improvements in domestic and in- ternational transportation infrastructures that have reduced per unit shipping costs. In the United States these have fallen nearly tenfold since the 1980s and, on average, now amount to just 1 percent of the final retail price of most products (see The Box that Launched 2000). Recent empirical research has pointed to other benefits of proximity, each of which offers not just the possibility of one-off cost reductions (as in the case of shipping costs), but the potential to accelerate the rate of innovation and productivity growth. These benefits include the following: * Large numbers of proximate firms engaged in similar economic activities are likely to attract specialized workers and professionals, resulting in the development of a "thick" urban labor market. Not only does this deeper pool allow firms to find better employment matches for their specific needs, but it also allows for the transmis- sion of tacit knowledge between firms as employees move from firm to firm. World cities such as London derive some of their suc- cess from providing specialized services sector firms with the large supply of skills they need to meet the great increase in outsourcing by MNCs and smaller companies. * Large numbers of proximate firms can sustain a considerable pool of specialized service suppliers that can customize services to meet purchasers' needs. Without a large potential market for specialist services, only generic service suppliers will find it profitable to es- tablish operations in a city. Such clustering also generates demand for upscale office and housing facilities, which in turn gives rise to THE WARP AND WEFT OF INNOVATION 49 demand for other service providers that can maintain and upgrade these facilities and cater to the needs of high-income occupants. * Proximity to suppliers and consumers is important when deciding whether and how to upgrade products and services. Such collabo- ration between firms and their suppliers and between firms and their customers is becoming the norm in high value added manu- facturing. * Proximity to firms in related but distinct industries can generate positive spillovers, as innovations in an industry provoke other firms to re-evaluate their own commercial practices. * Proximity to large, axial foreign firms or to universities and re- search institutes generates the potential for knowledge spillovers. Large foreign firms tend to be demanding purchasers, requiring constant improvements from suppliers, and often encouraging the latter to collaborate in innovation and to share best practices. With respect to academic institutions, in addition to the traditional ben- efits of training students, universities provide local firms with op- portunities to commercialize new findings, often in partnership with researchers. Each of these arguments had been advanced well before the late 1990s (see, for example, the well-known research of Jane Jacobs, Al- fred Marshall, and Edwin Mills). Only recently, however, have rich datasets been assembled to assess and confirm the empirical impor- tance of each of these factors.'4 Unsurprisingly, then, these findings have led to a reassessment of the growth potential of clusters, cities, and urban regions more generally. As Leamner and Storper (2001, pp. 3-4) observe: [R]outine coordination of standardized intellectual or physical tasks can bc done with markets that can be extended geographically with communica- tion technologies. But complex and unfamiliar coordination of innovative activities requires long term relationships, closeness and agglomera- tion.... [T]he successful transfer of complex and un-codifiable messages requires a kind of closeness between sender and receiver that the Internet does not allow.... It is a medium that may help maintain relationships but does not establish deep and complex contacts. Because knowledge is their lifeblood, clusters thrive on the density of informal networks made possible by proximity and further rein- forced by new communications technology. This helps build a special 50 CAN EAST ASIA COMPETE? type of social capital-horizontal relationships-and trust that induce the circulation of knowledge and are the basis of good governance within producer networks. The social capital that serves as the glue for Silicon Valley in California or Silicon Fen in Cambridge in the United Kingdom has little in common with what animates a dense civil society. Silicon Valley networks facilitate productive interactions among individuals, firms, educational institutions, financing houses, policy instruments, and other entities. In their analysis of U.S. clus- ters, Cohen and Fields (1999) found seven principal interactors: the great research universities; U.S. government policy, in particular, de- fense-related research was of vital significance; venture capital firms; law firms; business networks; stock options; and thick labor markets. More broadly, social capital also determines municipal governance and the supply of community services to firms. For example, the city government of San Jose, an agglomeration at the head of Silicon Val- ley, collaborates with existing area firms and helps new firms acquire testing facilities, laboratory space, and business services. Clusters come in many varieties. We focus on the large metropoli- tan variants that bring together a mix of industries and services and contain firms of varying sizes, but clusters can also be more modestly scaled and consist mainly of small networked firms, as in northern Italy, Brazil, and India. Clusters can also be of the hub and spoke vari- ety, in which activity revolves around a few large firms and their sup- pliers (Markusen 1996). Certain special economic zones in East Asia consist of the branch plants of MNCs plus components manufactur- ers. China and Korea have clusters that are anchored by a major state enterprise, such as an automobile or steel making firm, with numer- ous subcontracting affiliates. Single industry clusters, for example, the telecommunications clus- ter in Finland, generate localization economies arising from special- ization and focus that extend to suppliers, labor markets, infrastruc- ture, and logistics. They have their own dynamic and are capable of rapid growth. We are more persuaded by research optimistic about the prospects of the large, multifaceted agglomerations that stimulate intellectual spillovers. We see many significant innovations arising from the intersection of different disciplines and technologies, and believe that ICT will help catalyze even more interdisciplinary activi- ty. Furthermore, innovative clusters are also likely to be performing THE WARP AND WEFT OF INNOVATION 51 well with respect to other indicators, such as exports and employment (Department of Trade and Industry 2001). Clusters differ in the quality, variety, and nature of the relation- ships they induce among producers, and spillovers seem to be related to the size and depth of a cluster,"5 the specialization of the labor pool, and the mix of producers within a cluster. Innovation and the commercialization of new technologies is disproportionately concen- trated in clusters in a particular field (Porter 1998). For example, the Washington, D.C. clusters are specific sources of innovation in ICT and biotechnology; London leads in areas such as business services and finance; while Austin, Phoenix, Portland, and San Jose lead in electronics and software (Cortright and Mayer 2001; Markusen and others 2001). In some of the industrial subsectors productivity is aid- ed by strategic interaction between firms, often through buyer-driven supply chains, as well as suppliers of components, producer services, and now ICT services.'6 Such interaction, which embraces the provi- sion of venture capital, research, design, marketing, and other servic- es, is most effective in a compact coastal area well served by the full panoply of transport and telecommunications services. Only certain choice urban locations have the attributes conducive to cluster formation, and a good deal of research is being devoted to identifying these attributes. They range from the quality of municipal institutions and governance; to the caliber and research intensity of local universities; to the climate and the quality of the cultural ameni- ties,'7 housing, and schooling available (van den Berg, Braun, and van Winden 2001). Humphrey and Schmitz (2000) correctly point to governance as contributing to the competitiveness of clusters; their upgrading through strategic investments; and, most important, to their repositioning in the race to stay ahead. Most clusters started in an area favored by history and tradition with the appearance of a cluster engine, that is, a major university or research institute, a dynamic firm, or equivalent MNCs. Singapore, for example, brought in MNCs through the government's incentive regime and a stream of measures to supply skills and infrastructure targeted at these global firms' needs. The result has been that firms in the Singapore-Johore cluster have been able to move up the technol- ogy ladder (Mathews 1999). In other research, however, Mathews (1997) showed that governments can plant the seed for a cluster 52 CAN EAST ASIA COMPETE? through direct action, such as by setting up a free trade zone and then making it grow through well-timed interventions. The Baden-Wuerttemberg model of technology development and diffusion from Germany is an early example of how public initiative can create a multilayered system of institutes and research clusters that can work with private companies to meet a hierarchy of needs, ranging from identification of technological needs to adaptation, cus- tomized development, original research, and training. The system consists of the Steinbeis Foundation, which runs more than 100 tech- nology transfer centers, often in cooperation with a nearby technical college, university, or research institute carrying out basic or applied research. Complementing the transfer centers adjacent to universities are technical schools linked to between 10 and 20 firms closely asso- ciated with the universities. Support in the areas of natural and engi- neering sciences, jointly funded from private and public sources, is provided by Fraunhofer institutes. The institutes provide specific contract-based technology transfer and development services, mainly to large companies. Lastly, local chambers of commerce, supported by levies on companies, are responsible for educating, training, and organizing work groups in key areas such as biotechnology, laser de- velopment, and software. This multipronged system has helped se- cure Baden-Wuerttemberg's industrial leadership in Germany and has sustained the dynamism of Germany's several industrial clusters (Cooke and Morgan 1998). Hsinchu Park in Taiwan (China) was the outgrowth of deliberate government policy. Starting with the creation of the Industrial Tech- nology Research Institute in 1973, the government founded Hsinchu Park in September 1979 explicitly on the model of Silicon Valley. The government then set up the United Microelectronics Corpora- tion, which launched the electronics industry. The major advance, however, occurred when the semiconductor industry embarked on the so-called Very Large Scale Integration phase in the mid-1980s. Again, this was triggered by government initiative by requiring that the Industrial Technology Research Institute obtain technical capa- bility through Chinese-American firms in Silicon Valley. This was followed by a joint venture with Philips to form the Taiwan Semicon- ductor Manufacturing Corporation to fabricate silicon wafers. Philips not only transferred Very Large Scale Integration technology, but THE WARP AND WEFT OF INNOVATION 53 also its cross-licensing agreements with other companies, enabling the corporation to avoid intellectual property rights disputes. The success of the Taiwan Semiconductor Manufacturing Corporation initiated Hsinchu Park's virtuous spiral. The striking feature of such recent developments, especially the growth of leading-edge industries and services, is their tendency to concentrate in just a few localities. A small number of clusters in the United States-around Boston, Chicago, Los Angeles, New York, San Jose, Seattle, and Washington, D.C.-and in the United Kingdom- Cambridge, London, the southeast, the west Midlands, Scotland, and Wales-account for the bulk of the high-tech industry in those coun- tries (Markusen and others 2001; Department of Trade and Industry 2001). The same is true in Japan and in other East Asian countries. Once they have taken root, and if the area sustains its attractions,'" the cluster pulls in other activities and solidifies its advantages. As noted earlier, the high-tech clusters with the brightest prospects depend on large, diversified, and efficient labor markets. They must draw and retain people with the mix of skills that firms in the area re- quire and be a magnet for mobile and discriminating knowledge workers (Dumais, Ellison, and Glaeser 1997). A successful cluster ac- cumulates and continuously refreshes human capital. Penang's past success in attracting electronics-producing MNCs to Malaysia indi- cates how incentives, good governance, infrastructure, the environ- ment, and a flexible labor force can combine to yield success. Its cur- rent struggle to retain the companies that have made Penang their base and to attract new ones underscores the competitive nature of the regional environment and the importance of building tertiary-level in- stitutions to train workers in the requisite skills. The Malaysian au- thorities and the Penang municipality have worked hard to build the requisite university departments, and they are now ready to bring in professionals from overseas, but because a scarcity of skills continues, more MNC capital is flowing elsewhere, most of it to China and Ko- rea (see All Wired Up 2000; Kotler and others 2002; chapter 6 in this volume). The skills clusters need include those provided by specialty service entities, for example, firms that supply legal, headhunting, equipment leasing, accounting, investment banking, and venture capi- tal services. All these together compose the ecosystem that makes a cluster a world-class player. 54 CAN EAST ASIA COMPETE? In the lower-middle-income countries such as China, clusters are at an early stage and have yet to develop the full range of activities that contribute to virtuous spirals, although the Pearl River Delta re- gion is well served by Hong Kong (China), while clusters in Beijing and Shanghai increasingly have access to a wide range of services. More venture capital is becoming available in most East Asian economies, but research facilities and their links to business are both weak, and they have few ties with clusters in Japan, the United States, and Europe. If clusters are centers of growth, then understanding better how they might be created and managed in individual coun- tries is a matter of great significance for the future (see table 3.6 for a ranking of the perceived role of clusters by economy). Although government pressure on the corporate and research sec- tors backed by generous financial incentives can create an agglomera- tion of establishments devoted to research, the spillover effects of such deliberate action can be slow to materialize, moreover, clusters are often slow to acquire depth and to draw in the large MNCs that can lend momentum to a cluster and help link it more tightly to world markets (Survey of Cambridge 2001). Both Tsukuba Science Table 3.6 Perceived Role of Clusters, Selected East Asian Economies Economy Ranking Finland 1 United States 2 Germany 3 Singapore 4 Taiwan (China) 5 Japan 8 Hong Kong (China) 11 Korea, Rep. of 17 Malaysia 25 China 28 Philippines 35 Thailand 38 Vietnam 39 Indonesia 46 Note: Business people in 50 economies were asked to assess the extent to which they agreed with the following statement: "Clusters are present in most international industries and include not only suppli- ers, but specialized institutions such as university research programs and training providers"' The mean response for each economy was calculated. The lowest possible ranking is 50. Source: World Economic Forum (2000). THE WARP AND WEFT OF INNOVATION 55 City near Tokyo and the Daeduck Science Park in Taejon City in Ko- rea were the outcome of state directives, incentives, and infrastruc- ture building. Both now host a large number of research establish- ments; in the case of Daeduck 12,000 scientists and technicians work in 60 centers. But whereas in the case of Hsinchu Park a few govern- ment initiatives spawned a cluster of industries, the much more ambi- tious efforts in Japan and Korea failed to induce networking among the transplanted or newly created research centers and local industry or, at least until recently, the entry of new start-ups to feed off the re- search done in the parks (Asia's High Tech Future 2000; Castells and Hall 1994; Shin 2001). LOCAL-GLOBAL NETWORKING Clusters crystallize the new economics of industrial development on a municipal, or at most a regional scale, where the region might em- brace a number of contiguous centers that are all part of a local net- work. The Singapore-Johore region is one example, with Singapore serving as the hub. Another is the Hong Kong (China)-Pearl River Delta region, which comprises several somewhat specialized clusters, including Foshan, Dongguan, Guangzhou, Shenzhen, and Zhuhai (Yusuf and Wu 1997). Many of the firms in these clusters were initial- ly launched by entrepreneurs from Hong Kong (China), who began moving their factories out into the Pearl River Delta area to take ad- vantage of cheaper labor (Enright, Scott, and Dodwell 1997). Now the individual clusters concentrate on production and depend on Hong Kong (China) for management, marketing, consultancy, finish- ing, quality control, testing, certification, and shipping services-an efficient division of labor (Berger and Lester 1997). These networks are sufficiently diverse to internalize spillovers, but are also compact enough to capitalize on the benefits of proximity. The successful high-tech clusters, whether in East Asia, the United States, or Europe, have a local-global characteristic that is likely to be increasingly significant. Even lower-technology clusters producing buttons, cutlery, or frames for eyeglasses, whether in China orJapan, must be alert to changes in tastes, designs, and technology. High-tech clusters are no longer self-sufficient. Instead firms are continuously 56 CAN EAST ASIA COMPETE? acquiring, observing, or seeking alliances with firms in other clusters. Moreover, a trend toward production networking among clusters, frequently mediated by MNCs, distributes various activities in accor- dance with changing comparative advantage. Hsinchu Park outside Taipei and Silicon Valley are virtually coextensive because of multiple interfirm associations and the ceaseless movement of people to and from each location (Saxenian and Hsu 2000). This global dimension of clustering is now critically important for governments and firms. The overseas Chinese network, whose members are estimated to have an annual income of US$700 billion (not including Taiwan [China]), has long demonstrated the utility of informal, close-knit arrangements in the East Asian milieu.'9 It provides entrepreneur- ship, implements sanctions, deters opportunistic behavior, and gener- ally supplements weak legal systems in individual countries. In addi- tion, its members provide marketing information, matching, and referral services. This latter aspect of the network is declining in im- portance, however, especially for standardized products, as the infor- mation intensity of economic relationships increases, better commu- nications technology becomes available, contract enforcement is put on a more sound footing, and firms become part of international pro- duction networks in which formal market-based relationships begin to displace informal ties U. E. Rauch 1999a, 1999b). However, for new or differentiated products the Chinese network can still be valu- able for exchanging information and leveraging marketing skills (Rauch and Trindade 2002). GLOBAL CIRCULATION OF KNOWLEDGE AND OF KNOWLEDGE WORKERS The localization of knowledge goes hand in hand with the interna- tional mobility of knowledge workers and the need for leading uni- versities and research institutes to maintain global connections if they are to attract the best talent, stay abreast of the latest advances, and forge the alliances now intrinsic to cutting-edge research. While in- vestment in universities and in building research capacity is undoubt- edly necessary, East Asian countries have discovered that this alone does not result in research that generates marketable innovations and THE WARP AND WEFT OF INNOVATION 57 pushes development in desired directions. Research universities in East Asia need to establish links with institutions in the main research clusters in Europe, Japan, and the United States so that they can tap into the large pools of knowledge that sustain the world's main tech- no-industrial clusters. This is an integral part of knowledge network- ing in a world in which a high proportion of research skills remains concentrated in not much more than two dozen locations worldwide. Lucky (2000, p. 262) observes that In their influence on how science is transacted, the Internet and World Wide Web have had the greatest impact of any communications medium since possibly the printing press. The telegraph, telephone, and wireless were not different in concept from the postal system, except that the mod- ern technologies were so much faster. The postal system, telephone, and telegraph are also one-to-one topologics, connecting a single user to an- other single, predesignated user. On the other hand, radio and television are one-to-many topologies for the broadcast of a small number of com- mon channels to a great many receivers. The Internet and Web are some- thing else entirely. The beauty and power of these new media are that they allow the formation of spontaneous communities of unacquainted users. Their topology is neither one to one nor one to many, but rather many to many. They allow the sharing of infornmation in textual, graphic, and mul- timedia formats across these communities, and they empower users with- in these communities to build their own applications. It is this empower- ment of the periphery that has opened the floodgates of innovation to millions. In all previous communications technologies the ability to inno- vate and craft new systems and applications was confined to a small num- ber of industrial engineers who tended the centralized intelligence and functionality. ICT is helping boost networking by diasporas of professionals, in particular, African, Chinese, Indian, and Latin American. In the 18th and 19th centuries the transfer of industrial technology from the United Kingdom to the United States and Western Europe depend- ed on the movement of skilled workers (Jeremy 1981; Landes 1980). The experience of the 1980s and 1990s underscores this finding: the international movement of human capital is enormously fruitful in generating new knowledge and remains a major conduit for transfer- ring both codified and tacit knowledge. Such migration is linked to trade liberalization and to some easing of curbs on immigration in OECD countries (Globerman 2000). Japan, for example, will need to 58 CAN EAST ASIA COMPETE? import 30,000 skilled and professional workers to meet its require- ments over the next five years. Nowhere is the role of foreign-born science and engineering pro- fessionals more apparent than in the United States.20 The results of a study published in 1999 show that "individuals making exceptional contributions to S&E [science and engineering] are disproportion- ately drawn from the foreign born; individuals making exceptional contributions are also disproportionately foreign educated both at the graduate and undergraduate levels" (Are the Foreign Born 1999). Twenty percent of all those employed by the ICT sector in the Unit- ed States are foreign born (Devan and Tewari 2001). These flows lead to the concentrations of researchers in particular locations needed to promote usable research. They are responsible for the vitality of linked clusters within and between countries, as well as for the fruitfulness of collaborative research. Almost 40 percent of the companies in Hsinchu Park near Taipei were cofounded by at least one returnee from the United States (Valley of the Deals 1998). More than 320,000 mainland Chinese went overseas to study between 1978 and 1999, double the number that went abroad during 1872-1978, complementing substantial flows from Korea, Malaysia, and Taiwan (China). This was a large outflow from a fairly small pool of people with some tertiary-level skills. The brain drain is now being compen- sated for in part with brain gain, however, as 110,000 people have re- turned to their home countries. The trend in return migration as Chi- na develops is positive, aided by active recruitment by Chinese research bodies such as the Chinese Academy of Sciences (China's Leader Commits 2000; Scientists with Business Flair 1998). In several other countries, including India, Korea, and the Philippines, a small percentage of those leaving intend to return (figure 3.2). The continuous circulation of researchers among universities and research facilities in the United States is also a notable vehicle for hu- man capital augmentation, for cross-fertilization, and for the dissemi- nation of research. Similarly, the links established between U.S. West Coast clusters, Hsinchu Park near Taipei, and facilities near Seoul have begun to integrate research in those two economies with U.S. research. In large part this is due to the circulation of researchers, FDI, and the increasingly coextensive nature of the research in cer- tain fields on both sides of the Pacific. THE WARP AND WEFT OF INNOVATION 59 Figure 3.2 Likelihood that the Well Educated Will Remain at Home, Selected Economies (10 = the most likely to stay at home) 10 8 6 2 0 o,Q~~~e ;p C; and transferred Y1.4 billion of bonds billion billion by Danamodal billion by state trillion to AMCs n 0 Majority foreign ownership Allowed, but not for 1 1 sold with majority 13 foreign banks 4 completed, 2 of banks? domestic currency stake, 6 sold so that hold 30% of commercial pending m operations foreign owners hold bank assets significant stakes Note: GDP = gross domestic product; NPL = nonperforming loan; AMC = asset management corporation. Source: World Bank (2001). FINANCIAL AND BUSINESS SERVICES 87 Bank and the Bangkok Metropolitan Bank (Thailand, Bank Consoli- dation 2002). As the market becomes more competitive, such consoli- dation makes sense, not least because it enables banks to diversify their portfolios. Moreover profits increase with size, and larger banks are better at attaining risk-adjusted capital adequacy ratios (Reynolds, Ratanakomut, and Gander 2000). Consolidation of banks and the im- petus to supply insurance, brokerage, leasing, credit card, investment, and other services also stem from narrowing profit margins in tradi- tional lines of business. Despite competition from other providers of such services, banks find conditions favorable for such efforts because of economies of scale and the use of ICT Banks also find that their name brands and large client bases give them an edge over new Inter- net starts. Physical facilities, too, still confer some advantages over purely online providers (Online Finance 2000). This process of consolidation and heightened market contestabili- ty from local and foreign service providers is in its early stages, but it promises to gather momentum. The potential for innovation and better overall performance is there, but the risks are also increasing, as noted in chapter 2. The incidence of crises has risen in part be- cause of the interwoven nature of the global financial system (De Swaan 2000). The presence of extremely large, diversified financial institutions places additional burdens on regulators. They must be alert to collusive practices and pricing, conflicts of interest, excessive risk taking, and exposure in specific markets. Permitting such large entities to fail is much more problematic than allowing smaller banks to go under and constitutes a moral hazard to be monitored with care (Mishkin 1996; Mishkin and Strahan 1999). Banking reform is proceeding at varying speeds in virtually all the East Asian economies, and in some instances has been ongoing for close to two decades. As East Asia attempts to reinvigorate its growth, accelerating financial development will be an obvious policy option. With ICT and the possibilities created by consolidating fi- nancial institutions, the gains could be larger than in the past, and the costs of delaying bank reform and restructuring will be even greater. As indicated earlier, strengthened regulation must parallel financial change, but as Levine, Loayza, and Beck (2000) have point- ed out, financial development depends on the quality of legal and ac- counting services, a point we explore later. The emergence and spread 88 CAN EAST ASIA COMPETE? of venture capital are critical for an era in which technological change is bound to speed up. Venture Capital in East Asia The renaissance of venture capital in the United States and Europe is closely associated with the emphasis on commercializing innovative products and services (see World Bank forthcoming). Since the start of formal venture capital investing in 1946, the amounts invested worldwide have risen, increasing from US$100 million in 1965 to US$48 billion in 1999 (Kenney and Florida 2001). Venture capitalists are the handmaidens of an innovative economy, because they provide professionally managed, equity-like financing for young, growth-ori- ented companies (Hellmann 2000). Although the bulk of venture capital is in the United States, primarily in the Silicon Valley region, East Asia is attracting a trickle of American, Japanese (especially trad- ing companies branching into venture capital), and European compa- nies, for example, Warburg Pincus, Softbank, and Deutsche Bank. Some countries, such as China, Malaysia, and Singapore, are promot- ing home-grown venture capital from the public sector, for instance, Walden International and H&Q. While not significant in the near term, this development will promote the growth of technology-in- tensive firms that stimulate local innovation in the long run. Currently, most venture capitalist activity in East Asia is in Hong Kong (China), Singapore, and Taiwan (China).'" In part, this is be- cause innovation-driven entrepreneurship is concentrated in these lo- cations and because of government activism. However, these three economies also provide venture capitalists with the institutions that permit exit by way of an initial public offering or a merger with a larger company. Either option calls for well-developed capital mar- kets, a business-friendly environment, and listing rules hospitable to launching initial public offerings and to mergers and acquisitions. Venture capitalists are active in Taipei because of its institutions and multistranded links with Silicon Valley, and because, much like the San Francisco Bay area in the 1950s and 1960s, the region around Taipei has the large firms, wealthy individuals, and accumulated skills and experience that provide the soil from which venture capitalists spring (Kenney and Florida 2000). Elsewhere, true venture capital FINANCIAL AND BUSINESS SERVICES 89 activity is in its infancy. Unlike the United States and Europe, where venture capital is concentrated in a few high-tech sectors that hold promise of large gains in market capitalization and earnings, venture capital in most East Asian countries is distributed across a number of traditional and new sectors. Moreover, because of the tight hold of families, the venture capitalist rarely provides the inputs-technical and managerial guidance, contacts, networking skills, and strategy ideas-associated with venture capitalism in the West. Families are reluctant to sell their shares in firms. Thus much of East Asia is still not ready for venture capital in the institutional sense, and such insti- tutionaj preparation must precede less critical, but easier steps, such as setting up a second NASDAQ-style stock market. SALIENCE AND COMPETITIVENESS OF BUSINESS SERVICES The benefits of financial development increasingly extend beyond its contribution to macroeconomic stability and the efficient intermedia- tion of resources to industry. Currently, exports of services worldwide, direct and indirect, are about one-fifth of total trade and are growing faster than trade in manufactures. By 2001 international trade in serv- ices had reached US$1.4 trillion. Between 1990 and 1998 real global merchandise exports rose by 6.8 percent per year, while the growth rate for services was 8 percent. Financial services constitute a major component of the business services now coming to dominate East Asian economies in terms of employment, value added, and share of GDP. They also attract a significant percentage of inward FDI. For export-oriented economies, competitive financial and other business services can become a major source of earnings from direct exports (see table 4.3, which shows the still relatively modest share of exports of services by East Asian economies). In addition, because services are increasingly large inputs into the production of manufactures, every country now runs a large trade bal- ance in indirectly utilized services. Using input-output tables to com- pute the direct and indirect trade in services, Urata and Kiyota (2000) found that Singapore and Taiwan (China) run a net surplus in services (incorporated in manufactured exports). In contrast, China, Japan, Malaysia, and the Philippines run net deficits in services. Although 90 CAN EAST ASIA COMPETE? Table 4.3 Trade in Services, Selected Economies, 1987 and 1997 (share of services in total exports) Economy 1987 1997 Hong Kong (China) 22 58 Philippines 6 37 Singapore 6 30 Thailand 4 22 Korea, Rep. of 3 16 Malaysia 2 16 Japan 4 14 Indonesia 2 11 United States 9 25 Italy 8 23 United Kingdom 10 23 France 11 22 Germany 4 13 Source: Hardin and Holmes (1997). Japan is a net exporter of wholesale retail and transport services, over- all it emerges as a net importer of embodied and disembodied services. The cost competitiveness of banking, ICT, marketing, accounting, transport, and legal and other services influences the competitiveness of final goods. In the industrial countries, producers of high-tech items are gearing up to sell bundles of manufactured items and serv- ices to their customers, and because of the fierce competition in the market for manufacturers, differentiated services tailored for each customer can be a much larger source of profits than manufactured products alone. Companies such as IBM, Sun Microsystems, and General Electric derive most of their profits from the services they sell along with their computers, servers, and airplane engines. In the case of IBM, much of its growth and market capitalization comes from services (Getting the Most 2001). Boeing, General Motors, Dell, and Hewlett-Packard-Compaq are moving in the same direc- tion, with General Motors banking on telematic services provided to its automobile buyers generating a sizable slice of future profits (Can Compaq Escape 2001; Driving the Info Highway 2001). In the industrial countries intense competition in the retail banking market has sharply trimmed profit margins. Thus banks must reduce their costs; rely more on online banking; and diversify into credit card, FINANCIAL AND BUSINESS SERVICES 91 real estate, trust, and investment services, which offer greater scope for innovation. Supporting such moves is a large cohort of people ap- proaching retirement who demand a host of personalized services. Thus far, most of the increase in banking has been in electronic trans- actions, which in Asia, North America, and Western Europe and have now surpassed paper-based ones (figure 4.1). The era of online services provided by banks and other suppliers is less than a decade old, and many untapped possibilities remain as new demands are created and met. With Internet access via PCs and gen- eration (G) 2.5, and eventually 3G, wireless gadgets on the rise, a new era of service provision is opening up (see chapter 5). The opportuni- ties are immense, but they are greatest for those who establish new, widely adopted platforms and protocols that quickly capitalize on net- work economies and who invest in communications infrastructures. Figure 4.1 Electronic versus Paper Payments, Selected Years 140 - 1ZO 1994 -9720 03 Souce Texia 20) E >1 80 - 0 c 60 0 40 20 1994 1997 2000 2003' H Electronic U Paper a. Projected. Source: Teixeira (2000). 92 CAN EAST ASIA COMPETE? Ease of entry, a competitive environment, and integration with the world economy are the keys to a dynamic services sector, including the kind of financial industry that East Asian economies need as they enter the 21st century. This confronts policyrnakers with difficult choices. Banks and other providers of business services have traditionally been sheltered from external competition, and internal regulations have generally buffered them from domestic competition as well. Invariably the threat of having to face foreign suppliers elicits a strong negative response that politicians cannot easily ignore. But the status quo will prove untenable. The share of business services in GDP and their contribution to the efficiency and innovativeness of other industries is large and growing larger. Hence future growth depends on rapid gains in the competitiveness of the services sector-in terms of quality, timeliness, and convenience-and the liberalization of trade in servic- es. A survey of the business community in Hong Kong (China) indi- cated that business people view a freeing of trade in financial services as the highest priority, followed by other business services. We discuss how this might be accomplished on a regionwide basis in chapter 6.16 What is true of the banking and finance sector extends to other business services as well. East Asia has been slow to recognize the vi- tal role of such services in promoting productivity, directly and through the manufacturing sector; in inspiring innovation; and in contributing to employment and exports. In the area of legal and ac- counting services, for example, international law firms provide advice to local and international firms on expanding their businesses, han- dling regulatory matters, managing their finances, and resolving dis- putes. They constitute an important force supporting a more rules- based business environment. Furthermore, law firms can tap expertise from colleagues both in major financial centers worldwide and in innovative clusters such as Silicon Valley, thereby helping East Asian firms to raise funds overseas as well as to keep abreast of the le- gal consequences of the latest developments in business strategy and technological change (see, for example, the description of China's ex- pansion of legal services described in The Contender 2000). The cities with the most regional company headquarters and FDI are also those with the closest links to international law and accounting firms (figure 4.2). Hong Kong (China) is in the lead, but other cities have also entered the arena of globalized services. FINANCIAL AND BUSINESS SERVICES 93 Throughout the region, business services account for one-quarter to one-third of GDP, but remain relatively backward and inefficient, when they could be the equal of manufacturing industries. This can be traced in part to the still limited supplies of knowledge workers in many countries, such as China and Indonesia. The shortages of such workers are being reduced, but building a stock of experienced pro- fessionals will take time. Inadequate supplies of knowledge workers reflect past government inattention, except in economies such as Hong Kong (China) and Singapore, where governments have for some time consciously sought to groom service activities. Preoccu- pied with manufacturing, East Asian governments did not fully real- ize that the institutional environment for services is more complex than that for manufacturers, which at a pinch can be nurtured in en- claves. The bare-bones institutions adequate for manufacturing oper- ations and the types of workers appropriate for the assembly line are both unsuited for the business services now at the leading edge of the modern economy. Much like high-tech manufacturing, the modern services economy is on a different technical and institutional plane. Figure 4.2 Connections to Leading U.S. Law Firms, Selected Asian Cities, 1997 O 450 'L,> 400 u 350 2 -c 300 . 250 200 150 0 s 100 - Source:4World City Relational Database (http://www.lboro.ac.uk/gawc). Source: World City Relational Databnase (http://www.Iboro.ac.uk/gawc). 94 CAN EAST ASIA COMPETE? Both derive their impetus from inputs of skilled workers, large doses of ICT, continuous innovation, and international networking. The 1990s altered the perspective on development for both emerging and industrial economies. At the beginning of the decade they still viewed the manufacturing sector as the engine of growth and the principal source of transformative (or disruptive) innovations. The share of services in GDP was large and growing, and they saw services as crucial in support of manufacturing, but industry remained in the forefront. For East Asian economies banking and finance greased the wheels of industry and commerce, but their role was sub- sidiary to that of manufacturing. Thus policy focused on creating and sustaining conditions supportive of manufacturing. Labor market and banking sector policies sought to supply industry with low-priced in- puts, trade policies attempted to enlarge export markets, technology policies emphasized product and process innovation, and policies to- ward FDI sought to attract foreign capital to the manufacturing sec- tor. East Asia has not abandoned this mix of policies, and for most countries the significance of manufacturing remains basically undi- minished. However, a new, more broadly conceived approach to de- velopment is needed, and the winning combination will be a mix of manufacturing and services. Moreover, the productivity, profitability, and market penetration of manufactured goods are closely tied to the efficiency and innovativeness of service inputs. FUTURE ROLE OF BUSINESS SERVICES Three causes account for the increasing primacy of services. The first is the contribution of banking and finance to development beyond the mobilization of resources. The availability of risk capital is impor- tant to innovation, especially in ICT-sensitive sectors, because it pro- motes the entry of new firms and enables existing firms to commer- cialize new products and processes. A vast increase in trade in financial services has followed the liberalization of capital flows, and this in turn has given suppliers incentives to offer a wide range of new value adding services. In short, the financial sector is now at the core of a dynamic that entwines innovation and productivity growth. FINANCIAL AND BUSINESS SERVICES 95 A second underlying cause of the expansion in the services sector is the trend toward outsourcing services such as research, logistics, sup- ply chain management, fulfillment, back office functions, and systems integration (see Barthelemy 2001 for a discussion of the risks of such outsourcing without carefully screening vendors and drawing up de- tailed contracts). Since the 1980s firms have found that buying-in many services previously done in-house is advantageous, because it al- lows the firms to concentrate on their core functions.'7 Such special- ization enhances efficiency, and advances in ICT have made buying-in a more attractive and convenient prospect, as services can now be pur- chased wherever in the world firms find the lowest price. The third underlying factor, the subject of chapter 5, is ICT. The impact of ICT on business services cannot be overestimated. Not only has ICT become a leading growth sector in its own right, but its role in other commercial activities has allowed the effects of its diffui- sion to be much more widely felt. It has both induced and supported innovation in other service areas, such as finance, advertising, mar- keting, logistics, and supply chain management, while at the same time making it much easier for firms to buy-in services and weave them together with other in-house and contracted activities. As noted earlier, for many manufacturing firms the abundance of new services lowers costs and allows them to offer customers a tailored mix of goods and services that commands price premiums. For East Asia's economies, the beginning of the 21st century will almost certainly be remembered as the era in which services emerged as a leading component of a hybrid growth model. As this model takes shape in the coming decade, other economies will join Hong Kong (China) and Singapore in benefiting from the higher living standards created by services. Underpinning this development will be a change in the mind-sets of policymakers, who will move their coun- tries' economies forward by improving market regulations and by en- couraging improvement in the supply of skills and services that will define the tempo of growth for the foreseeable future. 96 CAN EAST ASIA COMPETE? NOTES 1. On the growth of services in East Asia, see Wirtz (2000), who underlines the significance of industrial services in East Asia. Profit margins in most manufac- turing industries are modest, stable, or falling. After six years of record growth the automobile industry has seen no change in its profitability, hence companies like General Motors and Boeing are preparing to diversify into telematic and fi- nancial services that they can provide joindy with their main products (Aviationz Week, "Boeing Widens Reach to Generate Growth," March 19, 2001; Oxford Analytica, "Auto Industry," April 6, 2001; Red Henring, "The New Thin Client," April 2001). 2. As Sicular (1998) has shown, however, and as is apparent from the large errors and omissions in the balance of payments accounts, capital flight from China is substantial and persists despite controls. 3. By comparison, the cost of bailing out U.S. banks at the time of the Great De- pression in 1929-33 was about 3 percent of GDP, and the cost of resolving seven major banking crises worldwide between 1870 and 1913 never exceeded 5 per- cent of GDP. This provides a perspective on how steep the cost of not managing and regulating East Asia's banking systems effectively has become (Calomiris and Mason 2001). 4. Cross-provincial data from China also support the view that the efficient alloca- tion of financial resources by the banking sector has a strong influence on growth, along with the ability of enterprises to raise funds by themselves (Liu and Li 2001). 5. Urata and Kawai (2001) provided empirical evidence forJapan that bank-direct- ed credit slows down the entry of new firms, because funds are diverted to in- cumbent firms. This has the effect of reducing the incentive for unprofitable in- cumbents to exit the industry, and not surprisingly, the net entry of new firms into Japanese industries is now zero. 6. The change in corporate governance is especially noticeable in Japan, long a stronghold of stakeholder power. There the increased shareholder influence arises from the buffeting endured by the economy over a decade, the banks' un- winding of share portfolios, the pressure on companies to diversify sources of funding, and the increase in shares held by non-Japanese investors (from 4 per- cent in 1990 to more than 14 percent in 2001), who are more forthright in ex- pressing their views and demanding a response from management (Far Eastern Economic Review, "Day of the Shareholder," September 13, 2001). 7. Moreover, cross-country experience during the 1990s suggests that a banking system sheltered from foreign competition and subject to the guidance of gov- emments and corporate clients will not become autonomous and efficient in a relatively closed environment. Regulators operating in such an environment, with its powerful traditions and exposed to the demands of public agencies, politicians, and business interests, have difficulty making banks adhere to strict rules. Forbearance is the rule. A sheltered environment also negates the influ- ence of other market entities-such as rating agencies, accountants, auditors, FINANCIAL AND BUSINESS SERVICES 97 financial media, and bank shareholders-from contributing to the level of mar- ket competition. 8. White (2001, p. 7) maintains that effective prudential regulation has the follow- ing components: "[A] transparent accounting system based on market value; ad- equate risk-based capital requirements that are based on the market value ac- counting framework and are forward-looking, sensible rules on risk limititions; insistence on the financial and managerial competency of owners and managers, a clear understanding that owners of insolvent institutions lose their invest- ments; clear authority for intervention and enforcement of the rules; and ade- quate numbers of well-trained and well paid regulators." 9. Calomiris and Mason (2001, pp. 3-4) observe that "too many influential people simply have too much at stake to allow banking to be reorganized efficiently. Banks are only able to channel favors if they themselves are recipients of subsi- dies from the government, hence the need to preserve the banks' exclusive rights and the need to offer banks subsidized deposit insurance, subsidized pur- chase of bad loans, or subsidized capital injections. Crony capitalists will appeal for bank assistance on the basis of 'credit crunch' motive while in fact hoping to channel government assistance for banks into their own coffers. Thus the cen- tral goal of bank bailout policy is to design bank assistance to meet the legiti- mate goals of mitigating credit supply contraction for value creating bank de- pendent borrowers while minimizing the potential 'abuse of assistance."' 10. De Swaan (2000) has doubts "whether the initiatives for cross border and cross sectoral cooperation can go far enough and fast enough in light of the globaliza- tion of the financial system. For example, the cooperation between independent supervisory bodies could result in conflicts of interest as well as overlaps or gaps in supervisory practices. Moreover differences between national supervisors in terms of resources, culture and legal interpretations could place serious road- blocks on the path to international harmonization or centralization of supervi- sion." 11. Nevertheless, significant NPLs remain. As a percentage of total loans, the re- maining NPLs in Indonesia, Malaysia, the Philippines, and Thailand were 48.7 percent, 19.5 percent, 17.4 percent, and 30 percent, respectively, toward the end of 2001 (see World Bank 2002 for details). In China Lardy (2001) states that central bank officials have reported that the four largest state-owned banks have NPLs equivalent to a quarter of the value of their loan portfolios. 12. The official estimates of the NPLs in the portfolios of China's big four banks are in the 30 percent range. Unofficial estimates put the ratio at 50 percent or more, with accumulation continuing (Financial Times, "China Survey," October 8, 2001). 13. Bankruptcy laws in some Southeast Asian countries, for example, are based on outdated commercial and corporate laws, enforcement by the courts is weak and unreliable, and the courts themselves are not free of corruption, which further reduces the laws' effectiveness (Oxford Anzalytica, "Southeast Asia: Bankruptcy Law," August 21, 2001). The low salaries of Indonesian judges is a problem, as is the extreme litigiousness of Filipino society, where business activity is constantly subject to nuisance and harassment costs. In mid-1997 a judge in a first-instance 98 CAN EAST ASIA COMPETE? court in Indonesia earned about US$380. By year's end this had shrunk to about US$160 (Backinan 1999). Backman points out that enforcement of bankruptcy laws was the main problem in Indonesia, with the law invoked only four times in the cases of insolvent companies in the 10 years prior to 1994. In Thailand lengthy windup and liquidation procedures plus high court fees discouraged re- course to bankruptcy procedures. In Taiwan (China) the Bankruptcy Law and the Company Law are designed, interpreted, and enforced so as to protect own- ers from creditors and to avoid letting companies go into receivership. This is achieved by processes that delay the appointment of an external administrator and by giving the creditor no final say in the restructuring plan, by the reluc- tance of the oversight agencies to decide whether a company is salvageable, and by the ill-defined nature of the audit process (Oxford Analytica, "Taiwan: Re- structuring Resistance," October 4, 2001). 14. Accounting firms must also be subjected to the discipline of strict rules and be held liable for connivance with corporate management and for gross oversight. 15. As of 2001, Singapore was second only to Japan in providing vcnture capital funding. The amount invested was US$379 million compared with US$634 million in Japan. The Singapore government's attempt to turn the island nation into a biotechnology hub has resulted in up to US$2 billion being provided to the government-backed Biomedical Sciences Group to invest in new start-ups and promising firms. This flood of money has also attracted other funds and drawn large pharmaceutical companies, such as Eli Lilly, Chiron, and Affymerix, to start research operations in Singapore (Red He77ing, "Singapore Gambles Big with Biotech Payouts," September 15, 2001). 16. Such services could be liberalized multilaterally, of course, and Mattoo (2001) discusses the prospects for this. 17. There are limits to the efficacy of outsourcing manufacturing, however. Compa- nies are finding out how to outsource advantageously, and are also discovering that sharing proprietary information with suppliers risks leaks to competitors. CHAPTER 5 ICT: BEYOND THE HYPE A n influential body of opinion now maintains that the pace of development in the industrial and middle-income coun- tries will be keyed to a range of general purpose technolo- gies spawned by ICT. Although its contribution to growth in most industrializing economies has been limited to date, the un- tapped potential of ICT can scarcely be doubted. Countries that do not embrace ICT at many different levels risk jeopardizing export shares in current markets; entry into new export markets; flows of in- ward FDI; and gains from production networking, technology trans- fer, and knowledge sharing. ICT will also be responsible for the growth of myriad high value added services, and it has already begun to transform the organization of firms. WHAT'S AT STAKE? The reach of ICT extends well beyond the economic realm. The end of the 1990s saw the rudiments of e-governance materialize in those countries where Internet use was most widespread (see the chapters on trends in the United States in Kamarck and Nye 2002), and both telemedicine and distance learning crossed major thresholds. Observ- ing these factors, many analysts are persuaded that ICT will be the 21st century equivalent of electricity and the automobile, slowly per- meating virtually all aspects of life and raising living standards far into the future. No doubt the U.S. experience in the second half of the 1990s has strongly colored both views on ICT and likely estimates of its role in 100 CAN EAST ASIA COMPETE? the growth process. A slowing of the pace of growth in the United States relative to the second half of the 1990s could shake faith in the power of ICT to enhance productivity. Even so, the ICT revolution will spread, and countries that do not assimilate the new technologies will lag behind in critical respects and will not fully realize the gains from participating in the global economy. The high expectations surrounding ICT derive mainly from the performance of the U.S. economy between 1995 and 2000 in three key areas: productivity growth, innovation, and improved business strategy. The U.S. experience can provide a lens for looking at East Asia. Thus in this chapter we first critically summarize the major findings of research on the United States, consider the relevance of these findings for East Asia, and offer a perspective on ICT penetra- tion in the region. The second part of the chapter concentrates on how government regulation could lead to more rapid and efficient ex- ploitation of ICT. The rest of the chapter is devoted to an analysis of the telecommunications sector, which paces the growth of ICT. The "New Economy" in the United States Between 1995 and 2000 the growth rate of American output per per- son was approximately 3.2 percent per year, well above the 1.9 per- cent per year that the American economy experienced between 1973 and 1990, and even higher than the 2.9 percent per year increase achieved during 1959-73 (Baily 2001). The difference between a 1.9 percent and 3.2 percent growth rate may seem small, until one appre- ciates that the former implies that output per person doubles in 37 years, while the latter implies a doubling of productivity in only 22 years. This fact alone calls for an assessment of the productivity ac- celeration and of whether it represents a phase shift from which to draw lessons that might apply to East Asia. Determining the contribution of ICT to this growth performance is central to this debate (for an overview of this debate see Landefeld and Fraumeni 2001). Spending on ICT-including software, hard- ware, and communications equipment-undoubtedly rose consider- ably during the 1990s. During this decade firms in the United States increased their investment outlays by 19 percent a year in real terms, and by the end of the decade more than half of this investment was in ICT: BEYOND THE HYPE 101 ICT. As a share of GDP, investment in ICT rose from 3 to 6 percent over the 1990s (Basu, Fernald, and Shapiro 2001). Other sources of demand were changing, too, however, clouding the picture. U.S. fed- eral government spending fell during this period, a fall partially offset by increased state and local expenditures (Vatter and Walker 2001). Statistical tools have been used to disentangle these different effects, and we now turn to these findings. Economists have employed growth accounting techniques to esti- mate the contribution to output growth of the greater use of the na- tion's productive inputs, such as capital, including ICT; labor; raw materials; and energy. Ideally these techniques enable us to see how much ICT has added to growth after taking into account the growth of the labor force, energy use, and other inputs into production. Ini- tially these studies were conducted on data for the entire U.S. econo- my, and the principal findings are reported in table 5.1. In all but one study, the contribution to growth of computer hardware-one form of ICT-doubled after 1995. This is the first indication that the spread of ICT throughout the U.S. economy quickened the pace of American economic growth. It is perhaps not surprising that all econ- omists agree with these findings. Analysis by Kiley (1999) suggests that computer hardware actually reduced growth every year (al- though at a slower rate after 1995), and Jorgenson and Stiroh (2000) found that investment in ICT has apparently not generated the spillover and networking effects that feed TFP. These disagreements reflect differences in the methods used to calculate actual inputs into Table 5.1 Contribution of Computer Hardware to U.S. Output Growth, Selected Years Before 1995 Since 1995 (contribution, (contribution, Before 1995 percentage Since 1995 percentage Study (years covered) points/year) (years covered) points/year) Oliner and Sichel 1991-95 0.25 1996-99 0.63 (2000) 1996-98 0.59 Whelan (2000) 1990-95 0.33 1996-98 0.82 Jorgenson and Stiroh (2000) 1991-95 0.19 1996-99 0.49 Kiley (1999) 1974-84 -0.34 1985-98 -0.27 Sources: Authors. 102 CAN EAST ASIA COMPETE? production as well as techniques employed to adjust for inflation. The number of people currently employed, for example, is not neces- sarily a good proxy for the total amount of labor used in production in a given year, especially if the number of hours worked varies be- tween years. Some variations in hours worked are recorded-and so can be taken in account-but no such records are kept for salaried workers. Making such adjustments is typically fraught with errors and controversy. One objection to these economywide studies is that they do not al- low for the differential effects of ICT in different sectors of the econ- omy. For example, knowing whether ICT had a greater impact in manufacturing industries than in service activities, such as banking, is useful. An analysis of U.S. productivity growth indicates that ICT- related inputs were responsible for about two-fifths to one-half of the acceleration in productivity growth between 1995 and 2000 (table 5.2). This is likely to be an underestimate, because it neglects the 0.3 percentage point gain from capital deepening. The bulk of the Table 5.2 Decomposition of Productivity Acceleration, United States, 1995-2000 Compared with 1973-95 (percent per year) Oliner Jorgenson and Sichel Gordon and Stiroh Baily Category (2000) (2002) (2000)' (2001) Labor productivity, nonfarm business 1.38 1.81 1.01 1.63 Cyclical effect n.a. 0.40 n.a. 0.04 ICT capital 0.70 0.60 0.44 0.62 Other capital -0.26 -0.23 0.09 -0.23 Measurement effects n.a. 0.14 n.a. n.a. Labor quality 0.04 0.01 -0.23 0.00 Multifactor productivity 0.90 0.89 0.71 1.19 Computer sector multi- factor productivity 0.31 0.30 031b 0.18 Other nonfarm business multifactor productivity 0.59 0.59 0.40 1.00 n.a. Not applicable. Note: ICT = information and communication technologies. a. Compares 1995-99 with 1973-95. Includes consumer durables and the farm sector. b. ICT-related multifactor productivity. Sources: Authors. ICT: BEYOND THE HYPE 103 productivity growth stems from gains in the sectors producing indus- trial machinery (mainly computers) and electronic machinery (mainly semiconductors), but wholesaling and retailing also registered sharp gains. Other services and construction performed relatively poorly, and most of the "non-new" economy industries turned in an average performance, with that of the food sector being unusually bad (Nord- haus 2001). Although Nordhaus's estimates of productivity increases are higher, his findings regarding the sources of productivity increases are not very different from those of Gordon (2002), who to date has estimated the smallest incremental returns from the spread of ICT. Gordon esti- mated that annual productivity growth in the United States from 1995 to 2000 was 2.86 percent, nearly double the 1.42 percent achieved from 1972 to 1995. Once a cyclical component of growth is removed (0.4 percent), however, the increase in trend growth is about 1.04 per- cent. After allowing for changes in measurement and labor quality, only 0.89 percent of the annual increase in the productivity growth rate needs explaining. Gordon ascribes this almost entirely to produc- tivity growth in the durable manufacturing sector, with 0.30 percent contributed by technical change in computer hardware and 0.22 per- cent coming from technological acceleration in the rest of durable manufacturing. Only 0.07 percent can be traced to gains in productiv- ity in the services sector, where much of the computer hardware is in- stalled. However, when Basu, Fernald, and Shapiro (2001) corrected for cyclical savings using average weekly hours as a proxy for factor utilization and for adjustment costs, they arrived at an estimate for the contribution of technological gains to growth well above that implied by Gordon (2002) and close to that of Oliner and Sichel (2000). Basu, Fernald, and Shapiro found that for the private sector as a whole, the contribution of technology in the second half of the 1990s rose from 1.2 to 3.1 percent per year. More surprising, and contrary to some other findings, nonmanufacturing technological performance in- creased from 0.9 to 2.7 percent. Because this sector accounts for four- fifths of production, technology gains in the nonmanufacturing sector were responsible for 1.3 percentage points of the overall 2.9 percent advance in annual technology-related benefits for GDP growth. Bear in mind that our understanding of sectoral productivity gains in the latter half of the 1990s is still evolving. Modification of the time 104 CAN EAST ASIA COMPETE? series data in 2001 revised U.S. GDP growth in 2000 to 4.1 percent, down from the earlier estimate of 5 percent. Annual growth in labor productivity between 1996 and 2000 is now estimated to have aver- aged 2.5 percent and not 2.7 percent (A Miracle Revised 2001; United States 2001). Moreover, contrary to the opinion of Basu, Fernald, and Shapiro (2001), the majority opinion remains that productivity in the services sector still lags for three reasons. First, the services sector accounted for the overwhelming part of the US$10 trillion plus U.S. economy in 2001, and the new investment in ICT was only a small share of the existing capital stock in the services sector.' (Recall that in the growth accounting framework, only increas- es in the stock of productive assets raise output. Investment adds to the stock of capital, and so affects growth through this channel.) Second, some of the ICT investments may have only diverted de- mand from other services rather than generated new demand for services, thereby providing additional growth impetus. Projections indicate that by the end of 2002, business to consumer sales will reach US$39 billion, or about 2 percent of retail sales in the United States.2 Some of those would probably have been made in the absence of ICT, and so the demand-stimulating effect of these technologies may have been quite modest to date. A third reason ICT effects appear to be so small in the services sec- tor is that growth accounting techniques do not include investments in difficult to measure intangible assets that complement the intro- duction of ICT. These intangible investments include outlays on training staff to use ICT and spending on organizational restructur- ing and the reorganization of decisionmaking power and incentives within the corporation. A recent analysis that incorporated these ef- fects into U.S. aggregate data found that ICT had contributed much more growth than had been shown by other methods (Yang and Brynjolfsson 2000). The techniques used to take account of intangi- bles comprising knowledge, skills, and learning acquired by workers are very much in their infancy, but they suggest the need to factor in the indirect role of ICT (Blair and Wallman 2001).3 From these findings, arguing that ICT has transformed the U.S. economy in five years is difficult. Nevertheless, enough positive evi- dence exists (including continuing productivity growth during 2001, which buttresses the claim that the benefits of ICT are continuing to ICT: BEYOND THE HYPE 105 work their way through the system) to merit a closer examination of how ICT affects firms, adopting a bottom-up analysis in contrast to the top-down statistical analysis reviewed here. Such a bottom-up ap- proach focuses attention on how ICT affects the way firms innovate and the rate of technical change, effects that are difficult to capture in the aggregate studies of economic growth discussed earlier. Research results are now also available on how ICT is impinging on European economies and higher-income East Asian countries. These findings, summarized in table 5.3, suggest that the United States de- rived a significantly greater growth stimulus than European countries and Japan: 1.71 percent per year in 1995-2000 compared with 0.21 to 0.62 in Europe. However, countries with large and dynamic ICT Table 5.3 Contribution of ICT Capital and TFP in ICT Production to Growth in GDP and Labor Productivity, Selected Countries and Years Contribution of ICT capital Country Years to annual GDP growth (%) Australia 1990-95 0.48 out of 3.4 Australia 1995-2000 0.68 out of 4.6 Canada 1990-95 0.30 out of 1.8 Canada 1995-2000 0.57 out of 4.2 Finland 1990-95 0.24 out of 0.7 Finland 1995-99 0.62 out of 5.6 France 1990-95 0.18outof1.0 France 1995-2000 0.35 out of 2.8 Germany 1990-95 0.30 out of 2.2 Germany 1995-2000 0.38 out of 2.1 Italy 1990-95 0.21 out of 1.4 Italy 1995-99 0.36 out of 1.9 Japan 1990-95 0.31 out of 1.3 Japan 1995-99 0.38 out of 1.1 United Kingdom 1990-95 0.27 out of 2.1 United Kingdom 1995-2000 0.48 out of 3.6 United States 1990-95 0.97 out of 2.6 United States 1995-2000 1.71 out of 4.4 Notes: iCT = information and communication technologies; TFP = total factor productivity; GDP = gross domestic product. All estimates refer to total economy GDP. All estimates on the contribution of ICT capital include software. Source: van Ark (2002). 106 CAN EAST ASIA COMPETE? Table 5.4 Contributions of Computer Hardware to GDP Growth in the Republic of Korea and Singapore, Selected Years (percent) Computer hardware's contribution Economy and years GDP growth rate GDP growth Own growth rate Korea 1980-95 7.9 2.54 39.1 Korea 1990-95 7.5 2.71 42.7 Singapore 1977-97 7.8 1.46 34.2 Note: GDP = gross domestic product. Source: Pohjola (2000). sectors, such as the Netherlands, Sweden, and the United Kingdom, have performed strongly, as have Korea and Singapore (table 5.4). In Japan, however, while a few leading companies with ICT-related products, such as Sony, Toshiba, NTT, Canon, DoCoMo, and Fujit- su, have continued to forge ahead, the effects of ICT on the economy as a whole were lower through the end of the 1990s. The rest of East Asia also presents a mixed picture. Hong Kong (China), Korea, Singapore, and Taiwan (China) have large ICT sub- sectors and have rapidly embraced the use of ICT in governance and transportation, for example.4 Internet connectivity is high and rising. As in the United States, the proportional contribution of computer hardware to growth is significant, but the full macroeconomic impli- cations for TFP are still in the future. Most firms, even the largest, have yet to overcome the skill and organizational hurdles on the road to e-business. This finding was recently reaffirmed by a study (Dutta and Biren 2001) of the use of the Internet, which showed East Asian companies lagging in the use of ICT for marketing and customer re- lations (figure 5.1). Impact of ICT on Business Strategy How quickly ICT can be deployed and begin to raise productivity substantially will depend in part on how readily firms alter their or- ganizations and strategies for innovation and customer service. They must, in essence, review their business models. Here the United States, with its ceaseless experimentation and prolific business schools, is very much the pacesetter. Accumulating evidence from corporations ICT: BEYOND THE HYPE 107 Figure 5.1 Use of the Internet by Fortune Global 500 Companies 30 0 0 25 0) 20- u5) > E 10 1 0 Customer Technology Product Placement Promotion Price relations setting * Asia-Pacific O Europe U North America Source: Dutta and Biren (2001). in the United States and elsewhere suggests that ICT will affect many dimensions of a firm's activities, making simple characterizations inap- propriate at best and misleading at worse. A single business model of ICT adoption for followers to emulate is unlikely to emerge. They will therefore have to experiment and choose. A readiness to be different characterizes most of the outstanding new captains of industry. Being different, however, does not represent a strength for the large East Asian conglomerates or for the smaller, family-owned firms. Rather, firms in the region excel at adroit imita- tion, producing at lower cost than rival producers, with at best some incremental innovation and, as in Japan, with close attention to opera- tional effectiveness. As a result companies can lose sight of the longer- term need to regularly transform the nature of a company by being alert to a wide range of possibilities and selecting the most promising from among them. Oracle, for instance, has sought to define its choic- es by asking its major customers what they want and then attempting to respond in a comprehensive and innovative way, for instance, it has integrated business systems through the E-business Suite III, thereby providing its customers with comprehensive, Internet-based services.5 108 CAN EAST ASIA COMPETE? Among the reorganization strategies brought to the fore, the fol- lowing seven are uppermost: * Use the scale, scope, and advantages ICT offers to consolidate and enlarge market share. * Join forces to create vertical or horizontal B2B exchanges tailored to the characteristics of the firm's products and to maintain compe- tition at healthy levels. * Focus on a niche, whether a final product or a component. This can be achieved by meeting specific client needs or by achieving control over a proprietary technology or process.6 * Build a brand image to differentiate the firm's products from those of its rivals. Few laptop owners ever see the Intel chips inside, but a carefully placed label on most laptops provides a permanent re- minder of their presence. * Tie in suppliers. Collaborating with suppliers and encouraging them to improve their components helps align their incentives with those of their customers. Toyota has done this to outstanding effect. Given the recent surge in outsourcing, collaboration re- minds suppliers that forward-looking parts and components pur- chasers care more about technology and satisfying their customers than they do about price, although sustaining price competitive- ness remains a crucial objective (Spear and Brown 1999). * Seek to mass customize certain kinds of products or services using the Internet and flexible manufacturing systems to individualize products. Dell Computer provides an example. Made to order, mass market products that allow customers their choice from among a fixed range of options drastically reduce waste and inven- tory costs. According to one estimate, such a step would be worth US$80 billion per year to automobile producers alone (A Long March 2001). * Add services. The idea here is to sell high-quality services along with a product, allowing the customer to differentiate the bundle of goods and services from that of rival suppliers. ICT opens several doors to firms seeking to refine their organiza- tions using these seven strategies. A company's primary goal, even for a fledgling dot-com, is to generate enough profits to grow. Such are the exigencies of financing and market competition that size is an ICT BEYOND THE HYPE 109 advantage. In many industries this dynamic favors the rapid growth of the leading firms through mergers and consolidations. Just five or six years ago the United States had six online services. The number has now shrunk to one. The number of workstation producers has dropped from three dozen in the mid-1980s to a handful today, only four disk drive makers selling directly to producers of computers sur- vive from the 75 that existed in the late 1980s, and most of the online booksellers that appeared in the mid-1990s are no longer in business. Although where services and technology are based on human inputs and intangibles, the significance of scale is by no means obvious (How Seagate 2002). By reducing the costs of intercompany transactions, ICT encour- ages the use of external B2B exchanges in certain standardized prod- ucts, resulting in greater specialization. Earlier fears of rampant out- sourcing, commoditization of products, and cutthroat competition appear somewhat unfounded, however.7 Current trends point to the emergence of B2B exchanges formed by large market incumbents that can generate sufficient volume. U.S. automobile manufacturers, for example, have come together to create Covisint, which manages automobile parts transactions with the major suppliers such as Delphi and Visteon (companies that were spun off as independent operations by the automobile makers themselves and retain close links with them), a marketplace worth US$250 billion (Lucking-Reilly and Spulber 2001). Aerospace companies, energy producers, grocery companies, and others have set up other exchanges. Some of what they trade are standardized items, the so-called indirect goods (for example, pencils or computers) used in many industries where suppli- ers are many and competition is fierce. However, 80 percent of the products traded in East Asia that feed the manufacturing sector are direct goods, generally customized for users to some degree. Online trading of these goods, using technology provided by Commerce One, Ariba, and i2, is likely to be of a different stripe. Because of the need for joint development of products; customization; and long- term, trust-based relationships, electronic transactions will most probably supplement more traditional forms of interaction (Dhawan and others 2000). The Web will facilitate supply chain management and reduce ordering, fulfillment, and payment costs, but it will not inevitably lead to generalized arm's-length transactions and intense 110 CAN EAST ASIA COMPETE? competition between atomistic producers except in certain standard- ized commodities. We can thus envisage both a greater degree of in- tegrated production structures as well as more cooperative interac- tion between buyers and suppliers using ICT to enhance efficiency and cut costs. Large multinationals such as Ericsson, Cisco, and IBM contract out a sizable share of their hardware production to specialized suppli- ers while concentrating on upstream activities and services. This opens niches for specialized contract manufacturers of electronic equipment such as Flextronics and Solectron, which in turn buy up producers in East Asia and organize regional supply networks so as to encourage competition between suppliers to original equipment manufacturers.8 Again, in theory, these developments can minimize governance and contracting costs, although it is too early to gauge the magnitude of gains likely to be realized, or even to assume that such forms of disintegrated subcontracting and reintegrating will persist. (For a novel analysis of the effects of globalization on vertical structure, see McLaren 2000.) In the coming years the effects of ICT on production networking will be influenced by the phasing out of local content requirements. This in turn will also profoundly affect FDI, component manufactur- ing, and the geographic distribution of assembly operations, especial- ly in the ASEAN countries.9 In the Japanese automobile industry, for example, much scope exists for rationalizing assembly in fewer loca- tions to take advantage of scale economies. In addition, component manufacturers, many of which are subsidiaries of Japanese firms and are increasingly independent of the automobile manufacturers, will be under pressure to restructure their operations."0 The growing presence of American and European producers in the region could intensify competition and the search for locations offering the largest mix of advantages. Taken together, these forces have the capacity to transform business structures in East Asia. These changes will occur as firms continue to "slice up the value chain," often with the inten- tion of moving labor-intensive production stages to labor-abundant locations that also offer low-cost transportation and communication infrastructures. By facilitating the development of networks of firms, ICT supports alternative business models to the large industrial groupings that have ICT: BEYOND THE HYPE 111 dominated the Northeast Asian commercial landscape. In the after- math of the 1997-98 crisis, the weaknesses of diversified, and often family-controlled, conglomerates-such as the chaebol and keiretsul arrangements-have become more evident, encouraging efforts at re- structuring, disposing of peripheral activities, and reducing overca- pacity (see Claessens and others 1999 for a discussion of the extent of family control over major businesses in many East Asian countries and how this may slow legal, regulatory, and corporate governance reform). The inefficiency and unwieldiness of state-owned enterpris- es, as in China, has been apparent for a couple of decades. In many industries the advantage appears to lie with specialized firms, big and small, that might be assemblers or suppliers of components or servic- es. The typical focused Taiwanese firm seems better placed to reap the gains from ICT and new multiplant, flexible production tech- nologies than does its larger, diversified Korean counterpart. The size of a Korean chaebol does, nonetheless, confer advantages mostly for the mature and slower-expanding company, enabling it to raise re- sources quickly to establish large-scale production facilities, conduct R&D, market on a global scale, and create an internationally recog- nized brand name (Claessens, Fan, and Lang 2002). ICT offers firms the means to refine other business strategies by reducing the cost of collecting, analyzing, and communicating infor- mation. Although media reports on the new economy have tended to focus on falling communication costs, the ease with which they can collect and mine data is critically important to firms. Firms now have available to them the extremely detailed information on consumer tastes and spending patterns stored in relational databases and coIml- piled, for example, through surveys and bar code readers in super- markets and elsewhere. The computing power to analyze these vast datasets now exists, allowing firms who retain staff or consultants with the requisite analytical skills to increase sales, improve margins, and identify niches of consumers willing to pay more for specialized or better-quality products (How Data Mining 2001). In the past such niches may have gone undiscovered or the decisions to supply them would be made on hunches-the infamous business intuition. One hurdle that is likely to remain for the smaller producer and the subcomponent manufacturer is developing a brand name and market reputation, a task likely to take on further significance as e-business 112 CAN EAST ASIA COMPETE? becomes commonplace not only for manufacturers but also, and to a greater extent, for service providers (Shapiro and Varian 1999). The intensity and cost of marketing on the Internet is one indicator of how critical the brand name might become. Moreover, advertisers are dis- covering that as the tempo of Web advertising grows, so does the diffi- culty of attracting the attention of potential customers. This might call for much greater attention to targeting customers and to using word of mouth before undertaking expensive saturation advertising." ICT will also expand the scope for mass customization of produc- tion. Most manufacturing is not conducted on a built to order basis. Instead firms must estimate future consumer demand and tailor cur- rent production schedules accordingly. When firms misjudge, de- mand shortages emerge, leading to consumer dissatisfaction or the buildup of inventories of unsold goods. Neither situation is good for the firm's long-term viability. Thus firms have every incentive to bet- ter match production output to demand. Clearly this incentive pre- dates the introduction of ICT, and indeed, automobile companies have long operated using buffer stocks of vehicles to cope with an un- planned expansion of demand. Internet technologies, however, add a new dimension to mass customization. Purchasers can now place or- ders for products that better meet their needs and consult with sup- pliers about potential modifications and varieties. Perhaps the best known example of this is the purchase of personal computers over the Internet from companies such as Dell Computers and Gateway. These companies pre-assemble modules that, when tailored to a cus- tomer's given specifications, can be quickly assembled into a final product. In this way the number of parts can be reduced, and yet economies of scale in the pre-assembly of modules can be retained (A Long March 2001). By expanding the ways in which services can be delivered, ICT has the potential to transform certain services sectors. In the past, for ex- ample, many customers had to travel to a bank or similar office to re- ceive financial services. Through ICT they now pay bills and perform other banking transactions online. These technologies have helped banks economize on the number of branches, reducing fixed costs and staffing levels. These financial services developments represent only the tip of the iceberg of potential future innovations in service deliv- ery. The entire communications industry-from telephone companies ]CT: BEYOND THE HYPE 113 through moviemakers-sees these new information technologies as a way to sell a wider variety of services directly to customers' homes. Arnazon.com, Ebay, and Priceline.com are already expanding aggres- sively in East Asian markets (Spending Spree 2002). Furthermore, firms and state bodies are experimenting with telemedicine and tele- education. The former offers an alternative to face-to-face consulta- tions between doctors and patients and between doctors themselves, and recent trials in Hong Kong (China) suggest that this approach could be widely applied, especially for the routine care of aged or rela- tively immobile patients. That new ser-vices often follow in the wake of ICT diffusion speaks to the importance of co-invention. Indeed, some observers have argued that these second-round service innovations make possible the realiza- tion of the full benefits of the primary invention: the Internet (Bresna- han 2001). Such obser-vers would argue as well, however, that many factors determine the pace of co-invention (including some of those discussed in chapter 3), not least the magnitude of any switching costs; the ease with which advances can be shared among users; and the net- work effects, through which the value of a service to any one purchaser depends on the number of other purchasers already using that service. The greater emphasis on innovation forces companies to pay much more heed to new possibilities arising within their own organizations, to technology available on the market for purchase or licensing, and to technology nurtured by start-ups that could be acquired."2 One of the more remarkable recent developments, one that the nimbler compa- nies are scrambling to use, is Web-based technology markets. A tech- nology exchange called yet2.com has allowed companies to shop for innovation for a modest membership fee and a 10 percent share of any completed transactions to a maximum of US$50,000. This exchange, and others like it, create more efficient markets for technology and help to monetize intellectual property. 'With bright new ideas an-d technologies being invented the world over, yet2.com reinforces the factors discussed in chapter 3 that unify the global market for ideas. Beyond that, Web-based knowledge and markets enable large, highly diversified companies such as General Electric and Procter & Gamble to identify and exploit the technologies already at hand within their own organizations that can be commercialized or licensed (New Rules 2001). 114 CAN EAST ASIA COMPETE? As the potential and current limitations of the Internet become better perceived, companies are beginning to see advantages in mar- rying the new approaches with the old in many different ways. For East Asian companies with a competitive edge mainly in the old-style industries, the road to success may lie in using ICT to leverage exist- ing strengths deriving from, for example, proprietary content, physi- cal location, product characteristics, and the use of catalogs (see Porter [2001], who endorses the exploiting of complementarities be- tween Internet use and more traditional corporate strategies). Simi- larly, combining "bricks with some clicks," that is, effectively harness- ing the Internet to enhance traditional marketing practices (see Willcocks and Plant 2001), might be the best way for most producers to enter the domain of electronic marketing. The pace can then pick up. Even so, for most firms self-reinvention as Internet-based entities may not be desirable, even if it were feasible. Gateway offers an example of a new-style computer supplier that found it profitable to use a combination of the telephone, physical stores, and the Internet to reach customers. It speaks to the enduring value created by physical stores that 25 percent of Gateway's sales are through its stores. Moreover, the stores are effective in building the brand, receiving feedback, and satisfying the large group of cus- tomers who wish to see a product before buying it (Bricks and Clicks 2000; How PC Makers 2000). ICT's deep penetration into old-line industries can be seen in the oil drilling industry's reliance on com- puterized seismic modeling, which has radically transformed dfilling techniques and the productivity of oil extraction (Rauch 2001). Companies that take the steps outlined earlier to differentiate themselves from their rivals reduce, but rarely eliminate, competitive pressures. For this reason the rapid shrinkage in the number of firms in some industries need not be a sign of diminishing contestability. On the contrary, even established firms, such as CISCO, EMC, and Intel, must continuously be on their guard against competitors with better products or superior ways of meeting customers' wants, appar- ent or latent. In the space of two years, for example, Juniper was able to eat into CISCO's apparently impregnable hold on the market for "core" routers. It reduced CISCO's market share by 11 percentage points in 2000 alone. Since then other challengers have emerged, such as Caspian Networks and Hyperchip, and CISCO has fought ICT BEYOND THE HYPE 115 back with its own "super routers" (A Central Theme 2002). Similarly, Hitachi Data Systems is nibbling into EMC's share of the market for storage devices. All the big storage system producers are being chal- lenged by start-ups offering huge 230-tetrabyte devices with greater capability, and new technological possibilities will continue to spawn others (Hitachi Is Now 2001; Storage Titans 2001). Examples of ICT penetration into a wide range of activities and its implications for business abound. From an East Asian perspective, two points are noteworthy. First, Internet access is rising throughout the East Asian region, most rapidly in the northeast. ICT-based products constitute a large proportion of total manufacturing output, especially in the Southeast Asian countries. So far the effects of e- business and e-commerce on productivity appear to be minimal, al- though online transactions in Southeast Asia alone ranged between US$20 billion and US$25 billion in 2000. Research in this area is still sparse, but casual empiricism suggests that e-business and e-com- merce are still in the early stages. Second, and more important, the Western experience to date shows that even though the Web can greatly facilitate market net- working, sales, and eventually business efficiency, retailing is still tied to the quality of logistics,'3 and manufacturing is tied to the cost and quality of the product. The power of the Internet is likely to grow, and its full potential gradually realized as many other complementary physical facilities, skills, and organizational capabilities fall into place. Companies and markets will need to be restructured, new business models assimilated, and the advantages of extraordinary amounts of information and computing power used to generate useful services. The next-generation Internet now taking shape under the initiative, once again, of the scientific community will create a global computer grid akin to the electrical grid and will provide instant access to a huge amount of computing power (IBM Joins Push 2001). Looking ahead-and probably well beyond this decade-the full impact of ICT on productivity and growth will be felt through a number of services that together will account for the lion's share of national output (box 5.1). These include finance, entertainment, health care, education, and government. In finance and entertain- ment the potential is already apparent, although a long way from be- ing realized. In the other three areas, a few countries have made a 116 CAN EAST ASIA COMPETE? start, including countries in East Asia, but technology, standards, reg- ulation, and content will take time to come together. In each case start-up costs will be heavy, and receptivity to electronic delivery of Box 5.1 Impacts of the Internet WHERE THE INTERNET MAY BE REVOLUTIONARY These information-intensive industries are good candidates for transformation by the Web: * Financial services. Most financial services can potentially be handled electroni- cally. Banks are struggling to find convenient ways to let people pay bills on- line with some success, although initial efforts, such as e-citi by Citicorp, have had to be scaled back. * Entertainment. Much entertainment can easily be digitized, but little of what is available is profitable, and the technological options are still evolving, as are business models that might sustain successful ventures. * Health ca7-e. The benefits of shifting health care transactions to the Web could be enormous, but so are the institutional barriers relating to, for instance, cer- tification, standards, privacy, and so on. * Educationz. E-learning could cut the costs of education, but only at the price of making education more impersonal, and with little guarantee that the end re- sults would improve on traditional approaches. * Governmenzt. Delivering information to citizens electronically and processing transactions, for example, tax applications, has enormous appeal, but requires massive investments. WHERE THE IMPACT MAY BE INCREMENTAL For these industries, information plays a relatively small role: * Retailing. Although Web sites and advertising have absorbed a lot of funding and attention, dot-com success turned more on the efficiency of logistics and warehousing and on pricing and marketing. * Manufactrining. Web-enabled supply chains and intranets are important, but ultimately a manufacturer's competitiveness depends on the uniqueness, pric- ing, and quality of its goods, as well as on associated services. * Tr-avel. Online travel sites are popular, but the real constraints on travel are the physical capacity of the air and road systems and the crowding of resort areas. * Power: Online energy exchanges attract publicity, but power generation and transmission capabilities will have the bigger economic impact. Sou7-ce: Adapted from Business Week, "Rethinking the Internet," March 26, 2001; see also Business Week, "How E-Biz Rose, Fell, and Will Rise Anew," May 13, 2002. ICT BEYOND THE HYPE 117 content, interaction, or transactions may be slow to materialize. Much will depend on how quickly businesses and other users assimi- late Internet-based technology. Much also rests on how regulatory policies shape the growth of the Internet and of the supporting telecommunications infrastructure. MAKING THE MOST OF THE INTERNET: THE STATE'S ROLE The growth of the Internet and the extent of its penetration into the business and household sectors in little more than a decade indicates how an environment with few entry barriers and regulatory impedi- ments can lead to rapid change. The attractions of e-mail, the devel- opment of browsers, the free sharing of software, and the readiness of private companies to capitalize on opportunities created by prior gov- ernment-sponsored research and investment have brought into exis- tence not just a new mode of communication, but what is almost a new lifestyle in middle- and higher-income countries. And this is only the beginning, because as the digital divide is closed, the Internet will multiply the sinews of globalization and come to embrace a majority of the world's population. As the scale, diversity, and value of Internet-based activities climbs, however, laissez-faire development is no longer either feasible or de- sirable. A framework of rules to regulate aspects of the Internet will be needed so that online transactions become as secure, orderly, and rule-bound as transactions made through other channels. Only through accepted rules can the opportunities that brought the Inter- net to its current level be preserved and its coexistence with different interpretations of state discretion and prerogatives be assured, but because the Internet is a new, rapidly evolving, and technologically protean phenomenon, much debate has arisen about what rules are desirable and how they can be harmonized among countries. Table 5.5, for example, indicates the level of disagreement within the U.S. population over the merits of state regulation of the Internet. Cross- country differences regarding the appropriate regulation of the Inter- net are likely to be even sharper. Uncertainty also prevails on how rules would be enforced and on whether the legal systems in place would be equal to the task, especially in the middle- and low-income Table 5.5 U.S. Public Attitudes toward Government Regulation of the Internet Percentage who think this issue is a problem and the Government Government Percentage who should do should not be think this issue is Percentage who Issue something involved not a problem don't know Dangerous strangers making contact with children 79 15 3 1 Availability of pornography to children 75 20 4 1 Availability of information about how to build bombs and other weapons 75 15 8 1 False advertising 62 20 12 4 Pornographyandadultentertainment 61 26 10 3 Ability to purchase guns 61 14 18 3 Loss of privacy 54 29 14 2 n Hate speech, information that attacks people based on their race, z religion, or ethnicity 53 27 15 5 r Violent games 51 31 15 3 uJ Source: Blendon and others (2001). > 0 0 m m ICT: BEYOND THE HYPE 119 countries. As indicated earlier, regulators and users share a desire to preserve many of the freedoms that have permitted and financed the extraordinary upsurge in innovation. On balance, the Internet's rede- finition of business activities and the social environment are viewed in a positive light, but in a number of areas consensus is emerging that a combination of public and private regulation would ensure better outcomes for the Internet's users. Any views regarding regulation of the Internet must remain highly tentative, not least because the scope of regulation and its feasibility are being constantly modified by technological advances occurring on a regular basis. Judging from the past decade's experience and from current concerns, six broad areas are likely to require some de- gree of regulation or legal rules. The need for such measures is likely to be universal, and it will become increasingly urgent in East Asia as countries increase their Internet use. These broad areas are the fol- lowing: * Rules governing the development of an ICT infirastru7Ictuzre that would support rapid growth throucgh mzultiple channels. These encompass ownership of physical telecommunications infrastructures, which in many countries are still in the public sector. Where they have been privatized, this has most often been done piecemeal, leaving those systems too in need of reorganization. The needed rules would also cover the assignment of bandwidth to providers, the li- censing of providers, and the setting of standards perniitting broad interoperability conducive to future advances. Finally, these rules would regulate the access to telecommunications services, the cost of placing calls, and the fees levied by Internet service providers (ISPs). This last is likely to be a function of an augmented compe- tition policy that takes into account the special characteristics of the ICT industry, including the contestability of its various seg- ments, based on start-up costs and the significance of scope and network economies. * Range and volume of infor-mation. The Internet is remarkable for the range of information available through Web sites and for the vol- ume of communications now conducted there. Inevitably a pro- portion of this traffic is repugnant to many, and some is considered deeply reprehensible by individual states (Hachigan 2001)." Thus 120 CAN EAST ASIA COMPETE? the regulation of access to certain information or messages com- municated via the Internet is a matter of concern to policymakers anxious to control the purveyance of pornographic material, for example. To forestall more intrusive state intervention, some firms respond to these concerns with self-policing mechanisms and tech- nology that permits content screening. Responsibility sharing re- mains fluid, however, with some states adopting a much stronger stance on content and the aggressive use of monitoring, filtering, and firewall and geolocation techniques. For instance, the Singa- pore government has an expansive set of regulations in place to manage the Internet and to screen content (Singapore: Internet Control 2001). Beyond a point, such control can inhibit Internet use, and the authorities in some countries walk a fine line in both promoting and inhibiting Web-based activity.'5 Upsutrge offrauid, cybercrime, and hacking. 16 This problem underlies the urgency felt about monitoring Internet traffic and Internet use. The terrorist attacks on the United States on September 11, 2001, sharpened concerns that the Web was becoming a tool used by ter- rorists to coordinate their activities and launder funds. If left unchecked, regulators fear, such clandestine use would inhibit Inter- net use for legitimate economic purposes; it could also substantially increase transaction costs. Major credit card companies in the Unit- ed States already adrnit to worries about fraud. Hackers who invaded computer networks in Southeast Asia and spread computer viruses such as the infamous Love Bug were responsible for losses equal to more than US$500 million in lost production and theft during 2000, an amount that does not even include the announcement effects on current or prospective users of Web-based services.'7 According to some estimates, unauthorized entry into computer systems doubles every 12 to 18 months (Southeast Asia: Cyber Threats 2001). Nu- merous technological fixes are being developed, such as better en- cryption techniques and special servers that sit beside routers and keep watch for traffic surges (One to Watch 2001), but regulators are also seeking to win agreement on rules that would require each computer to have a traceable identification, diminishing the degree of anonymity available to Internet users. Countries such as Malaysia and the Philippines are introducing laws making hacking a punish- able crime, but the real test will lie in the enforcement of laws and ICT: BEYOND THE HYPE 121 the kinds of punishment meted out to offenders. A survey conducted in the United States indicated that online credit card fraud was equal to 1.13 percent of transactions, 18 times the level on all other credit card transactions. This may be an underestimate, because credit card companies do not want to scare off customers (The Truth 2001). A survey by the Confederation of British Industry revealed that two- thirds of respondents had experienced a serious hacking, virus at- tack, or fraud, with only one-third of the companies surveyed believ- ing that business to consumer transactions are safe (Cybercrime Threat 2001). Regulators confront a dilemma, namely, too much privacy, and the sanctioning of rules and of associated technologies would open the door to more cybercrime, which the authorities would be powerless to tackle; too little privacy, and some of the users now contributing to the economic gains associated with the Internet will be driven away. Privacy. The previous point leads to the issue of privacy, arguably the single biggest worry of Internet users (Litan 2001). Continued rapid growth in Internet financial transactions and exchanges of sensitive information will depend on the assurance of privacy. Indi- viduals and companies must feel that information transmitted over the Internet is secure and only accessible to the parties for whom it is intended. If technology and government regulations permit eavesdropping, the Internet is less likely to thrive as a cyber mar- ketplace. In other words, if divulging information over the Internet increases the chances of fraud, theft, susceptibility to industrial es- pionage, or unwelcome attention from the authorities, fewer trans- actors will be attracted to the Web, and its utility as a fount for new business ideas will diminish. The U.S. Federal Bureau of Investiga- tion already uses the Internet surveillance DCS 1000 system Car- nivore, and the U.S. National Security Agency deploys Echelon to monitor e-mail traffic. Since the September 11, 2001, terrorist at- tacks, such surveillance is likely to have risen. * I7ztellectutal property and copyrights. Another privacy issue of concern to businesses is the protection of intellectual property and copyright (Litan 2001). The entertainment and software industries are most immediately affected by piracy, copying, and compression programs that make the sharing of music and videos possible. The eventual de- gree of regulatory oversight will depend on a combination of factors. 122 CAN EAST ASIA COMPETE? First, technological development will to some extent reduce the ease of copying or transmission or facilitate claims from producers for patent fees and copyright dues. Second, evidence is accumulating on the effects on sales of music and video copying. If, on balance, the ef- fect is positive, the case for regulation or encryption would carry little force. Third, developments in software might shift the main flow of earnings to customization and add-on services, which again would minimize the losses from piracy of packaged software. Fourth, legal action taken against Napster in the United States,'8 and international disputes being initiated under the Trade-Related Agreement on In- tellectual Property Rights in countries where piracy of software, mu- sic, and video is pervasive, demonstrates the relative power of legal remedies. In the process it is helping to map out the regulatory do- main, given the capacities of courts and enforcement agencies and the scope for international coordination of efforts to define and pro- tect intellectual property rights. * Taxation of services. The taxation of services traded by way of the Internet presents yet another reason for regulating Internet traffic. With services growing in volume, and with the possibility that us- ing the Internet to transact services will become even easier, gov- ernments are beginning to worry about the loss of tax revenue. Ac- cording to Tanzi (2001), in the United States, states stand to lose 4 percent of their sales tax revenues by 2004. The government's will- ingness to take unilateral action has been inhibited by the fear that attempts to monitor and tax business transactions would result in a flight of human and physical capital to other countries. Loss of revenue is a mounting concern, however, and some form of con- certed action is in the cards over the medium run, especially with the increasing availability of technologies that can identify the city from which a user is accessing the Internet. In sum, the phenomenal growth of the Internet since 1995 argues against heavy-handed regulation. Moreover, the experience of other networked industries in industrial and middle-income countries un- derscores the limits of regulation. In the United States, the advan- tages from increased entry following the deregulation of telecommu- nications and transport sectors altered choices for producers and strengthened incentives to innovate (Meyer and Menzies 2000). ICT: BEYOND THE HYPE 123 While markets often do need guidance, regulation is never a panacea. The dangers of corruption and stifled development are ever present, as is regulatory capture by the industry, which can lead to forbearance when trouble starts. As with the financial services industry, a judicious mix of market-based oversight and government regulation would seem desirable for the ICT sector, especially given the centrality of innovation and the uncertainty regarding its future evolution. Some of the technological advances aimed at enhancing Internet access will likely also address these concerns adequately.'9 Developing the software standards and equipment needed to provide transaction security and safety against hackers is in the industry's interests, but it is also becoming apparent that self-regulation and technology alone cannot allay the mix of concerns evinced by governments, producers, and users-again, concerns exacerbated by the September 11, 2001, terrorist attacks on the United States. In addition, the international harmonization of policies and standards will require sustained inter- governmental negotiation. The provisional and tentative lessons after just a decade of experi- ence are that governments should articulate a clear agenda of policy objectives. Rather than rushing in with poorly conceived, difficult to implement regulations, they must work with industry and users to find solutions that require the minimum of regulatory oversight and inter- vention. The situation is so fluid, the useful theory and empirical evi- dence so sparse, and the cost of undermining ICT development so substantial that countries will need to proceed cautiously. The ship of Internet regulation will have to be constructed on the high seas one plank at a time, with need and utility carefully weighed against the risk of a redundant superstructure that only adds to transaction costs. The least desirable approach for the East Asian countries-or, for that matter, for the industrial countries-to take would be to introduce broad and sweeping rules on the basis of limited information. These could become the source of numerous roadblocks in the future. Equally disadvantageous would be rules introduced under pressure from powerful, new, ICT-based industries looking to minimize com- petitive pressures. 124 CAN EAST ASIA COMPETE? FROM FIXED-LINE TO WIRELESS INTERNET ACCESS The last section discussed many of the policy questions that arise when individuals and firms access the Internet. Other areas of con- cern exist, however, of which the most central include the following: * The demand for Internet-based services, which in turn will deter- mine the scale and structure of a sustainable telecommunications infrastructure * The cost of Internet access, including not only the fees ISPs charge, but also local telephone rates * The financing and operation of the so-called Internet backbone, which connects networks of ISPs to one another * The financing and development of the last mile of distribution of high-speed, broadband services to households * The widespread use of different wireless technologies, the so- called second- and third-generation technologies (2G, 2.5G, and 3G), and within the decade, 4G.2") Before turning to each of these five issues in turn, recall that both private and state-owned firms have participated in running the Inter- net since it was commercialized in 1995, and that Internet use and ac- cess vary considerably across nations. By the end of 2001 the Internet had 500 million users (How E-Biz Rose 2002), with nearly 34 million in China alone, a 50 percent increase over the end of 2000 (Chronicle and Documentation 2002, p. 124). In East Asia, Korea and Singapore had Internet penetration rates exceeding 30 percent in January 2001, placing them 6th and 10th, respectively, in terms of national Internet connectivity (Iceland tops the list with a 60 percent penetration rate, followed by Norway and Sweden). In addition, active use of the In- ternet by East Asian individuals and firms ensured that 4 of the top 10 ISPs in 2000 were located in the region (Nifty in Japan and Chollian, Hitel, and Unitel in Korea). Worldwide, international Internet ca- pacity exceeded international telephone capacity for the first time in 2000. The former now stands at close to 300 gigabits per second, a fivefold increase over 1999, and yet another testament to the exten- sive investments in capacity that have occurred in recent years (ITU 2001b). ICT BEYOND THE HYPE 125 Demand for Telecommunications Services The Internet was initially seen as a means of sustaining communica- tions in the event of a nuclear attack, and later as a channel for com- munication among members of the scientific community (Naughton 2000). Its widespread appeal and use springs from the attractions and convenience of e-mail. This "killer app" produced a surge in demand for connectivity and for the associated telecommunications facilities. This prompted huge investments by telecommunications operators in fiber-optic transmission lines, wireless and satellite technology, and purchase of the spectrum, all in anticipation of increasing dce- mand for Internet access and for use of content. In the United States alone, 39 million miles of fiber-optic strand had been laid by 2001. However, demand, while rising, is not rising fast enough to absorb these investmients, and only 3 percent of the fiber was lit by 2001. Moreover, while digital subscriber line (DSL) and 2.5G wireless use is on the rise, barring the appearance of another killer app, the de- mand for broadband access and the utilization of capacity, both in the telecommunications sector and in the telecommunications manufac- turing industry, will continue to lag behind what now appear to have been greatly inflated market expectations. All this has had dire conse- quences for the profitability of surviving telecommunications and equipment suppliers. This is worrisome on four counts. First, to provide either broadband wireless or fixed-line service, a great deal of additional investment will be required above and beyond the hundreds of billions of dollars already poured into the telecom- munications business. Broadband wireless transmission and handsets and last-mile connections to the fiber-optic backbone will entail much additional spending. Whether these resources will be raised is unclear given the recent history of stock market losses of companies such as Qwest and Vodaphone and the collapse of WorldCom. Second, the likely contribution of broadband access to GDP growth or to productivity has yet to be discerned, let alone demon- strated, and until this is done convincing markets to provide addition- al billions in cheap capital will be difficult. Third, user surveys do not indicate that consumers have as yet re- vealed an inclination to surf the Internet using mobile phones. More- over, a survey of 2,400 users showed that only 4 percent thought that 126 CAN EAST ASIA COMPETE? they would ever use phones to spend money online, and only 2 per- cent had attempted to do so with their 2.5G phones (The Telecos Work 2001). Fourth, although NTT DoCoMo i-mode suggests that a demand exists for short messaging and the exchange of cartoons, news clips, and financial information, such traffic is scarcely the kind of killer app that will generate large profits. The current hope (and this changes on a monthly basis) is that voice-over DSL will give the telecommu- nications companies the demand they want, because most of their business is from voice services. Data, particularly financial data, pro- vide the bulk of traffic, but voice services are where the money is. Voice-over DSL would give small businesses multiple voice channels and Internet access all on a single line (FT Telecoms 2001). If telecommunications companies could combine a variety of attractive voice and data services with sophisticated billing technology that per- mits providers to charge optimally for type of service and customer, then, observers argue, the current squeeze on earnings would give way to an era of sustained growth (FT Telecoms' Billing Systems 2001). Costs of Internet Access The costs to consumers, firms, and governments of using the Inter- net to send and receive data messages-whether text, diagrams, or moving graphics-will be an important determinant of the competi- tiveness of clusters and the efficiency of logistics. The success of Sin- gapore as a transport hub and core of the fastest growing cluster in Southeast Asia, for example, has much to do with the creation of a state-of-the-art telecommunications infrastructure (Arun and Yap 2000).2' Internet access costs, including both the price of access and the costs of forgone time when downloading information, are them- selves in part determined by the quality and extent of a nation's com- munications infrastructure. Initially, fixed-line telephony was the principal infrastructure for accessing the Internet. Consequently, those East Asian nations with either little such infrastructure or highly regulated and inefficient telecommunications service providers saw slower diffusion of ICT. In addition, the emergence of the Internet has strengthened the case for ICT: BEYOND THE HYPE 127 moving away from state-owned telecommunications providers that face no competition, especially in the provision of local telephone services. The worldwide trend toward deregulation and privatization in telecommunications began in the 1980s and accelerated in the 1990s; however, East Asia still has much catching up to do.22 In 2000 only half of the telecommunications operators in the Asia-Pacific re- gion were privately owned, compared with 63 percent in Europe and 74 percent in the Americas (ITU 2000). Furthermore, just under 40 percent of local telephone service providers in the region experience competition. Finally, of the 96 new independent regulators created throughout the world in the 1990s, only 13 percent were in the Asia- Pacific region (for recent analyses of the Chinese and Korean telecommunications markets, see Lu 2000 and Jung 2000, respective- ly). This record is disappointing in light of cross-country empirical research suggesting that, even though the region could significantly increase telecommunications and Internet use by introducing compe- tition, privatization, or independent regulators, the biggest improve- ments occur after all three reforms are in place (Fink, Mattoo, and Rathindran 2001).23 Lowering the price of local telephone calls is perhaps the most im- portant step to reducing the cost of Internet access. Typically this re- quires allowing competition for local telephone services, with more than one firm using the telecommunications infrastructure to meet customers' needs. Setting the correct prices for access to this infra- structure is critical, as it also determines the funds available for fur- ther investments in infrastructure. Furthermore, without constraints on the pricing of access, a firm that both owns the telecommunica- tions infrastructure and offers local telephone service has a strong in- centive to charge high interconnection fees, thereby raising its rivals' costs. These rival firms wvill be further disadvantaged in the market- place if they pass their higher costs on to their customers. Not sur- prisingly then, pricing access to telecommunications infrastructure has tended to be placed in the hands of independent regulators. The specifics of pricing access-or of "interconnection" as the lit- erature refers to it-are extremely technical and involve several thorny issues. Only a sample of the tradeoffs involved can be conveyed here. Like most economic assets, access to telecommunications infrastruc- ture ought to be priced at the cost to society of the resources used at 128 CAN EAST ASIA COMPETE? the margin. What constitutes this marginal societal cost is the source of some debate. Should it be based primarily on the nature of the in- frastructure technology being employed, which is relatively easy to calculate, or should it also take into account the cost of the marginal increase in congestion created by the call? Policymakers appear to fa- vor pricing access based on congestion (OECD 1999). One advantage of doing so is that the infrastructure operator recoups some, maybe all, of its investment outlays. An alternative might be to adopt the for- mer noncongestion-based calculation for pricing access and to charge consumers a lump sum fee. Unfortunately, empirical evidence sug- gests that lump sum fees discourage telephone use among the poorest, and for the better-off, reduces demand for multiple household lines. However, pricing above technologically determined marginal costs, as in the congestion-based schemes, is likely to become less feasible over time (for an accessible account of pricing issues, see Wiseman 2000). The Internet Backbone As we noted earlier, global capacity to transmit information over the Internet has expanded rapidly. By and large, this investment has been conducted by profit-driven firms rather than state-owned corpora- tions. In addition to establishing ISPs, many firms have invested in the Internet backbone that connects networks of ISPs and enables the seamless transmission of data around the globe. To recover their investment outlays, backbone providers have moved away from so-called peering arrangements, in which the enti- ty sending messages across the backbone kept any revenues (such as monthly access fees to ISPs) derived from individuals or firms for transmission of their messages (for an account of the decisions of U.S. backbone providers in this regard, see Jew, Nicholls, and Floro 2000). This "sender keeps all" arrangement provided no compensa- tion to those providing the backbone infrastructure over which the messages were sent, and explains why access to the Internet backbone was said to be free.24 In the absence of any direct payments for access to the backbone, firms have little incentive to upgrade its quality. This has not yet become a critical issue; however, because recent in- vestment created a huge amount of unlit fiber-optic capacity, com- plete with optical circuits that, with suitable incremental investment ICT: BEYOND THE HYPE 129 in wavelength division multiplexing, could easily meet demand over the foreseeable future. The difficulty arises in deciding what alternative to the "sender keeps all" model should be adopted. A natural alternative would be for backbone providers to charge ISPs based on their use of a seg- ment of the backbone, with some adjustment for the quality of the backbone service. Thus, in principle, payment for backbone services could remain a matter for contracting between private firms. Fur- thermore, competition among backbone providers might create the impetus for upgrading the backbone and keeping access charges low. Unfortunately, private sector contracts alone may not lead to opti- mal outcomes. By moving away from sender keeps all toward charg- ing for backbone access, the cost of Internet access to individuals and firms will inevitably rise. Whether the charges are lump sum or per unit, Internet use will be affected.25 Essentially, however, this is the price to be paid for maintaining and upgrading the backbone. The central public policy question to be answered is what factors will de- termine access price. First, without intense competition between backbone providers-and recent consolidations among U.S. and Eu- ropean firms indicate that the market will be less than perfectly com- petitive-prices are likely to rise above socially optimal levels (how- ever, see Kende 2000, who argues that anticompetitive behavior is unlikely to emerge). Second, backbone providers that also operate ISPs may offer lower access charges to their own ISPs, perhaps enabling the latter to charge their subscribers lower prices than their competitors. Alternatively, backbone providers may charge much higher access prices to rival ISPs, again effectively undermining the rival's competitiveness. Given subscribers' sensitivity to ISP charges, these backbone providers can effectively tilt the playing field toward their own ISPs. Third, backbone providers may refuse access (or interconnection) to other backbone providers, hoping that subscribers will flock to their related ISPs (see Kende 2000 for a detailed discussion of these issues). (This actually happened in 1997 and 1998, when an Aus- tralian incumbent backbone provider refused to provide interconnec- tion to a new entrant.) The effect could be particularly powerful if- as is often claimed-the demand for an ISP's services is a function of the existing subscriber base. Such network economies provide a 130 CAN EAST ASIA COMPETE? strong incentive for incumbents to resist new entrants, whose at- tempts to lure away subscribers might quickly spiral into a substantial collapse in demand for the incumbent's services. Each of these three potential problems can be treated under stan- dard competition and regulatory policies. Oversight of mergers and acquisitions can tackle the first problem. Pricing access and refusals to deal are standard fare for regulators in network industries. In a sense, the provision of the Internet backbone raises new issues for policy- makers, but solutions lie in the well-established policies and best prac- tices that have evolved since deregulation began in the late 1970s. The Last Mile The last-mile challenge is to bring the full potential of broadband In- ternet access to every home user. To clarify what this means, we should start with what might be referred to as the first generation of Internet access: the telephone. The original means of Internet access for residential customers was to dial up an ISP over the telephone. However, the speed of such data transmission was extremely slow, around 9.6 kilobits per second, and as individuals sought to send ever more text, images, videos, and voice messages over the Internet, de- mand for faster connections grew. Initially integrated services digital network lines were introduced to meet this demand. This technology worked over conventional telephone lines and improved dial-up speeds to a maximum of 64 kilobits per second. Subsequently, asyn- chronous DSL and cable modems were introduced toward the end of 1999 in the United States, but take-up has been limited to 12 percent of ISP subscribers, a total of 8.1 million in 2001 (Broadband Internet Access 2001; ITU 2001a), rising to 10 million by early 2002. The last-mile challenge to bring the full potential of broadband to the home has a number of solutions, primarily wireless, cable, or fiber-optic lines. These technologies must be brought further into neighborhoods and combined with content-reducing or content- compressing technologies, especially for wireless devices (Chasing the Bottleneck 2001). The closer the fiber nodes are to the home, the easier it will be to provide service of 2 to 3 megabits per second, well above the 600 kilobits per second speed of current broadband modems. ICT: BEYOND THE HYPE 131 The U.S. experience with last-mile delivery of broadband service highlights the nature of the problems facing providers: the strength of demand and of business models, not of technology. In 1996 the United States created a class of competitive local exchange carriers to provide DSL service using the local networks of the Bell telephone companies. Because of low consumer uptake, poor business models, and resistance from the Bells, most of the competitive local exchange carriers have either been taken over or have gone bankrupt. The sur- vivors and the Bells have been left to struggle with the deployment of a mix of services at prices that will yield decent earnings. A good deal of market consolidation appears to be inevitable, with a small number of companies providing a mix of twisted copper wire, fiber-optics, ca- ble, and satellite-based services over the last mile. Currently, the ca- ble companies have the largest number of broadband subscribers and are best placed to offer concurrent video, voice, and data services (The Cable Guys 2001). Even though the number of subscribers is rising, the level of use and current basic rates do not generate suffi- cient earnings. For that to happen, people must be induced to sub- stantially expand usage: companies must provide content and services of a different order. In addition, pricing techniques and billing tech- nology have to evolve to generate incentives and to permit full ex- ploitation of the capacity to target content closely and to discriminate pricing (as achieved by NTFT DoCoMo in Japan). The physical aspects of the last-mile problem can be solved if the capital is forthcoming, whether by digging up streets or through wire- less means. The spread of broadband use, however, will depend on its perceived utility and on access pricing. For many East Asian countries, wireless technology might offer the easier mode of advance. Wireless Access Means of voice communication and the basic method of Internet access both changed substantially during the 1990s. Technological advances throughout the 1990s expanded the use of cellular phones and allowed the use of mobile telephones, television cable modems, satellite-based systems, and asynchronous DSLs to access the Internet. In 2000 about 3 percent of all voice traffic for new phone subscribers in the United States was wireless only, and this is expected to rise to 10 percent by the 132 CAN EAST ASIA COMPETE? end of the decade (Yen and Chou 2001). In Japan, by December 2000 two-thirds of Internet users employed mobile phones to access the In- ternet, compared with 30.7 percent for the older dial-up technology (ITU 2000). In mid-2001 China had 165 million fixed-line accounts and 120 million wireless subscribers (China Braces 2001). Almost 70 percent of Koreans had acquired mobile phones by the end of 2001, and of these more than 7 million were using them to access the Inter- net (Quantum Leap 2002; South Korea 2002). As figure 5.2 makes clear, even a relatively poor nation such as Indonesia has seen cellular phone subscriptions take off in recent years. Soon Indonesia will join the 40 other economies that had more mobile than fixed-line phones in use as of December 2000 JTU 2001a). The International Telecommu- nications Union estimates that in 2002 there will be 1.11 billion land lines and 1.4 billion mobile phone subscribers worldwide (ITU 2002). Between 1990 and early 2001 the number of mobile telephone subscribers worldwide increased 75-fold to 725 million people. These subscribers used 2G phones, based on either one of the competing global system for mobile telephony, time division multiple access, or code division multiple access (CDMA) standards, which allowed for much faster access to the Internet. An especially popular application of 2G phones is the short message service, which sends text from one Figure 5.2 Comparison of Number of Cellular Subscriptions with Fixed-Line Telephones, Indonesia, 1996-2001 8 b 6 0 1996 1997 1998 1999 2000' 2001' * Cellular El Fixed line a. Projected. Source: Business Asia, "Lifestyle in a Phone," May 29, 2000. ICT: BEYOND THE HYPE 133 mobile phone to another. In January 2001 alone, approximately 15 billion messages of this sort were sent around the world (ITU 2001 a). The use of mobile phones to browse the Internet was the next im- portant development. Some Web sites were adapted to fit on the small screens of mobile phones. In 2000 firms began using the Wire- less Application Protocol (WAP) and General Packet Radio Service to launch such browsing services, and by the end of the year 100 mo- bile operators were doing so. Take-up of WAP-enabled services was slow, however, with only 5 million subscribers by December 2000. This poor take-up rate arose from the shortcomings of the compres- sion technology, the pricing methods, the relatively slow speed of In- ternet connections to the World Wide Web, the small amount of content that was WAP accessible, and a shortage of handsets.26 The poor performance of WAP and General Packet Radio Service can be contrasted with the success of the i-mode service launched in Japan in February 2000. Adapting Hypertext Markup Language (HTML)-based Web sites for i-mode has been easier than doing so for WAP, and by opening its platform to third parties, NTT DoCo- Mo has encouraged the entry of content providers, including both registrants and informal users. By 2001 more than 40,000 i-mode- compatible sites were offering a wide range of content. An additional appealing feature of the i-mode service is that users pay only for in- formation received rather than for time online. Downloading streaming audio messages, still images, and videos still takes considerable time on 2G or 2.5G phones, and this has prompted the design of 3G services.27 While 3G services have a com- mon standard, International Mobile Telecommunications-2000 (ITU 2000), 3G mobile networks embrace three incompatible modes all based on the code division multiple access (CDMA) technology de- veloped during World War II: Universal Mobile Telecommunica- tions in Europe, CDMA-2000 in the United States, and wideband (W)-CDMA in Japan. These networks will support packet-switched, broadband Internet access and, with higher bandwidth, will provide downloading speeds of at least 144 kilobits per second and up to 2 megabits per second. Such services require the use of a higher-quality spectrum, which many nations (but not the United States) auctioned to potential 3G service suppliers in 2000. In light of the subsequent economic 134 CAN EAST ASIA COMPETE? downturn and the considerable technical difficulties in launching 3G services, the winners of these auctions appear to have vastly overesti- mated the commercial potential of 3G services, at least in the short run. In late 2001 NTT DoCoMo launched its Freedom of Multime- dia Access 3G service based on W-CDMA, with Korea following with its so-called IXRTT using CDMA-2000. Both have struggled to overcome teething difficulties and to deliver on their promises (3G by Any 2002). Providers in Europe and North America have faced similar obstacles. This delay reflects both technical difficulties and concerns about funding the deployment of 3G technologies, es- pecially since the lucrative mobile phone market has attracted sever- al new firms and profit margins have narrowed. Signs show that the Korean government, which had previously encouraged competition among mobile phone providers to build a sizable domestic market, now favors consolidation through mergers and acquisitions into a small number of large firms. The government is defending this change in policy on the grounds that only larger firms can afford to invest in 3G technologies; however, it risks sending the wrong sig- nals, dampening the entry of new firms and encouraging bigger firms to believe that the government will effectively underwrite their large investments in technology and service development. As with the Internet, state-sponsored, if not state-led, telecommunications development could easily deter technological change and drain ini- tiative from the industry. This is important, because the 4G technol- ogy currently being developed might, once deployed, save on invest- ment outlays while delivering far superior services. WIRELESS INFRASTRUCTURE, COMPETITION, AND POLICY The question, however, remains: Can a highly competitive market for telecommunications services attract the considerable funds needed to develop and maintain the infrastructures that competing suppliers need? Figure 5.3 illustrates the magnitude of these expenditures throughout the late 1990s. Part of the answer may lie in the success- ful bundling of Internet access with other services that would add to revenues. Furthermore, suppliers may be able to offer premium serv- ices that effectively enable them to price discriminate among their ICT: BEYOND THE HYPE 135 customers more effectively (see Greenstein and Spiller 1996 for evi- dence that U.S. households are willing to pay more for improved In- ternet access). The speed of data retrieval is likely to be the principal determinant of uptake, as those individuals and firms with high time opportunity costs will choose faster services. Another will be the quality and range of content that can be accessed, especially on imio- bile phones.2" A source of future funds and expertise for the East Asian market could be overseas telecommunications operators, once these have re- gained their financial strength. Concerns about national security- usually the stumbling block to allowing foreign investment-might be assuaged by creating joint ventures, including'separate sub- sidiaries, or other collaborative measures not involving takeovers by or mergers between foreign and domestic firms. One constraint on both domestic and foreign investments is legal restrictions that pre- vent private firms from owning telecommunications infrastructure assets. Providers of mobile phone services in Thailand, for example, can only set up networks for a fixed period of time, during which they Figure 5.3 Investment in Telecommunications as a Share of GDP, Selected East Asian Economies, 1990-99 C) J 0.025 _ = = _ _ _ _ _- 0 0.020 -__ E 0.01 0 - -t.< + = a S 0.005 _ _ E 0 I- 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 China Indonesia - r= Japan *e.e.. Korea --- Malaysia - Singapore Thailand Note: GDP = gross domestic product. Source: International Telecommunications Union database. 136 CAN EAST ASIA COMPETE? have to pay one of the two state telecommunications agencies, the Telephone Organization of Thailand or the Communications Au- thority of Thailand, 22 percent of their revenues. Furthermore, when the fixed period expires, the state agencies have the right to take over the private firm's network. Not surprisingly, this results in high monthly charges to Thai mobile phone users and reduced diffusion of the technologies, and has led to a reluctance to invest in marginally lucrative areas. As a consequence, e-commerce still accounts for just 0.04 percent of Thailand's GDP (E-commerce Emasculated 2000). The intensification of competition among telecommunications and other providers of ICT services will also call into question the fu- ture of unfunded mandates imposed on firms by regulators. The im- position of universal service requirements, for example, will further raise the need for public outlays, as telecommunications firms are un- likely to be able to continue to cross-subsidize loss-making services when competition in previously sheltered-and typically profitable- services intensifies. The funding of universal service requirements, effectively hidden before competition was admitted, would then be- come more transparent, in turn provoking a debate about whether other technologies can attain universal access more cheaply than fixed-line telephony.29 We turn now to the consequences for regulatory policy of the growing convergence in the services offered by telecommunications operators, other data transmission firms, broadcasters, and entertain- ment and publishing firms. Typically telecommunications and broad- casting firms have their own industry regulators and publishers do not. To the extent that the cost of supplying the same service differs across firms located in industries with different regulatory burdens, a degree of regulatory competition may emerge. Firms will prefer to supply the service in the industry with the least regulatory burden, and industry regulators will probably find themselves under pressure to re- spond by lowering regulatory standards. Such regulatory competition raises the question of whether devising common regulatory and com- petition policy principles for these three industries would be desirable. The implications of regulatory competition for the pricing of inter- connections are more stark. Reconciling the profitability of those firms that use a given network infrastructure (that are encountering consid- erable competition from sellers of similar services in other industries) with interconnection charges high enough to provide a market return ICT: BEYOND THE HYPE 137 on the investments made by the owners of that infrastructure may be- come increasingly difficult. As the latter's revenues depend on the ex- tent to which the network is used and the firms using the network can easily exit, the only viable option is to cut interconnection fees until enough firms supplying services over the network are profitable. If the final demand for services does not eventually rise enough so that inter- connection fees provide the infrastructure owners with an adequate rate of return, then perhaps policymakers should be prepared to let some network providers decline. The unpredictability of innovations in these industries-and their effects on the fortunes of firms-implies that plenty of shakeouts could occur in the digital transmission industries."(' Having adequate legal and financial institutions to facilitate exit (for example, bank- ruptcy laws) and entry (for instance, venture capital funds) will be an important prerequisite for benefiting from this technological conver- gence. A role for policymakers will be to ensure that the shakeout is not accelerated by the pricing and exclusionary practices of dominant firms, thus antitrust policymakers and enforcers will have to be vigi- lant. Even so, the end result may well be that a small number of firms-and possibly even one-will emerge as the dominant supplier of digital transmission services. Such an outcome could strengthen pressures to regulate dominant suppliers, but prior experience of the capture of similar regulators by corporate interests-resulting in an increased likelihood of frustrated entry by new competitors-should caution against taking this step. Instead, retaining vigorous competi- tion policy oversight of these dominant firms while ensuring that other firms can introduce new technologies probably offers the best chance for sustaining competitive pressures over the longer term. Looking ahead, an interesting area for research and policy debate will be the degree to which industry-specific regulatory structures can be replaced by cross-industry competition policies based on the experi- ence of leading OECD countries. INTERNET-BASED GROWTH IN EAST ASIA Computerization and the Internet, together with the development of telecommunications, are at the core of the ICT revolution. Invest- ment in computers and the telecommunications infrastructure has 138 CAN EAST ASIA COMPETE? already begun to contribute to GDP growth in East Asia; however, the real breakthrough in productivity is likely to come from the ex- pansion of e-business; e-commerce; and Internet use for finance, health care, entertainment, education, and governance. Transnation- al production networks and the initiatives of governments in coun- tries such as Singapore have accelerated the adoption of the Internet for these purposes, helped by investment in physical facilities, but as in the United States, business organizations and market institutions must adapt greatly before they can realize the full effects on cost re- duction and productivity. The trends are promising. Whether they will be sustained, however, depends on how these new developments surmount the slackening of economic growth in the medium term. For East Asia, regulatory action by individual governments will also determine how much is invested in telecommunications hardware, how widely the Internet is integrated into the business world, and how the standards and rules of the various economies will be harmo- nized across the region. NOTES 1. Taylor (2001) makes the general point that ICT is likely to take much longer than five years (the length of the boom years) to work itself through the US$10 trillion plus U.S. economy. 2. Electronic transactions amounted to US$336 billion out of the total of US$11.5 trillion in the U.S. economy at the end of 2000. B2B transactions reached US$90 billion in 1999 (Lucking-Reilly and Spulber 2001; Business Week, "Lessons of the Cyber Survivors," April 22, 2002). Worldwide B2B transactions reached nearly US$500 billion in 2001 and retail transactions climbed to US$112 billion (Busi- ness Week, "How E-Biz Rose, Fell, and Will Rise Anew," May 13, 2002). 3. Some of the adjustments are questioned by Kay (2001), who doubts their validity. 4. According to the Bank of Korea, 40 percent of the growth in Korea in the late 1990s was traceable to ICT (OECD 2000). 5. Rather than selling bits of software, Oracle now sees itself as providing Internet- based technology. Oracle's Chairman and Chief Executive Officer Larry Ellison claims that "what we are selling is cars. We have a marketing system that talks to the sales system, that talks to the service system that talks to the accounting sys- tem. All the pieces fit together. Our critics say 'But you haven't got the best windshield in the world.' That may be true, but our windshield fits in our car. That's the reason there's no air blowing on your face when you are driving" (Business Week, "The Oracle Speaks," February 26, 2001, p. 99). ICT: BEYOND THE HYPE 139 6. The unhappy experience and demise of Webvan.com, the online grocery busi- ness, highlighted the tentative nature of new business models. Webvan collapsed because it built up a huge infrastructure well ahead of demand and was not vig- orously frugal in managing inventory (Business Week, "Wcbvan Left the Basics on the Shelf," July 23, 2001). 7. Leamer and Storper (2001) maintain that because most transactions requirc long-term relationships, B2B exchanges will have to strugglc to survive. 8. More and more of the first-tier electronic component manufacturing and asseml- bly is becoming the prcservc of six large electronic manufactturing scrvicc providers, which actually make hardware for the likes of Sun, Intel, and IBM and now also for NEC and Canon. These are Celestica, Flextronics, Jabil Circuit, Sanmina, SCI Systems, and Solectron (Red Iei-ring, "Outsourcing Companies Benefit When Their Customers Are Under Stress," April 1, 2001). 9. For a selection of excellent papers on the effects of FDI in the region, sce Ito and Krueger (1999); for broader surveys of the latest research on international business networks see Feenstra and Rauch (1999) and Rauch (1999a); and for an extensive case study of one MNC's production network see Peking University, Tsinghua University, and University of South Carolina (2000). 10. For recent analyses of the behavior of Japanese MNCs, see Beechler and Bird (1999) and Mody, Dasgupta, and Sinha (1999). 11. Companies are reevaluating Internet advertising that has consuLimecd a large part of dot-coin budgets. Most of the effort is wasted, as click-through visitor rates are extremely low. 12. An extension of this is so-called expertise automation, which assists businesses by brokering connections between peoplc and by delivering specific skills when and where needed (Red Heoring, "From the Grouncl Floor," Marclh 6, 2001). 13. For instance, multiimodal transport capability and door-to-door delivery are still underdeveloped in the coastal areas of China, and logistics are weak throughout the interior of the country (Shaw and Wang 2002). 14. According to State Council Decree Number 292, issued in October 2000, each Internet content provider in China is to supply to the authorities on demand the content offered on their sites plus records of all those visiting their sites for up to 60 days prior to the date of request. In addition, Internet content providers are required to police their own sites and eliminate subversive material. In early 2001 China had 265,000 Internet content providers. For further details on the evolution of Internet use and its control see Harxvit and Clark (2001). 15. Filters and firewalls can be circumvented by dialing out to a foreign ISP, geolo- cation can be sidestepped by accessing sites using a computer in another cotin- try, and public key encryption can defeat eavesdropping (Economist, "The Inter- net's New Borders," August 11, 2001). Moreover, services such as Freenet promise anonymity by distributing files around the world on computers belong- ing to Freenet members, with the result that no file has a unique Internet ad- dress (Nezw Scientist, "Out of Control," March 25, 2000). Nevertheless, as Econzo- mist notes, most users are unable to employ stich measures, with the result that where governments desire, a high degree of control is possible. Interestingly, the use of much touted public key encryption software produced by companies 140 CAN EAST ASIA COMPETE? such as Entrust and Baltimore Technologies has bcen slow to catch on, because of the need for common architectures and new ways of doing business (Financial Times, "Information Technology Survey," September 5, 2001). 16. On the threat from hacking and computer viruses such as Code Red, see Scientific American, "Code Red for the Web," October 2001. The spreading use of Blue- tooth-enabled and Wi-Fi devices will increase the risks of unauthorized use and break-ins. 17. The denial of service attacks on Yahoo.com, Amiazon.com, and Buy.com in Feb- ruary 2000, with sites rendered inoperative by as few as 50 unwitting computers acting in concert, provided a foretaste of the vulnerability to attacks by hackers on businesses, public agencies, and key infrastructure nodes (OxfordAnalytica, "United States: Internet Insurgency," February 14, 2000; see also Information Security, "Fighting the Rising Tide," and "Deciphering the Cost of a Computer Crime," November 2000). 18. At its peak, Napster had 90 million users. Having had to pay damages for in- fringing songwriters' copyrights, Napster's future is uncertain. It also faces com- petition from new challengers and two industry-backed services: Music Net, backed by Warner Music, and Pressplay, backed by Sony and other companies (Financial Times, "Napster Closer to Music Deal," September 25, 2001). 19. Making predictions about technologies that will shape the future of the Internet and its communications infrastructure is risky. Certainly the 3G (and eventually 4G) technology, once it is widely deployed, will lead to a quantum leap in data transmission from today's 10 kilobytes per second to as much as 2 megabits per second, permitting streaming video and audio. But other technologies will criti- cally supplement advances in wireless networking. Voice browsers will make the use and navigation of the Web much easier. Short-range wireless coimlmunica- tion between equipment will be simplified by Bluetooth (short-range wireless technology permitting communication between electronic equipment, for ex- ample, phones and laptops) and other similar technologies. Peer-to-peer tech- nology is another genie that is set to grow and ramify. More complex e-com- merce applications and multitasking will stimulate demand for more powerful computers. Lastly, once they are widely adopted by individual industries as a standard comparable to hypertext markup language (HTML), extensible markup languages (XML), which describe each packet of information as a bar code does, will enormously ease the manipulation, sorting, and organizing of in- formation (see Economist Itelligence Unzit, "3G in Asia: Outlook and Opportuni- ties," January 10, 2001; New Scientist, "Everything Anywhere," October 21, 2000; Scientific American, "The Wireless Web," October 2000). 20. The 2G system uses the cellular structure and circuit switching of the first-gen- eration cell phones, but with the added advantage of digital transmission. The 2.5G technology employs packet switching and "always on" phones, but it is constrained by a narrow bandwidth that limits the transmission of graphics and video. The wide bandwidth of 3G phones will overcome this constraint and per- mit 2 megabit per second transmission of full-motion video. 21. For a recent exposition on the many ways in which telecommunications and transportation infrastructures influence the demand for one another in urban ICT: BEYOND THE HYPE 141 clusters, and more important, the spatial allocation of economic activity, see Wheeler, Aoyama, and Wolf (2000). 22. The next big step is the break-up of China Telecom into separate companies similar to the Baby Bells in the United States. These would provide a mix of services within regions (Red He77ing, "China Braces for a Break-Up," September 15, 2001). For a review of the extent and effects of telecommunications reform in developing economies see Noll (1999), and for its international antecedelnts see Drake (1999). The reform of local telephony is given special attention in Baumol and Sidak(1994).Joskow (1998) reviews the more general experience of reforming infrastructure industries in developing countries, many of which share the network economies feature of telecommunications. Holler (1999) pro- vides a review that is similar to that in Joskow, but focuses entirely on telecom- munications markets. 23. Boyland and Nicoletti (2000) providc an econometric account of complementa- ry reforms in the telecommliunications sectors of the OECD nations. Wallsten (1999) does likewise for selected Latin American ancl African countries. Li, Qiang, and Xu (2000) take a different approach and use statistical tools to locate the circumstances under which nations privatize and deregulate their telecom- munications sectors. 24. As many of the backbone providcrs were also ISPs, they earned revenues from the latter activity, effectively cross-subsidizing their backbone investmlients. 25. Recall that empirical evidence referred to earlier in this chapter suggests that higher lump sum access charges for the Internet reduce take-up among lower- income groups. 26. Aside from the slowness in transferring data, WAP is also hamiistrung by thc need to connect whenever calling up a new Web page. 27. Early attempts during 2001 to introduce 3G base station and handset technolo- gies by Nokia and NTT DoCoMo, for example, encountered teething difficul- ties that are being ironed out through software improvemiients. But the future profitability of providers rests on rapid increases in demancd for wireless data and e-mail services, availability of useful content, and advances in compression and transmission technologies facilitating the transmission of huge amounts of data. The latter are more likely to be solved quickly than are the former, with multi- path scattering using multiple antennae (each with its unique echo signatlre) of- fering a way of increasing long distance transmission of large volumes of data (Nature,."Talking Cheap in the City,"January 18, 2001). For a detailed account of 3G technologies and their commercial applications see Financial Times, "Un- derstanding 3G,"June 4,2001. 28. Korea, whose government has taken aggressive steps to integrate the nation into the knowledge-based economy, has funded relatively more of its investmcnts in ICT from public means than other East Asian economies. It is too early to say whether such public investments have attained a rate of return that more than. compensates for the forgone alternative use of these resources. 29. Crandall and Waverman (2000) thoroughly discuss the issues raised here in the context of telecommunications service provision in Canada, the Unitecd King- dom, and the United States. Although these authors do not consider the East 142 CAN EAST ASIA COMPETE? Asian experience, the technological and competitive pressures that they identify are at work on both sides of the Pacific. 30. Although the effect of shakeouts within a sector of the digital transmission indus- try will tend to raise the monopoly power of the surviving firms in that sector, the technological convergence across these sectors implies that the surviving firms will face greater competition from survivors in other sectors. Therefore, as shakeout and technological convergence have opposing effects on the market power of surviving firms, making strong claims for or against the futLre role of competition policy in the digital transmission industry is probably inappropriate. CHAPTER 6 MANAGING THE GLOBAL INTERFACE T he growing importance of services, the more intensive use of ICT, and the emergence of clusters as central poles of nation- al economic geographies all characterize the unfolding growth dynamic in East Asia. Previous chapters showed that this dynamic draws on the outside world by attracting FDI, skilled and highly educated professionals, some venture capital, and data and in- formation exchange. How East Asian nations can redesign their poli- cies toward the world economy-unilaterally and collectively-to take full advantage of this dynamic is the principal focus of this chapter. To make the most of the new growth dynamic, East Asian coun- tries must adopt a much broader set of market-opening policies. Without downplaying the importance of relatively open regimes for trading goods and investing in manufacturing, reliance on such arrangements is clearly no longer adequate for growth. Services sec- tors must also be exposed to greater international competition, in particular, the telecommunications sector, which is to international commerce in the 21st century what ports and railways were to com- merce in the 19th and 20th centuries. Another important change in mindset for policymakers and business people will be to consider a wider range of sources of value added in an information-intensive and globalizing economy. Many industries in the Group of Seven nations, which have faced considerable import competition (much of it from East Asia), have managed to reorganize themselves successfully, let- ting go of production stages dominated by lower-skilled workers. Ev- idence for such successful reorganization is apparent in global indus- tries such as textiles and apparel, electronics, and automobiles. By recognizing the existence of these opportunities and tailoring them to 144 CAN EAST ASIA COMPETE? the region's circumstances, East Asia can provide itself with a much needed antidote to the often heard pessimism about the effect of Chi- na's continued integration into the world trading system. REAPING THE BENEFITS OF OPEN TRADE AND INVESTMENT REGIMES While this chapter focuses on how East Asia can exploit its strong links with the world economy, we should not overlook the long- standing benefits that the region has garnered from its relatively open policies toward overseas trade and investment. The evidence of these benefits continues to pour in.' E. Kim (2000), for example, estimated the beneficial effects of trade liberalization on firm productivity in Korea and on reducing the market power of domestic firms as meas- ured by price-cost markups. In addition, Lawrence and Weinstein (2001) revisited the determinants of sectoral productivity growth in Japan during its growth spurt in the 1960s and 1970s, and found that import competition provided an additional stimulus to firms to en- hance their productivity. Such findings reinforce the case for trade reform. Another channel by which openness boosts domestic growth is through the importation of equipment. Mazumdar (2001) presents ev- idence that nontariff barriers reduce the importation of equipment and that this, in turn, depresses productivity growth in the importing nation. Unlike Lee (1995), a frequently cited study of the effects of equipment on growth in developing countries, Mazumdar adopts an approach that discriminates better between the contribution to pro- ductivity growth of domestically produced equipment and of import- ed equipment. Taking full advantage of cutting-edge equipment typi- cally requires well-educated and scientifically proficient managers and employees. Thus Miller and Upadhyay's (2000) confirmation that, among developing economies, the growth-promoting effect of open- ness is greater in nations with more human capital is not surprising. The amount of recent literature on the contribution of trade re- form pales compared with the mushrooming literature on the benefi- cial effects of FDI on East Asia's growth performance.2 Figure 6.1 in- dicates the scale of projected FDI inflows in 2001, as well as the stock MANAGING THE GLOBAL INTERFACE 145 Figure 6.1 Projected FDI, Selected East Asian Economies, 2000 and 2001 1,000 - 0~~~~~~1. 100- _ _1. 2.6 10.7 10.0 C 4. D 10 1&m <',X W (,5 9 '99 , m 3 X'~~~~~ NZ~X~ 29 O FDI stock in 2000 * FDI stock in 2001 Note: FDI = foreign direct investment. Source: Business Asia, 'Bucking the Trend," March 5, 2001. of existing FDI in several East Asian economies in 2000. In 2001 the total stock of FDI in mainland China was twice as large as that in any other economy in the region. Even though there has been some amount of "round tripping" of funds from the Chinese mainland through Hong Kong (China) (funds that are then invested back in the mainland and erroneously measured as FDI), the FDI stock in East Asia is now firmly concentrated in the economies of the northeastern part of the region, and this trend seems likely to continue. This concentration of FDI reflects not only China's opening up to foreign investment throughout the 1990s, but also liberalization in Japan and Korea after 1997. As figure 6.2 makes clear, the biggest changes in FDI inflows since 1995 occurred in Japan and Korea. Chi- na's FDI has risen approximately US$5 billion since 1995, while the rest of East Asia saw a slight fall in FDI inflows. This implies that the observed increase in the northeastern economies' share of the re- gion's total FDI inflows does not come at the expense of overseas in- vestment inflows into Taiwan (China) and the ASEAN economies. Nonetheless, some companies such as Seagate and Motorola are 146 CAN EAST ASIA COMPETE? Figure 6.2 Shift of FDI to Northeast Asia, Selected Years 90 80 70 ~n60 0 50- 40- IA 30 20 10 0 1988-93 1995 1999 2000 * Rest of East China *China EJapan rl Korea, Rep. of Note: FDI = foreign direct investment. Given the concern that much FDI into Hong Kong (China) has actually come from the Chinese mainland, this figure excludes FDI inflows into the former. Source: UNCTAD (2000; 2001, appendix B1). shifting some of their production facilities from Southeast Asia to China (East Asia Manufacturing Questions 2002). The relative importance of the numerous rationales for FDI differs across the economies in the region. Some foreign investors are moti- vated by a desire to supply the recipient nations' markets, whereas in many industries in China and in other lower-wage economies, such as Indonesia, FDI is principally directed toward export platforms (Feen- stra and Hanson 2001; Hill and Athukorala 1998). Although the lat- ter motivation is a long-standing one (see, for example, Wells 1993), in recent years a new development has emerged. As the examples in the next section will make clear, firms are systematically creating re- gional production networks that locate each stage of a production process in the economy in which that function is best performed (along some quality or cost consideration). This results in shipments of parts and components across more than one national border before the final product is assembled and delivered to customers, contribut- ing to greater intraregional trade flows. In Korea FDI is increasingly in the form of mergers and acquisi- tions, and such investments can be expected to introduce managerial MANAGING THE GLOBAL INTERFACE 147 innovations as well as potentially fusing Korean firms into MNCs' networks of activities throughout the world JETRO 2001; Mody and Negishi 2001; UNCTAD 2000). Likewise, foreign investment in Japan is playing a role in restructuring certain underperforming man- ufacturing and financial firms. These include GE Capital's acquisi- tion of the Japan Leasing Corporation, Renault's investment in Nis- san, and Daimler Chrysler's purchase of a stake in Mitsubishi Motors JETRO 2001).? In bothJapan and Korea increasing amounts of FDI are being directed toward the distribution sector, especially retailing, as evidenced by the aggressive investments made by the French su- permarket chain Carrefour, its British rival Tesco, and Wal-Mart. (On Wal-Mart's carefully orchestrated entry into the Japanese retail sector, see Wal-Mart Develops 2002.) These firms are bringing greater competition into one of the most inefficient sectors of these economies, much to the benefit of consumers (see, for example, A Hyper Market 2001, which describes the popularity of foreign retail- ers of everyday products among Asian consumers). Recent research has shed light on the factors underlying these FDI flows. Urata and Kawai (2000), for example, examined the differences in overseas investment behavior by small, medium, and large Japan- ese firms. Since the late 1980s smaller Japanese firms have invested abroad in larger numbers, joining the long-standing practice of the large Japanese conglomerates. Urata and Kawai confirm that, irre- spective of firm size, such overseas investment decisions are driven by the availability of low-wage labor, the quality of infrastructure in the potential recipient, the presence of good governance, and a sizable local market. They found evidence, however, that smaller Japanese firms were more sensitive to each of these factors than the larger firms. In addition, smaller Japanese firms place a greater premium on investing in areas with existing agglomerations of simllilar firms, rein- forcing the importance of clusters as the emerging paradigm of eco- nomnic activity in East Asia.4 The desire to control costs is an important determinant of overseas FDI by Japanese multinationals, according to evidence presented by Mody, Dasgupta, and Sinha (1999).5 Given that firms in East Asia fre- quently invest in each other's economies, this suggests that govern- ments in the region have strong incentives to keep down the costs of raw materials, of intermediate inputs such as business services, and of 148 CAN EAST ASIA COMPETE? transportation. This is best achieved by stimulating competition in the markets that supply these goods and services through trade re- forms, reassessing the benefits of long-standing domestic regulations, and relaxing restrictions on FDI. Evidence is also growing that both workers and local firms acting as suppliers gain from FDI. In their study of Indonesian manufacturing firms, Lipsey and Sjoholm (2001) found that, controlling for the dif- ferent characteristics of manufacturing plants, foreign firms pay some- what higher wages. Moreover, the competition for talent induces local firms to raise their wages as well. This beneficial finding should not be taken too far, however, as much research has shown that attracting FDI in the first place requires the availability of a large pool of skilled and educated workers (see Noorbakhsh, Paloni, and Youssef 2001). This research reinforces the importance of improving secondary and tertiary education systems, a point discussed in chapter 3. In sum, even though East Asian nations have reaped considerable benefits from relatively open regimes toward FDI, the magnitude of these benefits depends on the presence of the complementary fac- tors-human capital, competitive factor and service markets, and a focus on excellence in both service and manufacturing innovation- that have been the traditional focus of East Asian firms. These factors will be increasingly important as the region adjusts to China's inte- gration into the world economy, whose export growth parallels that of the rest of the region (figure 6.3). Even though China's integration into the world economy is unlike- ly to decimate the manufacturing industries of the rest of East Asia, firms elsewhere in the region will certainly be called on to adjust their industrial structures, to coordinate their commercial activities across borders, and possibly to relocate their labor-intensive production stages to China. Low tariffs and transportation costs, especially via air transport, make shipping parts and components across many borders before they are sent to consumers as finished products a reasonable option. More important, perhaps, is the quality of ICT, which facili- tates the coordination and organization of activities over large dis- tances. In the 1990s production networks or global supply chains, formed partly as a result of the development of ICT, allowed Western firms to respond better to rising import competition from every source, including East Asia. Understanding how these Western firms MANAGING THE GLOBAL INTERFACE 149 Figure 6.3 Growth of China's Low- and Medium-Tech Exports, 1990-2000 450- 400 - 350 300 0 250- x -0 200 - ___ ___ 150 - _ _ _ _ 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 * Developing East Asia except China [1 China Source World Bank Systemwide Initiative on Malaria and Agriculture database. responded to increased import competition may shed light on the op- portunities for East Asian firms and governments, but using this infor- mation will require that the East Asian firms take a broader view of commercial organization than has typically been the case to date. The first step in developing this enriched view is to recognize that ICT and efficient transportation systems permit more rapid respons- es to changes in consumer demand and the customization of prod- ucts. The U.S. textile industry provides perhaps the best example of the use of these advances to respond to import competition (Aber- nathy and others 1999, 2001). Producers handle consumers' wide range of tastes by offering a large variety of products, the fashion component of which yields greater profits than basic apparel prod- ucts. To avoid being saddled with lots of unsold stock, however, which is costly to stores and that must eventually be heavily discount- ed, thereby lowering profits, retailers and wholesalers are making greater use of ICT advances-principally bar code technology and computer programs that analyze sales data-to better predict de- mand. This permits making the product in smaller quantities and de- tecting changes in demand more easily. To meet particularized de- mand and minimize their inventories, retailers and wholesalers now 150 CAN EAST ASIA COMPETE? place a larger number of smaller orders with their suppliers, and they expect fast turnaround. Likewise, suppliers demand rapid responses from the firms supplying them with fabrics, buttons, dyes, and so on, as well as from logistics firms. Thus, for this section of the clothing market, the cost of production is not the only major determinant of a firm's competitiveness. Inventory management, accurate assessment of demand patterns, and rapid delivery (also referred to as rapid re- plenishment of stocks) are at a premium. Analysts argue that a firm's capacity to integrate all the relevant technologies (information, com- munications, and production) is critical to competing effectively in this industry (Abernathy and others 1999). For some East Asian firms, acquiring that capacity may initially require partnering with other firms closer to the cutting edge. This entire strategy, however, will be fatally undermined if governments have not provided ade- quate ICT and transportation infrastructures in the first place. Thun (2000) provides evidence that the Taiwanese counterparts of U.S. textile and apparel firms have ensured their survival by concen- trating value added in commercial activities other than the produc- tion process. To be sure, these firms have increasingly relocated pro- duction stages to the Chinese mainland to take advantage of the lower wage costs there, but they have also concentrated on earning their profits from what is becoming a large, distinct branch of man- agement practice, namely, supply chain management. That is, they locate suppliers in China, monitor the quality of their work, transmit orders and deal with buyers, organize shipments, and sometimes en- gage in rudimentary marketing. Their expertise in managing supply chains has grown to such an extent that some such firms have depart- ments that provide technical expertise and computer programs that enable other firms to better manage chains of suppliers and respond to demand shocks. The shoe industry in Taiwan (China) has evolved in a similar way. Hsing (1999) describes how product cycles in casual shoes grew from two to eight per year in the mid-1980s. This forced Taiwanese shoe manufacturers to fulfill smaller orders placed at more frequent intervals. Reinforcing this shift was the deliberate choice of Taiwanese shoe manufacturers to offer a greater variety of shoes, in- cluding fashion shoes made with synthetic materials, special orders, and work shoes.6 These firms responded by creating networks of shoe manufacturers in lower-wage East Asia economies and specialized in MANAGING THE GLOBAL INTERFACE 151 coordinating these networks, introducing new production techniques and materials and, where possible, exploiting economies of scope. The foregoing examples demonstrate what is often called value migration, that is, the process whereby the principal source of value added in a commercial activity shifts from one stage to another. In the examples, value added derives increasingly from organization, co- ordination, marketing, logistics, and the ability to forecast demand accurately, rather than from the actual production process. Feenstra and Hanson (2001) provide striking evidence of the profitability of being an intermediary in the supply of differentiated products. They calculated, industry by industry, the average percentage increase or markup in unit prices of goods first imported from China by Hong Kong (China) merchants, then processed or marketed, and finally re- exported. From 1988 to 1998 the median markups were relatively sta- ble, ranging from 28 to 34 percent, suggesting that substantial rents exist from coordinating economic activity in and beyond China. Although many of the same factors are at work in both the textile and automobile industries, one difference is apparent: the ability to integrate an enormous number of different parts and components in a seamless fashion and to continue to do so even when innovations in one part call for changes in others is a critical determinant of a suc- cessful automobile company. Dyer and Nobeoka (2000) analyzed how Toyota-a world leader in this respect-has accomplished this by fa- cilitating knowledge sharing among firms in a network. To encourage suppliers to share knowledge about innovations, Toyota invests heavi- ly in supplier start-ups and discourages short-term opportunism by retaining key suppliers over many years. In this case both the network organizer and the suppliers earn profits principally from innovations, and network formation and communication are themselves facilitated by ICT. The evolution of the East Asian HDD industry affords a coin- pelling example of how firms can compete on dimensions other than cost. As McKendrick, Doner, and Haggard (2000) have documented, during the 1990s suppliers of HDDs also competed on time to mar- ket schedules; time-to-volume (the time needed to attain a sufficient scale of production); and reliability.(or a reduction in defects). Per- sonal computer manufacturers demanded enhanced performance from HDD producers across all these dimensions. Locations such as 152 CAN EAST ASIA COMPETE? Penang and Singapore-which in the 1980s were the initial recipients of overseas investment in HDD production by American firms- found they could retain this industry even while wages rose as their economies grew. This was facilitated in part by the desire of U.S. HDD firms to maintain a portfolio of different production sites, which acted as insurance against shocks to any one economy or plant, whether caused by exchange rate movements, labor unrest, or inter- ruptions in supply or outbound shipments. This led to the creation of what McKendrick, Doner, and Haggard (2000) call a regional pro- duction system throughout East (principally Southeast) Asia. An important key to the continuing success of the Singaporean HDD industry, however, was the product's increasing technical com- plexity. McKendrick, Doner, and Haggard (2000, p. 165) argued that [NIo other location possessed the depth of engineering resources to make them. Singapore also assumed a more explicit role in developing and man- aging the regional production network, functioning as a transfer station for the introduction of new product(s). Finally, the country began to diver- sify into new niches, including media, drive design, and other branches of data storage. In short, Singapore was where new disk drives were tested, errors were corrected, and initial production runs were executed. Only after development in Singapore would the production of newer versions of disk drives move to different production sites in Southeast Asia, such as Malaysia and Thailand. Even then, some purchasers with specific or technically demanding needs maintained connections with Singa- porean firms for their small production runs.7 Competition from lower-wage locations is thus neither new nor a specifically China-related phenomenon. Many industries in Western nations have been successfully adjusting to such competition for decades. The dimensions of time, customization, and reliability have proven to be as important to competition as labor costs. Because of ICT, firms are better able to customize products, analyze demand patterns, and organize production and suppliers. Moreover, integrat- ing these different ICT-related functions is increasingly critical to commercial success. Government policy has played a role by keeping borders open, facilitating the creation and flow of ideas within and across borders, and ensuring that low-cost and high-quality logistics and communications facilities are open to firms. MANAGING THE GLOBAL INTERFACE 153 As discussed in chapter 3, efficient international air freight services are extremely important for production networks and supply chains. Without these services, regions effectively remove themselves from contention as potential production locations. In this context, for an East Asian economy to protect its national air carriers from competi- tion in its domestic market is highly disadvantageous. Protection lessens the airlines' motivation to improve its reliability and reduce transit and transshipment times. Developments in ICT have made processing orders and scheduling pickups and deliveries more effi- cient, as demonstrated by Federal Express's operations throughout East Asia. Indeed, part of the pressure on Federal Express to enhance its services comes from the competition it faces from UPS through- out the region, and now also in China. The Impetus for Regional Integration Whether driven by FDI, strategic alliances, or the formation of sup- ply chains or production networks, the forces described in the last section point to a thickening web of commercial ties across the Pacif- ic Ocean and within East Asia. These forces have not exerted them- selves in a policy vacuum. On the contrary, throughout the 1990s governments have taken a number of steps toward regional integra- tion. At the turn of the century, however, the taste for grander initia- tives has given way to far less ambitious proposals to form bilateral trade agreements. This section accounts for the current state of af- fairs and evaluates whether bilateral agreements are the best way for governments to take advantage of the commercial dynamics discussed earlier. Although our focus here is regional initiatives, bear in mind that East Asian nations could pursue trade and investment reforms either unilaterally or through negotiations at the WTO. The latter has tra- ditionally been the more politically palatable, principally because do- mestic export interests are often willing to support proposed WTO agreements that offer greater market access abroad. This path to trade reform lost considerable appeal, however, after the signing of the Uruguay Round. That round proved to be a monumental accom- plishment, bringing in new disciplines for trade and investment in services, intellectual property rights, and agricultural protection 154 CAN EAST ASIA COMPETE? rules, in addition to the usual reduced tariffs on manufactured goods. Countries signed this agreement as a single undertaking, which pre- vented them from cherry picking only those constituent agreements that best suited their interests (Hoekman and Kostecki 2001). More recently, industrial and developing economies were able to agree on an agenda for a new multilateral round. In November 2001 participants at the WVTO Ministerial Meeting in Doha, Qatar, agreed to begin negotiations on traditional market issues, including agricul- ture; regulatory issues; and matters relating to the implementation of previous agreements, an important item for many developing economies. Whether these negotiations will succeed, and whether the ghost of the failure to launch a trade round in Seattle in 1999 has been exorcised, remains to be seen. Furthermore, the lukewarm re- ception many observers gave to the Doha Ministerial Declaration no doubt reflects the growing dissatisfaction with the WTO as a source of trade reforms, dissatisfaction that continues in the run-up to the 2003 WTO Ministerial Meeting in Mexico. Unfortunately, neither ASEAN's nor the Asia-Pacific Economic Cooperation's (APEC) trade initiatives have lived up to prior expecta- tions. ASEAN was formed in 1967 with five founding members: In- donesia, Malaysia, the Philippines, Singapore, and Thailand. The ini- tial objective was to foster regional stability and promote political and economic cooperation. Throughout the 1990s, the latter goal came to dominate, and the original ASEAN members agreed in 1993 to form an ASEAN Free Trade Area (AFTA). Trade reform at first fo- cused on tariff reductions on manufactured goods. In 1995, however, ASEAN members agreed to expand AFTA to include services, intel- lectual property rights, investment negotiations, and nontariff barri- ers. Furthermore, ASEAN members agreed in 1998 to accelerate the rate at which tariffs were reduced to between zero and 5 percent, with the overall target for zero tariffs on manufactured goods set for the early part of this decade (around 2003). ASEAN's membership expanded throughout the 1990s to include Vietnam in 1995, the Lao People's Democratic Republic and Myan- mar in 1997, and Cambodia in 1999. (Brunei Darussalam joined in 1984 after gaining independence from the United Kingdom.) These new members are not expected to phase out tariffs and other trade barriers as quickly as the original members, generating fears that a MANAGING THE GLOBAL INTERFACE 155 two-speed ASEAN has emerged (Problematic Newcomers 2001). Vietnam has until 2006, while Cambodia, the Lao People's Democra- tic Republic, and Myanmar have until 2008 to reduce their tariffs be- low 5 percent. Another wrinkle has been the use of exclusion lists and similar methods to avoid phasing out tariffs on sensitive agricultural and manufacturing items. This ambitious program of trade reform has, however, been called into question in recent years. In January 2000 Malaysia stepped back from its commitments to reduce tariffs on automobile parts. Even though in May 2000 ASEAN ministers had agreed to allow Malaysia to retain these tariffs until 2005, confidence in the movement toward freer intraregional trade was shaken. Other members, notably Thai- land, have subsequently announced that they too are reconsidering their commitments to liberalizing their automobile industries in light of Malaysia's move. This backsliding effectively undermines plans to allow automobile parts to move across ASEAN's borders tariff-free, a scheme known as the ASEAN Industrial Cooperation Scheme that has attracted applications from 50 foreign investors, including Volvo, Honda, and Toyota (Free Trade Retreat 2000). Malaysia's move was blamed, in part, on the aftereffects of the East Asian financial crisis, which analysts claimed undermined the viability of its automobile industry. Given the severe downturn in Southeast Asian exports in 2001 and the slowdown in the world economy in 2001 and 2002, pressures for protecting national firms in the ASEAIN regions are likely to intensify, further postponing AFTA's likely com- pletion date. This gloomy outlook may explain why one of AFTAs most vocal supporters, Singapore, has turned to negotiating its own bilateral trade agreements. Looking beyond Southeast Asia, ASEAN's attempts to forge closer economic and political ties with three of its northeastern neighbors (China, Japan, and Korea) are still in their infancy. At the meeting of finance ministers of these so-called "ASEAN Plus Three" in Chiang Mai, Thailand, in May 2000, the ministers agreed to expand regional currency swaps and-despite the lack of progress in completing AFTA-to explore the possibility of a free trade area encompassing all their economies. To date, progress on the former has been greater than on the latter, and several currency swaps were signed at a meeting of finance ministers of the ASEAN Plus Three group in Honolulu in 156 CAN EAST ASIA COMPETE? May 2001. These included a US$2 billion dollar-won arrangement between Japan and Korea, a US$1 billion dollar-ringgit arrangement between Japan and Malaysia, and a US$3 billion dollar-baht arrange- ment between Japan and Thailand. These arrangements are in addi- tion to a previously agreed on US$1 billion intra-ASEAN swap arrangement, an existing US$5 billion dollar-won arrangement be- tween Japan and Korea, and a US$2.5 billion agreement between Japan and Malaysia. Thus in total, the East Asian nations have created a US$14.5 billion web of commitments to help stabilize their curren- cies and to ward off speculative attacks. More such arrangements are expected in the near term, especially between Japan and the Philip- pines (Currency Swaps 2001). While these arrangements sound impressive, for several reasons they represent only tentative steps toward regional monetary cooper- ation. First, their scale (US$14.5 billion) is tiny compared with the daily turnover in East Asian currency markets, thus whether these arrangements could prevent a stampede out of any one East Asian currency is an open question. Second, the US$6 billion of arrange- ments agreed on in Honolulu require International Monetary Fund conditionality before funds can be mobilized. This effectively renders these intraregional arrangements hostage to a Washington-based agency, circumscribing a regional response to potential future crises. Third, these arrangements do not require countries to cede any sov- ereign policy-setting rights, a concession that has been central to Eu- rope's success with intraregional exchange rate management. Diplomatic and strategic concerns seem to overwhelm trade and in- vestment priorities in ASEAN and its related forums, as evidenced by the cautious reception given to China's proposal for a free trade area with the ASEAN economies. Concerns that this potential free trade area might become a vehicle for greater Chinese influence in South- east Asia prompted suggestions that it be expanded to include Korea and Japan. In sum, ASEAN Plus Three appears to be going through a teething stage, and the scope of any durable regional free trade initia- tive is unclear in terms of both membership and of sectors covered. The forum that advanced intraregional cooperation and liberaliza- tion the most in the Pacific region during the 1990s was APEC. It was founded in 1989, when representatives of 12 nations, meeting in Can- berra, prepared a work program for future economic cooperation MANAGING THE GLOBAL INTERFACE 157 across the Asia-Pacific region.8 At subsequent meetings the group made progress on trade facilitation. China, Hong Kong (China), and Taiwan (China) joined APEC in 1991, and APEC's profile was further raised when the United States invited the heads of state of APEC economies to a summit in 1993, which marked a turning point in U.S. attitudes toward APEC. Afterwards the United States issued a "vision statement" that recognized the growing economic interdependence across the Pacific and within East Asia. Cooperation flourished on a range of trade, investment, and other issues, and commitments were fleshed out in greater detail at the subsequent 1994 and 1995 summits held in Bogor, Indonesia, and Osaka, Japan, respectively. Recognizing deeper historic divisions and rivalries within the re- gion, APEC has been founded on three core principles: consensus, voluntarism, and unilateralism. APEC members tend to move to- gether, by consensus, or not at all. Reforms agreed to in this forum are not codified in explicit regional agreements with rules, enforce- ment mechanisms, or formal ex post monitoring. Nations implement reforms unilaterally, principally through published individual action plans of measures taken in 15 regularly updated policy areas, but they frequently extend those benefits to nonmembers as well. This formu- la has been useful in encouraging nations to reform their economies and to shore up domestic support for their reforms. At the Bogor summit APEC leaders declared that their reforms are meant to move APEC members toward regional free trade: by 2010 for industrial na- tions and by 2020 for developing nations. Even those sympathetic to APEC's vision, goals, and methods, however, have become increasingly concerned about the organiza- tion's lack of progress toward its stated goals. Central among these concerns are the slow adoption of APEC's reform agenda, the mem- ber nations' failures to set specific and observable goals beyond their Uruguay Round commitments, and the weak mechanisms for evalu- ating members' actions and the lack of incentives for encouraging members to align their actions with collectively agreed on goals. In its first policy report the APEC International Assessment Network argued that APEC had developed a form of "soft institutionalism" where "[d]uring its first decade, APEC has created a set of norms, procedures and structures that define its essence" (APIAN 2000, p. 4). Their argument ran as follows: 158 CAN EAST ASIA COMPETE? [W]e believe that this soft institutionalism served APEC well during its in- fancy. Many of those who criticize APEC for not accomplishing more fail to understand the nature of soft institutionalism and why the region's real- ities allowed no other choice. We also believe that as APEC enters its sec- ond decade, it must constantly engage in self-examination. It must consid- er whether its soft institutionalism is facilitating decision making, whether the vision and mandates of the leaders are being transformed into tangible actions, and whether APEC officials are receiving the critical feedback in- tegral to sound governance. They also argued (pp. 4-5): "What may have been realistic at the out- set may have become an avoidable obstacle to further achievement. WXhat may have seen hopelessly idealistic at the beginning may have become more feasible as members gain confidence in APEC and in each other. What seemed dangerous may now appear comfortable and desirable." The Trade Policy Forum of the Pacific Economic Cooperation Council expressed concern about APEC's lack of progress. It found that "[g]enerally, the Individual Action Plans lack detail, or do not shed sufficient light on the medium and longer term policy develop- ments. Progress is more significant in areas where APEC has focused on collective actions" (Pacific Economic Cooperation Council 1999). This analysis revealed that the individual action plans contained only moderate coverage of nontariff barriers and few unambiguous com- mitments to liberalize services. In contrast, it found, by and large, clear commitments and programs in relation to reducing tariffs, lib- eralizing investment, removing impediments to the movement of business people, and implementing Uruguay Round reforms. Where APEC has sought to liberalize a particular sector, however, it has quickly accomplished concrete reforms. The first VWTO agree- ment on trade in information technology products, which reduced barriers to trade in this critical component of the knowledge-based economy, was originally negotiated among APEC members in 1996 before being quickly extended to include all WTO members at the Singapore WTO Ministerial Meeting. This agreement provided for the elimination of tariffs on all ICT-related goods by January 2000. Signatories to this agreement now represent more than 90 percent of the US$500 billion plus global ICT trade (Wilson 1998). A central question, however, is whether this liberalization of a leading sector of MANAGING THE GLOBAL INTERFACE 159 the East Asian economy provides a model for future negotiations within APEC. In sum, while APEC reform initiatives have accomplished tangible improvements in selected aspects of the trans-Pacific business envi- ronment, given its current schedule, future reforms are likely to fall short of the organization's potential. A greater prioritization of re- form initiatives, a willingness to codify commitments and to back them up with strengthened evaluation and enforcement mechanisms, and, perhaps most important of all, a move toward collective rather than unilateral commitment to implementation will be needed to turn APEC's long-term goals into reality. In what is perhaps a testament to current difficulties in stimulating regional and multilateral trade reform, several nations on both sides of the Pacific have turned to a third option: negotiating bilateral re- ductions in trade barriers (for an economic analysis of some proposed and potential bilateral and subregional initiatives, see Scollay and Gilbert 2001). Even Japan and Korea, those longtime skeptics of the benefits of preferential trade reform, have begun exploring potential bilateral arrangements with other East Asia partners (see, for exam- ple, details of Japan's change in trade strategy in Japan Ponders Free Trade Alliances 2001). From a regionwide perspective, the most optimistic scenario is that these bilateral agreements will grow in size, eventually yielding an East Asian or pan-Pacific arrangement. Without recounting the now familiar, but still relevant, debate about whether such preferential deals are stumbling blocks or building blocks to the completion of multilateral trade deals, a few points ought to be made (for contribu- tions to this debate, see Bhagwati and Panagariya 1996; Panagariya 1999; Winters 1999; World Bank 2000). First, the recent enthusiasm for signing codified agreements with bilateral trading partners is dif- ficult to square with the frequently heard reluctance to sign such agreements in APEC forums. Second, the amount of trade reform involved in some of the com- pleted deals is actually quite small, suggesting that bilateral arrange- ments will have a limited impact. The Closer Economic Partnership signed by New Zealand and Singapore in November 2000, for exam- ple, will eliminate tariffs on trade in goods. In reality, this means that Singapore must remove its one remaining tariff on New Zealand's 160 CAN EAST ASIA COMPETE? imports (on beer), and the only significant tariff barriers that New Zealand must eliminate are on textiles, clothing, and footwear, which account for a trivial share of their bilateral trade (New Zealand 2000). In contrast, freer trade in services is to be phased in over 10 years, with loopholes built in to delay implementation. In other bilateral negotiations, notably between Japan and its trading partners, agricul- tural trade tends to be excluded. The bilateral trade agreement nego- tiated between Japan and Singapore, for example, focuses on liberal- izing investment regulations and services sectors, and not on freeing trade in goods.9 Third, given the limited scope of some of the proposed arrange- ments, the estimated gains for the nations signing them are small. In an assessment of many potential bilateral agreements, Scollay and Gilbert (2001) estimated that the gains of tariff elimination to bilater- al signatories are almost always worth less than 0.5 percent of their GDP. As these findings include the effects of eliminating tariffs on agricultural goods, which few nations appear willing to countenance, the actual gains are likely to be much smaller. Meanwhile, the effects on countries not party to such arrangements are typically adverse: their exports suffer as a result of preferential liberalization by the sig- natories. Thus such agreements tend to have a deleterious "beggar thy neighbor" aspect to them. Fourth, bilateral arrangements may well spread only to those na- tions with sizable markets, and hence the greatest potential for export growth from such arrangements. Remarkably, certain members of the Japanese business community are unenthusiastic about the prospects of a free trade agreement with Korea because of the relatively small size of the latter's markets for certain products Japan/South Korea 2000). Furthermore, the inevitable exclusion of some sensitive sectors from bilateral arrangements suggests that the size of a nation's trad- able sector is a poor indicator of the size of the markets that might be opened up to competition. These points imply that the web of bilateral arrangements that might eventually develop may link only a small number of economies on both sides of the Pacific, creating opposing sets of "insiders" (who have signed bilateral deals) and of "outsiders" (who have not). Such a development would represent a step back from the goals and vision underlying the APEC initiative. MANAGING THE GLOBAL INTERFACE 161 By the end of the 1990s, clear differences were apparent between business-led and policy-led economic integration in East Asia. Busi- nesses across the region were expanding their activities across bor- ders, investing proportionally less in export-oriented manufacturing and more in services. Moreover, to the extent that they were engaged in manufacturing, they were increasingly fragmenting production across national borders, with labor-intensive activities concentrating in China. Firms increasingly built links with foreign counterparts through strategic alliances, joint ventures, and in some cases outright mergers and acquisitions. This proliferation of international com- mercial contacts contrasts with policymakers' stalled attempts to lib- eralize trade and investment flows across the region, and as we have seen, whether the recent interest in bilateral arrangements will even- tually lead to tangible regional reform is at best unclear. Faced with this dichotomy, a reassessment of the priorities underlying regional integration initiatives is in order. Toward a New Regional Commercial Initiative The slow progress in accomplishing reforms through APEC's indi- vidual action plans in recent years can be attributed in part to their overly ambitious scope, both in terms of sectors covered and of mem- bership. The bilateral response has been to narrow the scope along one dimension-membership-while preserving a fairly broad sec- toral coverage. An alternative response could be to narrow the scope of sectoral coverage, that is, to focus the sectoral coverage on those policies that directly affect the dynamic elements of the business landscape outlined in this study. In essence, this proposal would in- volve replicating APEC's successful formula with the Information Technology Agreement on a larger scale. Such a proposal would not only concentrate negotiating effort, but it would present an opportu- nity to tailor regional initiatives to the likely growth poles of the 21st century. Such an approach does not imply leaving tariffs on manufac- tured goods unaltered; as argued earlier, lower tariffs on parts and components underpin the profitability of production networks throughout the region. A new regional commercial initiative should give more prominence to services than has been the case in the past. As we have seen in previous 162 CAN EAST ASIA COMPETE? chapters, the economic role of educational, health, logistical, and trans- portation services will likely grow in importance in the near future. These services should therefore be considered in all negotiations on services, not just the routinely considered communications and finan- cial services, though the benefits of reforming the latter are sizable. Mattoo, Rathindran, and Subramanian (2001) found that economies with fully liberalized telecommunications and financial services sectors can grow up to 1.5 percent per year faster than those with more restric- tive policies. As well as reducing the costs to firms of purchasing these services, a reduction that in turn bolsters their competitiveness also leads to an increase in direct exports of services. Considerable scope exists for reducing barriers to trade and FDI in services in East Asia (Ito and Krueger 2000). Recent econometric analysis reveals the restrictiveness of national nontariff barriers in several services sectors (see Findlay and Warren 2001 for details). Table 6.1 estimates these barriers on a 100-point scale for several East Asian economies, and shows just how far Korea, the Philippines, and Thailand are behind world best practices. In the case of telecom- munications, figure 6.4 shows that Japan is the only economy in the region whose communications sector comes close to approaching the efficiency of its U.K. and U.S. counterparts. Chapter 4 emphasized the importance of a competitive national fi- nancial system in underpinning this new growth dynamic. Also of considerable interest is the role, if any, played by external reforms. Figure 6.5 summarizes the findings of one recent study of the effects on the margins charged by domestic banks of restrictions on overseas trade in banking services and on investments in banks (Kalirajan and others 2001). These border barriers can be expected to ease the com- petitive pressures on domestic banks, enabling them to charge higher margins and earn supranormal profits. Figure 6.5 shows just how dis- torting those barriers can be, for instance, Malaysian domestic banks charge interest margins 60 percent higher than would be the case in the absence of such barriers. The flip side to these higher margins is reduced loans to firms, some of which may be in dynamic industries. The objective of negotiations pertaining to the regional commer- cial initiative should be to devise horizontal disciplines that apply across all the services sectors. This avoids the tendency for service MANAGING THE GLOBAL INTERFACE 163 Table 6.1 Indexes of the Restrictive Effect of Policies toward FDI, Selected East Asian Economies and the United States (no restrictions = 0, maximum value = 100) Business Financial Transportation Economy services Communications Distribution services services China 36 82 28 45 46 Hong Kong (China) 2 35 5 23 9 Indonesia 56 64 53 55 53 Japan 6 35 5 36 11 Korea, Rep. of 57 69 63 88 57 Malaysia 32 42 8 61 12 Philippines 48 76 48 95 98 Singapore 26 52 25 38 25 Thailand 78 84 78 88 78 United States 1 35 0 20 3 Note: FDI = foreign direct investment. The economies are ranked by the simple mean of the sectoral indexes of FDI restrictiveness. Hence, for East Asia, mean restrictiveness is the lowest in Hong Kong (China) and the highest in Thailand. Source: Hardin and Holmes (1997). Figure 6.4 Degree of Protection in the Telecommunications Sector, Selected Economies 90 , 80 - 81 70 68 -s = n * C 70 > 60 0 44~~ 0 30 30 20 14 21 21 U 10 U, 0 3~ 9 6 6 C ( C 5 C Source: Warren (2001). 164 CAN EAST ASIA COMPETE? Figure 6.5 Effect on Net Interest Margins of Trade and FDI Barriers to Banking Services, Selected Economies 70- 60 60.8_ 50 47 49.3 40 ~~~~~~31.5 33.1 36~ 20 48 7 5 10 70U . . . eO