Industrial Policy in an Era of Globalization: Lessons from Asia (Policy Analyses in International Economics)

Industrial Policy in an era of Globalization

Lessons from Asia

Industrial Policy in an era of Globalization

Lessons from Asia

MARCUS NOLAND HOWARD PACK

Institute for International Economics Washington, DC March 2003

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page iv

Marcus Noland, senior fellow, was a senior economist in the Council of Economic Advisers and held research or teaching positions at the Johns Hopkins University, the University of Southern California, Tokyo University, Saitama University, the University of Ghana, the Korea Development Institute, and the East-West Center. He received fellowships sponsored by the Japan Society for the Promotion of Science, the Council on Foreign Relations, the Council for the International Exchange of Scholars, and the Pohang Iron and Steel Corporation. He won the 2000–01 Ohira Masayoshi Award for his book Avoiding the Apocalypse: The Future of the Two Koreas (2000). He is the author of Pacific Basin Developing Countries: Prospects for the Future (1990); coauthor of No More Bashing: Building a New Japan-United States Economic Relationship with C. Fred Bergsten and Takatoshi Ito (2001), Global Economic Effects of the Asian Currency Devaluations (1998), Reconcilable Differences? United States-Japan Economic Conflict with C. Fred Bergsten (1993), and Japan in the World Economy with Bela Balassa (1988); coeditor of Pacific Dynamism and the International Economic System (1993); and editor of Economic Integration of the Korean Peninsula (1998).

INSTITUTE FOR INTERNATIONAL ECONOMICS 1750 Massachusetts Avenue, NW Washington, DC 20036-1903 (202) 328-9000 FAX: (202) 659-3225 http://www.iie.com

Howard Pack has been a professor of economics and professor of business and public policy at the Wharton School since 1986, and professor of management there since 1995. He was a consultant at a number of institutions including the World Bank, the Asian Development Bank, the InterAmerican Development Bank, the Agency for International Development, and the Overseas Development Council. He was a fellow at the Harry S. Truman Institute for Peace Research, the Hebrew University, Jerusalem, and the Jerusalem Institute for Israel Research at the same university. He is on the editorial boards of World Bank Research Observer, World Development, Journal of Development Economics, and World Bank Economic Review. He is the author of Productivity, Technology and Industrial Development (Oxford University Press, 1987) and Structural Change and Economic Policy in Israel (Yale University Press, 1971).

Printed in the United States of America 05 04 03 5 4 3 2 1

C. Fred Bergsten, Director Valerie Norville, Director of Publications and Web Development Brett Kitchen, Director of Marketing and Foreign Rights Printing by Kirby Lithographic Company, Inc. Typesetting by BMWW Copyright © 2003 by the Institute for International Economics. All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by information storage or retrieval system, without permission from the Institute. For reprints/permission to photocopy please contact the APS customer service department at CCC Academic Permissions Service, 27 Congress Street, Salem, MA 01970.

Library of Congress Cataloging-inPublication Data Noland, Marcus, 1959– Industrial policy in an era of globalization : lessons from Asia / Marcus Noland, Howard Pack. p. cm. Includes bibliographical references and index. ISBN 0-88132-350-0 1. Industrial policy—Asia. 2. Asia— Economic policy. 3. Economic stabilization—Asia. I. Pack, Howard. II. Title. HC412 .N5747 2002 338.095—dc21

2002032236

The views expressed in this publication are those of the authors. This publication is part of the overall program of the Institute, as endorsed by its Board of Directors, but does not necessarily reflect the views of individual members of the Board or the Advisory Committee.

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page v

Contents

Preface

ix

Acknowledgments

xi

1

2

Introduction

10 11 15

Industrial Policies in Japan, Korea, and Taiwan

21 21 23 37 51 57 63

Industrial Policies: Some Preliminaries Industrial Policy in Japan Industrial Policy in Korea Industrial Policy in Taiwan Exports, Growth, and Productivity Appendix 2.1: Total Factor Productivity Growth

3

1

Definition Initiation and Maintenance of Industrialization The Case for Selective Industrial Policy

Unintended Consequences Political Economy Long-Term Use May Be Hazardous to One’s Health Corruption

67 67 69 72

v

00--Front Matter--iv-xii

4

3/14/03

10:10 AM

Page vi

Replicability Stability and Complexity Irreproducible Initial Conditions? Geography Is Not Destiny—Policy Matters Does the International System Now Constrain Industrial Policy?

5

Conclusions Final Thoughts

77 77 78 83 88 93 100

References

103

Index

111

Tables Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5

Table 2.6 Table 2.7 Table 2.8

Table 2.9 Table 2.10 Table 4.1 Table 4.2 Table 4.3 Table 4.4

Figures Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 vi

Human Capital Index and per capita income, mid-1950s Capital subsidy-investment ratio in Japan Government subsidy share of total research and development in Japan Effective rates of protection for Japan Impact of industrial policy on sectoral growth, capital accumulation, and total factor productivity in Japan, 1960–90 Normalized sectoral tax rates in Japan Science and engineering students Impact of industrial policy on value added, growth rate of capital stock, and sectoral total factor productivity in Korea, 1963–83 Intersectoral purchases, Korea, 1985 Purchases of domestically produced and imported machinery, Korea and Japan Distributional indicators, circa 1960 Export similarity, 1968 Rates of growth of total factor productivity Rates of growth of fixed capital stock

Sectoral composition of on-budget subsidies in Japan, 1955–80 Uses of FILP funds in Japan, 1955–98 Human capital accumulation, 1960–86 Rates of return on capital in Korea Export loan subsidy in Taiwan

25 27 30 31

35 37 40

45 50 51 83 84 86 87

26 27 39 43 53

00--Front Matter--iv-xii

Figure 2.6 Figure 2.7 Figure 3.1 Figure 3.2 Figure 4.1a Figure 4.1b Figure 4.1c

3/14/03

10:10 AM

Page vii

Average tariff burden in Taiwan Research and development in Taiwan, 1978–2000 Corruption rankings for Japan, Korea, and Taiwan Governance rankings for Japan, Korea, and Taiwan Endowment triangle: Labor, physical capital, land Endowment triangle: Labor, human capital, land Endowment triangle: Land, physical capital, human capital Figure 4.1d Endowment triangle: Labor, physical capital, human capital

54 56 73 73 80 80 81 81

vii

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page viii

INSTITUTE FOR INTERNATIONAL ECONOMICS 1750 Massachusetts Avenue, NW, Washington, DC 20036-1903 (202) 328-9000 Fax: (202) 659-3225 C. Fred Bergsten, Director BOARD OF DIRECTORS *Peter G. Peterson, Chairman *Anthony M. Solomon, Chairman, Executive Committee Leszek Balcerowicz Conrad Black Chen Yuan George David *Jessica Einhorn Stanley Fischer Maurice R. Greenberg *Carla A. Hills Nobuyuki Idei Karen Katen W. M. Keck II Lee Kuan Yew William McDonough Donald F. McHenry Minoru Murofushi Karl Otto Pöhl *Joseph E. Robert, Jr. David Rockefeller David M. Rubenstein Renato Ruggiero *Stephan Schmidheiny Edward W. Scott, Jr. George Soros Lawrence H. Summers Peter D. Sutherland Jean Claude Trichet Laura D’Andrea Tyson Paul A. Volcker *Dennis Weatherstone Edward E. Whitacre, Jr. Marina v.N. Whitman Ernesto Zedillo Ex officio *C. Fred Bergsten Nancy Birdsall Richard N. Cooper Honorary Directors Alan Greenspan Reginald H. Jones Frank E. Loy George P. Shultz

*Member of the Executive Committee

ADVISORY COMMITTEE Richard N. Cooper, Chairman Isher Judge Ahluwalia Robert Baldwin Barry P. Bosworth Susan M. Collins Wendy Dobson Juergen B. Donges Barry Eichengreen Jeffrey A. Frankel Jacob A. Frenkel Daniel Gros Stephan Haggard David D. Hale Dale E. Hathaway Takatoshi Ito John Jackson Peter B. Kenen Anne O. Krueger Paul R. Krugman Roger M. Kubarych Jessica T. Mathews Rachel McCulloch Sylvia Ostry Tommaso Padoa-Schioppa Dani Rodrik Kenneth Rogoff Jeffrey D. Sachs Nicholas Stern Joseph E. Stiglitz Alan Wm. Wolff

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page ix

Preface

In a 1990 Institute study, Latin American Adjustment: How Much Has Happened?, Senior Fellow John Williamson coined the phrase “the Washington consensus” to encapsulate the reigning agreement in policy circles about the desirability of a package of policies including macroeconomic stability, respect for property rights, sectoral tax neutrality, and trade openness. In the intervening years, the failure of the transitional economies to meet expectations, the Asian financial crisis, disappointing performance in Latin America, and falling per capita incomes in much of sub-Saharan Africa have spurred widespread disillusionment with these prescriptions and a reassessment of alternative development strategies. One such impulse has been a revival in interest among middle-income and least developed countries alike in the more dirigiste policies pursued by three Asian success stories: Japan, Korea, and Taiwan. In this volume, Marcus Noland and Howard Pack analyze the industrial-policy experiences of these three Asian economies to discern what, if any, lessons they might have for contemporary developing countries. They ask whether the selective intervention policies pursued by these countries were an important part of their successes, whether changes in the international economic system impede the ability of contemporary developing countries to implement similar policies, and whether prospective emulators would be likely to achieve the same outcomes that the Asians did. The conclusions of the authors are largely cautionary. Industrial policies, at best, enhanced growth modestly in the Asian economies. At the same time, they contributed to unintended negative side effects in the Asian economies. Differences in the economic fundamentals of contemporary developing countries in comparison with Japan, Korea, and Taiwan, as well as changes in the international economic system, constrain the ability of today’s developing countries to implement industrial policies in ix

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page x

the same way that the Asians did during their heyday. Hence their successes are unlikely to be reproduced in other countries today. The Institute for International Economics has done extensive research throughout its history both on “the Washington consensus,” especially its applications in Latin America, and on the economies of East Asia. On the former, one of our initial studies was Toward Renewed Growth in Latin America (1986) by Bela Balassa and three leading Latin American economists. Williamson produced a series of volumes, including the one cited here and The Political Economy of Policy Reform (1994). We will shortly release his new coedited volume on the topic After the Washington Consensus: Restarting Growth and Reform in Latin America (with Pedro-Pablo Kuczynski). On East Asia, major releases included Adjusting to Success: Balance of Payments Policy in the Asian NICs (1987, rev. 1990) by Balassa and Williamson; three major studies on Korean development, Korea in the World Economy (1999) by Il SaKong, Dynamics of Korean Development (1994) by Cho Soon, and Avoiding the Apocalypse: The Future of the Two Koreas (2000) by Noland; and several studies that addressed Japanese industrial policy including Japan in the World Economy (1988) by Balassa and Noland and Reconcilable Differences: United States-Japan Economic Conflict (1991) by C. Fred Bergsten and Noland. The Institute for International Economics is a private nonprofit institution for the study and discussion of international economic policy. Its purpose is to analyze important issues in that area and to develop and communicate practical new approaches for dealing with them. The Institute is completely nonpartisan. The Institute is funded largely by philanthropic foundations. Major institutional grants are now being received from the William M. Keck, Jr. Foundation and the Starr Foundation. A number of other foundations and private corporations contribute to the highly diversified financial resources of the Institute. About 31 percent of the Institute’s resources in our latest fiscal year were provided by contributors outside the United States, including about 18 percent from Japan. The Korea Foundation supported this study. The Board of Directors bears overall responsibility for the Institute and gives general guidance and approval to its research program, including the identification of topics that are likely to become important over the medium run (one to three years), and which should be addressed by the Institute. The Director, working closely with the staff and outside Advisory Committee, is responsible for the development of particular projects and makes the final decision to publish an individual study. The Institute hopes that its studies and other activities will contribute to building a stronger foundation for international economic policy around the world. We invite readers of these publications to let us know how they think we can best accomplish this objective. C. FRED BERGSTEN Director February 2003 x

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page xi

Acknowledgments

This monograph has its origins in a paper presented at a conference organized by the Central Bank of Chile. We would like to thank the organizers of that conference, Norman Loayza, Klaus Schmidt-Hebbel, and Raimundo Soto for inviting us to participate, and to the attendees, particularly Ronald Fischer, for helpful comments on that paper. We subsequently benefited from valuable comments provided by participants at the ERC/METU international conference in economics in Ankara, Turkey and at an IIE study group meeting in Washington, DC. Arthur Alexander, Fred Bergsten, Edward Lincoln, Ian Little, Daniel Tarullo, Larry Westphal, and Jungho Yoo provided instructive written comments on the draft version of the monograph. Scott Holladay provided exemplary research assistance.

xi

00--Front Matter--iv-xii

3/14/03

10:10 AM

Page xii

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 1

1 Introduction

The term “globalization” reigns supreme as a description of recent economic transformation—and it carries many meanings. In a technical sense, it is usually associated with an increasing proportion of national income devoted to international trade, hosting or providing larger amounts of foreign direct investment (FDI), and being the recipient or, less frequently, the source of portfolio investment. At the same time, globalization carries less precise, yet nevertheless important, connotations related to loss of national control, sovereignty, or identity in the organization of economic and cultural life. In the policy realm, the orthodox terms of engagement for the process of globalization have been enshrined in the “Washington consensus,” a shorthand for a set of policies including secure property rights, fiscal discipline, sectorally neutral tax and expenditure policies, financial liberalization, unified and competitive exchange rates, openness to foreign trade and investment, privatization, and deregulation, as described by Williamson (1990). Since then, however, disappointing results in Latin America, lackluster performance in the transitional economies, and the Asian financial crisis of the late 1990s have all contributed to a crisis of faith in these policy prescriptions, in Washington and elsewhere. One response has been to augment the consensus with “second-generation” reforms such as strengthening prudential supervision of financial markets or competition policy. Another impulse has been to reconsider the more statist policies that had been marginalized. As former Treasury Secretary Paul H. O’Neill toured Africa with pop star and development economics enthusiast Bono, Ugandan President Yoweri Museveni, who is deservedly known for his com1

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 2

mitment to improving economic policy, raised a question “that both African and American officials are just starting to talk about: what brand of capitalism should Africa pursue as it wades into global competition. . . . Africans sometimes look more to Asia than to the United States in thinking about how to advance their development plans. The model for development in much of Asia has included heavy government involvement in nurturing industries and companies that could become globally competitive and stimulate job growth.”1 In this regard, Japan, Korea, and Taiwan are promoted as the poster nations that have derived great benefits from increasing integration with the international economy without surrendering national autonomy in the economic or cultural spheres—in effect beating the West at its own game. As one commentator put it, “The East Asian tigers . . . were free to do their own thing, and did so, combining trade reliance with unorthodox policies—export subsidies, domesticcontent requirements, import-export linkages, patent and copyright infringements, restrictions on capital flows (including direct foreign investment), directed credit and so on—that are largely precluded from today’s rules” (Rodrik 2001, 28). The mechanisms partly foreclosing these policy options include the World Trade Organization (WTO) and “the ubiquitous role of the World Bank and IMF [which] make it harder for governments dependent on these institutions to depart from the orthodoxy” (Rodrik 1999, 8). This encapsulates the fundamental questions addressed in this monograph: Was industrial policy defined as choosing individual manufacturing sectors to foster indeed a major source of growth in these three economies? If so, can it be replicated elsewhere under current institutional arrangements—and if so, is it worth replicating, or would developing countries today be better off embracing the suitably refined orthodoxy? Much of the dominant view was based on the perceptions, circa 1990, of the policies that had led to rapid growth in a few Asian countries. Although many developing countries had experienced some spurts of growth between 1950 and 1973, often in conjunction with the end of colonial status, the period after 1973 was generally dismal. Sub-Saharan Africa, with the exception of Botswana, experienced a significant decline in per capita income. The 1980s saw no growth in per capita income in Latin America, partly as a result of extensive borrowing to pay for more expensive oil during the 1970s and the consequent default on sovereign debt in the early 1980s that led to contractionary policies. Although by 1990 China had undergone a significant acceleration of growth, much of

1. Robert W. Stevenson, New York Times, June 2, 2002. The allure of intensive government guidance was apparently not diminished by the disastrous experiments in neighboring Tanzania and many other African countries.

2

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 3

it had been due to the liberalization of the agricultural sector and the impact of foreign direct investment that had been enticed into special economic zones.2 The focus on the smaller Asian economies and Japan as potential models for growth was partly an outcome of a sometimes desperate search for a coherent formula for growth. The Washington consensus crystallized an emerging understanding among both academics and policymakers of the factors underlying Asian successes. Some important aspects of Asian policy were disregarded, particularly the considerable intervention that had occurred in product and factor markets and the political stability—initially authoritarian, then progressively more democratic—in Korea and Taiwan. The emphasis was on the role of what were thought to be relatively unhindered market forces. Empirical analysis of countries that had earlier attempted to intervene extensively had shown that it would be very difficult for governments to successfully intervene in a continuous fashion. This skepticism reflected a huge empirical literature documenting the ubiquitous failure of the strategy of import-substituting industrialization (ISI) that had been pursued by many countries, including early attempts in Korea and Taiwan.3 To draw an analogy from finance, an investment bank, a brokerage house, or an actively traded mutual fund might be able to identify one winning sector or set of companies for a few years, but over long periods such efforts rarely improve on the performance of a passively managed index fund. Likewise, the Washington consensus reflected doubt that a government could systematically intervene in a manner that accelerated growth, even though it might make the occasional lucky pick. It is important to underline that in this volume we attempt to evaluate one part of the overall economic growth strategy of Japan, Korea, and Taiwan, namely, the effects of their attempts to discriminate among various industrial sectors. As will be seen, we are skeptical that this was the major key in their success. We do not question that all three countries were committed to sustained economic development and pursued this goal in many ways, from investing in infrastructure to improving the education system—perhaps skewing it toward a general favoring of the industrial sector. These instances of intensive devotion to economic development often produced what might be called generic infrastructure, for example, roads that could be used to ship any product, not just those of a specific sector. Similarly, the education received by chemical engineers enabled them to enter any of a large number of sectors, ranging from food processing to industrial chemicals. The effort in these directions may indeed

2. Small rural (“township and village”) industrial enterprises also played a significant role. 3. For comprehensive surveys of major studies see Balassa et al. (1982), Bhagwati (1978), Krueger (1978), and Little, Scitovsky, and Scott (1970).

INTRODUCTION

3

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 4

have been greater in these nations, but they are not synonymous with the favoring, say, of heavy chemicals over the canning of jams and jellies. This latter aim requires narrowly targeted, sector-specific policies. Chemical engineers could have supervised the production of either jams or sulfuric acid, and either product could have been shipped along the same roads and through the same ports. More generally, Japan, Korea, and Taiwan pursued policies similar to those of many newly independent countries after World War II. Compared with most, however, they had much better macroeconomic policies (World Bank 1993). In addition, although all three countries protected their domestic markets, they also relied more intensively than other developing countries on exports, an emphasis that required exporting firms to improve their productivity and product quality in order to compete in international markets. Efforts to improve the competitiveness of exported products inevitably affected the productivity of output destined for the protected domestic market as well—firms did not establish separate production facilities for exports and domestic production. Moreover, many of the numerous interventions to encourage exports were designed to offset some of the protection, including rebates for the higher price of protected intermediate inputs employed in the production of exports.4 Many have argued that this aspect of the system was unnecessarily cumbersome and that the same result—neutrality of incentives between production for the domestic market and for export—could have been achieved by low uniform protection combined with a subsidy on value added in exports.5 The three Asian nations intervened fairly heavily for varying lengths of time, as did all other developing countries with the exception of Hong Kong. Thus we are afforded the opportunity to observe perhaps 100 experiments involving systematic government intervention in developing countries, or more, given that many countries halted such intervention only to resume it during difficult periods. After half a century, three nations may have succeeded by using extensive selective intervention—four if Singapore is included, though its reliance on FDI requires a different analysis. If a country were nevertheless to insist today on pursuing selective industrial policy (defined in detail below), it would have to identify those unique aspects of it in Japan, Korea, and Taiwan that may have led to suc-

4. In addition to rebates on intermediates, lower interest on loans that financed export production were used, as were export targets. 5. In fact, there may have been some slightly positive net incentive given to exports, as the package of special incentives slightly more than offset the protection of intermediates used in the exporting sectors (Westphal and Kim 1982).

4

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 5

cess. Since almost every policy pursued by those countries was adopted at some point in one or more ISI countries, it would be necessary to identify the combination of policies that led to success.6 Even if that could be done, it would still be necessary to avoid the long-run costs that have revealed themselves in the past decade, particularly in Korea and Japan, which are discussed in chapter 3. While the odds of achieving successful development through detailed interventions are much better than those of winning the lottery, they are not particularly favorable. Before delving into industrial policy itself, we should take note of the danger of a “post hoc, ergo propter hoc” fallacy with respect to Japan, Korea, and Taiwan. They have all intervened, Japan and Korea very intensively, and they are among the few developing countries to have become relatively rich. It is thus tempting to conclude that industrial policy was a key to their growth. But as many analyses have noted, these countries also differed in a number of fundamental determinants of per capita growth, including investment relative to GDP, public and private education expenditures, stable real exchange rates, low inflation (relative to other developing countries), and initial endowments.7 Their investment rates, for example, were 30 percent or more for much of the post-1965 period, roughly twice the typical investment rate in less-developed countries (LDCs). While high investment does not guarantee rapid growth— the former Soviet Union comes to mind—it is unlikely that a country can achieve 9 percent GDP growth year after year without it. An alternative view of these Asian success stories is that they are largely the result of getting macroeconomic policies right: responsible government monetary and fiscal policy, low inflation, and maintaining the correct real exchange rate were key to their success, as was the con-

6. Most observers agree that a distinguishing characteristic of Asian industrial policy was its use of export performance to monitor the success of individual firms. Growth in exports was a demonstration of quality and price competitiveness, although at any given moment this competitiveness may itself have stemmed from extensive subsidies, including those provided by consumers in the protected domestic markets. Insofar as subsidies were not growing over time as a percentage of the f.o.b. price of exports, subsidies could not account for the growth of exports. There were other differences with respect to the ISI countries as well. 7. Some would argue that an important dimension of their success may have been undervalued exchange rates, particularly for Japan and Taiwan. Insofar as such undervaluation led to continuing export surpluses, they imply that domestic saving (private and public) exceeded domestic investment. But this pattern would normally be attributed to macroeconomic policy rather than sectorally selective policy. On a related point, these countries experienced less real exchange rate volatility than other major developing countries; lower levels of uncertainty about relative rates of return across alternative activities in turn reduced the riskiness of investment, which contributed to growth. Again, the maintenance of stable real exchange rates is really an issue of macroeconomic policy rather than industrial policy.

INTRODUCTION

5

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 6

siderable investment in the education system.8 Growth was propelled largely by physical and human capital accumulation and the growth rate of total factor productivity (TFP), while not spectacular, was high by LDC standards. It is difficult to sustain the synthesis of these views—namely, that the industrial policies were the source of superior macroeconomic performance. Some argue that policy in the Asian countries has had a “proproducer” bias and that this may have contributed to their growth performance by increasing incentives to save, providing their firms with a ready supply of low-cost capital.9 This argument is seldom if ever formalized, however, and while it has some superficial plausibility, it is hard to reconcile with the life cycle hypothesis and with research on Japanese saving behavior, which, for example, has not uncovered links between industrial policy and national saving.10 That said, an interesting paper by Yano (2001) demonstrates that in a dynamic two-country model, a lax competition policy with respect to the nontraded sector of a large tradesurplus economy can act as a “beggar-thy-neighbor” policy, shifting real income to itself from its trade-deficit partner. Some fraction of this income may find its way into either government or private saving. Some of the more general issues associated with the interaction of micro- and macroeconomic policies are discussed in chapter 3. The disagreement between those who believe in the efficacy of industrial policy and those who maintain that economic fundamentals were critical is, at one level, unbridgeable, as it would require agreement on the evolution of sectors and productivity in the absence of industrial policy. Nevertheless, the considerable body of evidence available from attempts to empirically assess the impact of industrial policy brackets most of the plausible counterfactual scenarios. The neoclassical view that success was the result of getting the fundamentals right may be correct, but it must deal with the abundant evidence that the Asians were indeed interventionist (Komiya, Okuno, and Suzumura 1984; Pack and Westphal 1986; Wade 1990; World Bank 1993). For the revisionist view, the issue is whether the documented use of industrial policy can be shown to have 8. The second-generation reforms that have been widely urged since the onset of the financial crisis in 1997, such as better prudential bank regulation, more transparent corporate governance, and more vigorous competition policy, were clearly not part of the institutional structure of these nations, although some interventions, such as regulation of the capital market, had impacts that to a limited extent led to some desirable results, such as limiting risk taking by borrowers (Stiglitz 1993). 9. A largely closed capital account up through the mid-1980s facilitated the maintenance of a pool of captive saving, although this is not absolutely necessary if there is home bias in portfolio allocations. 10. See Balassa and Noland (1988, chapter 4) and Horioka and Watanabe (1997) on this point.

6

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 7

been a quantitatively significant contributor to welfare. Thus the contribution of industrial policy can be understood as its impact on the level and productivity of physical and human investment. If a country would have achieved a growth rate of 9.7 percent a year from conventional policies that encouraged the accumulation of physical and human capital and from TFP growth, and instead realized 10 percent as a result of its industrial policy, the latter, while not cumulatively unimportant, is a relatively small part of the growth story. If the growth rate would have been 7 percent in the absence of selective industrial policy because of lower factor accumulation and TFP growth, the benefits are clear.11 Our reading of the extensive evidence we present in chapter 2 is that the interpretation in which selective industrial policy is seen to have accelerated growth only slightly, at best, is the most persuasive one. Despite the important macroeconomic dimensions in which Japan, Korea, and Taiwan differed from other developing countries, some analysts suggest that the Washington consensus may have been wrong and that sufficiently competent governments could identify profitable sectors, channel resources to them, and accelerate growth rates. If this is indeed the case, some of the pillars of the Washington consensus are undermined. Sectorally neutral tax and expenditure polices may be wrong, financial institutions should perhaps be encouraged to lend to targeted sectors rather than aim for sectoral neutrality, and tariff protection to allow firms to move down their average cost curves may be justified. Such intervention is viewed sympathetically in many poorer countries where the program embedded in the Washington consensus, sometimes derided as “neoliberalism” or “the Anglo-American model,” is viewed with deep suspicion as part of an attempt to institute “neocolonial” policies or worse. Often the suspicion is expressed in terms of doubts about “globalization,” interpreted as an emphasis on growing international trade in goods and services: opponents are convinced that local firms can neither compete with efficient imports or export to other countries without a sustained period of government fostering of the industrial sector. Thus it is important to understand the role of industrial policy in Asia, to obtain some indication of its success or failure, and, if it is found to have been successful, to identify both the economic and political components that contributed to its success. As noted earlier, this skepticism finds support in an extensive literature on the widespread failure of ISI policies in many countries. The World Bank’s East Asian Miracle volume (World Bank 1993) confirmed the emerging evidence that considerable intervention had occurred in Japan, Korea, 11. This implies that the 35 percent national saving rates and the passion for education reflected improved profit and wage opportunities that were generated by industrial policy or the lower risk attached to a given prospective rate of return. We discuss this in chapter 4.

INTRODUCTION

7

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 8

and Taiwan—they were hardly the conventional laissez-faire economies imagined by some. On the other hand, it was difficult to document a significant quantitative contribution of these policies after taking account of the fundamentals, such as high investment rates. Partly catalyzed by publication of The East Asian Miracle, an enormous amount of empirical research on the effect of selective industrial policy has since been conducted. This literature has generated a considerable amount of new data and provided many new tests of the impact of industrial policy. Our purpose in this volume is to provide a road map for policymakers in contemporary developing countries on the potential costs and benefits of industrial policy by incorporating this new evidence into a coherent framework. We do not attempt to sort out doctrinal disputes. Our point of departure is to assume that the broad facts of intervention are correct even though some of the interpretations have recently been disputed.12 We do not start from the position that industrial policy cannot work. On the contrary, one of us coauthored a paper that interprets early Korean industrial development as partly the result of clever industrial policy (Pack and Westphal 1986). While this seemed a possible explanation of the earliest Korean successes, the evidence from later periods is decidedly less favorable for intervention, raising the question of how policymakers would know when to phase out a previously successful policy—an issue not unfamiliar to investors who rode the stock market bubble of the late 1990s. Thus the evidence of the limited or perverse effects of industrial policy in the past quarter century is important. Once such policies had been implemented, governments found them hard to adjust, let alone to abandon. This volume does not address any of the influential journalistic approaches, which we view as untestable. For example, in the pessimistic atmosphere that characterized the United States in the 1980s, a large number of books (e.g., Vogel 1979; Magaziner and Reich 1982; Prestowitz 1988) were published that argued that “Japan Inc.” was a reflection of the attempt of an entire nation to excel economically and that Japan had devised new approaches to economic success that the United States (in particular) ignored at its own risk. As in many paeans to the emergence of a new, unbeatable form of social organization,13 these volumes and many others focused on a small segment of the national economy, namely, 12. For example, Wade (1990) presented a highly detailed account of extensive intervention in Taiwan but did not include an evaluation of its impact in terms of productivity growth or allocative efficiency. In an exhaustive reexamination of the evidence from Taiwan, Smith (2000) questioned the accuracy of Wade’s depiction of the government’s fostering new sectors with emerging comparative advantage; she provides careful tests of their economic effects, casting doubt on the growth effect of the policies. On Korea, see Yoo (1990, 1994). 13. Panegyrics to the superiority of the US model of the 1990s and the “new economy” have disturbing parallels to this literature.

8

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 9

Japan’s success in a limited range of manufactured exports.14 The extent of market penetration in the United States was emphasized, but the welfare consequences in Japan were ignored. No discussion was given of the fact that purchasing power parity estimates of GDP per capita showed that Japan was still much poorer than the United States despite a quarter century of investment to GDP rates twice as high as in the United States and the implied lower level of consumption in Japan. And despite the frequent use of anecdotes about just-in-time delivery, social cohesion, and the brilliance of the Ministry of International Trade and Industry (MITI) and the Ministry of Finance in correctly assessing dynamic comparative advantage, other revealing facts were missing. In particular, there was no evaluation of the welfare of Japanese consumers and producers in sectors other than large-scale manufacturing (Balassa and Noland 1988). The exceptionally long work hours, by Organization for Economic Cooperation and Development (OECD) standards, and the absence of lifetime employment in the small-scale sector that supplied components to the large exporters was ignored, as were the $25 cantaloupes, the tiny apartments of the upper middle class, and the extraordinarily long commuting times in Tokyo due to inadequate infrastructure that made Los Angeles and New York seem commuter friendly. Needless to say, after 13 years of unprecedented stagnation for a developed economy, Japan Inc. is now viewed more skeptically. Protectionists in the United States who examine the impact of protection on the welfare of relatively small numbers of American workers (and shareholders) make a similar error when they ignore the consumer surplus accruing to 280 million consumers. The justifiable concern about displaced American workers (and the need for improved government aid to them) led to a disregard of the fact that many of the same workers purchased Japanese TV sets and cars that had lower prices and were of higher quality than American-made ones while their peers in Japan paid much higher prices for the same products. The workers of the world were united but not quite in the way Marx envisioned. Another dubious proposition is articulated by those who argue that Japan (or Korea or Taiwan) would not have succeeded in any sector without intervention in all sectors. Hence it is necessary to examine Japan’s aggregate growth rather than parse the impact of industrial policy sector by sector, as do all of the studies we discuss in chapter 2. Implicit in this argument is that there were extensive externalities flowing in unpredictable directions and that a simultaneous promotion policy was necessary in large numbers of sectors. Of course, it is not quite clear what externalities were derived by electronics firms from the protected small-scale Japanese 14. Contemporaneous Japanese accounts were much less sanguine. See Komiya, Okuno, and Suzumura (1984).

INTRODUCTION

9

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 10

rice sector that produced rice selling at six times the world price, nor the externalities flowing from Toshiba to inefficient Japanese construction firms that could not match competitive international bids for domestic construction projects. The oft-heard argument that one must take a “holistic” approach and look at the entire economy rather than the impact of industrial policy on individual sectors is irrefutable. But the burden on those who hold such views is to demonstrate that externalities existed and were large. Similarly, some analysts suggest that Japan and Korea generated “deep” industrial structures—that is, they produce most types of industrial goods—and take this as a positive benefit of industrial policy.15 However, India also achieved a thick industrial structure, but one that was extraordinarily inefficient and resulted in much lower per capita income than would have resulted from more attention to efficiency and less to encouraging all sectors. Hong Kong, with a much less deep industrial structure, is considerably richer than India. Not incidentally, Hong Kong also has a higher per capita income than Japan and Korea. All of this suggests that the efficiency of the industrial path chosen and its general equilibrium consequences on welfare are the critical issue rather than industrial depth. Consumers cannot eat industrial “depth” for breakfast but do value increases in their real standard of living.

Definition We define selective intervention or industrial policy briefly as an effort by a government to alter the sectoral structure of production toward sectors it believes offer greater prospects for accelerated growth than would be generated by a typical process of industrial evolution according to static comparative advantage. Used without more specificity, all developing countries, excluding perhaps Hong Kong, have used and continue to use industrial policy.16 Credit directed at specific sectors with below-market interest rates for long-term and working capital, sectorally differentiated profits taxes, subsidized electricity rates, research and development subsidies, control of the entry and exit of firms, export targets, and highly differentiated tariffs and nontariff barriers are all forms of industrial policy. Japan, Korea, and Taiwan used most or all of these instruments as well as others for varying lengths of time and at different intensities. Given their 15. In technical terms they are arguing that the input-output tables will generally have few empty cells and that the individual coefficients are larger than they would have been absent government efforts. 16. Contrary to some popular impressions, Hong Kong engaged in some efforts to improve the prospects of the industrial sector. In particular, it tried to keep food and housing prices low to avoid increases in the nominal wage. 10

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 11

rapid growth in per capita income, it is understandably tempting to conclude that the industrial policy played a decisive role in their success.

Initiation and Maintenance of Industrialization In analyzing the impact of industrial policy, it is important to distinguish between the initiation of industrialization and its continuance once a higher level of growth had been achieved. The former is of interest to many of the least-developed countries, particularly in sub-Saharan Africa, and the latter to countries that have implemented many basic economic reforms but whose growth rate has responded slowly, for example, Chile. The recovery in Japan between 1945 and the Korean War was probably accelerated by the government’s efforts to restore prewar levels of capacity and productivity in sectors such as autos, aluminum refining, steel, textiles, and shipbuilding (Mutoh 1988; Tanaka 1988; Yamawaki 1988; Yamazawa 1988; Yonezawa 1988). In some ways this was the easier part of postwar growth in Japan, as the knowledge on which the prewar structure was based had not been destroyed. Protection of sectors that encouraged investment in them, the direct allocation of foreign exchange to acquire critical equipment and technology licenses, and investment coordination almost surely served a positive role, although this proposition is not easily tested given the lacunae in the data for this period. One interpretation of the impact of these policies is that they compressed the time it would otherwise have taken to restore prewar industrial output levels, but given high investment and the largely intact stock of human capital, this would inevitably have occurred anyway. When one considers the Japanese or foreign image of Japan Inc. popular in the 1970s and 1980s, the issue is not the contribution of industrial policy to the immediate postwar recovery in basic industry but the role of government in fostering the entry of firms into new sectors such as televisions and electronics and whether such policies were the source of the rapid growth in living standards between the Korean War and 1990 (Johnson 1982). Most of the scholarly literature that we review has a skeptical evaluation of this view. Similarly in Korea and Taiwan, government may have played some positive role in the initiation of industrial development from around 1960 to 1970 in largely labor-intensive sectors, although the relative contributions of industrial policy and of fundamentals has not been demonstrated quantitatively. Little (1994) emphasizes that Korean and Taiwanese industrial growth in their first years of rapid development was concentrated in sectors such as clothing, textiles, and sporting goods, laborintensive sectors in which their relatively large supplies of low-wage, fairly productive labor conferred a comparative advantage. Once macroeconomic reforms in the 1960s provided a relatively stable environment, INTRODUCTION

11

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 12

the expansion of exports in these sectors was not surprising, in Little’s view, given that a similar phenomenon had occurred in Hong Kong despite the absence of government policy. These efforts were complemented by measures undertaken by the governments to facilitate growth, and several efforts have been made to understand the potential contribution of these policies (Pack and Westphal 1986; World Bank 1993). Given the patchiness of the data for this early period, the relative impact of government policy and of the improved general economic environment resulting from reforms is impossible to establish quantitatively. Nations such as Uganda that are anxious to jump-start their development process cannot seek guidance in the experience of these countries alone. But the enormous literature from other nations does show, rather definitively, that government policies that favor individual sectors generally have not been successful. By the early 1970s Korea and Taiwan had both achieved considerable growth in per capita income largely on the basis of labor-intensive industries, and, as in Japan, efforts were then made to move into more capitaland technology-intensive sectors, which is the issue of interest for semiindustrialized nations. It is this period for which considerable evidence is available and on which we concentrate in chapter 2. Nevertheless, the distinction between these emphases should not be drawn too sharply. For example, many of the sectors that were encouraged in Taiwan in the 1980s were labor intensive (Pack 1992a; Smith 2000). A low-income country that is trying to initiate industrialization has several options from which to choose. Assuming (perhaps heroically) that it can raise its investment and education levels and maintain a good macroeconomic policy,17 its options include the following: 1. Follow a laissez-faire policy with no detailed intervention—roughly the policy of Hong Kong, in which local firms become important subcontractors within international supply networks. Although the conventional wisdom is that Hong Kong’s skill base was very high as a result of the immigration of Shanghai industrialists in 1949, it is not obvious that the modern industrial skills they possessed were particularly great. Small clothing factories in the economically backward China of the 1930s were unlikely to be the repository of internationally competitive management competence. Rather the small firms on which Hong Kong’s industrial success was built were chosen to be-

17. This does not imply these are the only prerequisites. This is used as a shorthand for many basics, including the now-hackneyed but nevertheless important tasks of a well-regulated financial system, the provision of widespread education, the construction of social overhead capital, a workable judicial system, and so on. Obviously, a country as poor as Uganda cannot do all of these at once. Certainly Japan, Korea, and Taiwan did not achieve all of them instantaneously. 12

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 13

come part of international supply networks because of the stable government and good macroeconomic environment.18 The continuing search by wholesalers and retailers in the OECD countries for lowcost international suppliers offers a poor country considerable opportunity provided it can emulate Hong Kong’s policy environment.19 2. Actively seek foreign direct investment. Singapore, which was similar to Korea in endowments (discussed in chapter 4) took an entirely different route to augmenting local skills by actively recruiting multinational corporations and by building infrastructure ranging from ports to airports to telecommunications systems in order to make location there attractive. Like Hong Kong, Singapore fostered a successful industrial sector and has achieved a higher per capita income than Japan and Korea. While skeptics argue that Singapore’s success is fragile, as footloose multinational corporations can relocate to still less expensive countries, the infrastructure and educational stock are not easy to replicate, and the formula has worked for 40 years. 3. Follow a detailed industrial policy of the type pursued by Japan, Korea, and Taiwan, the focus of this study. One broad rationale for this approach is the need to encourage purportedly missing entrepreneurship; the other is to correct market failures. Before moving to a detailed discussion of market failures that may require the government to implement corrective policies, it is useful to analyze explicitly the question of the need to encourage entrepreneurship. Much of the rationale for industrial policy has been based on this perceived requirement, but the discussion has largely ignored the huge literature on the rationality of small farmers and households in developing countries that has become, to a large extent, synonymous with development economics in the past two decades.20 The question of entrepreneurship has a long history in economics, and much of the discussion has amounted to a catalogue of characteristics—the conditions necessary for

18. For a detailed discussion of the recent expansion of international supply networks and the incorporation into them of many heretofore marginal countries, see Sturgeon and Lester (2002). Hong Kong was one of the early beneficiaries of this development, but the new data suggest that this is an ongoing phenomenon of considerable importance to many nonindustrialized nations. 19. Morawetz (1982) has shown that maintaining a place in this international system is not easy. Colombia was producing clothing for American retailers long before the Asian countries were, but its position was undermined by, among other things, a poor real exchange rate policy. 20. See, for example, the last two volumes of the Handbook of Development Economics, edited by Behrman and Srinivasan (1994), and recent textbooks in development economics, such as Ray (1998), in which more than half of the volume is devoted to these issues. INTRODUCTION

13

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 14

its emergence.21 In the “modernization” literature of the 1950s and 1960s, much was made of the need to instill entrepreneurship into societies in which it was alleged to be missing; efforts were undertaken to measure the personality traits of entrepreneurs who would then become the protégés of aid programs, foundations, or nongovernmental organizations that would convert them into local Henry Fords (McClelland 1961; Inkeles and Smith 1974). This early literature assumed that the critical bottleneck to development was the absence of attitudes of calculation, risk taking, and the Schumpeterian ability to perceive and exploit profitable opportunities. Largely ignored was the research of Theodore Schultz (1964) and others who argued that small farmers were extraordinary in their rational calculation, a contribution cited by the Nobel Prize committee honoring Schultz in 1979. Two decades of intensive empirical effort have confirmed that households and farmers with limited education have made quite rational choices, in one case approximating the optimal solution found by economists using dynamic programming models executed on a supercomputer (Rosenzweig and Wolpin 1993). Households routinely deal with risk, for example, by having children choose spouses from families in a region in which weather patterns have a limited correlation with their own. Peasant farmers change their planting patterns as relative prices of crops change (and implicitly take account of the standard deviation of these prices over time). Managers of township and village small industrial enterprises responded with alacrity to the opportunities in post-1977 China despite three decades without rewards or incentives, finding new product niches and appropriate production methods. It might be argued that these demonstrated abilities are not those necessary in large-scale industry, but advocates of industrial policy appear to be unaware of these results and hence have not engaged them. But even if we assume that the relevance of these skills for large-scale industrial firms is limited—as should be argued by proponents of industrial policy— is a policy directed to fostering large individual firms the only choice for a very poor country? Is the model of the Korean chaebol or the members of the Japanese keiretsu the only path? Or is it possible to envision an industrial trajectory such as that followed initially in Hong Kong and to a lesser extent in Taiwan, to say nothing of the rural Chinese township and village enterprises, that initially relies on small-scale firms that gradually become larger and more efficient?22 Moreover, even the usual characteri21. Perhaps the richest and most insightful discussion is that of Leff (1979). 22. For a formal model of the implications of such growth for the aggregate economy, see Nelson and Pack (1999). It can also be correctly argued that such a small-scale strategy requires centralized technical support, as was provided in Taiwan (Dahlman and Sananikone 1997). This is much closer to the first best imperative of targeting specific market failure rather than providing broad support such as tariffs and quantitative restrictions. 14

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 15

zation of the Korean model is open to question, as it had experienced some industrialization under the Japanese occupation, although the eventual geographic distribution of managers between North and South is not clear. None of this is to argue that it is inconceivable that entrepreneurship needed fostering. Rather it is to indicate that invoking the need to encourage missing entrepreneurship as a defense of industrial policy (and its earlier incarnation in import-substituting industrialization) rests on an assumption that is belied by almost all modern development economics. Thus, even focusing on the three Asian nations that have succeeded, the model pursued by the Asians is hardly the only option for very poor countries attempting to initiate growth, and the other two options need to be considered as serious alternatives. A nation attempting to derive lessons from them should worry about the replicability of their experience. A large number of conditions must be present, including significant government competence and an overriding interest by the government in economic success measured in growth in income per capita, rather than the enrichment of specific groups at the expense of the society. The interest of more advanced developing countries, such as Chile, looking to the Asian experience is not their ability to export wigs, baseball gloves, or shirts—important products in the initial growth of manufacturing—but their later transition into more complex sectors. Much of the evidence we examine in this monograph considers the three Asian countries’ success in developing more sophisticated industries in the period after higher growth began; thus, for these more advanced nations, the evidence for the late 1970s and 1980s is of great relevance. There are two issues to be addressed: the impact of industrial policy on the initiation of growth, and its effect after the initial stage. In fact, these issues are not easily separable. As alluded to earlier, if industrial policy is pursued intensively in the initial phase, altering its substance, course, or pace may quickly become politically difficult. Moreover, as we discuss in chapter 3, some economic damage may be done if one of the instruments of industrial policy is the suppression of the banking sector that is used as a passive conduit of funds to chosen sectors and firms. In principle, we would like to answer the following questions. During each period of growth, initial and follow-on, was industrial policy the source of growth, or was it a mild accelerant, improving the growth rate slightly given the high growth of capital, education, and gains in total factor productivity (TFP) realized from borrowing technology from abroad? Second, are any of the problems that have been encountered in Japan since 1990 and in Korea since 1997 partly the legacy of one aspect or another of industrial policy?

The Case for Selective Industrial Policy For selective government intervention or industrial policy to be welfare improving, policymakers must identify market failures that would proINTRODUCTION

15

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 16

vide the scope for welfare-enhancing interventions, design and implement the appropriate interventions, and correct or terminate the applied policy as changing circumstances warrant.23 To use a construction metaphor, to build an addition to a house there must be room for the extension, a feasible blueprint must be drawn up, and the right tools must be used correctly to get the job done. In technical terms, economists have identified numerous circumstances in which market failures could provide scope for welfare-enhancing industrial policy. These include: 1. Real external economies. These provide benefits to firms that generate improvements in their productivity (or a reduction in their costs) as a result of the activities of other firms. In a familiar example, assume that Microsoft develops new software and that other firms informally learn some of the details of the development process (through after-hours conversations at the local Starbucks) and are thus able to avoid blind alleys. Competitors such as Symantec may be able to produce a new software program more quickly. This reduction in development time will lead to greater profitability for Symantec and perhaps to lower consumer prices. Such knowledge spillovers may occur from informal exchanges in both professional and social contexts. In the case of goods that enter international trade, real externalities improve the welfare of the entire nation only if they allow goods to be produced at less than the imported c.i.f. price.24 Thus if the new software package produced by Symantec is already available at a lower price from SAP, a German software producer, no benefit is realized by the United States as a whole. 2. External economies that arise as the size of a competitive industry increases, permitting a falling long-run supply curve. For example, machine tool manufacturers typically obtain specialized inputs from foundries. The machine tool firms benefit from a specialist foundry that can operate a large piece of equipment for three full shifts as it pools orders from a large number of machine tool manufacturers. If the industry is not sufficiently large to support a specialist, each ma-

23. We use the terms welfare-enhancing and growth-accelerating interchangeably in this discussion. Most of the theoretical models are explicitly static, hence the normative results are expressed in terms of welfare enhancement, not growth acceleration. While it is possible that industrial policy could generate a one-step increase in welfare that would not lead to an acceleration in the secular growth rate, we believe that focusing solely on explicitly dynamic models would be too limiting in this context. 24. This is not sufficient to justify intervention, however. A socially successful intervention depends on whether the present discounted value (PDV) of future producer surplus exceeds the PDV of the social cost of subsidies.

16

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 17

chine tool firm will have to invest in its own foundry, operating it for a small fraction of 168 hours per week, thus raising production costs substantially. The gains in productivity in a competitive sector in which individual firms exhibit constant or increasing costs are attributable to “economies of scope” in the use of specialized equipment and greater specialization of individual labor skills, in this case, in the foundry. Encouragement of many firms to enter the machine tool sector could generate economies of scope.25 3. Pecuniary economies that may occur in some industries in which large-scale economies exist and only one firm can be established in each sector, for example, a steel mill in a relatively small economy. If a steel manufacturer perceives only a domestic market, it will construct a larger plant only if a potential purchaser, such as in the auto sector, also establishes a large plant that generates extensive demand. The market failure is that at a given point in time, current prices may not convey the information about prospective expansion that is relevant to attaining a lower cost of production through larger plant size (Scitovsky 1954; Chenery 1959). This generates an argument for coordination of planned investment as delineated by Murphy, Shleifer, and Vishny (1989), who formalize Rosenstein-Rodan’s (1943) idea of the “big push.” There are multiple equilibria because of pecuniary externalities generated by imperfect competition and large fixed costs. Murphy et al. argue that industrial policy that “encourages industrialization in many sectors simultaneously can substantially boost income and welfare even when investment in any one sector appears impossible” (p. 1024). Such arguments critically depend on the costs of importing some of the inputs or difficulties in exporting the resulting output (Pack and Westphal 1986). Growth of the size of the economy will eventually preclude the need for policies to obtain the productivity gains from economies either of scope or of scale. Government encouragement of individual sectors and coordination of these sectors could potentially yield greater productivity, providing that international trade options could not be exploited. 4. Externalities that may be conferred on other firms in an industry by the first entrant. These include the demonstration that the sector is physically and economically feasible (Pack and Westphal 1986; Rob 1991) and the diffusion of information about technology and market-

25. This is usually classified as a pecuniary external economy, as the machine tool firms benefit from a lower price in specialist foundries. In contrast, in the preceding paragraph, Symantec does not purchase a cheaper or improved input from Microsoft. The increase in profitability is not mediated through the price system.

INTRODUCTION

17

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 18

ing conditions to other domestic firms that are considering establishing production.26 5. Results of research by one firm that benefit other firms. Xerox’s demonstration of the feasibility of a dry photocopying process led many other firms to seek methods of dry copying that did not infringe Xerox’s patent. The incomplete appropriability of the results of R&D may provide a strong argument for governments’ encouragement of research. Without such efforts, too little R&D may take place. 6. Externalities that could arise from the interaction of suppliers and buyers on the design or method of production of a product leading to a better or cheaper good than is available internationally. Thus a local machine producer may be capable of designing equipment that is responsive to local humidity, temperature, or variation in input quality. The cost saving to textile producers might justify the establishment of a loom manufacturer that was not as efficient as foreign competitors. In this case, the source of the externality is the nontradability of some types of inputs or knowledge—otherwise the improved method or product could be obtained from international suppliers. 7. Finally, firms that anticipate learning-by-doing reduce their costs, and eventually become competitive. In principle, they could self-finance this process or raise funds from friends and family if their capital needs are limited and resort to bank finance if large amounts are required. However, if bankers take a myopic view, require large amounts of collateral that new firms do not have, or are excessively risk averse, some subsidy might be appropriate. In all of these cases, industrial policy can be directly welfare enhancing by improving the competitiveness of domestic industry, leading to both higher national and world output.27 There are additional cases in which

26. Okuno-Fujiwara (1988) provides a formal example of this phenomenon in the form of a model of the interdependence of two industries. One industry, which produces an intermediate product, is assumed to be oligopolistic because of underlying scale economies and engages in Cournot competition. The other industry, which produces a final product from an intermediate product, is perfectly competitive. In this situation there may be multiple equilibria with one equilibrium Pareto-superior to the others. Industrial policy has a positive role in the form of preplay communication to generate a superior coordinated equilibrium. For the intervention to convey some purely national welfare enhancement, there has to be some nontraded aspect of the externality. Otherwise, foreigners have access to the same low-cost inputs, and the pattern of production in the downstream industry is indeterminate without additional assumptions. 27. As is well known from the Bhagwati-Ramaswamy (1963) contributions, the optimal strategy is to address the source of the externality, for example, by providing R&D subsidies rather than by less targeted policies such as tariffs. 18

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 19

industrial policy can be welfare enhancing or growth promoting through the capture of rents or terms of trade effects associated with international trade.28 In these cases, national industrial policies have a zero-sum element at the global level and hence could be thought of as containing a strategic or predatory element. Similarly, the endogenous trade and growth literature that links the cross-national pattern of international trade specialization to differential cross-national growth rates provides numerous theoretical possibilities for growth-enhancing industrial policy at the national level, for example, through R&D or output subsidization schemes that could accelerate new product development, innovation, and growth (Grossman and Helpman 1991).29 (Of course, these results are assumption-specific: if in reality the subsidized R&D-intensive sectors are characterized by low levels of TFP relative to world best practice, the policy could be welfare reducing at both the national and global levels.) The point is simply that the theoretical possibility exists—we have room to build that addition to the house; whether the blueprint makes sense or the contractor can build it is another issue. In this discussion we have established the theoretical possibility of welfare- or growth-enhancing industrial policies. However, it is beyond the scope of this monograph to comprehensively map the advisable policy interventions to the specific market failures or strategic opportunities identified in the literature. Nevertheless, we think it worthwhile to point out a few general caveats for the successful implementation of industrial policy. First, the appropriate policy response may be highly case specific. For example, in the well-known Brander-Spencer model, the optimal intervention changes from an export subsidy to an export tax, if price competition rather than quantity competition is assumed.30 In the case of the international trade models, multiple policy tools may be necessary to pursue do28. Early formalizations of arguments along these lines are provided by Spencer and Brander (1983) and Itoh and Kiyono (1987). Helpman and Krugman (1989) offer a synthesis of the subsequent literature on strategic trade policy. Kang (2000a,b) shows that the degree of intellectual property rights protection can have a strategic effect similar to export subsidization in the earlier literature. For a skeptical view of the relevance of strategic trade theory for developing countries, see Corden (1991). 29. It might at first seem surprising that the normative results of these models to a large extent turn on conventional differences in factor usage across industries. Consequently they do not appear to yield robust policy inferences. Empirical work has focused on modeling international spillovers arising from research and development activities (e.g., Coe and Helpman 1995; Coe, Helpman, and Hoffmaister 1997) rather than on the implications of industrial policy. 30. Similarly, the presence of increasing returns to scale decreases the likelihood that the optimal policy is a subsidy, since a subsidy may encourage the entry of additional firms into the market and reduce efficiency by reducing plant size or output. See Helpman and Krugman (1989) for more such examples. INTRODUCTION

19

01--Chapter 1--1-20

3/14/03

10:11 AM

Page 20

mestic and international goals in concert if the good in question is not purely importable or exportable—almost surely the case in the real world. Second, with the exception of some policies that might be accomplished through pure informational or coordination effects, industrial policies require scarce resources. It is not sufficient, for example, to show that in a partial equilibrium sense a particular production or export subsidy might be potentially growth enhancing if the necessary resources are mobilized at the expense of even more worthy sectors (Dixit and Grossman 1986).31 In other words, while adding an additional bedroom to a house may increase its resale value, adding another bathroom may increase it even more. This, of course, suggests a more general informational problem, namely, that even if policymakers identify the possibility of a growthaccelerating intervention and the appropriate policy package, they still have to calibrate the appropriate magnitude of, say, a tax or subsidy; after all, it is as possible to intervene too much as too little. Third, in the case of globally zero-sum strategic policies, policymakers must consider the possibility of retaliation. As a general proposition, one would expect that the possibility of retaliation would reduce the likelihood of growth-accelerating industrial policies, and indeed, retaliation may make both parties worse off relative to their initial positions.32 The basic lesson from the strategic trade literature is that the possibility of retaliation further complicates the problem of identifying optimal policies.33 Finally, in the cases discussed thus far, intervention may be effective if the government itself does not suffer from deficiencies leading to government failure. One of the notable lacunae in the industrial policy literature is the general absence of discussion of political economy factors, in particular the possibility of rent-seeking behavior by self-interested firms and policymakers and the concomitant degradation of policy. A good blueprint is of little use if your contractor is a crook. One of the important aspects of Asian industrial policies was the relative lack of corruption, as we discuss in succeeding chapters.

31. Almost all of the literature advocating industrial policy not only ignores the immediate opportunity cost of funding a preferred sector but also the deadweight loss often imposed on other sectors, for example, by taxes that are used to pay for subsidies. 32. However, as demonstrated by Johnson (1953–54), the possibility of retaliation does not eliminate the possibility that the introduction of a tariff by a large country would necessarily be welfare reducing even allowing for retaliation. 33. For example, in the Brander-Spencer model with retaliation, the previously optimal export subsidy policy is welfare reducing, and the optimal policy is a coordinated export tax by both national governments. 20

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 21

2 Industrial Policies in Japan, Korea, and Taiwan

Industrial Policies: Some Preliminaries In this chapter we set out the available evidence on the effect of industrial policy and our interpretation of it. Many scholars have expended considerable effort to detail the various aspects of industrial policy—surely no objective observer can any longer hold the view that the Asian success stories, with the exception of Hong Kong, were the result of laissez-faire policies. The main question that we address is the impact of these policies on welfare. It is important to note that much of the empirical evidence on the impact of industrial policy is available only for the periods after the initial industrialization drive.1 For Japan, the coverage of the 1945–60 period is limited. But during this early period Japan’s efforts were mainly reconstruction of sectors that had already been advanced before World War II.2 This was the easy part: the knowledge and human capital were largely intact, as were networks of industrialists from the prewar and wartime periods. What was additionally necessary were the financial resources to 1. As noted in chapter 1, there are exceptions, such as the studies of the recovery of individual industries in Japan (Mutoh 1988; Tanaka 1988; Yamawaki 1988; Yamazawa 1988; Yonezawa 1988). 2. As Nariai (2002) observes, at the end of the war, industrial production was down to 1926– 27 levels. It was not until the mid-1950s that textile, steel, and machinery industries, boosted by the Korean War procurement boom, reattained their prewar peaks. 21

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 22

reestablish the physical capital stock (often foreign equipment), to purchase intermediate inputs, and to obtain knowledge of improvements in technology that had occurred during and immediately after the war. In an economy with low saving and limited foreign exchange, the role of government allocation was important. This role was fairly easy to implement, however, as the goal was easy to grasp and required no crystal ball to forecast the evolution of new sectors and to pick winners. When popular writers in the 1980s declared the Japanese model to be the standard of success, its glories were not in the restoration in the production of aluminum back to 1939 levels but to Japan’s “victory” in commercializing 64-kilobyte memory chips and color TV.3 Thus the fact that most of the empirical research reported in this chapter covers the post-recovery period is meant to address the issues raised by those who claim that the fostering of sectors new to Japan was successfully orchestrated by the government. Middle-income countries that want to learn the Japanese secret are not seeking answers about how to restore previously robust sectors but about how to initiate new ones. The experience of Japan in the postrestoration period is thus of great interest. While Japan’s postwar performance can have little interest for the leastdeveloped countries embarking on industrialization, those of Korea and Taiwan in the years from about 1960 to 1973 may be relevant, although even here, as we show, the transferability of experience would have to take account of the unusually high level of education in Korea and Taiwan at the beginning of their industrialization effort as well as the considerable industrial history during the Japanese occupation.4 Unfortunately, although the effort at government direction for the years 1960 to 1973 has been carefully documented, the effect on the growth of the economy has not been. For some authors the two are synonymous, but of course they aren’t. Most of the empirical evidence assessing the impact of industrial policy covers the post-1970 period and misses the years of the initiation of rapid industrialization. For those looking for guidance from Korea’s and Taiwan’s early experience, caution is warranted. First, assuming that market failures were identifiable, there is the question of whether a government contemplating such policies today has the ability to duplicate the extraordinarily complex methods of intervention and to avoid being captured by its beneficiaries. A second issue is whether 3. Japan of course also increased production and sales of autos and steel. In these “old” sectors industrial guidance was fairly straightforward. The size of foreign market and income elasticities of demand, especially in the United States, were viewed as the most relevant information in choosing sectors to stimulate. The evolution of new products and the technology to produce them were not an issue. This can be contrasted with the much more difficult task of forecasting new product areas and the nature of new technologies and to anticipate changing comparative advantage. 4. On Korea’s history, see Kuznets (1977) and Chung (2002). On Taiwan, see Ranis (1979). 22

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 23

it is likely that the policies adopted could be withdrawn without considerable political upheaval. The latter is important, as most of the evidence that we review suggests that after the initial period of intervention, continuing efforts to intervene had few beneficial effects. Moreover, this is assuming that benefits are indeed to be derived from the initial intervention, an assumption that may be unwarranted given the complexity of the process. Even for countries such as Chile and Costa Rica that have had serious macroeconomic reforms and significant growth and are seeking guidance about alternative trajectories that would move them into more complex industrial sectors, the lessons may be difficult to extract. But the relevant experience for such nations would be that of Korea and Taiwan from the mid-1970s onward, and, as we will show, there are good reasons to be cautious about the potential benefits of industrial policy as implemented in these countries in this period.

Industrial Policy in Japan The roots of contemporary industrial policy in Japan go back to the Meiji Restoration of the mid-19th century and the state-led development under the slogans Shokusan-Kogyo (industrialization) and Fukoku-Kyohei (a wealthy nation and a strong army).5 Ironically, the unequal treaties concluded between Japan and Western powers, which greatly circumscribed Japan’s ability to protect its domestic industries through tariffs, encouraged Japanese policymakers to develop other tools, such as targeted subsidized lending through state-controlled banks, to achieve the same effect. Conceptually, the Japanese took their cues from Prussia (a curious precursor of the Axis alliance of World War II), not Britain, and it was Friedrich List, the proponent of infant industry promotion, not Alfred Marshall, the father of neoclassical economics, who developed a following in Japan.6 Japan developed a dual economy, exporting labor-intensive products such as tea, textiles, and apparel in which it had a comparative advantage,

5. The government built model factories, including state-run cotton and silk spinning factories, and provided imported machinery to private producers at subsidized prices. It built a state-run iron and steel works, initiated a rail system, and provided subsidies to shipbuilding (Nariai 2002). 6. Neoclassical economics remained weak in Japan, and until quite recently the bulk of Japanese academic economists were Marxist in orientation. This is relevant to the extent that there was a general coincidence between the neomercantilist orientation of many of the socalled modern economists and the Japanese Marxists, who regarded industrial policy as the manifestation of state monopoly capitalism, arguably a progressive development from their perspective. Economics, like politics, makes for strange bedfellows. INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

23

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 24

while at the same time developing considerable heavy industry, much of it organized by family-dominated conglomerates (zaibatsu) and oriented toward military production. Japan defeated China in 1895 (annexing Taiwan), then Russia in 1905 (eventually annexing Korea), and established itself as a formidable military power as recognized by Great Britain in the Anglo-Japanese Alliance of 1902. The state’s role in the economy, which had waned in the early part of the 20th century as the private sector expanded, revived with the political radicalization of the late 1920s, the Great Depression, and the onset of World War II in the Pacific. Many of the institutional features often thought of as uniquely Japanese have their origins in the wartime economy (Okazaki 1993; Noguchi 1995). The devastation of World War II destroyed a considerable portion of Japan’s capital stock and left Japan’s per capita income at barely half its prewar level (Nariai 2002). However, the contemporaneous level of per capita income was surely a misleading indicator of Japan’s underlying technological capacity—Japan, after all, had produced aircraft carriers and fighter airplanes in the 1930s and, as shown in table 2.1, the human capital embodied in Japan’s labor force was quite high relative to per capita income. In the aftermath of the war, the Japanese government, together with American occupation authorities, implemented an economic reconstruction plan characterized by a considerable amount of direct state resource allocation, multiple exchange rates, extensive quantitative controls on imports, foreign exchange, inward foreign investment, and royalties for technology licensing.7 After the withdrawal of US occupation forces in 1950, Japan continued to implement sectoral industrial policy through (1) direct subsidies; (2) tax policy and off-budget finance; (3) subsidized credit, including the channeling of capital to specified sectors (which required repression of the financial system), supplemented by research and development incentives; (4) controls on international trade, investment, technology imports, and the direct allocation of foreign exchange; and (5) tolerance (and often encouragement) of cartels and other kinds of anticompetitive behavior on the part of domestic firms. In addition to these formal policy tools, government officials also sought to exercise influence through informal administrative guidance (gyosei shido), coercing recalcitrant firms if necessary. The focus of these efforts was largely oriented toward rebuilding heavy industries such as steel and transportation equipment that had been destroyed during the war. 7. For histories of early postwar economic policies, see Shinohara (1982), Morishima (1982), Johnson (1982), and Calder (1993). The classic work on Japanese industrial policy is Komiya, Okuno, and Suzumura 1984/1988. Okazaki (2001) provides a highly informative description of the institutions through which postwar Japanese industrial policy was carried out. See also Johnson (1984) and Patrick (1986). 24

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

Table 2.1

3/14/03

10:13 AM

Page 25

Human Capital Index and per capita income, mid-1950s

Country

Year

Human Capital Index

Japan Korea The Philippines Israel Thailand Greece Malaysia United States Italy Turkey Argentina Mexico Spain

1955 1955 1956 1954 1955 1956 1957 1955 1956 1955 1955 1955 1955

1,673 494 738 1,200 302 693 334 2,293 787 267 760 352 389

Per capita income

Ratio of Human Capital Index to per capita income

519 217 277 609 181 468 351 2,443 971 365 1,059 637 652

3.2 2.3 2.7 2.0 1.7 1.5 1.0 0.9 0.8 0.7 0.7 0.6 0.6

Notes: Human Capital Index is educational expenditure embodied in the labor force. See Psacharopoulos (1973). Values for Japan, Mexico, Spain, Turkey, and the United States are interpolated from 1950 and 1960 observations; values for Greece and Italy are interpolated from 1951 and 1961 observations; values for Argentina and Thailand are interpolated from 1947 and 1960 observations. Per capita income is the purchasing-power-adjusted figure in international dollars from the Penn World Tables.

1.

Direct subsidies. The conventional wisdom among economists is that direct subsidies have at best had a minor role in fostering changes in Japan’s industrial composition. As shown in figure 2.1, the declining sectors of agriculture, forestry, fishing, and coal mining have typically accounted for 90 percent or more of direct on-budget subsidies in the period after 1955. One study by the Japanese government found that only one sector, food processing, received direct subsidies exceeding 0.1 percent of GDP originating in that sector (Saxonhouse 1983).

2.

Tax policy and off-budget finance. Another possibility would be indirect subsidies through the tax system and off-budget finance. The primary source of subsidized capital is the Fiscal Investment and Loan Program (FILP), which is under the control of the Ministry of Finance Trust Bureau. The FILP is an off-budget program around half the size of the general account budget. It has been a powerful policy tool, allowing bureaucrats to address priorities not met in the general accounts budget with this second or shadow budget. Funds for the FILP come mainly from the postal savings system and social security payments. In addition to financing the activities of public corporations, private-sector investments are financed through public financial institutions such as the Japan Development Bank, the Export-Import Bank, and the Housing Loan Corporation. Support for industry briefly accounted for more than 20 percent of FILP funds in the late 1950s, but the allocation to implement industrial policy fell below 10 percent in the early 1960s and never reattained a double-digit share, INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

25

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 26

Figure 2.1 Sectoral composition of on-budget subsidies in Japan, 1955–80 sector percent of total 100 90 80

Agriculture, forestry, and fisheries

70 60 50 40 30 20 10

Coal mining Sea transport

High tech

Small business, textiles, etc.

0

1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980

Source: Ogura and Yoshino (1988, table 1).

while social welfare and infrastructure almost always absorbed more than half of spending (figure 2.2). Unlike the case of direct subsidy, agriculture’s share never reached 10 percent. Perhaps the bureaucrats at the Trust Fund Bureau were more insulated from political pressure than the elected politicians in the Diet, or maybe they just had a different set of clients. One source of indirect subsidies is the public financial institutions that offer loans at rates below the prevailing market interest rate. A second source of implicit capital subsidy is the accelerated depreciation allowed under the tax system.8 Although some countries allow instantaneous depreciation of new investment, the only method that does not distort profitability of new investment, most require depreciation to be taken over the life of the asset. Insofar as legal asset life and the structure of assets differ among sectors, there may be implicit differentiation among them in the present discounted value of depreciation allowances. In addition, an export-based special depreciation system was in place in Japan from 1961 to 1972. An indication of the quantitative significance of the implicit capital subsidies is given in table 2.2, which reports the ratio of the implicit 8. This discussion follows that of Ogura and Yoshino (1988). Special depreciation schemes have existed in Japan throughout the postwar period. The most important of these had the effect of subsidizing certain classes of investment goods. 26

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 27

Figure 2.2 Uses of FILP funds in Japan, 1955–98 percent of total FILP funds 45 40

Infrastructure

35 30 Health, education, and welfare

25

Small business

20 15

Industry and technology Export and ODA

10

Postal saving/ own investment

Agriculture

5 Regional development

0

1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 ODA = overseas development assistance

Source: Fiscal and Monetary Statistics Monthly (FILP Volume), Ministry of Finance, Japan.

Table 2.2

Capital subsidy-investment ratio in Japan 1968

1976

1984

Industry

Loan

Tax

Total

Loan

Tax

Total

Loan

Tax

Total

Mining Food processing Textiles Pulp and paper Chemicals Petroleum and coal products Nonmetallic products Iron and steel Nonferrous metals Metal products General machinery Electrical machinery Transportation machinery Precision instruments

9.38 0.65 0.66 0.01 0.71

1.36 0.49 1.60 0.26 0.54

10.74 1.14 2.26 0.27 1.25

13.28 1.24 2.59 0.03 1.63

1.48 0.81 0.88 0.66 0.39

14.76 2.05 3.47 0.69 2.02

3.83 0.51 0.22 0.03 0.44

1.29 0.46 0.51 0.42 0.17

5.12 0.97 0.73 0.45 0.61

0.00

n.a.

n.a.

n.a.

n.a.

n.a.

2.83

0.14

2.97

n.a. 0.50 0.48 0.85 0.35 0.37

n.a. 0.87 0.46 1.16 0.50 0.84

n.a. 1.37 0.94 2.01 0.95 1.21

0.72 1.39 8.40 1.52 2.02 1.25

0.11 0.58 0.34 0.75 0.43 0.47

0.83 1.97 8.74 2.27 2.45 1.72

0.44 1.52 0.62 0.57 0.28 0.39

0.13 0.96 0.35 0.63 0.20 1.45

0.57 2.48 0.97 1.20 0.48 1.84

2.95

0.79

3.74

3.76

0.71

4.47

0.56

0.20

0.76

n.a.

n.a.

n.a.

0.54

0.47

1.01

0.05

n.a.

n.a.

n.a. = not available Note: Figures are in percentages. Source: Noland (1993a). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

27

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 28

capital subsidy to investment for 14 industries in 1968, 1976, and 1984.9 In general, the low-interest rate loans have been of greater quantitative significance than the special depreciation provisions. With the exception of mining, where investment has been weak and the involvement of public financial institutions high, the ratio of implicit capital subsidy to investment has been low, generally less than 5 percent. After mining, the greatest beneficiary of the reduced interest burdens has been the transportation machinery industry, which includes shipbuilding, motor vehicles, and aircraft.10 Certain tax and budget policy provisions beyond the relatively uniform low subsidy ratios reported in table 2.2 have been used to promote high-technology sectors. There are special depreciation provisions for the purchase of numerically controlled machine tools, computers and terminals, computer-aided design equipment, and industrial robots. Additional tax incentives exist for the use of these products by small businesses, although the amounts appear to be relatively small. Other special tax provisions exist for the software industry.11 The Japanese computer and robotics industries have been further assisted by the Japan Development Bank and Small Business Finance Corporation funding, including the establishment of special leasing corporations to encourage the leasing of Japanese computers and robots, especially by small firms.12

9. The implicit subsidy provided through the provision of these low-interest loans has been calculated as the difference between interest rates charged by private- and public-sector financial institutions multiplied by the amount of government financial institution loans. In the case of the tax provisions, the special tax depreciation can be thought of as an interestfree loan, and thus the subsidy value of the special depreciation provisions is the implicit interest-burden reduction associated with the loan. 10. Japanese policymakers also have access to off-budget funds for industrial promotion through revenues of quasi-public organizations such as the Motor Boat Racing Association and the Japan Bicycle Rehabilitation Association (Prestowitz 1988). The amounts of these funds that went to industry do not appear to be particularly large, however. Saxonhouse (1983) cites the Wall Street Journal to the effect that no more than $500,000 a year from these sources was made available to the Japan Machine Tool Builders Association. 11. The tax benefits are not contingent on the origin of the purchased software or equipment, so the impact of these provisions has been to expand the Japanese market for these products, not to assist Japanese manufacturers per se. Likewise, special provisions that allow computer manufacturers to deduct expected losses on the return of equipment offered to users on a trial basis do not discriminate by origin and thus in principle could be used by domestic manufacturers, local subsidiaries of foreign manufacturers, or importers. 12. Unlike the tax provisions, which are justified on the grounds of promoting the diffusion of new technologies and do not discriminate between domestic and foreign products, the leasing schemes specifically apply to Japanese-made equipment. The amounts involved appear relatively small, however.

28

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3.

3/14/03

10:13 AM

Page 29

Subsidized credit and R&D policy. The government has also promoted high-technology sectors through direct subsidies to R&D activity, special deductions for R&D costs, and reduced interest burdens through the provision of low-interest loans by public financial institutions. Tax preferences were provided through a variety of schemes. In addition, there have been direct subsidies to R&D activity. The most important channel in quantitative terms has been the system of research contracts on large-scale industrial technology R&D established in 1966. Of particular significance were subsidies to promote the development of computers in the 1970s, and research contracts on next-generation industrial technology, including new materials, biotechnology, and new electronic devices, in the 1980s. Lastly, private R&D has been subsidized through the provision of low-interest loans by public financial institutions for “financing development of new technology.” Private R&D activities are provided indirect support by a number of government-supported institutions. These include national and public research institutes, private nonprofit research organizations, special public corporations, and the mining and manufacturing technology research associations, such as the Very Large Scale Integration Research Association. In quantitative terms, the direct subsidies are the most important component of government R&D support, about twice as large as the tax provisions in most years. Implicit subsidies through the provision of low-interest loans have been relatively unimportant. Government support for research organizations is approximately as large as direct subsidies. Assessing the sectoral pattern of R&D is difficult. Direct subsidies from the government, public corporations such as Nippon Telephone and Telegraph (NTT), and special R&D tax deductions are only reported at the aggregate level. Sector-specific indirect support through the research associations is difficult to ascertain, partly because individual associations frequently encompass more than one sector and partly because the budgets of these organizations include private as well as government funding. Data on the government subsidy share of total R&D expenditures are reported in table 2.3. As can be seen in these figures, government support of R&D activities is low, with total government support, allowing for nonsubsidy financing, certainly less than 5 percent of private R&D expenditures for the economy as a whole, far less than the comparable figure for the United States. If one looks at individual sectors, government R&D as a share of total R&D is seen to have been highest in the declining mining industry. After mining, support has been highest in the energy-related sector of petroleum and coal products and, as in the case of the capital subsidies, the transportation equipment industry, which includes aerospace.

INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

29

02--Chapter 2--21-66

Table 2.3

3/14/03

10:13 AM

Page 30

Government subsidy share of total research and development in Japan

Industry Mining Food processing Textiles Pulp and paper Chemicals Petroleum and coal products Nonmetallic products Iron and steel Nonferrous metals Metal products General machinery Electrical machinery Transportation machinery Precision instruments

1968

1976

1984

3.2 0.0 0.7 0.8 0.5 1.0 1.0 0.2 0.8 0.1 1.4 1.7 1.0 1.8

3.2 0.1 0.2 0.3 0.3 0.3 0.8 0.6 1.5 0.2 2.2 1.5 4.4 0.3

14.0 0.4 1.1 0.0 0.8 7.2 1.8 1.7 2.9 0.2 1.2 1.4 4.7 0.1

Source: Kagaku Gijutsu Kenkyu Chosa Hokoku [Report on the Survey of Research and Development], various issues.

4.

Controls on international trade, investment, technology imports, and foreign exchange. With respect to external relations, some have emphasized the government’s role as a “doorman,” “determining under what conditions equipment, technology, and manufactured products enter and leave Japan” (Borrus et al. 1986, 98), directly allocating foreign exchange to be used for the importation of goods and services until the mid-1960s. Effective rates of protection (ERPs), computed from tariff data and the Japanese input-output table, are listed in table 2.4.13 In 1968, ERPs exceeded 10 percent in all manufacturing sectors except publishing, where the ERP was negative. The highest ERPs, in excess of 40 percent, were in food processing, textile products, and transportation machinery. The estimates for food processing and textile products are probably upwardly biased indicators of the true ERPs, however, since in these cases major inputs were subject to quota protection that was not included in the ERP calculation. By 1975, ERPs had fallen for most manufacturing categories. The reductions were most dramatic in the machinery sector, where the ERPs for transportation and precision machinery fell by approximately 40 and 20 percentage points, respectively. The final column of table 2.4 presents estimates of ERPs for 1987 based on tariff cuts agreed to under the Tokyo Round negotiations. With the aberrant cases of food processing and textiles excluded, the

13. The ERPs for the primary product sectors are misleading, because they do not take into account quotas in agriculture and subsidies in agriculture and mining. Likewise, the ERPs for manufacturing do not take into account “informal” barriers that certainly affected the availability of competing imports in the manufacturing sector. 30

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

Table 2.4

3/14/03

10:13 AM

Page 31

Effective rates of protection for Japan

Industry

1968

1975

1987 (est.)

Traded goods Primary Agriculture Forestry Fishery Mining

24.9 5.9 7.6 –1.0 13.9 –0.6

19.3 5.5 9.4 –0.1 8.2 –0.7

15.8 4.5 7.6 –0.1 6.7 –0.5

Manufacturing Food processing Textile spinning Textile weaving Textile products Wooden products Pulp and paper Publishing Leather and rubber Chemicals Petroleum and coal products Nonmetallic mineral products Iron and steel Nonferrous metals Metal products General machinery Electrical machinery Transportation machinery Precision machinery Miscellaneous products

26.7 45.4 21.0 33.6 41.0 18.7 21.9 –3.4 26.0 18.9 10.9 17.7 28.9 31.0 18.7 17.9 21.0 45.4 27.3 28.0

20.6 55.6 10.8 92.6 35.4 8.9 21.9 –3.3 23.5 15.7 6.7 8.8 20.8 32.2 8.6 8.2 13.4 5.4 8.7 20.4

16.9 54.1 12.5 94.2 35.1 6.6 13.5 –2.3 22.0 12.3 7.0 6.4 14.9 20.1 6.3 6.2 6.5 1.4 7.2 9.9

Source: Shouda (1982).

ERPs are under 10 percent for most manufacturing categories, indicating a general fall in rates of protection over a 20-year period. (Additional tariff cuts were subsequently made during the Uruguay Round negotiations and today Japanese tariff rates, outside of agriculture, are de minimis [Bergsten, Ito, and Noland 2001, table 4.3]). Again, it should be noted that these calculations are based on tariff protection only; they do not take nontariff barriers into account, and the sectors are relatively aggregated. Nonetheless, barring a dramatic increase in nontariff protection, a distinct impression of a gradual liberalization in most manufacturing sectors emerges.14 The Japanese government also bargained with foreign technology suppliers acting as a monopsonist. Goto and Wakasugi (1988) provide the example of royalty payments on the importation of a particular Austrian steel production technology that were held down to 1 cent per ton for Japan through an agreement between MITI and the industry, while US firms paid up to 35 cents per ton for the licensing of 14. See Sazanami, Urata, and Kawai (1995) for alternative estimates of Japanese protection. INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

31

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 32

the same technology (p. 190). Borrus et al. (1986) provide examples from the microelectronics industry in the 1960s and 1970s of how the Japanese government used its monopsonistic power to extract very low prices for technology transfers from US firms. Nevertheless, the saving thus achieved was minuscule compared with either export revenues or GDP. Government procurement is another channel through which the Japanese government could seek to tilt the playing field. Bergsten and Noland (1994), for example, calculated that if in 1990 Japanese public entities had exhibited the same purchasing patterns with respect to supercomputers produced by Japanese and US firms (the only nonJapanese producers) as did purchasers in the EU market (the only third market), US producers would have increased their sales by $30 million annually, supporting nearly $5 million in additional R&D.15 Similar, and quantitatively larger, results were obtained for public procurement of computers other than supercomputers. In another public procurement case, the 1980s dispute over the FSX fighter agreement could be interpreted as an attempt by the US government to use its market power to counterbalance the Japanese government’s monopsony position vis-à-vis General Dynamics. What is common to these cases—steel, numerically controlled machine tools, microelectronics, and possibly aircraft—is a pattern of selective protection, strict regulation of inward foreign direct investment and technology transfer, and preferential tax treatment and access to capital until an industry has achieved international competitiveness. Rosovsky has called this pattern “the denial of the profits of innovation.”16 5.

Tolerance of anticompetitive behavior. Another possible channel of industrial policy was the tolerance of anticompetitive behavior by private firms, explicitly by the sanctioning of “recession” or “rationalization” cartels, or implicitly through lax or absent enforcement of competition policies. While there is considerable evidence that anticompetitive behavior, typically facilitated by government regulation, has impeded the access of foreign goods and investment to the Japanese market, the scope for using such practices as an industrial promotion policy is less clear: the most obvious examples are related to declining natural resource sectors such as soda ash or to the construction industry and appear to be driven by narrow parochial political concerns and not forward-looking strategic policies (Bergsten, Ito, and Noland 2001).

15. The same calculation found that the US government appeared to discriminate reciprocally against Japanese supercomputer producers in its procurement decisions. 16. Henry Rosovsky, “Trade, Japan, and the Year 2000,” New York Times, September 6, 1985. 32

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 33

Assessment A number of researchers have attempted to model the impact of Japanese industrial policy on output, trade, and welfare in a cross-industry framework.17 Lee (1993) used a computable general equilibrium model to examine its impact in Japan. Unfortunately, the high degree of aggregation (only three traded-goods sectors) and the calibration assumption (in the 1950s it had no impact) render his results suspect. It is of course impossible to evaluate the policies one by one, as each of them has some impact on the structure and performance of industrial sectors. Several researchers begin by using the extent of international trade protection as the focus. For example, Noland (1993a) attempted to evaluate the impact of these policies on the Japanese economy. The results obtained indicate that trade protection as measured by the ERPs in table 2.4 was generally associated with worse-than-expected performance in net exports, apparently contradicting the notion that Japanese policymakers had successfully promoted infant industries.18 Indirect subsidies, however, were associated with the expansion of output and better-than-expected trade performance. In fact, the estimated effects were so large as to buttress the argument that Japanese industrial policy had acted as a signaling device to private investors, either because the government was better able to process information than private agents or because government participation in a sector or project created a moral hazard or one-way bet. Industrial policies were effective in the sense that market interventions did appear to have an impact on sectoral resource flows and affected the composition of output. However, in the context of the Itoh-Kiyono model, which runs off terms of trade effects, on the whole they did not appear to be welfare enhancing: the actual interventions undertaken did not succeed in capturing contested industries from foreign competitors thereby generating improvements in the terms of trade. Welfare-enhancing interventions appeared to be the exception, not the rule. There is considerable evidence supporting the unsurprising notion that during the postwar period, Japan’s comparative advantage shifted into R&D-intensive activities (Balassa and Noland 1989; Vestal 1989; Grossman 1990). Evidence on the impact of public policies is scarcer. Noland (1996) disaggregated R&D into basic, developmental, and applied activities and separated public and private sources of funding. At the end of the sample period of 1969–89, Japan had a comparative advantage in goods 17. See Baldwin and Krugman (1988) and Flamm (1996) for examples of models of single industries. 18. Noland (1997a) obtained more ambiguous results for a more detailed menu of Japanese trade policies. Audretsch and Yamawaki (1988) investigated the impact of Japanese industrial policy by including a dummy variable for “favored industries” in a regression analysis of US-Japan bilateral trade. The coefficient was significant and had the expected sign. INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

33

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 34

that were intensive in total, privately funded, and applied R&D activities, and a comparative disadvantage in publicly funded and basic R&D intensive goods. Sony and Toshiba were world-beaters. Japanese pharmaceutical firms were invisible internationally. However, the change in coefficient values over the course of the sample period suggested that publicly financed R&D had a large positive impact on sectoral trade competitiveness through the late 1970s and early 1980s. This result could be interpreted as being consistent with the notion that the relative impact of public support could be relatively high at early stages of development before the private-sector R&D capacity was significantly developed and during the period of technological catch-up when R&D priorities could be relatively well defined on the basis of existing technologies.19 Likewise, Branstetter and Sakakibara (2002) uncovered a positive impact on both participant and nonparticipant firms from government-supported research consortia in Japan over the period 1980–92, particularly in the targeting of basic research, although the impact of direct financial support appears modest.20 Beason and Weinstein (1996) directly confront the connection between industrial policy and sectoral TFP growth. Working with a 13-sector sample for the period 1955–90, they uncover no evidence that subsidies, taxes, or preferential policies as measured by the effective rates of protection reported in table 2.4 targeted sectors with increasing returns to scale or that they contributed to TFP growth (table 2.5). They do find some evidence of a positive impact on TFP through subsidized loans. However, the positive impact of the subsidized loans appears to be offset by the negative impact of tax relief, for no net effect. They also find some evidence that before the first oil shock, industrial policy targeted sectors with high labor usage. Lawrence and Weinstein (2001) extend this work on a slightly different data set and find that differential corporate tax rates had an impact on sectoral TFP growth, whereas direct subsidies and subsidized loans did not. Moreover, they find that the effective rate of protection measure is negatively associated with sectoral TFP growth and that imports, not exports, are positively associated with TFP growth. There are at least two channels through which imports could contribute to increasing productivity. The first is by allowing domestic producers to use new, improved, or highly specialized intermediate inputs to which 19. Kim and Oh (1999) analyze annual data on research and development expenditures for 1971–97 and find that public R&D expenditures Granger-cause private R&D in Japan during this period. Unfortunately, their limited sample size precludes the testing of this result for subperiods. 20. However, in earlier work, Sakakibara (1997) casts doubt on even this modest formulation, arguing that participation in publicly supported R&D consortia was concentrated in slow-growth sectors and that sharing fixed costs was not an important factor in determining participation. 34

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

Table 2.5

3/14/03

10:13 AM

Page 35

Impact of industrial policy on sectoral growth, capital accumulation, and total factor productivity in Japan, 1960–90 Dependent variables

Independent variable

Growth

Capital accumulation

Total factor productivity

Tax relief, lagged 2 years

–0.004 (–1.66) 0.000 (0.28) 0.004 (2.04) –0.001 (–0.56) –0.004 (–3.40) –0.000 (–0.69) 0.012 (1.99) 0.007 (1.50) 0.005 (4.22) 0.000 (0.15) 310 0.14 0.08

–0.003 (–2.06) –0.002 (–1.4) 0.003 (2.77) –0.001 (–0.35) –0.003 (–3.34) –0.000 (–0.74) –0.001 (–0.18) 0.013 (4.37) 0.003 (4.14) 0.000 (0.20) 310 0.18 0.13

–0.004 (–2.88) 0.000 (0.02) 0.004 (3.97) 0.000 (0.25) –0.001 (–1.12) –0.000 (–0.08) 0.003 (0.88) 0.001 (0.394) 0.001 (1.12) –0.000 (–0.05) 310 0.06 0.01

Tax relief, lagged 5 years Japan Development Bank loans, lagged 1 year Japan Development Bank loans, lagged 5 years Excess tariff protection, lagged 1 year Excess tariff protection, lagged 5 years Subsidies, lagged 1 year Subsidies, lagged 5 years Effective rate of protection, lagged 1 year Effective rate of protection, lagged 5 years N R2 Adjusted R2 Note: t-statistics in parentheses.

Source: Beason and Weinstein (1996, tables 5 and 7).

they would not otherwise have access. The second is by competing with domestic products: that is, the availability of imports acts as a constant spur to domestic producers to cut costs and improve quality. Lawrence and Weinstein divide imports into “competitive” and “noncompetitive” imports and in the case of Japan find evidence to support the second hypothesis. From this they conclude that Japan’s growth would have been even faster if the government had cut tariffs and exposed a greater share of its domestic producers to foreign competition.21 This contradicts Rodrik’s (1995, 1999) interpretation of the experience of Korea and Taiwan, as discussed below. 21. Japan’s Ministry of Finance apparently agrees. In a June 2002 report issued by its Policy Research Institute, it opines that “the Japanese model was not the source of Japanese competitiveness but the cause of our failure” and specifically argues that sectors sheltered by MITI became bloated and inefficient, while those exposed to international competition tended to be more market-aware, efficient, and profitable (Issei Morita, Financial Times, June 27, 2002). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

35

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 36

Tests of the impact of industrial policy using TFP growth as a major indicator of sectoral performance are dependent on the appropriateness of the TFP calculations, as noted in appendix 2.1. Most investigators use growth accounting and assume that the underlying production function is Cobb-Douglas. Moreover, the use of observed factor shares implies that factor markets were competitive and that despite major changes in the structure of the economy, the factor markets were in equilibrium, with marginal productivities equated to factor costs. Clearly these are strong assumptions, and as the debate over the role of TFP growth in the Asian economies indicates, there are considerable discrepancies in interpretation (Pack 2001a). Nevertheless, overturning the regression results shown in table 2.5 would require very large changes in assumptions about the parameters used. It is more difficult to assess the impact of the informal policies, if for no other reason than that they are less amenable to formal modeling. For this reason, it would be desirable to develop better descriptions of the workings of the industry councils (shingikai) and the process of setting targets. It would be equally desirable to develop better accounts of the penalties and rewards used to encourage adherence to informal guidance. The one study that attempted to model the impact of administrative guidance (Weinstein 1995) found that administrative encouragement of cartels had only a minor impact on prices, margins, and sectoral resource allocation during the period 1957–88. Sakakibara and Porter (2001), who examine the impact of tolerance of cartels on domestic competition and international trade performance, interpret their results (cartels are negatively associated with domestic competition which, in turn, is positively associated with international competitiveness) as undercutting what they perceive as the conventional wisdom that sectorally targeted policy has promoted Japanese competitiveness. In summary, there is considerable evidence that industrial policy has influenced the sectoral composition of output and trade in Japan. However, rather than being the forward-looking driver that proponents of selective promotion envision, at least in terms of measurable interventions, the evidence suggests that such policy was aimed overwhelmingly at internationally noncompetitive natural-resource-based sectors. Indeed, once general equilibrium considerations are taken into account, in all likelihood the manufacturing sector as a whole experienced negative net resource transfers. This supposition appears to be borne out by table 2.6, which reports sectoral tax rates normalized for the overall corporate tax level. The normalized tax rates for the manufacturing sector are almost uniformly negative—that is, the sector was paying higher-than-average taxes. Within manufacturing, industrial policy might then be regarded as a compensatory policy toward some favored activities or firms. Thus the empirical estimates of many types reviewed here do not reveal robust evidence that selective intervention was welfare or growth en36

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

Table 2.6

3/14/03

10:13 AM

Page 37

Normalized sectoral tax rates in Japan

Industry Electrical machinery General machinery Transportation machinery Fabricated metal Petroleum and coal products Precision instruments Ceramics/stone/glass Pulp and paper Chemicals Basic metals Food processing Mining Textiles

1955–90

1955–73

1974–90

Normalized Industry tax rate rank

Normalized Industry tax rate rank

Normalized Industry tax rate rank

–0.403 –0.403

8 8

–0.26 –0.26

8 8

–0.56 –0.56

10 10

–0.403 –0.069

8 7

–0.13 –0.26

7 8

–0.56 –0.35

10 8

–0.009 –0.403 –0.009 –0.891 –0.009 –0.069 –0.736 6.658 0.719

3 8 3 13 3 6 12 1 2

0.30 –0.26 0.30 –0.13 –1.72 0.30 –1.52 0.92 11.68

3 8 3 6 13 3 12 2 1

0.14 –0.35 –0.56 0.00 0.04 –0.35 0.00 1.04 0.50

3 7 10 5 4 8 5 1 2

Source: Beason and Weinstein (1996, table 1).

hancing in the period after postwar reconstruction. This could be due to the inability of policymakers to identify market failures and design appropriate interventions. However, the evidence that most resource flows went to large, politically influential “backward” sectors suggests that political considerations may have been central to this outcome. Middleincome countries that are roughly where Japan was in 1955 and that have solved most of their macroeconomic problems may be tempted to follow some of the strategies of Japanese industrial policy. Yet the evidence presented here about the post-recovery efforts does not provide grounds for optimism despite the fact that Japan had a large and well-trained group of bureaucrats whose motivation was to restore Japan’s economy and polity after the disaster of World War II.22

Industrial Policy in Korea The Korean experience with industrial policy has drawn significantly less attention than the Japanese case—Korea is a smaller economy, Korea has posed less of a competitive threat to US industry and hence has attracted less attention from US-based scholars, and finally, limitations of Korean data on the relevant policy instruments have constrained the ability of re22. As testimony to the intensity of that motivation, Saburo Okita, who would become Planning Director at the Economic Planning Agency and later Foreign Minster, recounts in his memoir a May 1945 conversation in which he expressed the view that “An army in uniform is not the only sort of army. Scientific technology and fighting spirit under a business suit will be our underground army. This Japanese-American war can be taken as the khaki losing to the business suits” (Okita 1985, 26). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

37

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 38

searchers to carry out the kind of applied work on Korea that they have on Japan.23 Like Japan, Korea went through an extended period of relative isolation from the rest of the world, which came to an end in the final quarter of the 19th century. Korea attempted, with variable success, to import foreign technology and social innovations; at the turn of the century, the capital, Seoul, was the only East Asian city to have electricity, trolley cars, a water system, telephones, and telegraphs at the same time (Chung 2002). Korea was occupied by Japan in 1905, and formally annexed in 1910. Japanese economic institutions and practices were transferred to the peninsula, and considerable industrialization and technological learning occurred during the Japanese colonial period. The number of people employed in industry increased from a few thousand in 1904 to 280,000 in 1942, industrial output grew at around 11 percent a year, and industry’s share of GDP reached 29 percent during the period 1939–41. Initially industrialization was concentrated in light industry. By 1943, heavy industry accounted for nearly half of industrial output, and the chemicals complex at Konan was by far the largest in East Asia. Most of the industry was located in the northern part of the peninsula, with the southern part serving as the breadbasket.24 Japanese colonial rule ended with Japan’s defeat in 1945, and the peninsula was divided into US and Soviet zones of military control. The partition of the peninsula was formalized in 1948. As in the case of Japan, operation of the economy during the period of US military occupation was characterized by a high degree of state control and use of quantitative allocations. The Korean War (1950–53), which involved the armies of both sides traversing the peninsula twice, destroyed much of the capital stock. (The head of the US Air Force’s Far Eastern Bomber Command, General Eugene O’Donnell, Jr., memorably testified during the MacArthur hearings that so much destruction had been wreaked on the peninsula that there were no targets left to bomb.) Mass population movements presumably resulted in a net flow of human capital from the North to the South. As in the case of Japan’s emergence from World War II, the data in table 2.1 suggest that in the aftermath of the Korean War, South Korea’s endowment of human capital was high relative to its contemporaneous income level.25 Moreover, South Korea continued to accumulate human capital rapidly after the war (figure 2.3). Its students were relatively concentrated in sci23. Nevertheless, a significant body of literature is available on the subject—for example, Amsden (1989), Jones and SaKong (1980), Mason et al. (1980), Pack and Westphal (1986), and Westphal (1990). 24. See Noland (2000a) and Chung (2002) for additional details and references to the relevant literature. 25. Rodrik (1995, 1999) makes the same point. 38

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 39

Figure 2.3 Human capital accumulation, 1960–86 education (years) 8 7 Argentina 6

Korea

Mexico Egypt

5 4

India Brazil

3

Pakistan

Turkey 2 1

1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 Note: Mean years of education for population over age 15.

Source: Nehru and Dhareshwar (1993).

ence and engineering, though not remarkably so—indeed, the striking thing is how similar the Asian and other major developing countries looked in this regard (table 2.7). What is remarkable is that in the succeeding three decades the qualitative transformation of education was very rapid—by the 1990s almost 40 percent of Korean tertiary students were in these fields, well above the OECD mean. The maintenance of negative real interest rates until the 1960s inhibited the development of the banking sector, which was permitted little freedom from government control, and encouraged the channeling of capital to large politically influential borrowers. As the prominent South Korean economist Cho Soon observed, “the most notable feature of the [South] Korean economy during the 1950s was its dependence on US economic aid” (Cho 1994, 13).26 26. This assistance was not entirely without merit, however. South Koreans were able to expand their skill base through cooperation with the United States. American aid directly contributed to the rapid expansion of education within South Korea and made overseas training and education possible for thousands of Koreans (Westphal et al. 1981), including some of its future economic policymakers. Some transfer of technical skills and management techniques undoubtedly occurred through close contact with US military forces, but its significance is difficult to assess. Likewise, local firms certainly benefited from participation in local military procurement programs, and later from offshore procurement programs during the Vietnam War (Rhee 1994). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

39

02--Chapter 2--21-66

Table 2.7

3/14/03

10:13 AM

Page 40

Science and engineering students

Country

Year

Total

Share of science and engineering students in tertiary education

Argentina Chile Egypt India Japan Korea Malaysia Mexico Pakistan The Philippines Turkey

1955 1957 1957 1964 1955 1956 1967 1961 1957 1957 1957

142,522 18,185 11,231 245,482 589,903 80,935 8,455 94,073 9,686 224,988 4,027

0.161 0.214 0.114 0.062 0.152 0.206 0.142 0.255 0.224 0.145 0.162

Source: UNESCO, Statistical Yearbook, various issues.

The orientation of Korean policy changed significantly in the mid-1960s after a military coup that brought General Park Chung-hee to power.27 Export performance was seized as a barometer of success—as one observer put it, “They were the only statistics that couldn’t be faked.” But apart from the accuracy of the data available to the government, exports were viewed as a measure of the ability of local firms to compete on the world market, although this was an imperfect measure in some instances given that protection of the local market allowed some cross-subsidization of exports. But given the very rapid growth of exports, it was not possible to continuously increase their subsidization—firms did have to improve their efficiency to compete on the world market. Multiple exchange rates were unified and the currency devalued in 1964. Export targets were formulated in considerable detail by product, market, and exporting firm. Firms that did not achieve them were not subject to penalty; however, the targets were sometimes negotiated jointly with wastage allowances (a form of subsidy), and there is some evidence that firms that achieved their targeted goals could expect more favorable tax treatment (Westphal and Kim 1982). At the same time the government began to introduce a wide range of export promotion measures. A government-subsidized organization, the Korea Trade Promotion Corporation (KOTRA), was established to promote exports and perform market research. Exporters were provided exemptions from duties on imported intermediates, tax incentives, preferential access to capital, special depreciation allowances on imported 27. See Bae (2001) for an overview of the evolution of the Korean industrial policy regime from 1962 to 1997. 40

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 41

capital equipment, and a variety of nonpecuniary awards. Exporters also received generous wastage allowances on duty-free imports and reduced prices for electricity and rail transport.28 The export-import-link system allowed exporters to earn rents through the importation of restricted items. Overall, the trade regime could be characterized as modestly biased toward exports, with proexport policies superimposed to offset the preexisting protectionist, antiexport regime. Established industries received roughly neutral effective incentives, while a few infant industries were actively promoted (Westphal and Kim 1982).29 Economic policy began to change in the 1970s in response to a variety of internal and external political developments. Korea initiated the heavy and chemical industry (HCI) drive in an attempt to steer the composition of industrial output toward more capital- and technology-intensive sectors (to reduce reliance on low real wage levels) and engineering-intensive products with the aim of upgrading its export profile and reducing its reliance on imported arms. Efforts at selective industrial policies were intensified, and in contrast to the relatively rules-based policies of the 1960s, greater policy discretion and selectivity were introduced. The financial liberalization policy was reversed in 1972, when interest rates were lowered and direct government control of the banking system was increased in order to channel capital to preferred sectors, projects, or firms. In order to finance large-scale projects, special public financial institutions were established, and commercial banks were instructed to make loans to strategic projects on a preferential basis. By the late 1970s, the share of these “policy loans” had risen to 60 percent (Yoo 1994). These loans carried, on average, negative real interest rates, and the annual interest subsidy grew from about 3 percent of GNP in the period 1962–71 to approximately 10 percent of GNP on average between 1972 and 1979 (Pyo 1989). Capital channeling policies were augmented by extensive tax incentives for the priority industries. It is estimated that the effect of the special tax measures was to reduce the marginal corporate tax rate from 28. The excess wastage allowances on duty-free imports for export production allowed export-oriented firms to divert these duty-free inputs into the production of goods for local sale to their competitive advantage in the domestic market. 29. While the trade regime was being recast toward greater export orientation, reforms were also implemented in other areas of economic policy. In 1963 the military government revised the labor laws to discourage the establishment of independent labor unions and encouraged the organization of unions within a centralized system, established so as to facilitate government control. This system was tightened further in 1971 by the introduction of legislation banning strikes, which made virtually any form of collective bargaining or action illegal (Haggard 1990; Cho 1994). Nevertheless, real wages grew at over 5 percent per annum over this period, roughly the rate of growth of labor productivity. Financial reform began in 1965, when interest rates were raised, encouraging saving and financial deepening as well as more efficient use of capital. The national saving rate doubled in five years, and the ratio of M2 (a broad definition of the money supply) to GNP nearly tripled over the same period. INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

41

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 42

50 percent to 20 percent for the targeted industries. These industries also received trade protection. This era came to a close in April 1979 when the government announced a Comprehensive Stabilization Plan that effectively ended the HCI drive. The assassination of Park and the second oil shock later that year precluded any turning back. Subsequent Korean governments have attempted to scale back industrial policy, with varying degrees of enthusiasm and success.

Assessment For industrial policies to be successful, the market equilibrium must be suboptimal. Governments must be able to identify these opportunities for welfare-enhancing interventions based on the market failures noted in chapter 1, formulate and implement the appropriate policies, and prevent political failures from leading the interventions astray. In the case of Korea, industrial policies clearly affected the cross-sector allocation of resources. Yoo (1994) estimates that as a consequence of the heavy and chemical industry credit, tax, and trade policies, during the late 1970s around 80 percent of fixed investment in the manufacturing sector went to these industries. During the first three years of the Fourth Five-Year Plan (1977–81), investment in basic metals and in chemicals amounted to 130 percent and 121 percent, respectively, of the targets for the entire period, while textiles and other light industries received only 50 percent and 42 percent, respectively, of their planned investment (Balassa 1990). Whether this resource channeling was welfare-enhancing or growthpromoting is less clear. There are three sorts of evidence that can shed light on this issue. The first attempts to evaluate the effect of industrial policy either across different time periods or in comparison to other countries; the second examines its impact on industry TFP growth rates directly; and the third derives economywide estimates of the effect of intervention by examining the potential quantitative impact of interindustry externalities. Kim (1990) surveys the fiscal, credit, tax, and trade policies undertaken during this period and concludes that the policy was unsuccessful: it had the predictable result of generating excess capacity in favored sectors while starving nonfavored sectors for resources, as well as contributing to inflation and the accumulation of foreign debt. Moreover, “the government [was] reckless in its selection of launch enterprises and in its almost haphazard provision of generous incentives, and [its] direct, unlimited role in industrial promotion placed it in the position of an implicit, de facto risk-partner, thus complicating the efforts at market-determined adjustment” (p. 44). Yoo (1990) covers similar terrain, distinguishing between the less selective efforts at export promotion in the 1960s and the more aggressive in42

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 43

Figure 2.4 Rates of return on capital in Korea percent 0.25 Light industry 0.20

All manufacturing 0.15 HCI 0.10

0.05

0 1970

1975

1978

1983

HCI = heavy and chemical industries

Source: Yoo (1990).

dustrial promotion efforts of the 1970s. Yoo also directly addresses the argument that the HCI policy was a success inasmuch as the industries favored by the HCI policy became major exporters in the 1980s. He addresses this argument by posing two counterfactuals: what would the Korean economy have looked like in the absence of the policy, and how would the Korean trade structure have looked in its absence? Using reasoning similar to Kim’s, Yoo concludes that in macroeconomic terms the Korean economy would have been better off without the HCI policy. Figure 2.4, based on Yoo’s data, shows the rate of return on capital in the heavy and chemical sectors, light manufacturing, and all manufacturing. The static allocative costs of the HCI policy can be seen from the substantial differences between the initial rate of return in the HCI sectors versus those in light industry. A standard infant industry argument would contend that these were an investment in future gains that would be obtained from more rapid growth in TFP. But as will be seen below, a variety of estimates of TFP growth suggest that TFP growth was not in fact higher in the favored sectors. But what about industrial upgrading? Yoo compares the Korean experience with those of other, similarly endowed economies (in particular Taiwan) and concludes that on the basis of upgrading or trade performance, the heavy and chemical industry policy was not a success. In 1987, about 15 years after it began, Taiwan had a larger share of OECD imports INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

43

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 44

of both light industrial products and heavy and chemical products, although Korea’s share of basic metals, largely reflecting the performance of POSCO, the major iron and steel producer, was much larger than Taiwan’s. Even in this sector, the ratio of Korean to Taiwanese shares had declined significantly. This evidence does not support the notion that selective intervention had a decisive (or even necessarily a positive) impact on the Korean economy. But the virtue of industrial policy, like beauty, can be in the eye of the beholder. How about attempts to measure its impact on industry TFP directly? Park and Kwon (1995) conclude that during the heavy and chemical industry drive, the establishment of oligopolistic positions by the chaebol retarded technological change. They argue that once scale economies were taken into account, TFP, correctly measured, actually turned negative, although the disentangling of scale economies from TFP is not straightforward, and in any event the capture of scale economies itself could be regarded as an accomplishment.30 Similarly, Kwon and Paik (1995) use a computable general equilibrium model calibrated to 1978 to investigate the potential magnitude of these directions. They conclude that resource misallocation reduced GDP by less than 1 percent if capital is assumed to be immobile, and more than 3 percent if it is mobile. The welfare impact they calculate is higher. The paper that considers most thoroughly the linkage between industrial policy and sectoral productivity growth is Lee (1996). It examines a panel of 38 Korean industries over the period 1963–83. Critically, this sample period encompasses the initiation and the greatest intensification of these policies—unlike the case of Japan in the immediate postwar period, there can be no credible argument here that positive effects occurred in an earlier, untestable period. However, it is nevertheless possible that for a short period of time, seven or eight years after the first year in the time series, industrial policy had a positive effect in stimulating growth, although the growth in the colonial period cited above suggests that an important base in knowledge, if not in physical capital, may have existed. Lee finds that trade protection in the form of tariff or nontariff barriers is negatively associated with the growth rate of labor and total factor productivity, while tax incentives and subsidized credit were uncorrelated with sectoral productivity growth (table 2.8), leading him to conclude that

30. See Zeile (1990, 1991). Kwack summarizes 23 studies of Korean TFP growth (2000, tables 7 and 8). In Kwack’s own estimation he finds, like Park and Kwon (1995), that Korean TFP growth declined over time, and in the case of light industry, actually turned negative. It is hard to understand how resource misallocation driven by favoritism toward heavy industry could result in light industry TFP turning negative. This simply underscores the point that these TFP estimates are unlikely to be very robust, depending on both theoretical specifications and delicate estimates of capital stocks during a time of both rapid capital accumulation and technological obsolescence. 44

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

Table 2.8

3/14/03

10:13 AM

Page 45

Impact of industrial policy on value added, growth rate of capital stock, and sectoral total factor productivity in Korea, 1963–83a Dependent variables

Independent variableb Log (initial value added) Log (initial capital) Nontariff barrier coverage ratio Tariff rate Tax incentives Subsidized bank loans Estimation techniquec Number of observations

Value added

Capital accumulation

Total factor productivity

–0.163 (–11.64) 0.172 (7.17) –0.092 (–3.07) –0.118 (–3.03) –0.084 (–0.50) 0.092 (10.76) 3SLS 146

0.025 (1.79) –0.209 (–12.30) –0.159 (–5.68) –0.029 (–0.76) 0.499 (4.75) –0.019 (–.10) 3SLS 146

–0.153 (–10.2) 0.078 (4.33) –0.167 (–5.06) –0.113 (–2.35) 0.074 (0.65) –0.123 (–0.68) 3SLS 146

Note: t-statistics in parentheses. a. The dependent variable for the first regression is annual growth rate of real value added per workhour over each five-year period from 1963 to 1983. The dependent variable for the second regression is the annual growth rate of the capital stock over each five-year period. It is derived from the growth accounting value added. The dependent variable for the third regression is the annual growth rate of total factor productivity over each five-year period. It is derived from the growth accounting value added. b. The independent variables are as follows: Initial value added is real value added per workhour in the initial year of each period; initial capital is real value of net capital stock per workhour in the initial year of each period; nontariff barrier is the ratio of tariff lines subject to discretionary import approval to the total number of lines in the midyear of each period; tariff is the output-weighted average of legal tariff rates in the midyear of each period; tax incentive is the period average of ratio of difference between legal and effective marginal corporate tax rates; bank loans is the period average of ratio of subsidized bank loans to total assets; growth rate of capital stock is the annual growth rate of capital stock per workhour over each period. c. The three-stage least-squares (3SLS) technique uses log values of initial value added and capital, and one-period lagged policy variables as instruments. Industry- and period-specific effects are controlled. Source: Lee (1996, tables 4–6).

“the Korean data present evidence that less intervention in trade is linked to higher productivity growth” (Lee 1996, 392). The fact that labor productivity growth was slower in protected sectors is very important, as disagreements about the correct method for calculating TFP growth are moot—labor productivity is much less problematic. Similarly, Pack (2000) finds that TFP growth in the heavy and chemical industry sectors was not sufficiently large to have exerted a major impact on aggregate growth. Finally, as in the case of Japan, a paper by Yoo (1993) that analyzed the determinants of the cross-sectional pattern of trade protection did not obtain INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

45

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 46

robust results but was suggestive of political economy rather than efficiency considerations as determining the pattern of protection. These results cast further doubt on the efficacy of resource channeling. What about the line of argument of Pack and Westphal (1986) and OkunoFujiwara (1988) that has focused on interindustry linkages and the potentially welfare-enhancing coordination role for the government? Pack and Westphal suggested that Korea’s selective intervention policy might have been successful in fostering industry without significant losses in efficiency in the early stages of development, the mid-1960s to the early 1970s, a period that constitutes part of Lee’s study of 1963–83. The key was to capture latent interindustry pecuniary and nonpecuniary externalities. “The Korean government can be seen as having achieved integrated decision-making by acting as a central agent mediating among market agents, forcing and facilitating information interchange and insuring the implementation of decisions reached; weighing costs and benefits from a collective standpoint and often intervening to reward cooperative players and punish uncooperative ones” (p. 99). In both this model and that of Okuno-Fujiwara, the same outcome could presumably be attained through organizational integration. Pack and Westphal argue that in the case of Korea this was not feasible: “The externalities may flow in complex and inseparable patterns among (actual and potential) agents covering most if not all of the industrial sector” (p. 99), necessitating government intervention.31 Investment coordination may have helped to overcome these patterns in the early stages of industrialization. However, by the 1970s the growth of the chaebol had undoubtedly reduced the importance of government coordination. While none of these giant firms produced the entire range of industrial products, the owners of the firms knew each other and private coordination became feasible and undoubtedly occurred. While government intervention might have reduced some interpersonal transaction costs, many of the potential externalities were presumably dealt with by Coasian agreements among the firms.32 The key to welfare-enhancing industrial policies through government coordination activities to capture interindustry externalities lies in the existence of interindustry externalities that, when captured, expand the

31. Indeed, Auty (1991) provides detailed descriptions of indivisibilities and other entry barriers in the HCI sectors. Even after assessing possible pecuniary and nonpecuniary externalities, however, he concludes that from an economywide perspective, resources were misallocated. 32. If anything, this argument seems more applicable to the Japanese case. In Japan vertical integration is less complete: the keiretsu, networks of affiliated firms, strike a balance between the coordination advantage of full integration, and the maintenance of competition among suppliers. In this more loosely organized system the government’s coordinating role could be larger. 46

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 47

production set of the economy. It is difficult to model this rigorously. However, it would seem that the likely scope for growth-enhancing interventions would be increased if the industries targeted for intervention met three criteria. The first is that they have strong interindustry linkages to the rest of the economy. Second, they should be leading in a causal sense, so that growth stimulus would be transmitted forward through the economy. One might think of an input supplier industry in the OkunoFujiwara model as an example. Finally, variations in output should have a strong industry-specific component; otherwise variations in output might simply be due to common macroeconomic shocks, with little scope for industry-specific stimulus. The existence of industry-specific variation in output suggests the possibility for industry-specific technical change or scope for industry-specific policy interventions to increase output. Noland (1993b) examined data on 26 Korean manufacturing industries over the period 1960–89. He identified four sectors that possibly met these criteria—wood products, paper, petroleum and coal products, nonferrous metals—and a fifth, nonmetallic products, that arguably did. These are not the typical sectors that one would associate with heavy and chemical sectors, nor, with the exception of nonferrous metals, were any of them promoted in this period. An alternative interpretation of Korea’s development experience during this period is provided by Rodrik (1995, 1999), who argues that industrial policy was key, but avers that it worked by engineering an investment, not an export, boom. Rodrik’s argument is that changes in the Korean export regime occurred in the 1950s, before the boom, and that the increase in export incentives in Korea was insufficient to spur the huge increase in exports that Korea experienced. He argues that the observed increase in exports was the by-product of an investment boom induced by the government. With respect to timing, Rodrik ignores the fact that in the late 1950s Korea was in the process of recovering from wartime devastation. Much of the behavior of the Korean economy (and similarly the Japanese and Taiwanese economies) during this period should be viewed through the prism of postwar reconstruction. Given the incentive systems in place, households, for example, exhibited surprisingly high rates of saving, presumably in an attempt to rebuild household wealth. The macroeconomic behavior of these economies in the 1950s in large part can be regarded as a convergence back to prewar steady states. On the issue of incentives, Rodrik uses data from an unpublished paper by K. S. Kim that indicates that the increase in the export real exchange rate in the early 1960s was not dramatic. The data, however, appear to be from an earlier, published paper (Westphal and Kim 1982, table 8.3) that catalogues the ways that their series understates the true improvement in incentives. Westphal and Kim state that “the rise in the overall inducement to export resulting from the incentive policies adopted during the INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

47

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 48

first half of the 1960s was certainly greater than is indicated by the comparison of the gross real effective exchange rates for exports in table 8.3” (p. 219). They cite three reasons: First, the devaluation and unification of the exchange rate reduced real exchange rate volatility. Second, their estimates ignore two incentive mechanisms that increased in importance in the 1960s, namely wastage allowances and the export-import-link system. Finally, Westphal and Kim indicate that their estimates do not adequately reflect the increase in credit subsidies associated with increased credit ceilings for exporters in the early and mid-1960s. Moreover, there is no reason to believe that an export boom amounting to 30 to 40 percent of GDP and much of it consisting of nontraditional pure exportables (e.g., color television sets) would not have required a substantial increase in the investment rate. Rodrik’s argument would be more plausible if he provided evidence that the increase in investment went primarily into nontradables, but he does not. Bhagwati (1999) makes precisely this point about investment in Korea and Taiwan in the 1950s, arguing that it went into infrastructure that was critical for the export drive. It would seem to us more plausible that much of the demand for investment was fueled by either simultaneous or anticipated increases in exports (see appendix 2.2). The exports that resulted did then generate the foreign exchange necessary to pay for imported raw materials, industrial intermediates and components, and, to a lesser extent, imported machinery. Another test of potential interindustry externalities is provided by Pack (2000). Industrial policies could have generated benefits in other sectors as a consequence of three developments: (1) domestic production of intermediate goods with special characteristics that were not available internationally but improved productivity in the local purchasing firm; (2) movement by workers and managers from firms in promoted sectors to firms in other sectors, the movers bringing with them uncodified knowledge; or (3) direct interactions on equipment design between producers and local buyers of machinery that led to adaptations to machinery that were particularly suitable for local firms. All three externalities could potentially increase TFP growth in the neglected sectors in addition to any benefits accruing to the directly promoted sectors. The potential quantitative importance of specialized nontraded intermediate inputs and uncodified knowledge transmitted by workers depends on how much the neglected sectors interact with the promoted ones. One way to gauge the potential benefits is to measure the purchases of inputs from a favored sector per won of gross output in the neglected sector (the won is the South Korean currency). The larger the purchase, the more likely it is that the neglected sector may derive some benefits from the existence of local producers. The neglected sector may also derive greater benefits if there are few imports, which constitute an alternative source of specialized inputs.

48

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 49

We assume that items (1) and (2) depend on the magnitude of interaction with the promoted sectors. Such interactions can be measured by Leontief input-output coefficients. The nxn input-output coefficient table, A, consists of two sets of flows: the domestic intersectoral flows, AD, and the import flow matrix, AM, A=AD+AM. aij is a typical coefficient of the domestic flow table, while mij denotes elements of the import matrix. The extent of interaction between favored and neglected sectors is given by the domestic input-output coefficient afn, which measures the purchases of an input from a favored sector, f, per won of gross output of the neglected sector, n. The larger afn is, the more likely the neglected sector may derive some benefits from the existence of local producers.33 The neglected sector may derive greater benefits if there are few imports that constitute an alternative source of specialized inputs. Thus, the lower mij is relative to aij, the larger the potential impact of the availability of local production. Several measures of the magnitude of interaction between the promoted and neglected sectors in Korea in 1985 are presented in table 2.9. The average input-output interaction between favored and neglected sectors is quite small. The favored sectors account for a very small portion of the domestically purchased inputs of most neglected sectors. Second, the heavy industries purchase extensively from one another. Third, the imports of the neglected sectors in Korea are, on average, twice the size of the combined purchases from the favored domestic sectors (.134 versus .068). Given the fact that input-output tables typically become more dense as industrialization occurs, the interactions were almost certainly smaller in, say, 1965, at the beginning of the pursuit of industrial policy. These patterns suggest the following probable effects of industrial policy on nonpromoted sectors in the mid-1980s and a fortiori earlier:  It is unlikely that the promoted sectors were quantitatively critical in increasing the range of available inputs to nonfavored sectors. Although industrial policies may have encouraged the domestic production of some unique, nontraded inputs, the overall impact was small relative to all domestic and foreign purchases. Unless there was very low substitutability between local and foreign inputs, the quantitative effect of local supply of such inputs was limited.  Insofar as movement of workers and managers might provide important knowledge to the neglected sectors, the small purchases from the promoted sectors imply that such knowledge transmission would

33. It is possible to test whether indirect interactions mediated through other sectors have an effect by using the inverse coefficients of the Leontief matrix. But the sources of real external economies enumerated above are not easily extended to indirect interactions.

INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

49

02--Chapter 2--21-66

Table 2.9

3/14/03

10:13 AM

Page 50

Intersectoral purchases, Korea, 1985 Purchases from:

Purchasing sector Neglected sectors Food Beverages Tobacco Textiles and clothing Leather Wood and wood products Paper Printing and publishing Petroleum Rubber products Nonmetallic minerals Miscellaneous manufacturing Average Favored sectors Chemicals Heavy industries Iron and steel Metal products Nonelectrical machinery Electrical machinery Transportation machinery Heavy industry average

All domestic sectors

“Heavy” industries

Chemical industry

Foreign suppliers

.147 .290 .048 .522 .319 .240 .422 .408 .053 .373 .293 .402 0.293

.007 .025 .002 .007 .003 .026 .019 .017 .003 .025 .029 .096 0.021

.021 .012 .006 .125 .055 .043 .044 .042 .003 .121 .020 .087 0.047

.029 .019 .009 .099 .355 .060 .183 .039 .009 .124 .029 .123 0.134

.357

.010

.249

.209

.542 .412 .387 .324 .388 0.411

.466 .335 .334 .245 .332 0.342

.009 .031 .016 .034 .015 0.021

.131 .143 .163 .272 .173 0.176

Source: Calculated from input-output tables contained in Bank of Korea, Monthly Statistical Bulletin, various issues.

have been limited. Any tacit knowledge brought by worker mobility, about the special properties of purchased inputs or how to use them more effectively, would affect only a small component of total costs.  Promoted sectors are substantial purchasers of one another’s inputs, at least in the metal-based sectors. Thus the necessary condition for investment coordination benefits existed in 1985 although the interactions may have been less two decades earlier. However, by 1985 these sectors were already large exporters and importers of products within the sector (see column 4). While investment coordination may have had a role in Korea’s early industrial development, the extensive international trade suggests that it was unlikely to have been a decisive one during the intensive HCI effort of the 1970s. Some interactions are not captured by the input-output transactions shown. In particular, the interactions between the producers and the final purchasers of machines are not shown, as investment is a final demand. Table 2.10 shows the ratio of imports to domestic production of machinery. In Korea, imports of nonelectrical machinery amounted to three times the level of domestic production. It is difficult to argue 50

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

Table 2.10

10:13 AM

Page 51

Purchases of domestically produced and imported machinery, Korea and Japan Ratio of imports to domestic production

Sector

Korea, 1985

Japan, 1980

3.04 .27

.06 .04

General machinery Electrical machinery Source: See table 2.9.

that there were no imported substitutes or that special adaptations to local conditions were quantitatively significant. Even if locally produced equipment conferred some cost reductions on its users that would not have been available from internationally available equipment, it would have affected only one-quarter of annual general machinery investment as late as 1985. On the other hand, the extensive import of machinery provided considerable potential for Korean firms to take advantage of the foreign technology embodied in equipment.

Industrial Policy in Taiwan Like Korea, Taiwan is a former Japanese colony, and like Japan and Korea, it also had an Olsonian upheaval, in this case associated with the conclusion of the Chinese revolution and the decampment of the Nationalist government and thousands of its supporters to Taiwan in 1949. There has been considerable analysis of Taiwan’s experience with industrial policy. The standard neoclassical interpretation (Little 1979) has been that Taiwan’s development was primarily attributable to a low level of trade protection, the availability of inputs to exporters at international prices, a conservative macroeconomic policy manifested in limited inflation, and factor markets that were competitive. The last points are suggested by positive real rates of interest and the absence of duality in the wage structure, either by size of firm or by sector. In detailed analyses, Wade (1990) and others contend that a critical component of Taiwan’s success was its industrial policy, which helped to establish new and successful manufacturing sectors.34 These studies have documented the extensive employment of tariffs, quantitative restrictions, and selective credit policies and argue that Taiwan’s success in the period considered was attributable to an intensive effort by the government to direct the sectoral evolution of the economy. This was implemented by a variety of means:

34. See, for example, Clark (1989), Gold (1986), and the papers in Winckler and Greenhalgh (1988). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

51

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 52

(1) the establishment of public enterprises when private initiative was not forthcoming or the capital markets were reluctant or unable to fund very large projects; (2) extensive employment of tariffs and quantitative restrictions on imports; and (3) the directing of credit to preferred industrial sectors through the highly controlled financial sector. The view that Taiwan approximated the laissez-faire environment of Hong Kong is untenable in light of the carefully accumulated facts. Moreover, the data on which earlier interpretations were based cited fairly low (but by no means single-digit) effective protection rates that were estimated in the late 1960s. As in the case of Japan and Korea, Taiwan’s industrial policy may have helped to jump-start the economy from its low 1950 levels, and much of the evidence on the role of the government focuses on the 1960s and early 1970s, a period in which the manufacturing share of GDP surged from 19 percent to 29 percent. However, the benefits from industrial policy in the succeeding years are not easily shown, although there were continuing efforts.35 The basic fiscal incentive program was the Statute for the Encouragement of Investment, which was in place from 1961 to 1990. Available to both foreign and domestic firms, it targeted specific industries, although the focus shifted over time from labor-intensive exports (1960s) to capitalintensive sectors (1970s) to technology-intensive sectors (1980s). Under this program participating firms could choose tax exemptions or accelerated depreciation on capital equipment; most firms took the former option. In 1987 the government commissioned a study to examine the program. It concluded that although the program might have contributed to economic development at the outset, by the 1980s it had outlived its usefulness. Specifically, the report cited four problems: (1) the program was contributing to highly uneven tax burdens across firms; (2) the complexity of the law was creating insurmountable difficulties in administration; (3) the creation of groups with a vested interest in the continuation of the status quo was inhibiting the adoption of more rational policies; and (4) the correlation between participating firms and productivity or profitability was low. In response, the government discontinued the statute in 1990, replacing it with the more functionally oriented Statute for the Upgrading of Industry. Under this statute, firms were eligible for tax relief based on their expenditure on socially favored activities such as R&D or pollution control, although some industry-specific incentives in the high-technology sector were retained. Interestingly enough, Taiwan’s president at the time, Lee Teng-hui, was a US-educated economist who had coauthored the first study of effective incentives in Taiwan (Lee and Liang 1982).

35. A thorough survey of the policies and their effects in the 1980s in both Korea and Taiwan is provided by Smith (2000). Wang (1998) updates the Taiwan story through the mid-1990s. 52

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 53

Figure 2.5 Export loan subsidy in Taiwan loan subsidy as a share of export value (percentage points)

interest rate differential between export and nonexport loans

0.30

8 7

0.25

6 0.20

5 Interest rate difference

4 3 2

0.15

0.10 Export loan subsidy

1

0.05

0

0 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987

Source: Smith (1997, table 4; 2000, table 2.7).

A second industrial policy tool was directed credit. Like Japan and Korea, Taiwan maintained a relatively repressed financial system and channeled credit, though not to the extent that the Koreans did. Smith (2000) shows that in the 1980s public utilities were the largest recipient of loans to promote strategic industries, followed by chemicals. The most important subsidized credit program was export financing to cover certain preshipment costs and the importation of raw materials. As shown in figure 2.5, the differential between the interest rate for export loans and nonexport loans was significant in the 1960s and 1970s. However, the volume of these loans was rather small, and at least since 1971, the first year for which data are available, the subsidy component was less than onequarter of 1 percent of the value of exports. High-technology industries were not major recipients of either strategic or export loans. The third major tool of industrial policy was trade controls. Taiwan pursued import-substituting industrialization (ISI) policies in the 1950s, and its trade regime in the 1960s and 1970s was characterized by relatively high nominal tariffs, especially in agriculture; ubiquitous nontariff barriers—again, especially in agriculture; restrictions on inward foreign investment; and the promotion of state-owned firms. However, beginning in the 1960s, policies were adopted to limit the inefficiencies associated with trade protection. With respect to domestic sales, producers seeking protection had to justify protection on the basis of their ability to compete against imports and were subject to time-phased price controls that forced INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

53

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 54

Figure 2.6 Average tariff burden in Taiwan tariff revenue divided by total import value (percent) 45 40 35 30 25 20 15 10 5 0 1955 1965

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

Source: Smith (1997, table 2; 2000, table 2.2).

them to reduce prices in the local market to within 5 percent of comparable imports by 1973 (Lee and Liang 1982). With respect to foreign sales, the impact of trade protection on final producers was partly offset by various tax rebate schemes, duty drawbacks (as in the case of Korea), and the creation of export-processing zones and bonded manufacturing warehouses; the latter institutions eventually accounted for a significant share of Taiwanese exports. As a consequence of these policies, actual tariff collections were well below statutory rates (figure 2.6). Beginning in 1989, the government undertook a far-reaching trade liberalization that brought the level of trade protection down to developed-country levels, at least in the manufacturing sector.36 Putting the tax, subsidy, and trade components together, Smith (1997, 2000) calculates effective rates of assistance. Among her results is the sur36. As in the case of Korea, Rodrik (1995, 1999) argues that it was investment, not exports, that drove Taiwan’s boom. However, the evidence does not support this view in the case of Taiwan. Rodrik avers that export real exchange rate data do not exist, and argues, on the basis of a nominal exchange rate index, that the increase in export incentives in the late 1950s was insufficient to generate the dramatic increase in exports that began around 1960. However, Lee and Liang (1982) calculated an export real exchange rate index for Taiwan. It shows a 65 percent increase between 1956 and 1959, and rough stability thereafter (Lee and Liang 1982, table 10.3). Contrary to Rodrik’s view, there was a substantial increase in export incentives immediately preceding the Taiwan export boom. 54

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 55

prising one that in 1989, assistance was so great in miscellaneous food products, nonalcoholic beverages, wool and worsted fabrics, certain chemical products, cement products, industrial fertilizers, other artificial fibers, medicines, and motor vehicles that these sectors were producing negative value added at world prices. Although there is no documentation of this phenomenon for the 1960s and 1970s, the fact that it existed so late in Taiwan’s evolution suggests that it could have occurred in the earlier period. Beyond these standard tools of selective intervention, there was also another set of policies conducive to the development of the manufacturing sector, namely, the establishment of a large number of institutions that were designed to identify, transfer, diffuse, and efficiently absorb foreign industrial technologies and then to undertake innovation. These latter policies were largely introduced in the late 1970s and 1980s, although precursors existed in the 1960s, including the Hsinchu Science Park and the Industrial Technology Research Institute (ITRI).37 These efforts reflected the fact that Taiwan’s policies were more neutral than Korea’s and Japan’s with respect to firm size (Haggard and Cheng 1987). Much of its industrial development was based on firms with fewer than 100 employees. Centralized research could be justified on the standard grounds that social rates of return to R&D exceed private returns and economies of scale exist in R&D; the science park could be viewed as a means of generating economies of scope in the use of critical services such as accounting and consulting that were provided by the park. Moreover, part of the rationale of the science park was to demonstrate to expatriate Taiwanese, residing largely in the United States, that Taiwan was committed to a serious effort in high technology. Whether this was as important as the high salaries in luring engineers back to Taiwan is unknown. Although the state’s role in supporting research was prodigious, the amount of R&D performed by the private sector grew rapidly over time, increasing ninefold between 1980 and 1990. And although the state-financed share declined commensurately, at the start of that decade the government accounted for just under half of Taiwan’s R&D expenditures—still a significant share (figure 2.7). In addition to the relatively generic centralized research efforts, through a funds-matching policy, the government provided direct R&D subsidies to private enterprises in targeted industries. (Not surprisingly, once the program was initiated in 1991, the list of eligible industries steadily lengthened.) In an economywide sense, the amount of funding appears relatively minor, although for the favored projects support was not insubstantial, with private-sector expenditures, matching funds, and direct subsidies each accounting for about one-third of the budgets (Wang 1998). The government also fostered the creation of venture capital funds to provide capital for these start-ups. Intellectual property rights protection, 37. The most thorough analysis of these institutions is contained in Dahlman and Sananikone (1997). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

55

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 56

Figure 2.7 Research and development in Taiwan, 1978–2000 R&D expenditure as a percentage of GDP

government share of total R&D expenditure (percentage)

2.5

70 60

2.0

Government share of total R&D expenditure

50

1.5

40 R&D expenditure as percentage of GDP

30

1.0

20 0.5 10 0 1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

0 2000

Source: Taiwan Statistical Databook (various issues) and National Statistics of Taiwan, Republic of China.

which had been notoriously lax in Taiwan, was tightened in the 1990s in response to internal factors (the growth of activities producing intellectual property domestically) and external pressures for better intellectual property right enforcement, bilaterally from the United States, regionally through the Asia-Pacific Economic Cooperation forum, and multilaterally through Taiwan’s World Trade Organization accession negotiations.

Assessment As in the case of Japan and Korea, a variety of studies have been carried out on the effectiveness of policies in stimulating more rapid growth. Smith (2000) presents work conducted by herself and by others that, as in the case of Japan and Korea, generally fails to find links between these interventions and sectoral TFP growth or trade performance in the 1980s. Rather the pattern of intervention appears to be driven more by political economy considerations, such as sectoral employment, the presence of large firms, or the degree of sectoral concentration, than by dynamic comparative advantage. Other evaluations using a variety of strategies support her results (Little 1994; Pack 1992a; and the chapters by Thorbecke and Wan in Thorbecke and Wan 1999). Pack and Lin (2001) follow a different strategy, assuming there are nonmeasured forms of stimulation such as the subsidy equivalent of the 56

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 57

establishment of industrial parks, centralized research institutes, and centralized productivity centers. These may be large and have a limited correlation with the effective rates of protection (or subsidy). Their method then assumes that any exceptional growth in the favored industrial sectors was due entirely to industrial policy and that the TFP growth rate in such sectors was doubled. With these assumptions, which are very favorable to finding a positive role for industrial policy, they find that industrial policy could have added one percentage point of TFP growth in manufacturing. Given that manufacturing accounted for about 30 percent of GDP, this would have increased aggregate TFP by roughly 0.3 per year out of a total gross domestic product growth rate of 10 percent per year in the period 1962–89—not trivial but hardly the entire story of Taiwan’s development. The high rate of TFP growth in all sectors, even neglected ones, the high rate of saving and investment, even apart from the higher levels induced by industrial policy, and the acquisition of skills through education all played a significant role. Industrial policy may have played a more significant role if one accepts the most optimistic assumptions. The preceding assumes that the impact of selective industrial policies benefited only the promoted sectors and that the high rate of productivity growth in the neglected sectors was not affected by spillovers. If, however, the rate of TFP growth in neglected sectors was increased indirectly by the growth of the favored sectors, the calculated increment to TFP may underestimate the impact of industrial policy. Indeed, proponents of the benefit of industrial policies often argue that some of its major effects are manifested indirectly in other sectors and dismiss evidence about the limited impact in the targeted sectors as inconsequential. Employing the input-output approach similar to that described above for Korea to obtain a measure of the potential indirect impact of the promoted sectors, Pack and Lin (2001) find similarly small potential gains accruing to the neglected sectors. Finally, Wang’s case study examination of the information technology sector reaches a similar conclusion about the modest impact of Taiwan’s selective interventions: “Rather than the government’s specific sectoral policy and preferential treatment of the strategic industries, it was the government’s creation of a favorable climate and environment that contributed the most to the blossoming development of Taiwan’s information industry” (Wang 1998, 85).

Exports, Growth, and Productivity At a number of points we have suggested that a characteristic differentiating Korea and Taiwan—and Japan to a smaller extent—from other countries that have intervened intensively was their emphasis on exports. Export growth played a role, we believe, in a number of dimensions: it faINDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

57

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 58

cilitated macroeconomic management; it was used as a standard for success of firms that precluded the types of inefficiency that allowed firms selling in protected domestic markets to avoid efforts to improve their productivity; and it had some productivity-augmenting effects on the domestic production, although some of such productivity augmentation resulted from efforts that were undertaken for domestic sales.

Macroeconomic Management While it is possible to implement a monetary-fiscal policy that would permit 10 percent growth in GDP over an extended period, a domestically oriented economy will have greater difficulty doing so than one with growing exports. While international demand for imported goods may fluctuate, relatively small changes in real exchange rates may offset fluctuations. Increasing a small country’s share of the huge world market by 0.1 percent per year can generate a 17 percent growth rate in exports over a long period. It is considerably more difficult to pursue a policy that stimulates growth rates in investment of this magnitude without quickly encountering diminishing returns in the domestic market. To take one example, much of the early growth in Japan and then Korea and Taiwan consisted of simple, labor-intensive products such as clothing, toys, sporting goods, and very simple electronics. Clearly a country could not simply increase production of such items at these rates for a purely domestic market—which is the point of RosensteinRodan’s emphasis on a big push in which simultaneous expansion of the output of many industries would provide sufficient growth of demand so that the complementary income elasticities would prevent diminishing returns. But such diffuse growth would have been technically difficult to obtain in many industries at once—certainly in Korea and Taiwan, which had relatively little industrial experience outside of these products. Exports permitted economizing of relatively scarce industrial expertise as well as providing a vent for surplus unskilled labor. If an attempt at more domestically oriented rapid growth had occurred, we conjecture that it would have incurred diminishing returns to capital and discouraged the maintenance of high investment. As we note in chapter 4, perhaps the main “miracle” in the countries in question was the maintenance of a high rate of investment and its relatively efficient absorption, something we doubt could have transpired with a big push. Thus, even had the governments of these countries possessed the skills (which we doubt) to successfully generate a fiscal surplus combined with a monetary policy that led to low interest rates to encourage investment, that effort would have floundered because of the microeconomic problems of entering many industries simultaneously.

58

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 59

Another benefit of growing exports is their conferring on policymakers the potential for greater latitude in setting policy. After the oil price increases in the 1970s, Japan, Korea, and Taiwan were able to pursue a policy of reducing domestic absorption and altering the real exchange rate between tradables and nontradables with some confidence that exports would respond to the new price structure and a serious recession would be avoided—contractionary devaluation could be avoided. Countries that had pursued ISI failed to pursue this policy package in the 1970s, possibly because of export pessimism, given the previous domestic orientation of firms.

Exports as a Measure of Success of Government Intervention We have suggested in this chapter, following Jones and SaKong (1980), that in Korea the government used exports as a numeraire that informed its decisions about whether firms deserved continued favorable treatment. It is less clear that this standard was used as bluntly in Taiwan or in Japan. The Korean process provided a standard and an implicit threat of a reduction in firms’ benefits if they did not meet export targets. The export standard made policy implementation easier given the complex policies that were pursued, although it could plausibly be argued that a less cumbersome system such as subsidies on export value added would have worked as well.

Exports and Productivity There are two types of benefits from export growth that support more rapid demand growth: national and industry specific. The national benefits accrue from the more rapid reallocation of labor from lower productivity uses in the agricultural and urban informal sector to higher productivity in the formal industrial sector, which is in turn facilitated by the rapid growth in manufactured exports.38 De Melo (1985), using a standard input-output-based demand decomposition, calculated that in Korea in the period 1955–73, the more rapid expansion of exports than of other components of final demand accounted for roughly three-quarters of the 28 percent faster growth of the manufacturing sector than would have occurred had exports expanded by the same rate as other components of GDP. Similar results were obtained for Taiwan. This “disproportionate” expansion of exports generated rapid employment

38. The magnitude of gains is discussed in Pack (1997).

INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

59

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 60

growth in these sectors. It has been calculated that the reallocation of labor in the decade of the 1960s contributed between 12 and 18 percent of the growth of economywide labor productivity (De Melo 1985). In addition, industry may gain as the more rapid growth of output permits scale economies to be realized and learning by doing, a function of cumulated output, is compressed into a shorter period of time. But both reallocation and the exploitation of static scale economies are exhaustible sources of productivity gain, and unless there is a change in product, learning by doing is also exhaustible. Are there other, more sustainable productivity gains that are accrued from exporting? Several types have been suggested, including gains from knowledge obtained simply by being forced to compete in international markets and unintended spillovers generated by customers. A number of studies have examined the impact of exports on productivity, and the results have been inconclusive.39 Some authors claim that the sequence goes from increases in productivity to growth in exports (Clerides, Lach, and Tybout 1998). However, considerable case study evidence suggests that such econometric exercises may be misleading. Consider the productivity-enhancing efforts of Samsung as it tried to initiate microwave oven production in the 1970s: It took the team a year of 80-hour weeks to complete the first prototype (in 1976) but the plastic in the cavity melted in a test. . . . Finally in June 1978, after two years, the team developed a model that survived the test; but it was too crude to compete in the world market. Samsung incrementally improved the product and developed a makeshift production line, producing one over a day, then two, which it placed in local bakeries for feedback from users (Kim 1997, 137).

This intense activity occurred in response to a large order from the nationwide retailer J.C. Penney. In this case the export order catalyzed the increase in productivity, with the growth in TFP (or product quality) preceding the actual export by several years. The causality goes from an export order to productivity. In principle, a domestic order of comparable size could have warranted the investment in product development. However, the size of the foreign order and its prospective magnitude if there was an initial success dwarfed any potential domestic sales, as Korean per capita income was then still far from levels at which microwave ovens are purchased. Although a production subsidy could have been offered by the government and achieved a similar result, officials would have had to be implausibly well informed to have chosen both the right product and the appropriate firm. Hobday (1995), who examines the development of the electronics industry in several Asian newly industrializing countries, also underlines the importance of the contracts the local firms obtained as original equip39. A thoughtful review of the issues and empirical evidence is given in Westphal (2002). 60

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 61

ment manufacturers to companies such as Radio Shack. The firms not only received technical aid from their customers but also, given the large size of the orders, were more willing to undertake their own R&D. Thus the original finding of Rhee, Ross-Larson, and Pursell (1984) about “externalities” from exporting is supplemented, in newer industries, by the role that large foreign sales play in encouraging R&D. Such effects are likely to be difficult to sort out in large multiproduct firms, but nevertheless the case studies are quite uniform in suggesting this link between exports and productivity growth mediated by R&D. Kim (1997) also reports another phenomenon that suggests externalities from exports that manifest themselves in ways in which a standard production function-oriented approach will not capture. The success of Samsung in producing microwaves and its ability to fulfill export orders had a significant benefit for another Korean firm, Lucky Goldstar (LG). When LG had earlier attempted to reach a technology licensing agreement with Hitachi, a Japanese firm, Hitachi had declined, not wanting to provide knowledge to a potential competitor. After Samsung’s demonstrated success at technological development and exporting, Hitachi agreed to a licensing agreement with LG, concluding that it might as well obtain some royalties rather than watch as LG duplicated Samsung’s success or hired workers from Samsung (p. 139). While it is difficult to classify this externality, it is clear from this and other case studies that potential technology licensors are more willing to enter agreements once there has been a demonstration of local technical ability, particularly if it meets the market test of sales in export markets. Given the interest of the foreign technology developer—say, Hitachi—in extracting profit in one form or another, its earlier and more complete divulging of the technological process will benefit the recipient firm, LG. LG obtained a decrease in costs, and presumably an increase in profits, as a result of Samsung’s innovation cum exporting. Such instances of changed willingness of foreign technology suppliers to offer local firms proprietary technologies are not found in case studies in Latin America and India, presumably reflecting the absence of such phenomena given the inability of local firms to offer a credible threat to international licensors. In these regions, most of the innovation efforts were directed to redressing various handicaps imposed by the ISI regime itself. Finally, early retrospective studies of the acquisition of technology by firms in Korea and Taiwan found that OECD importers often provided extensive advice on production methods, design, and quality control that improved the exporters’ productivity (Hou and Gee 1993; Rhee, RossLarson, and Pursell 1984).40 While such advice might have been purchasable through licensing, some of it was akin to trade secrets that could 40. For a theoretical explanation of the willingness of purchasers to provide such knowledge that might have benefited their competitors, see Pack and Saggi (2000). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

61

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 62

not have been obtained otherwise. While the actual benefits of such free transfers has not been estimated, the qualitative evidence suggests that they were of some significance. It is clear from these examples that testing the relationship between productivity and exports can be difficult. Existing econometric studies of this relationship must be viewed as tentative in light of case studies that delineate a very complex relationship between export orders and efforts to enhance productivity. None of this is meant to imply that the sole source of productivity growth was the need to improve performance in export markets—on the other hand, the effects of exports on productivity suggested by many case studies may have been of importance.

62

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 63

Appendix 2.1 Total Factor Productivity Growth Total factor productivity (TFP) growth is the increase in real output once labor and capital inputs are taken into account. TFP can be measured at the firm, industry, or national level. It is frequently interpreted as a proxy for technology, but may be affected by other factors such as the quality of management, the regulatory environment, and, depending on precisely how it is measured, economies of scale. In colloquial terms, TFP is how much output bang you get for an input buck. Slightly more formally, the rate of growth of TFP measures the rate of growth of output not accounted for by the rate of growth of capital and labor. There are two alternatives for calculating TFP growth—growth accounting and the estimation of production functions. Growth accounting weights the rate of growth of factor inputs by factor shares and subtracts the weighted growth of inputs from that of output growth to arrive at a measure of TFP growth. Production function estimates regress output on a usually complex functional form of capital and labor to obtain estimates of TFP growth. Most growth accounting implicitly invokes a number of restrictive assumptions, including constant returns to scale, Hicks-neutral technical change, an elasticity of substitution equal to unity in the case of two inputs, and competitive factor markets. If these assumptions are not correct, growth accounting estimates of TFP are less than robust, the question being the extent to which the assumptions are violated and the resulting quantitative impact on the TFP estimates. Production function estimates do not invoke restrictive assumptions. Few of the studies examining the impact of industrial policy on productivity growth differences among sectors within a country have used an estimated production function; most have relied on growth accounting.41 The critical question for our purposes is whether there is any a priori reason to believe that growth accounting estimates of TFP growth in sectors that received extensive promotion are systematically biased downward relative to those that were not targeted sectors.42 Consider how TFP growth measures are obtained. Growth accounting employs observed factor shares of each sector, usually from the national accounts, to estimate partial output elasticities.43 The change in the aggregate amount of inputs is calculated using the Tornqvist index, as follows:

41. There have been production function estimates for the national economy for the Asian countries, for example, Kim and Lau (1994). 42. This section relies on Nelson and Pack (1999). 43. Some analysts, such as Yoo (1990), adjust the shares to correct various deficiencies, such as those dealing with the remuneration of unpaid family members, and the factor shares are derived from a combination of sources including input-output tables. INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

63

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 64

T = i [1/2(Si,t + Si,t–1) (ln xi,t – ln xi,t–1)]

(1)

where Si,t is the observed share of factor xi in period t. This is subtracted from the log difference in output to obtain TFP growth, A* = log(Yt/Yt–1) – T

(2)

Several critical factors determined the estimate of A*. In particular, the Si,t may be endogenous, reflecting technical or structural change (Nelson and Pack 1999). Existing growth accounting exercises assume Hicksneutral technical change. But this assumption cannot be supported by independent production function estimates for any industry in one country given the impossibility theorem of Diamond, McFadden, and Rodriguez (1972), which shows that for a general neoclassical production function, the elasticity of substitution and the bias of technical change cannot be estimated simultaneously.44 Nelson and Pack (1999) show that the Si,t are endogenous by assuming a neoclassical production function, Q = f(K,mL), in which m represents Harrod-neutral (labor-augmenting) technological advance. The rate of change of factor shares Si,t is a function of the elasticity of substitution, , and m, or S*K = [S0L (1 – )/] (m – k*)

(3)

S*L = [S0K (1 – )/] (k* – m)

(4)

where k* is the growth rate of the capital-labor ratio. Equations (3) and (4) show that the factor shares utilized in calculating the Tornqvist index are affected by the elasticity of substitution, any bias in technical change, and changes in capital intensity. In growth accounting exercises, the Si,t are assumed to provide information about the elasticity of output with respect to factor inputs. But the Si,t are “uncontaminated” measures only if the assumed underlying production function exhibits Hicks-neutral technical change. If technical change was, in fact, labor augmenting as in (3) and (4), SK,t used in (1) would have been lower, and hence the calculated value of T would have been smaller (as k* was greater than 0) and the calculated value of A* would have been greater. The critical question for our purposes is whether the reduction in S*K (and the increase in S*L) is likely to have been systematically related to promoted or neglected sectors. Most studies of sectoral production functions cannot reject the null hypothesis that the long-run elasticity of substitution is identical across sec-

44. Kim and Lau (1994) estimate an aggregate production function and solve this problem by using several countries in their pooled cross-section time-series analysis. 64

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 65

tors and is, in fact, unity.45 Even if there were some sectoral differences among sectors in Japan, Korea, and Taiwan during the periods considered, the impact on the capital and labor shares is likely to have been small unless there were much larger differences than were observed in the values of k*. Nevertheless, some relatively small understatement of the value of A* may be present in the various tests of the effect of industrial policy, but these should not vary systematically between promoted and nonpromoted sectors. Perhaps of greater concern is the possibility that input markets were distorted (for example, because of the suppression of unions), and the factor shares may not yield good estimates of the elasticity of output with respect to the factor in question. If output markets were not competitive, for example, as a result of high rates of effective protection or high concentration, markup pricing could have given rise to distorted values of the Si,t. In either case, it cannot be assumed that factor shares represent competitive imputations derived from Euler’s theorem. Factor markets distortions may have been important in Korea and perhaps in Japan. In both it is widely believed that wages were suppressed during most of the period. There was a significant decrease in the capital share during the political liberalization of the late 1980s, when previously docile unions became more assertive. While the decline could have been due to a very large decrease in FK, the rapidity of the change and the confluence with greater union autonomy suggest an end to wage repression. This would imply that SL,t was artificially low and that SK,t was thus overstated. Given that capital was the fastest-growing factor in all sectors in Korea, this would overstate the value of T and thus understate the value of A*. However, it is not clear that there should be any specific pattern of this understatement between the heavy and chemical industries (HCI) that were promoted and other sectors. On the other hand, the HCI sectors had a much lower cost of capital because of preferential lending to them, and thus their observed capital share may have had a downward bias. This implies that the index of inputs, T, in the HCI sectors was too low, and hence their values of A* as typically calculated are too high. In addition, the protected sectors had typically higher rates of effective protection, which increased value added at domestic prices. The precise distribution of value added between labor and capital is unknown. If, however, it allowed firms to pursue some form of markup pricing that favored profits, then Hall’s (1990) results show that the correct value of SK,t is less than that observed in the national accounts, introducing a source of upward bias in T and of downward bias in A*. While domestic competi-

45. Different issues arise for the aggregate economy insofar as shifts among sectors and changes in the size composition of firms can lead to a variety of results. See Nelson and Pack (1999). INDUSTRIAL POLICIES IN JAPAN, KOREA, AND TAIWAN

65

02--Chapter 2--21-66

3/14/03

10:13 AM

Page 66

tion could have reduced or eliminated rents, the high levels of concentration in Japan and Korea militated against this. In sum, then, there are many potential pitfalls in the use of TFP estimates as the sole test of the impact of industrial policy. It is not obvious given the above that there is any systematic bias between promoted and neglected sectors, nor that the magnitude of any bias is likely to be large. Nevertheless, this suggests that other forms of evidence such as the concentration of lending to declining sectors should be used in assessing the impact of industrial policies, as is done in the body of this chapter.

66

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 67

3 Unintended Consequences

An important dimension of selective intervention is its distributional impact: by design it supports some sectors to the detriment of others. As a consequence there is a natural tendency for policymakers, whether elected or not, to favor incumbent firms or industries that constitute existing political pressure groups. Indeed, for this reason, absent extraordinary insulation of decision makers from political pressure, selective intervention will tend to target declining sectors (Baldwin and Robert-Nicoud 2002). Probably the single best illustration of this phenomenon of political capture is the lion’s share of direct subsidies going to declining natural resource sectors in Japan (figure 2.1). Paradoxically, the contemporary political histories of Japan, Korea, and Taiwan all conveyed certain advantages in this respect, even if they were not entirely successful in avoiding this pitfall. In this chapter we consider two related questions: How successful have the three nations been in the political management of their selective intervention policies? Have there been negative unintended consequences that should be weighed against any positive contributions of these policies?

Political Economy For reasons that differed in each case, after World War II the governments of Japan, Korea, and Taiwan had little legitimacy. Japan had suffered a traumatic defeat after initiating the war in the Pacific. Korea had gained independence from its Japanese colonial ruler but had then been partitioned and suffered a devastating war that destroyed much of the infrastructure and caused enormous casualties between 1950 and 1952. Taiwan 67

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 68

was the base of the defeated Kuomintang government that had hastily left the mainland in 1949. In each case, the government eventually tried to establish its legitimacy by emphasizing economic growth—in the 1950s in Japan and in the early 1960s in Korea and Taiwan. In all three, land reform overcame one set of opponents to policies that were conducive to growth with equity; in turn this sharing in rapid growth may have led to a perception that government policies benefited the general population.1 The combination of political trauma that would disrupt existing ties and the need to deliver better living standards to generate legitimacy may have weakened the tendency toward political capture. On a more mundane level, during this phase the governments of Japan, Korea, and Taiwan had weak legislative branches dominated by bureaucracies, so the logrolling problems associated with legislatures were minimal. Within governmental ministries, bureaucratic hierarchy could presumably deliver internally consistent plans, with any possible inconsistencies across ministries resolved by executive fiat. Of the three countries, Japan had the weakest chief executive, and, perhaps not surprisingly, conflicts between competing ministries are a recurrent feature of Japanese politics. An example is the perennial clash between the Ministry of International Trade and Industry (or its successor, the Ministry of Economy, Trade, and Industry) representing the interests of the electronics firms and the Ministry of Posts and Telecommunications (or its successor, the Ministry of Public Management, Home Affairs, Posts, and Telecommunications) representing the interests of Nippon Telephone and Telegraph.2 Lack of policy coherence is another interpretation of the result that policy interventions were not on the whole welfare enhancing. In the case of Korea, during the most intensive phase of industrial policy implementation in the 1960s and 1970s, coherence was achieved through enormous centralization of power in the Blue House (Korea’s White House) under Park Chung-hee. With Park’s death, however, succeeding governments were less able to insulate economic policymaking from either bureaucratic or popular politics. Kang (2002a,b) argues that under the Park regime there was a relative balance between the strong Korean state and a limited number of large, diversified chaebol. In Kang’s analysis, the rough balance of power between the state and the firms

1. Rodrik (1995) makes a similar argument. See also Campos and Root (1996). 2. Inevitably what is at issue is the desire of the electronics firms to use telecommunications reform to encourage the growth of electronic data transmission and other activities that could be expected to increase demand for electronic equipment such as computers. These disputes can result in protracted periods of uncertainty and policy paralysis until the interministerial conflict is resolved. In 2000–01 Japan undertook a number of telecommunications reforms. Nevertheless, the principal theme of the Ministry of Economy, Trade and Industry’s 2001 White Paper, which was released after the telecom reforms were enacted, was the need for further reform of the telecom sector—the purview of another ministry. 68

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 69

along with the diversified nature of the firms tended to prevent the struggle over rents from degenerating into predatory state behavior on the one hand or intensive rent-seeking by narrow sectoral interests on the other. The change came with democratization in the late 1980s. The development of more genuine competition among the political elite increased the importance of political funding in Korean politics (as in all democracies) and created a seller’s market, so to speak, shifting the balance of power toward those chaebol that could supply the cash. In Kang’s telling, this shift, combined with the weakness of Korea’s underlying financial system, gave rise to the debt-fueled crisis of 1997. One implication of this model is that more difficult times may lie ahead for Taiwan as it experiences its own development of more genuine political competition. In the case of Taiwan, industrial policy was implemented by a competent bureaucracy subject to political control by an authoritarian leader and a strong political party in the form of the Kuomintang. Yet even in this situation, the actual implementation of industrial policies has been subject to considerable parochial political pressure. Whatever the extent of interventions in the 1950s and 1960s, by the 1970s and especially the 1980s, the government, convinced that its industrial policies were having only a modest impact at considerable cost, was actively attempting to scale back incentives. This attempt to rationalize policy ran into political constraints, however. Ironically, the effective rate-of-assistance estimates calculated by Smith (2000) exhibit greater cross-sectoral dispersion at the end of the 1980s than at the beginning of the decade, as politically influential sectors were able to preserve their perquisites in the context of overall shrinking support to industry. Similar problems occurred as the Korean government, recognizing some of the problems with the heavy and chemical industry effort, tried to change course. While it is often noted that policies to encourage import substitution create constituencies that make trade liberalization difficult to put into practice, long-sustained export-oriented selective policies may generate similarly powerful interest groups. Many policies appeared to be determined more by narrow political considerations than by the pursuit of economic efficiency, circumstances that imply considerable scope for intervention to go seriously awry.

Long-Term Use May Be Hazardous To One’s Health Setting aside the question of why the outcome of these political struggles among the political and business elite was a relatively efficient set of policy choices, the issue arises as to whether those choices were inevitably unsustainable—that is, whether even in countries where selective intervention policies are well implemented by international standards, such policies can be counterproductive because of the difficulty of reversing UNINTENDED CONSEQUENCES

69

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 70

them with changing circumstances. In this regard, the financial sector appears to be the key point of vulnerability because of its (until recently) underappreciated impact on other sectors of the economy. The rents that formed the basis for politicized transfers were in substantial part generated by the repression of the financial system that was a prerequisite for some of the basic industrial policies that the three countries pursued. This is problematic because of the potential for negative knock-on effects in terms of industrial performance. Financial repression is a requirement for capital channeling and can be expected to have two deleterious effects. One is simple corruption as firms compete through bribery to obtain rationed credit. The second and possibly more insidious effect is the dumbing down of the financial system as business acumen is made subservient to bureaucratic dictates and risk is socialized. The chaebol could use capital from favored projects to cross-subsidize other ventures, confident that the government would not allow them to fail. The result was investment without regard to rates of return and weak corporate balance sheets. Without workable bankruptcy or “exit” policies to discipline failures, management strategy amounted to unlimited expansion—what Yoo (1999) called “survival of the fattest.” Statistics on chaebol do not exist for the 1960s (because of lack of balance sheet data), but SaKong (1993) documents that the share of the top ten chaebol in South Korean GDP rose from 5 percent to 23 percent in the decade between 1973 and 1982. According to the OECD, “shareholder value was systematically destroyed from the late 1980s onwards” (OECD 1998, 23). The events of recent years are a testimony to this weakness—the $73 billion bankruptcy of Daewoo, the country’s second-largest chaebol, at the time was the largest corporate failure in world history, and Hyundai, the country’s largest, literally disintegrated under market pressure. In Korea the suppression of the financial system and the use of directed credit to individual firms discouraged the accumulation of normal financial evaluation skills and affected the quality of financial intermediation. Low-cost loans were used to encourage many industrial firms to expand beyond their core competence; capable manufacturing firms entered the resort industry or others that were quite different from their existing capabilities, both in production and marketing. Even in early 2002, four and a half years after the onset of the crisis, government officials contend that Korean banks were not capable of dealing with the weakness of highly indebted firms, a problem that has plagued Japan since 1990. The soft budget constraints imposed on the banks by government support were then reflected in the inability of banks to discipline their borrowers.3 This may also have encouraged trade unions to resist necessary changes in the labor market. 3. Don Kirk, New York Times, February 23, 2002. 70

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 71

While a full scholarly understanding will take some time to emerge, it seems likely that many of the putative benefits of industrial policy were eventually partly offset by the unforeseen consequences the policy set in motion. A strong case can be made that a significant mistake in Korea was the opening of the capital account with inadequate supervision, a phenomenon that is not related to industrial policy. Nevertheless, one major problem that led to harmful long-term effects—and would have done so even if the capital account had not been liberalized—was the impaired ability of banks to evaluate loans. The government’s motivation to prod firms into new sectors stemmed from an often-justified concern that laborintensive activities would succumb to still lower wage countries such as China. But choosing the precise sector, for example, machine tools, and directing loans to this sector—or, more precisely, to a few firms—gave banks little incentive to improve their ability to evaluate business models. Even in the case of nondirected credit, the banks exhibited considerable laxness in their assessment of loan quality. According to the Korean government’s official estimate, the net cost of the financial sector cleanup will total 87 trillion won, or about 16 percent of 2001 GDP. In the case of Taiwan, the recapitalization of the banking sector is expected to cost NT$1.05 trillion, or about 11 percent of GDP. In Japan, private estimates of bad loans in the banking system are on the order of 35 percent of GDP; even if the recovery rate were 50 percent, this would put the net cost on the same order of magnitude as the cases of Korea and Taiwan.4 By comparison, the US savings and loan cleanup cost less than 2 percent of GDP. On the other hand, not all (or even most) of the blame for the financial crisis in these economies can be attributed to industrial policy. Many of the protective policies had been largely dismantled, and other policies had been reduced in scope. Nevertheless, the decline of 7 percent in GDP in Korea in 1997–98 was not due to a sudden decrease in productivity in promoted sectors, nor to the collapse of crony capitalism. Most analyses suggest that the sources lay elsewhere—in inadequate supervision of the banking sector combined with the liberalization of the capital account that was urged on Korea by external advisers (Radelet and Sachs 1998; Furman and Stiglitz 1998; World Bank 1998). Yet the laxity of the domestic banking system in many dimensions compounded the dangers inherent in capital liberalization, and this lack of skill in the private sector was exacerbated by the government’s continuing to view these banks mainly as a passive conduit of funds to favored firms. The slow growth of GDP per capita in Japan in the 1990s, extraordinary for a developed nation, may

4. Chosun Ilbo, June 28, 2002; Jason Dean, “Seize the Day,” Far Eastern Economic Review 5, September 2002: 48; Dan Nystedt, “Taiwan rebuffs criticism by IMF,” Financial Times, September 13, 2002; Bergsten, Ito, and Noland (2001). UNINTENDED CONSEQUENCES

71

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 72

have had even deeper roots in the utilization of the banking system in Japan as a key element in the industrial policy system (Lincoln 1999). Having pursued early industrial policies with careful attention to implementation, Japan, Korea, and Taiwan did not suffer and may have extracted some small benefit for an initial period, although some would argue that they could have done still better given their high saving and investment rates. But the experiences of Korea and Japan in recent years suggest that long-term harmful consequences were not entirely avoided.

Corruption In the extreme, the rough-and-tumble of political competition deteriorates into corruption that degrades public policy. The benign experience in Japan, Korea, and Taiwan during the heady days of intervention and growth may have had unfortunate side effects. In particular, it is hard to reconcile the account of careful monitoring with the widespread impression that all three economies have been subject to corruption. Numerous surveys have been conducted that attempt to measure bribery, corruption, nontransparency, and associated ills. Results from some of these surveys are summarized in figures 3.1 and 3.2. Figure 3.1 displays the percentile rankings of Japan, Korea, and Taiwan in three surveys of perceptions of corruption; the higher the ranking, the worse the corruption.5 The surveys refer to the late 1990s; comparative data do not exist for earlier periods. Although with one exception the countries do not fall into the bottom half of the rankings in any of the surveys, it is fair to say that their rankings are not outstanding. Among the three, their rank ordering in all the surveys (Japan, Taiwan, Korea) is consistent with their respective levels of per capita income. Corruption problems appear to be worst in Korea, which, in addition to being the poorest of the three, has the greatest concentration of political power in the executive, relatively unconstrained by other institutions such as the bureaucracy or a political party (historically, the Liberal Democratic Party in Japan and the Kuomintang in Taiwan). One would expect that industrial policies, which inherently discriminate among firms and industries, would contribute to corruption by expanding the set of favors available from the government, so this result is 5. The sample sizes of the three surveys vary enormously, from 35 countries in the case of the PricewaterhouseCoopers survey to 155 for the World Bank survey. In figures 3.1 and 3.2 all percentile ranks are based on the PricewaterhouseCoopers sample to ensure comparability across the three sources. The countries in this sample are Argentina, Brazil, Chile, China, Colombia, the Czech Republic, Ecuador, Egypt, Greece, Guatemala, Hong Kong, Hungary, India, Indonesia, Israel, Italy, Japan, Kenya, Korea, Lithuania, Mexico, Pakistan, Peru, Poland, Romania, Russia, Singapore, South Africa, Taiwan, Thailand, Turkey, the United Kingdom, Uruguay, the United States, and Venezuela. 72

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 73

Figure 3.1 Corruption rankings for Japan, Korea, and Taiwan Transparency International Japan Korea Taiwan Pricewaterhouse Coopers Japan Korea Taiwan World Bank Japan Korea Taiwan 0

0.1

0.2

0.3

0.4

0.5

0.6 percentile

Note: Larger values indicate higher corruption. Sources: Transparency International: Global Corruption Report 2001. World Bank: Kaufmann, Kraay, and Zoido-Lobatan (1999). PricewaterhouseCoopers: The Opacity Index 2001.

Figure 3.2 Governance rankings for Japan, Korea, and Taiwan Time spent with government bureaucracy Japan Korea Taiwan Government effectiveness Japan Korea Taiwan Competence of public officials Japan Korea Taiwan 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4 0.45 percentile

Note: Smaller values indicate more effective government, less time spent with bureaucracy, and more competent public officials. Sources: Time spent with government bureaucracy: World Competitiveness Report 2000. Government effectiveness: World Competitiveness Report 2000. Competence of public officials: Kaufmann, Kraay, and Zoido-Lobatan (1999). 73

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 74

not surprising. Again, the question is why it has not been more debilitating. Why do the three countries more closely resemble Denmark in outcomes than Nigeria? One possibility is that while the implementation of selective intervention policies at the highest levels was politicized and prone to corruption, once decisions were made at the top, they were implemented in a relatively efficient manner. There are reasons to believe that such systems impose less of a deadweight burden on an economy than systems characterized by cascading corruption all along the line (Shleifer and Vishny 1993). Some support for this notion is provided by the data shown in figure 3.2. The first panel reveals that in terms of the World Competitiveness Report indicator of time spent by senior management interacting with government officials, Korea looks considerably better (in the 17th percentile) than in the corruption indices (43rd to 51st percentile). Japan is in the 6th percentile. In the second panel of figure 3.2, showing the World Bank index of overall government effectiveness, Japan and Taiwan do well, and Korea less so. Taiwan in particular places noticeably higher in the World Bank government effectiveness ranking (11th percentile) than in the corruption rankings in figure 3.1. Finally, the third panel of figure 3.2 reports the percentile rankings from the World Competitiveness Report on the competence of government officials relative to their private-sector counterparts. All three countries are in the top quartile of the sample (with Japan in the top decile), perhaps reflecting the high status of civil service jobs and their relatively high rate of remuneration.6 Taken together, the data presented in figure 3.2 suggest that government performance in Japan, Korea, and Taiwan is stronger than the simple corruption perception indices displayed in figure 3.1 might suggest. A second possibility is that although the overall level of corruption has been high in these countries relative to other countries, corruption is largely concentrated outside the industrial sector, in particular in construction and finance, and hence has not had a major impact on competitiveness in traded goods. In Japan corruption is heavily concentrated in construction, where the losses to bid-rigging, especially on public contracts, have been estimated to exceed the national education budget (Bergsten, Ito, and Noland 2001). Financial repression has encouraged corruption in financial institutions as well, although the impact on the allocation of capital in a macroeconomic sense may be small. Taiwan had similar problems in public works construction and finance (the construction of the Taipei subway is a well-known example) but is distinct in that the ruling political party during the relevant period, the Kuomintang, was organized along Leninist lines and maintained an ex6. See World Bank (1993, chapter 4) and Campos and Root (1996). 74

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 75

tensive commercial network. While party-affiliated firms may have been the beneficiaries of favoritism in various forms, they did not receive onbudget subsidies, and unlike typical state-owned enterprises elsewhere, they were run along commercial lines and expected to make money for the party. Corruption in public works construction is ubiquitous globally, and if Japan, Korea, and Taiwan are unique in this phenomenon, it is only in scale. In any event it does not appear to be related to industrial policy per se. This hypothesis of narrowly circumscribed corruption is less clearly true in the case of Korea, which appears to have the worst problems of corruption of the three. There have been notable corruption scandals in the industrial sector (Noland 2000a), and sons of President Kim Dae-jung, and his immediate predecessor, Kim Young-sam, have been jailed for influence peddling while their fathers were in office. This lack of transparency imposes a penalty on financial transactions in the Korean market, increasing investor hurdle rates and inhibiting the ability of good firms to access capital. The transparency risk premium, separate from and in addition to conventional country and currency risk, inhibits investment in the Korean economy.7 Korea still lacks viable exit mechanisms for failing firms, and business-government relations remain nontransparent and steeped in corruption. The observations developed in this chapter—that even in Japan, Korea, and Taiwan, the implementation of industrial policy has proved politically problematic and contributed to financial sector problems—present a cautionary counterbalance to the possible benefits of selective intervention policies. Having examined potential costs and benefits of the policies from a normative perspective, in the next chapter we look at whether the experiences of the Asian countries are likely to be replicated by contemporary developing countries.

7. See Noland (2001) for more details. UNINTENDED CONSEQUENCES

75

03--Chapter 3--67-76

3/14/03

10:14 AM

Page 76

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 77

4 Replicability

The preceding chapters have examined the potential benefits and costs of a selective industrial policy strategy. This chapter addresses the issue of whether the positive experiences of Japan, Korea, and Taiwan would likely be replicated by contemporary developing countries if they were to pursue a similar development strategy. Three topics are examined: the political prerequisites for successful industrial policy, the possibility that the initial conditions of the three Asian nations were unique and irreproducible, and the extent to which today’s international system forecloses options that were open a generation ago.

Stability and Complexity At the outset, it is worth stating the obvious: for the past 50 years, Japan, Korea, and Taiwan have experienced a degree of political stability that is foreign to most contemporary developing countries. While Korea and Taiwan have faced external tensions and Korea experienced localized internal unrest in 1980 that resulted in considerable loss of life, none of the three have faced any serious domestic opposition to the state itself.1 There have been no civil wars, no armed insurrections, and no secessionist groups; this characteristic alone separates these countries from much of the rest of the developing world. 1. There has been tension and occasionally violence between “native” Taiwanese—those who arrived from the mainland before the 20th century—and those who fled it in 1949. However, this never manifested itself in an opposition to the state itself. 77

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 78

Moreover, as shown in figure 3.2, all have had highly competent government bureaucracies—and bureaucratic competence should not be underestimated given the complexity of the policy interventions undertaken. In terms of the difficulty of the process, consider just one aspect of the trade policies applied in Korea. Sectors were afforded protection from competing imports. As tariffs and quantitative restrictions increased the cost of production in industries that used domestically produced intermediates, firms were given rebates based on the fraction of their output that was exported. This in effect allowed them to face international prices in both product and factor markets. But the rebates were established on the basis of an industrywide standard of intermediate input usage that required something close to a full input-output table, at a time when Korea had only a handful of trained economists. In addition, firms were given wastage allowances, under the assumption, for example, that some of the cotton imported for use in producing exported shirts was wasted in the production process. Insofar as wastage was intentionally overestimated, the allowance provided firms with a duty-free import that could be used in production destined for the domestic market. Such a complex system required substantial numbers of qualified government officials to administer and was obviously subject to the potential for corruption, which in general did not occur. The effect of the complex system was to replicate (roughly) a free trade system, at least for exporters. There was nevertheless a deadweight loss in terms of the use of skilled labor that, at least initially, was probably quite large given the small number of highly trained people during the 1960s and early 1970s. Any country trying to replicate this system would have to consider this cost as well as the large potential for corruption, which in fact has been generated in countries that have followed related though not identical policies in implementing import-substituting industrialization.

Irreproducible Initial Conditions? The contrast between the relatively benign experiences of Japan, Korea, and Taiwan and the more deleterious contemporaneous outcomes elsewhere suggests that perhaps the relevant question for contemporary lessdeveloped countries is not whether selective intervention in Asia worked, but rather why it was not more debilitating. Were initial conditions so favorable that it was more likely to work (at least initially) or more likely to do little harm? Were factor proportions somehow more favorable to its success? Some would point to better initial conditions. Yet the relevance of Japan for developing countries is problematic. While it is true that the level of per capita income in Japan was low in the immediate postwar period, as

78

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 79

we argued in chapter 2 this datum is deceptive on two counts. First, it is a misleading indicator of technological sophistication or societal capacity: Japan, after all, refitted battleships and defeated Russia in the 1904–05 war, and later built aircraft carriers and scored major victories against the military forces of developed Western countries during World War II. This was no typical developing country. Moreover, the conditions obtaining in Japan in the immediate postwar period were surely transitory: like other war-devastated economies, a period of rapid capital accumulation and rebuilding followed the cessation of hostilities. As shown in table 2.1, in the mid-1950s Japan had the highest ratio of human capital to per capita income in the world. By the 1960s, if not before, Japan’s trade pattern more closely resembled those of Western European countries than those of the rest of Asia or the developing world (Noland 1997b, table 4). In this respect, Korea and Taiwan are better comparators, though similarly problematic in certain respects. As has been emphasized in numerous studies, Korea and Taiwan exhibited higher literacy rates than countries with similar per capita income and arguably better infrastructure such as roads and ports at the beginning of their high-growth episode in the mid-1960s.2 Even a brilliantly designed economic program would have floundered if internationally traded goods could not have been moved to and from ports and if the ports had themselves not been fairly efficient. On the other hand, too much can be made of such differences and of the purported benefits of the long Japanese occupation that had been responsible for considerable investment in education and infrastructure, even if only for the colonialists’ intended benefit. Neither country had the university education levels of, say, Argentina, Greece, or the Philippines (table 2.1). And all of the latter had sufficiently good transportation and ports to have engaged in significant primary product exports that would have enabled the importation of industrial intermediates and capital goods. Paradoxically, however, this relative lack of natural resources may actually have facilitated the implementation of reasonably rational industrial policy in Japan, Korea, and Taiwan. Scatterplots of data on labor, physical capital, human capital, and arable land endowments for a number of countries in 1968 are shown in figure 4.1. Each panel of the figure shows a barycentric projection of three endowments. Every endowment point on a ray emanating from one corner of the triangle has the same ratio of the other two factors; points that lie closer to the corner of the triangle have a larger relative endowment of that factor. The point at which the three rays emanating from each vertex intersect in the middle of the triangle indicates the average endowment bundle of the sample. 2. In the case of Korea, this reflected considerable effort at reconstruction after the devastation of the 1950–53 war.

REPLICABILITY

79

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 80

Figure 4.1a Endowment triangle: Labor, physical capital, land (1968 data) Land

4

0.25 Argentina

2 Tunisia Brazil Turkey

1 Pakistan Thailand

Philippines

0.25

Labor

Peru China India

0.25

Canada

1

Spain Greece US

Mexico

Denmark Sweden Finland Malaysia France Italy Austria Israel UK Norway Germany Japan Korea Taiwan Hong Kong Singapore Total

1

2

2 4

Physical capital

4

Figure 4.1b Endowment triangle: Labor, human capital, land (1968 data) Human capital

4

0.25

Japan 0.25 Germany Taiwan Singapore Korea Austria 2 Norway Hong Kong UK Sweden Israel Italy Denmark 1 Philippines France Peru US Finland India Greece Canada Total Thailand Malaysia China Spain Mexico Brazil Turkey Pakistan

1

2 4

Tunisia Argentina

Land

0.25

Source: Noland (1997b). 80

1

2

4

Labor

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 81

Figure 4.1c Endowment triangle: Land, physical capital, human capital (1968 data) Human capital

4

0.25

2 Taiwan Korea

1

Philippines

1

Germany

Sweden Peru UK China Japan Austria Denmark Hong Kong India Norway Total Thailand Israel Italy France US Greece Singapore Canada Mexico Finland 0.25 Spain Malaysia Turkey

Physical capital

0.25

2

1

2 Pakistan

4

Brazil Tunisia

4

Argentina

Land

Figure 4.1d Endowment triangle: Labor, physical capital, human capital (1968 data) Human capital

4

0.25

2 1

0.25

Labor

Korea Denmark Germany Taiwan Philippines Canada Peru Austria Sweden US Greece Japan Argentina UK China Italy Total Thailand Brazil France Turkey Spain Israel Tunisia Finland Hong Kong Norway India Singapore Pakistan Mexico Malaysia

0.25

1

2

1

2 4

4

Physical capital

Source: Noland (1997b). 81

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 82

Thus, for example, in figure 4.1a, Taiwan, Korea, Hong Kong, and Singapore are arrayed across the bottom of the triangle, far from the land endowment vertex, in order of increasing physical capital-labor ratios. The point is that the land-scarce countries of East Asia tend to cluster in each scatterplot—across the bottom in figure 4.1a, near the human capital vertex in figure 4.1b, and so on. In contrast, other major developing countries tend to reveal relatively large endowments of land and low endowments of physical capital, with Argentina, Tunisia, Brazil, and Turkey the most dramatic examples in figures 4.1a and 4.1b. In figure 4.1c, this group is augmented by Pakistan, and in figure 4.1d by Thailand, India, the Philippines, Peru, Chile, and Mexico. These multifactor starting points are important. Leamer (1987) has shown that there is some econometric evidence that land-scarce countries (such as those of East Asia) will tend to specialize in manufactures earlier (i.e., at lower levels of per capita income) and more intensively (i.e., producing more output per worker) than economies with more diversified resource bases. Moreover, economies along the bottom of figure 4.1a will almost surely experience rising wages as physical capital is accumulated and capital-labor ratios rise, generating “growth with equity.” In contrast, in economies with larger natural resource bases, the rents generated by resource extraction will retard specialization in manufacturing and increase the likelihood that capital accumulation might not be accompanied by rising wages (“growth without development”). Both of these implications are supported by data. As shown in table 4.1, Japan, Korea, and Taiwan had relatively equal initial distributions of income and land, the latter reflecting the land reforms that each had undertaken.3 Furthermore, as the multifactor perspective would suggest, the two poorer nations specialized in manufactures relatively early. As shown in table 4.2, by the 1960s they had made considerable strides in exporting manufactures, and their export profiles most closely matched—after one another’s—those of poorer countries in the European sphere such as Austria, Spain, and Israel; over the next two decades their patterns of export specialization closely tracked the evolution of Japan’s (Noland 1997b,

3. The land reform carried out in US-occupied Japan in 1947–48 was one of the most effective in modern times, and it had an important demonstration effect in Korea, where the US military government carried out one reform involving the distribution of Japanese-owned land, and the Korean government subsequently undertook another of Korean-owned land, in part in response to a land reform in North Korea. After the Chinese revolution and the decampment of the Nationalist government to Taiwan, the Kuomintang undertook a similar reform under American tutelage. The three cases have certain similarities: all were undertaken by governing entities that had only weak political ties to the landowners, and all were encouraged by the United States. Moreover, in the cases of Taiwan and the second Korean reform, they were arguably in response to reforms undertaken in rival communist states.

82

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

04--Chapter 4--77-92

3/14/03

Table 4.1

10:15 AM

Page 83

Distributional indicators, circa 1960 Gini coefficient for income

Gini coefficient for land

Japan Korea Taiwan Mean average

0.40 0.34 0.31 0.35

0.47 0.39 0.46 0.44

Argentina Brazil Egypt Hong Kong India Indonesia Kenya Malaysia Mexico The Philippines Singapore Thailand Turkey Mean average

0.44 0.53 0.42 0.49 0.42 0.33 0.64 0.42 0.53 0.45 0.40 0.41 0.56 0.46

0.87 0.85 0.67 n.a. 0.52 n.a. 0.69 0.47 0.69 0.53 n.a. 0.46 0.68 0.64

n.a. = not available Source: Adapted from Rodrik (1994).

table 4). In this regard, they differed fundamentally from the other developing countries of the time (except perhaps the city-states of Hong Kong and Singapore), including those of Southeast Asia, which did not exhibit the rapid and complete move into manufactures that Korea and Taiwan did. Instead, as Leamer’s argument would imply, the more land- and natural-resource-abundant economies specialized less in manufactures, so that, for example, in the 1960s, the export profile of Thailand most closely resembled those of Brazil, Mexico, and Turkey—not those of its Asian neighbors Korea and Taiwan.

Geography Is Not Destiny—Policy Matters None of this is to argue that an early shift toward manufacturing guarantees success. Clearly many countries that attempted to industrialize quickly by protecting industries against competitive imports failed. In these nations manufacturing did not generate the modernization in attitudes or labor skills or the magical externalities predicted by its early advocates. Industrialization is not a fail-proof strategy—it must be implemented successfully. The Philippines began the postwar period with many advantages, including high levels of education (tables 2.1 and 2.7), a large number of English speakers, which is conducive to trade relations, and close

REPLICABILITY

83

04--Chapter 4--77-92

84

Table 4.2

Export similarity, 1968

3/14/03 10:15 AM

n pa a Ja a re si Ko 0.36 ay al o M ic 0.11 0.20 ex ay M 0.24 0.33 0.30 rw an No st 0.44 0.24 0.36 0.44 ki a P 0.22 0.40 0.11 0.28 0.12 ru Pe 0.05 0.19 0.36 0.48 0.31 0.16 0.05 0.24 0.34 0.38 0.12 0.11 0.53 0.27 0.26 0.52 0.34 0.27 0.17 0.23 0.45 0.33 0.30 0.60 0.42 0.21 0.24 0.56 0.29 0.20 0.28 0.55 0.10 0.14 0.47 0.53 0.17 0.49 0.22 0.27 0.17 0.11 0.22 0.40 0.59 0.25 0.19 0.38 0.20 0.20 0.29 0.38 0.31 0.13 0.52 0.09 0.22 0.19 0.59 0.19 0.41 0.37 0.62 0.30 0.16 0.33 0.40 0.18 0.12 0.48 0.29 0.22 0.39 0.40 0.16 0.25

Page 84

a in nt ge r ria A st il Au Austria 0.19 az da Br Brazil 0.57 0.20 k na ar Ca Canada 0.20 0.42 0.30 nm d e D Denmark 0.52 0.48 0.22 0.33 an nl ce Fi Finland 0.15 0.45 0.18 0.43 0.34 y an an Fr France 0.32 0.67 0.27 0.51 0.55 0.35 m er e ng G Germany 0.18 0.62 0.13 0.44 0.50 0.31 0.72 ec Ko re G a g Greece 0.54 0.25 0.55 0.37 0.30 0.19 0.41 0.26 n si ne Ho Hong Kong 0.13 0.43 0.10 0.16 0.29 0.18 0.33 0.32 0.15 do n I Indonesia 0.27 0.11 0.26 0.24 0.14 0.08 0.20 0.09 0.31 0.05 el ra Is Israel 0.40 0.39 0.37 0.28 0.32 0.21 0.48 0.35 0.47 0.39 0.21 ly Ita Italy 0.25 0.67 0.18 0.46 0.51 0.29 0.75 0.71 0.30 0.43 0.18 0.45 Japan 0.15 0.67 0.12 0.33 0.47 0.34 0.62 0.63 0.19 0.46 0.05 0.35 0.64 Korea 0.19 0.41 0.28 0.30 0.29 0.28 0.32 0.25 0.28 0.60 0.14 0.35 0.37 Malaysia 0.15 0.19 0.21 0.35 0.18 0.15 0.19 0.12 0.26 0.10 0.51 0.17 0.18 Mexico 0.58 0.30 0.60 0.47 0.35 0.23 0.44 0.28 0.66 0.17 0.27 0.47 0.34 Norway 0.20 0.45 0.21 0.52 0.44 0.47 0.46 0.41 0.34 0.18 0.15 0.28 0.36 Pakistan 0.23 0.20 0.20 0.13 0.17 0.08 0.22 0.14 0.26 0.29 0.09 0.20 0.24 Peru 0.36 0.11 0.36 0.34 0.14 0.08 0.21 0.09 0.46 0.05 0.31 0.22 0.13 The Philippines 0.36 0.13 0.42 0.30 0.13 0.11 0.18 0.07 0.41 0.07 0.32 0.29 0.13 Singapore 0.31 0.33 0.30 0.36 0.33 0.23 0.41 0.29 0.35 0.22 0.75 0.35 0.40 Spain 0.47 0.46 0.44 0.44 0.48 0.33 0.60 0.44 0.53 0.29 0.30 0.57 0.58 Sweden 0.15 0.65 0.22 0.59 0.45 0.59 0.60 0.64 0.24 0.25 0.13 0.26 0.57 Taiwan 0.44 0.49 0.45 0.31 0.31 0.30 0.43 0.32 0.41 0.52 0.18 0.52 0.48 Thailand 0.46 0.13 0.70 0.28 0.16 0.11 0.23 0.12 0.50 0.10 0.38 0.38 0.16 Tunisia 0.30 0.28 0.32 0.39 0.24 0.16 0.39 0.22 0.39 0.10 0.49 0.26 0.28 Turkey 0.51 0.15 0.56 0.22 0.18 0.13 0.26 0.12 0.72 0.09 0.29 0.37 0.16 United Kingdom 0.20 0.62 0.14 0.48 0.52 0.31 0.77 0.81 0.30 0.35 0.11 0.40 0.75 United States 0.31 0.52 0.31 0.54 0.51 0.31 0.71 0.70 0.39 0.25 0.27 0.40 0.67

s ne pi lip i e Ph or e ap Th g n n Si 0.26 ai en Sp 0.32 0.50 ed an Sw 0.17 0.28 0.40 iw nd Ta la 0.32 0.36 0.58 0.32 ai a h T si 0.32 0.37 0.45 0.14 0.39 ni ey Tu 0.46 0.47 0.39 0.27 0.22 0.23 d m rk ite do Tu 0.36 0.26 0.42 0.15 0.37 0.50 0.30 Un ing K 0.08 0.31 0.51 0.60 0.35 0.15 0.23 0.13 0.25 0.41 0.49 0.60 0.36 0.27 0.31 0.29 0.72

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 85

affiliation with the United States. Nevertheless, despite predictions in the 1950s that it would be the success story in Asia (Morawetz 1980), its dismal performance reflected import substitution policies similar to those of Latin America. (Indeed, most of the standard empirical studies of the impact of import substitution include the Philippines with Latin American countries.4) A latitude and longitude placing a country in Asia was hardly a guarantor of growth—correct basic policies matter.5 Hence to point solely to initial conditions is inadequate: the differences in the nature of the industrial policies and their implementation are critical. Extensive protection was given to many sectors in India, in Latin American countries, and many of the more advanced African countries, as evidenced by the high rates of effective protection calculated for all of the countries for which such estimates were made. While the general characteristic was that protection rates were highest for consumer goods and lowest for machinery, they were nevertheless high for most sectors. Firms in inefficient sectors could earn significant profits and their employees high wages (paid out of the rents collected from consumers) and faced little credible prospect that protection would be contingent on improved efficiency. There was no monitoring mechanism; once protection was granted, reductions in its level were limited until crises occurred in the 1980s and later. In contrast, in Japan, Korea, and Taiwan there was continuous monitoring of the progress of firms. The clearest example is provided by Korea, in which subsidized credit and protection in the domestic market were contingent on export performance (Jones and SaKong 1980). Exports became the numeraire by which the progress of individual firms was measured. Current data on exports of individual firms were presented at quarterly meetings at the Blue House, the seat of the executive, with all of the firms in a given promoted sector. The information was obtained not from companies but from the difficult-to-falsify bills of lading at Korean ports. Realized exports were compared with targets set by the Economic Planning Board for each firm. As the export targets were constantly increased, firms were forced to improve their productivity in order to lower marginal costs, the alternative being lower profits over time. While many firms initially subsidized their unprofitable exports by cross-subsidies from their profitable (protected) domestic market, clearly this could not

4. See, for example, Little, Scitovsky, and Scott (1970), Balassa et al. (1982), Bhagwati (1978), and Krueger (1978). 5. Ironically, as detailed in Noland (2000b), in the 1990s the Philippines undertook considerable reforms, especially in the financial sector, and under a variety of indices weathered the Asian financial crisis better than the other heavily affected economies. Of course, its per capita income remains far below that of Korea and Taiwan, partly reflecting these countries’ generally better policies for the prior 30 years. REPLICABILITY

85

04--Chapter 4--77-92

Table 4.3

3/14/03

10:15 AM

Page 86

Rates of growth of total factor productivity Contribution to growth of output per worker Growth of output per worker

Physical capital per worker

Education per worker

Total factor productivity growth

Korea 1960–73 1973–84 1984–94

5.6 5.3 6.2

3.2 3.4 3.3

.9 .8 .6

1.4 1.1 2.1

Taiwan 1960–73 1973–84 1984–94

6.8 4.9 5.6

3.9 3.0 2.3

.5 .9 .5

2.2 .9 2.8

Latin America 1960–73 1973–84 1984–94

3.4 .4 .1

1.3 1.1 .1

.3 .4 .4

1.8 –1.1 –.4

South Asia 1960–73 1973–84 1984–94

1.8 2.5 2.7

1.4 .9 1.0

.3 .4 .3

.1 1.2 1.5

OECD other than United States 1960–73 1973–84 1984–94

4.8 1.8 1.7

2.3 1.1 .8

.4 .6 .2

2.2 .2 .7

Country and period

OECD = Organization for Economic Cooperation and Development Source: Collins and Bosworth (1996, tables 6 and 7).

work for long, as the export targets were increased considerably faster than the growth of domestic sales.6 Firms were thus forced to concentrate on improving productivity, which led to intensive efforts to import and assimilate foreign technology (see Kim 1997 on Korea; and Dahlman and Sananikone 1997 and Pack 2001b on Taiwan). Despite controversies about the precise rates of TFP growth in Korea and Taiwan, it is clear that they were above those in Latin America during its import substitution phase (table 4.3), and that these were sustained despite extraordinary rates of capital accumulation (table 4.4). In contrast, elsewhere there was no attempt to combine a stick of control with the carrot of protection. We are aware of no instances in the lit-

6. In the period of interest, 1965–97, Korean exports increased by a compound rate of 15 percent per annum, and Taiwan’s by a similar percentage. Japan’s growth and its share of exports in GDP were much lower than those of the two smaller economies. 86

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

04--Chapter 4--77-92

Table 4.4

3/14/03

10:15 AM

Page 87

Rates of growth of fixed capital stock

Countries

1950–60

1960–70

1970–80

1980–90

Asian newly industrialized countries Indonesia Korea Malaysia Singapore Taiwan Thailand

.055 n.a. n.a. n.a. n.a. .089

.030 .125 .097 .166 .146 .133

.113 .147 .109 .144 .146 .096

.098 .108 .082 .095 .082 .075

OECD France Germany Japan

.062 .067 .117

.077 .062 .145

.054 .037 .092

.029 .023 .053

Africa Ghana Kenya Nigeria

.032 .046 .071

.067 .020 .070

.024 .047 .142

.011 .016 .007

Southeast Asia India Pakistan

.044 .078

.058 .138

.045 .052

.048 .057

South America Argentina Brazil Mexico

.043 .068 .082

.047 .062 .082

.047 .099 .084

.000 .037 .037

n.a. = not available OECD = Organization for Economic Cooperation and Development Note: The growth rate of capital is the fitted growth rate of the fixed capital stock from Nehru and Dhareshwar (1994).

erature on ISI of a government’s actually reducing protection to sectors that did not perform well. As noted earlier, it is impossible to confirm that industrial policies in Asia generated substantial benefits. But as contrasted with the ISI experience elsewhere, no major short-term damage was caused by industrial policies. Korea and Taiwan did experience fairly high TFP growth rates compared with other developing countries, although much of this would have accrued without selective intervention. The major difference, we believe, is the use of some numeraire, particularly exports, to measure success rather than the provision of open-ended protection for inefficient sectors.7 7. Aside from any theoretical reasons preferring exports as a measure of success, as a purely practical matter, performance in world markets was probably the criterion least amenable to rigging by the firms or their bureaucratic counterparts. As mentioned in chapter 2, as a former Korean official put it, “[the international trade statistics] were the only numbers that couldn’t be faked.” REPLICABILITY

87

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 88

Does the International System Now Constrain Industrial Policy? It is frequently claimed that the rules of the World Trade Organization (WTO) and the behavior, singly or in combination, of the World Bank and the IMF foreclose the possibility today of the kind of activist policies pursued by Japan, Korea, and Taiwan in earlier decades. As seen in the previous chapter, at times all three imposed barriers to foreign trade and investment; subsidized production and/or exports directly and indirectly through state-owned financial institutions and the tax system; subsidized R&D and encouraged the growth of knowledge-intensive industries through the creation of science parks, research consortia, and other general incentive programs; and permitted anticompetitive behavior on the part of private firms. The WTO agreement affects the ability of countries to implement industrial policy through two channels: the extension of rules and disciplines to new areas, and the improvements in dispute settlement and enforcement mechanisms over the earlier General Agreement on Tariffs and Trade (GATT) system. With respect to the rules, the Uruguay Round agreement that created the WTO tightened strictures on international trade protection in two ways. The first was through the traditional tariff-cutting exercise. In the WTO system countries commit themselves to keeping product-specific tariffs at or below their “bound” rates. Once “bound,” tariffs cannot be raised, except in certain emergency circumstances under the so-called “safeguard” provisions. As part of the WTO agreement, signatories agreed to lower tariffs and some developing countries increased the number of their bound tariff lines. (In these cases, the previously applied tariffs were “unbound” and hence could be raised unilaterally without violating the GATT.) Nevertheless, many of these tariffs were bound at levels well above the actual applied rates. So while the Uruguay Round agreement increased the discipline to which developing countries were subject in tariff setting, in actuality plenty of room for maneuver remains if countries seek to increase protection for the purpose of industrial promotion. Possibly more important from the standpoint of selective intervention was the Uruguay Round ban on voluntary export restraints (VERs). While this was justly applauded as a prohibition on a nontransparent and discriminatory “gray area” instrument, VERs had been used by Korea and Taiwan toward Japan as an industrial promotion measure, most notably in the case of Korea’s “import diversification program,” which banned many Japanese products (such as automobiles) from the Korean market. All in all, though, the increased restriction on trade protection embodied in the Uruguay Round agreement would not appear to have been a major blow to countries’ ability to implement industrial policy.

88

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 89

Perhaps more important than its circumscribing the scope for trade protection was the agreement’s expansion of disciplines on subsidies. Under the WTO, subsidies—which can include the sort of indirect support through financial institutions and fiscal incentives that Japan, Korea, and Taiwan practiced—are classified into one of three categories. Under the GATT system, export subsidies were prohibited. The WTO agreement extends this prohibition to include production subsidies designed to displace imported intermediate inputs. A second set of subsidies is considered “actionable” if they demonstrably distort trade, and the agreement sets out a specific set of criteria to establish a presumption that the subsidies “seriously prejudice” the interests of another signatory.8 Countries with per capita incomes below $1,000 are exempted from these restrictions. Once they cross that threshold, developing countries are given eight years to comply with the rules.9 A third category of subsidies is explicitly permitted under the WTO agreement. These include subsidies for research and precompetitive development, regional development, and the retrofitting of existing facilities to meet environmental requirements. The agreement permits subsidies of up to 75 percent for certain costs of industrial research and up to 50 percent of precompetitive development activity. In addition, unlimited subsidy is permitted of noncommercially linked research conducted in universities or national laboratories or institutes, and subsidy of research in the civil aviation sector is exempt from any restraints. The other permitted class of subsidies is for the development of depressed regions, providing that the subsidies are general and the targeted territory meets the income and unemployment criteria for a depressed region specified in the agreement. Taken together the exemptions for research and development and regional development would appear to create fairly large loopholes for a government committed to selective intervention. As one comment noted, the definitions embodied in the subsidies rules are so broad that “governments will find it all too easy to justify subsidies” (Whalley and Hamilton 1996, 49). Thus the WTO rules do not appear to foreclose the practice of industrial policy. What is true, however, is that in the post-Cold War world, devel8. For an excellent overview of the Uruguay Round agreement see Schott (1994). Benitah (2001) contains a detailed technical analysis of the subsidy rules. 9. In addition to this phaseout period, developing-country exporters are granted more lenient standards in countervailing duty (CVD) cases: the subsidy must be 2 percent of the unit value of exports (rather than the developed-country standard of 1 percent) to be actionable, and a de minimis provision excludes exports from developing countries with less than 4 percent market share from the CVD rules, unless such developing countries in total account for more than 9 percent of the imports in question.

REPLICABILITY

89

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 90

oped countries are more likely to pursue their narrow economic interests in a less diplomatically constrained way, and the WTO’s improved dispute settlement mechanism gives them a more effective instrument to do so. Thus, for example, in 1996 the United States and Japan together successfully challenged certain Indonesian policies undertaken to promote its automobile industry. In 1997 Japan threatened to lodge a complaint against Korea with the WTO over its aforementioned “import diversification program,” which was clearly inconsistent with the WTO agreement; the policy’s termination in 1999 as part of Korea’s IMF program relieved Japan of the necessity of doing so. Similarly, in 2002 the United States has threatened to take Korea to the WTO over its indirect financial subsidies to Hynix, a manufacturer of semiconductors, through the Korea Development Bank. Whether it does so may depend on whether Hynix merges with a US company, as it had planned to do in an earlier agreement with US-based Micron Technology that failed. A related issue is whether World Bank and IMF policies effectively discourage countries from pursuing the sorts of industrial policies that are permitted under WTO rules. This argument is sometimes made on the basis of the alleged ideological predilections of the Bretton Woods institutions (see, e.g., Rodrik 1999) and sometimes attributed to their status as tools of self-interested developed-world governments (e.g., Khor 1996). Here one must differentiate poor countries with weak economic policies that lead to repeated borrowing from the international financial institutions on the one hand, and middle-income countries that are less dependent on them insofar as they have access to private sources of capital. With respect to the latter, it is doubtful that the practices of the Bank and the Fund impose significant constraints on countries’ ability to pursue selective intervention. First of all, to have any impact, they have to have lending operations in place, and few of the relevant middle-income countries have such programs; Malaysia, for example, is currently the recipient of neither Bank nor Fund lending and is free to construct its hightechnology media corridor whether doing so is sensible or not. Second, even in countries where the Bank and Fund lend, the bulk of the evidence suggests that conditionality has only a marginal impact on policy (Goldstein 2000). The degree of policy independence derives from institutional capacity to formulate policy independently of advice from international financial institutions, which in turn is presumably largely a function of a country’s size and its per capita income. Third, today the multilateral development banks are sufficiently incoherent ideologically that it is hard to imagine the banks’ digging in their heels against industrial promotion policies short of policies that are blatantly unsound. The IMF is more coherent ideologically and, in Nobel laureate Joseph Stiglitz’s pejorative, more “market fundamentalist,” but after its foray into micromanagement during the Asian financial crisis, the trend has been away from micro-

90

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 91

economic policy conditionality in Fund programs (Stiglitz 2002). The ideologically contested space is sufficiently large that “reasonable” industrial promotion policies undertaken by middle-income countries are clearly not beyond the pale. The situation is more ambiguous with respect to small, poorer countries that are more reliant on the international financial institutions for financing and have less institutional capacity for independent policy formation. For better or worse, they are more dependent on advice coming from the staffs of the Bretton Woods institutions. For example, policymakers in at least one developing country were told erroneously by their advisers at these agencies that the WTO agreement banned export subsidies, despite the fact that the country was well below the $1,000 threshold, and that the country’s ongoing programs would be imperiled by the modest industrial promotion policies that it was considering. The shift toward grants instead of loans and direct poverty alleviation through nongovernmental service providers, if it eventuates, will attenuate the Bank-Fund policy influence in these countries. The real issue is whether the current system imposes significant constraints on the sort of monitoring disciplines that we have argued were an important component in the application of industrial policy (in Korea especially) and helped prevent such intervention in all three countries from devolving into the open-ended and ultimately unproductive support that it did elsewhere. In this regard, the use of export targets was important for two reasons. First, success in the world market was a clear and neutral standard against which the performance of domestic firms could be measured. The world market standard had the added attraction of being less subject to manipulation than domestic performance criteria might have been. Second, the emphasis on exports, whatever positive externalities they may or may not have conveyed, helped finance imports, which clearly spurred increases in domestic productivity. The WTO provides an enforcement mechanism against export subsidies that did not exist in the earlier period. But no one argues that export subsidies (largely used to offset protection on imported inputs) per se were key, as they could have been subverted. Thus what was important was the ability to monitor exports to ensure performance, preventing firms from resting on their privileged rent-earning positions. The WTO subsidy rules distinguish between subsidies that are generally available to all firms—such as a general accelerated depreciation tax provision— and those that single out particular enterprises, the former being permissible and the latter not, according to a “specificity test.” It is doubtful that selective incentives, even if provided neutrally with respect to the destination of sales but triggered by export performance, would pass the specificity test embodied in the WTO rules. While the loopholes in the subsidies agreement relating to research and development and regional aid

REPLICABILITY

91

04--Chapter 4--77-92

3/14/03

10:15 AM

Page 92

are large, the rules would appear to make it difficult to pursue the kind of export-performance-conditional and firm-specific subsidy policies pursued by, say, Korea in the 1960s.10 It might be possible to devise conditional support policies based on a nontrade criterion, such as achieving specified cost reductions, that would pass WTO muster, however. In summary, the international environment is modestly less amenable to the pursuit of selective industrial policies than it was a generation ago. Yet, it is a country’s internal characteristics, not imposed externally constraints, that are likely to be the decisive determinants of its degree of success in the use of such policies—though again we are not optimistic about the likelihood of success. The initial conditions embodied in most contemporary developing countries are more likely to generate conflict over the allocation of policy-generated rents and “growth without development” than the benign “growth with equity” experienced by our three countries a generation earlier. Whether the governments of prospective followers of selective strategies can maintain the political stability and bureaucratic competence achieved by the three Asian exemplars during their developmental phase is a critical question, and can only be answered on a case-by-case basis. Even if they can, the import of the studies reviewed in chapter 2 as well as of the second-round effects surveyed in chapter 3 suggests that net benefits will not be easy to extract.

10. Of course, a government acting in bad faith could pursue such a policy through the guise of an R&D subsidy, for example, with the hope that none of its trade partners would challenge it, or if they did, that nontransparent implementation would frustrate the complainant’s establishment of the burden of proof. 92

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 93

5 Conclusions

In this volume we have addressed the experiences of Japan, Korea, and Taiwan with selective intervention or industrial policies. We believe that, on balance, the weight of the evidence derived from both econometric and input-output studies of these economies that we presented in the preceding chapters indicates that industrial policy made a minor contribution to growth in Asia. Skepticism is often voiced about such empirics. One question is whether the sectoral total factor productivity (TFP) growth rates, which are key to demonstrating that industrial policy boosted income and welfare, are reliable measures. We have noted the problems with their calculation (appendix 2.1), particularly the possibility that these calculations may understate TFP growth rates in countries with rapid growth in capital-labor ratios if the elasticity of substitution is below 1 or technical progress is not Hicks neutral but labor augmenting. However, as noted in the discussion of Korea in chapter 2, even labor productivity growth was not positively associated with the extent of promotion. Furthermore, in addition to the TFP-based studies, we have introduced a much broader array of evidence that runs counter to the view that industrial policy was a major source of growth. For example, in our discussion of Japan we examined sectorally differentiated subsidies and noted that their small size and the fact that so much was directed to agriculture suggest that their sectoral impact in industry was likely to have been small. Another widely cited preferential activity, Japan’s R&D subsidies, was also small. Unless the relevant elasticities of the R&D decision with respect to subsidies were implausibly large, the amounts involved would have been too small to have fundamentally altered private-sector decisions. 93

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 94

Some proponents of industrial policy do not simply voice skepticism about one or more of the measures employed to gauge its impact but rather object to the entire effort to measure its quantitative effect. For example, it is held that the mere commitment of the governments to intervene established a general atmosphere conducive to growth whose beneficial effect is difficult or impossible to measure, particularly for individual sectors. Industrial policy may have led to an increase in “animal spirits” or the willingness to take risks and hence led to riskier investment with higher returns, which in turn encouraged saving. Rapid growth of physical capital could thus have been the result of industrial policy applied to selected sectors, and the rapid accumulation of human capital as well may have been a rational response to growing opportunities generated by the atmosphere of confidence and by increasing amounts of complementary physical capital. Such views could of course be correct; as noted in chapter 1, the high savings and investment rates have yet to be completely understood. There are many possible explanations, and as of now there is no generally accepted explanation of the extraordinary savings rates in the three countries, but it suffices to say that there is no convincing case linking industrial policies—as opposed to demographics, bequest motives, and a host of other hypothesized explanations—to the observed behavior (Deaton and Paxson 1994). The experience in other countries calls into question whether either the high investment or savings rate is plausibly attributable to sectorally targeted policies as opposed to favorable macroeconomic policies. Many countries have attempted systematic import-substituting industrialization to foster growth. They, too, discriminated among sectors and in general pursued many policies that were very similar to those employed in Japan, Korea, and Taiwan. The major difference was the greater emphasis on exports in the Asian countries. Yet the types of protection afforded under import-replacing policies, particularly quantitative restrictions, reduced risk even more than did the industrial policy in the Asian countries, as it precluded any imports above the quotas: there was no possibility that sufficiently efficient foreign competitors could still compete with local firms despite the imposition of tariffs. Yet the import-substituting countries never approached the 30 percent or greater investment rate in Asia or experienced the rapid growth rates of capital shown in tables 4.3 and 4.4. Nor did overall higher education enrollments or those in science and engineering respond in the manner that they did in Japan, Korea, and Taiwan, perhaps because the growth of employment opportunities was much lower, as output was largely sold in slowly growing domestic markets.1 1. Both Korea and Taiwan experienced considerable emigration of university graduates in the 1950s and 1960s. Only with the rapid growth of export-oriented industries and concomitant job opportunities did this stop and eventually reverse. 94

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 95

The phrase “Asian Miracle” in the World Bank’s volume (1993) was intended ironically, as most of the volume was devoted to cataloguing the readily explicable sources of growth. If there was any “miracle,” it resided in the incompletely explained rates of accumulation. Those hewing to a strongly positive view of industrial policy would then have to explain why no such miracle occurred in India or Latin America. In contrast to those who have argued that the high investment rates were due to industrial policy, Bhagwati (1999) argues that they were due to the public investment in infrastructure in the 1950s and 1960s that made future private investment profitable.2 Once in place, this infrastructure facilitated growth, including that in export earnings, which were critical to the financing of imported equipment whose marginal social product exceeded its cost in international prices. Although it is impossible to be certain that “atmospherics” have not been important and that neglecting their effects results in an underestimate of the impact of selective intervention, the fact that acceleration in investment in physical and human capital were not observed in other interventionist countries requires some suspension of the usual canons of evidence unless only the East Asian brand of industrial policy is conducive to such accumulation. Thus South Asian and Latin American countries that pursued interventionist policies had less than half the growth rate of education per worker in their periods of intensive import substitution (table 4.3). Moreover, these rates were lower even in the 1960– 73 period, when they were moving into more capital- and technologyintensive sectors. In this period Korea and Taiwan were still not committed to a high-tech route, and students could not have been sure of the payoffs to their newly acquired skills. Indeed, in the 1960s the high rates of tertiary enrollment in Korea and Taiwan were accompanied by significant brain drain, hardly suggestive of a policy-led shift in the demand for educated labor. Another doubt often articulated, discussed in chapter 2, is that intervention generated significant externalities whose impact is not captured by the tests reported earlier that concentrate on individual sectors within manufacturing. This may indeed have been an achievement of industrial policy in the Asian countries, but the sector-specific evidence, admittedly for a later period in the case of Japan, such as the absence of correlation between incentives and the TFP growth rate, places the entire burden of demonstrating the benefits of policy on the existence of externalities (in the absence of an economywide impact on capital accumulation—which has not been substantiated). One cannot of course reject the possibility that externalities were generated. But given the evidence from detailed sectoral analysis, the argument that externalities had significant beneficial 2. In this he differs from Rodrik (1995), who ascribes the investment boom to a response to investment subsidies in the 1960s and 1970s. CONCLUSIONS

95

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 96

effects on growth would require that any externalities offset the negative intersectoral effects of industrial policies, such as loans going to sectors with low TFP growth. The argument can be made that the entire manufacturing sector, and perhaps the entire economy, is the beneficiary of widely diffused external economies regardless of what occurs in individual sectors, a view espoused by early postwar advocates of industrialization. Such a flow of externalities should be reflected in manufacturingwide or aggregate TFP measures. However, some of the TFP growth that did occur in individual sectors was explicable by conventional explanations that were not related to externalities, such as learning-by-doing at the firm level and the importation of foreign technology that firms undertake without government prodding.3 As noted in appendix 2.2, rapid export growth may have accelerated both of these phenomena, but such growth could have been generated by general rather than selective export promotion. If extensive government intervention generates benefits external to the promoted sectors,4 then it should show up in high rates of TFP growth for the entire manufacturing sector—and perhaps for the entire economy, assuming that manufacturing confers externalities on the agriculture, construction, and service sectors. However, such gains should have been manifested in all economies that undertook industrial promotion, not simply the Asian countries. Yet the evidence shows that in India, for example, almost every one of the two-digit manufacturing sectors experienced negative TFP growth in the period from 1960 to 1980 (Ahluwalia 1985), suggesting that there was no significant effect from cross-sector externalities, and the economywide TFP growth rate was close to zero.5 Many Latin American countries experienced similarly dismal performance, though not in the 1960s. In the cases of Korea and Taiwan, nationwide TFP growth was not usually absolutely high in any period, being only slightly above the OECD values for most of the subperiods shown in table 4.3.6 And surprisingly, in the period 1960–73, when it might be argued that in Korea and Taiwan industrial policy may have had its major benefit, their TFP growth rates were at or below those in the OECD countries. 3. On the role of foreign technology in Taiwan’s development, see Pack (2001b). More generally, on the interest in obtaining new technology, see Hobday (1995) and Kim (1997). 4. The evidence presented in chapter 2 on input-output linkages indicates that such external benefits would have to stem not from intersectoral purchases of inputs, and most likely not from movement of workers, but from general “atmospherics.” 5. Of course intrasector performance could have been still worse, and externalities may have raised TFP growth rates from –2.0 to –1.5, but this is presumably not what advocates of industrial policy have in mind. 6. Using comparable methods for each period to avoid any biases inherent in the calculations, Collins and Bosworth (1996) calculated the TFP growth rates and output per worker growth rates for Korea and Taiwan for several periods. 96

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 97

However, it is possible that these calculations of aggregate TFP growth across countries understate its true rate (Nelson and Pack 1999), and thus measures of TFP levels may be useful.7 In 1975, at the end of the initial spurt of industrialization, the average TFP level in Korean manufacturing relative to that in the United States was 0.18 (Pilat 1994). Can this evidence be reconciled with the conjecture that in the earliest stage of development, say, 1960 to the early 1970s, industrial policy helped initiate Korean or Taiwanese growth (Pack and Westphal 1986)? Growth in output per worker is a function of the growth rate of TFP and of the growth rate of the ratio of capital to labor. As shown in table 4.3, in the critical period 1960–73, TFP growth rates accounted for a substantially lower share of growth in output per worker than did growth in capital per worker. The TFP growth rates were below those in the OECD countries despite the very large absolute gap with respect to this group and the possibilities for narrowing it.8 An interpretation favorable to the impact of industrial policy is that it was successful insofar as the investment ratio and TFP growth rates did not fall in the face of extraordinary rates of capital accumulation (see tables 4.3 and 4.4). The capital stock growth rates were double or triple those of other countries, even Germany, yet TFP growth remained positive. The metric of achievement is then the ability to stave off rapidly diminishing returns to capital that would have brought the accumulation process to a halt. Especially in the earlier years, through the mid-1970s, the various governments were successful in avoiding major unprofitable investments and in avoiding the scale of corruption that undermined the productivity of investment in other developing countries. Moreover, the allocation of capital that occurred could not have been particularly misguided, as even the most pessimistic assessment of TFP growth finds it to have been positive. In Korea and Taiwan in the 1960s and early 1970s, TFP growth was impressive especially given that they were still very poor countries, recovering from enormous political and economic dislocation in the preceding decades. From this perspective, the success of selective intervention can be measured as the positive TFP growth rates achieved by initially poor countries that were absorbing extraordinary increases in productive factors. The productive absorption of enormous increases in capital presumably could have been achieved by generally responsible macroeconomic and banking regulation without a detailed industrial policy. But if one takes this view, the role of industrial 7. In terms of equations (3) and (4) in appendix 2.1, either m or  could vary among countries, although reasons for such intercountry variation are not immediately obvious. 8. All of these calculations are subject to the caveats noted in appendix 2.1 about the underlying elasticity of substitution and the bias of technical change. However, there is no reason to believe that the values of these variables differed systematically across countries, thus the ranking of TFP across countries is unlikely to be affected though the absolute levels may not be correctly calculated. CONCLUSIONS

97

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 98

policy, in part, would be that it substituted for the absence, at this early juncture, of the regulatory institutions that come with greater economic development.9 Similar arguments might be made for Japan for the period between 1945 and the Korean War boom. There are, however, many equally plausible alternative explanations that could account for the avoidance of diminishing returns to capital. Nelson and Phelps (1966) and Schultz (1975) argue that education has a high payoff only if there is an inflow of new technology. The typical example would be that greater education would not generate much increase in productivity for a sewing machine operative using a simple sewing machine. In contrast, if workers and managers are called upon to utilize new machine tools or new intermediate goods, such as new metal alloys, those with more education are likely to learn to utilize these inputs more effectively. This view of the determinants of productivity growth suggest an important potential interaction that almost certainly took place in Japan, Korea, and Taiwan. The growth in capital spending financed considerable growth in imported capital goods and intermediate products. These imports had an immediate effect insofar as they embodied new technology (Pack 1992b, 2001b).10 They also led to a higher marginal productivity of recent investment in education. Thus the interaction of education and new inputs is likely to have been a critical component in forestalling a decline in capital’s marginal product. Not only was education growing more rapidly than in South Asia, Latin America, and other large developing countries (see figure 2.3 and table 4.3), but the education also had a much greater productivity because of the inflow of new technology, a phenomenon that could not occur in the import-substituting countries of the other regions given their inward orientation. The increasing education levels were not a result of industrial policy: as can be seen in table 2.1, Japanese and Korean human capital in the mid-1950s was already high relative to their per capita income. Another characteristic that contributed to the high marginal product of capital, especially in Korea and Taiwan, was the considerable flexibility of labor and the efforts of firms to improve their productivity once the easiest parts of labor-intensive growth had occurred (Fields 1984). Labor flexibility implied that workers moved without impediment to expanding areas rather than try to maintain their positions within sectors that were coming under increasing competitive pressure. In Taiwan, for example, workers moved to sectors in which the skills obtained in their previous 9. Stiglitz (1994) has argued that it was precisely the absence of such institutions that undercut many of the market reforms introduced in postcommunist societies in Eastern Europe. 10. Bhagwati (2000) also emphasizes the importance of the inflow of new producers’ goods and its effective use by the highly educated labor force. The productivity-raising effects of a greater range of inputs has been stressed in the endogenous growth literature, for example, Grossman and Helpman (1991) and Rivera-Batiz and Romer (1991). 98

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 99

industry of employment were relevant as measured by the input-output coefficient between the sector of previous employment and that of new employment (Pack and Paxson 1999). Such fluidity would improve the productivity of capital invested in these sectors, as the complementary labor was of a higher quality. The absence in much of the period of either union- or government-imposed rules was critical to this pattern. An additional explanation of the difference in performance points to the more competitive atmosphere that resulted from the reliance on exports as a measure of performance or from the practice of competitive “tournaments” in Japan (World Bank 1993).11 In addition, a critical difference was the relative openness of the Asian countries to disembodied technology imports obtained through technology licenses, whereas the Latin American countries and India were both restrictive, fearing monopolistic pricing and excessive foreign exchange costs. And, again, given the highly educated labor force, this inflow of technology was used effectively. While it is well documented that the Japanese government achieved better terms for firms that imported technology, the absolute cost savings relative to the enormous productivity of such technology was necessarily tiny.12 It is possible that for the initial phases of industrialization in Korea and Taiwan, government efforts had a significant role, as suggested by Pack and Westphal (1986) as well as by other authors. Unfortunately, it has not been possible to confirm this interpretation empirically except in the cases of a few specific sectors for the period in question. As we have shown, the evidence on the benefits of industrial policy for later periods has been generally negative. These empirical findings do not preclude the possibility that for some firms and subsectors government intervention had positive effects. POSCO, a government-owned steel plant initiated by the Korean government because of imperfections in the capital market and now among the most efficient in the world, is often cited. On the other hand, even this view is now in question; for example, Little (1994) calculates that its rate of return on equity capital has been a paltry 2 percent. Today, from the perspective of four decades of growth, it is very difficult to support the view that industrial policy was the decisive source of growth after the initial spurt—and the initial spurt itself may simply have reflected the exploitation of comparative advantage in the labor-intensive sectors of newly liberalized economies. Industrial policy was merely a secondary boost to this evolution. Suppose a policymaker is still not convinced by the evidence we have marshaled to suggest that industrial policy had a limited effect on growth and maintains that something that cannot be measured conventionally oc11. The monitoring role provided by exports was first articulated by Jones and SaKong (1980). 12. For figures and a description of Japanese technology licensing activities, see Nagaoka (1989). CONCLUSIONS

99

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 100

curred in the three nations that was attributable to its industrial policy. Would this constitute a justification for a very poor nation to embark on such a program? Obviously we would recommend caution. First, the policies deployed were exceptionally complex and were implemented in the Asian countries under conditions of political stability by highly competent bureaucracies. Attempts to emulate their experiences in the absence of these political precursors could easily result in counterproductive interventions and corruption. Second, a proper evaluation must include the second-round effects that may have manifested themselves, at least partly, in the financial crisis, particularly in Korea in the late 1990s and in Japan since 1990. Third, the external environment is modestly less hospitable than it was a generation ago. The WTO’s tightened rules and the greater willingness of developed countries to use its strengthened dispute settlement process in the post-Cold War world make it somewhat more difficult to employ some of the instruments used by the three early starters. Fourth, there were alternatives to the government’s “picking winners,” as the experiences of both Hong Kong and Singapore show.

Final Thoughts A large part of the “Asian Miracle” was attributable to nonmiraculous good macroeconomic policy, including limited government deficits, low rates of inflation, and very stable real exchange rates. These were conducive to high rates of saving and investment, important components of the growth story. Another component was the Asian countries’ slight bias toward exporting, which may have generated some benefits that would not have accrued from domestic sales (Pack 1997). The path of Japan, Korea, and Taiwan is more likely to generate “growth with equity” as capital is accumulated, and less likely to run into problems with allocating natural-resource-derived rents. The politics of industrial policy are likely to be less contentious, and since they are implemented in the manufacturing sector, they are more likely to be “leaning with the wind” of comparative advantage. In any event, the strategy may be irreproducible: some of the subsidies carried out by the Asians in the past can no longer be pursued as openly. The end of the Cold War and the concomitant willingness of the United States and other major trading powers to assert their economic interests, together with the existence of a stronger subsidies code and dispute settlement in the WTO, foreclose some, though not all, of the options that existed in the past. Countries that have experienced slower growth than expected despite relatively good macroeconomic policies may be tempted to pursue industrial policies. The large number of experiments with import-substituting industrialization suggests this strategy has not been very successful. The Asian experience, especially in Korea and Taiwan, provides some guide100

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 101

lines to avoiding some of the potential harmful consequences if industrial policy is nevertheless pursued. Yet even in these successful nations, the evidence suggests that the benefits were limited. Countries with less dedicated and less competent bureaucracies and policymaking apparatuses that are more amenable to lobbying pressures could expect even smaller net benefits. While it is understandable that countries that have gotten the basics right are impatient that growth has not accelerated, it is nevertheless particularly problematic to identify broad sectors of growth, let alone specific ones. The difficulty of demonstrating that the major source of either manufacturing or aggregate economic growth was sectorally targeted industrial policies is not equivalent to denying the importance of a significant government role other than macroeconomic management in stimulating economic growth. Growth-enhancing measures in the Asian countries, which did not differentiate among sectors, included large expenditures on primary and secondary education, the building of large and efficient social infrastructure, a favorable attitude toward international technology transfer, including both technology licensing and direct foreign investment, and a substantial investment in public technology institutions. Governments seeking a more active role in accelerating growth should consider these policies rather than selective industrial policies.

CONCLUSIONS

101

05--Chapter 5--93-102

3/14/03

10:16 AM

Page 102

06--References--103-110

3/14/03

10:16 AM

Page 103

References

Ahluwalia, Isher Judge. 1985. Industrial Growth in India. Delhi: Oxford University Press. Amsden, Alice H. 1989. Asia’s Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. Audretsch, David B., and Hideki Yamawaki. 1988. R&D Rivalry, Industrial Policy, and U.S.Japan Trade. Review of Economics and Statistics 70, no. 3: 438–47. Auty, Richard M. 1991. Creating Comparative Advantage: South Korean Steel and Petrochemicals. Tijdschrift voor Economie en Social Geografie 82, no. 1: 15–29. Bae, Jin-Young. 2001. Incentive Structure and Its Changes in the Korean Industrial Policy Regimes from 1962–1997. Journal of the Korean Economy 2, no. 2: 297–331. Balassa, Bela, et al. 1982. Development Strategies in Semi-industrial Economies. Baltimore: Johns Hopkins University Press. Balassa, Bela. 1990. Economic Policies in the Pacific Area Developing Countries. London: Macmillan. Balassa, Bela, and Marcus Noland. 1988. Japan in the World Economy. Washington: Institute for International Economics. Balassa, Bela, and Marcus Noland. 1989. The Changing Comparative Advantage of Japan and the United States. Journal of the Japanese and International Economies 3, no. 2: 174–88. Baldwin, Richard E., and Paul R. Krugman. 1988. Market Access and International Competition: A Simulation Study of 16K Random Access Memories. In Empirical Methods for International Trade, ed. Robert Feenstra. Cambridge, MA: MIT Press. Baldwin, Richard E., and Frederic Robert-Nicoud. 2002. Entry and Asymmetric Lobbying: Why Governments Pick Losers. NBER Working Paper 8756. Cambridge, MA: National Bureau of Economic Research. Beason, Richard, and David E. Weinstein. 1996. Growth, Economies of Scale, and Targeting in Japan (1955–1990). Review of Economics and Statistics 78, no. 2: 286–95. Behrman, Jere, and T. N. Srinivasan, eds. 1994. Handbook of Development Economics 3a. Amsterdam: North Holland. Benitah, Marc. 2001. The Law of Subsidies Under the GATT/WTO System. The Hague: Kluwer Law International. Bergsten, C. Fred, and Marcus Noland. 1994. Reconcilable Differences? United States-Japan Economic Conflict. Washington: Institute for International Economics.

103

06--References--103-110

3/14/03

10:16 AM

Page 104

Bergsten, C. Fred, Takatoshi Ito, and Marcus Noland. 2001. No More Bashing: Building a New Japan-United States Economic Relationship. Washington: Institute for International Economics. Bhagwati, Jagdish. 1978. Anatomy and Consequences of Exchange Control Regimes. Cambridge, MA: Ballinger Publishing. Bhagwati, Jagdish. 1999. The “Miracle” That Did Happen: Understanding East Asia in Comparative Perspective. In Taiwan’s Development Experience: Lessons on Roles of Government and Market, ed. Eric Thorbecke and Henry Wan. Boston: Kluwer Academic Publishers. Bhagwati, Jagdish, and V. K. Ramaswami. 1963. Domestic Distortions, Tariffs, and the Theory of Optimum Subsidy. Journal of Political Economy 71, no. 1: 44–50. Borrus, Michael, Laura D’Andrea Tyson, and John Zysman. 1986. Creating Advantage: How Government Policies Shape International Trade in the Semiconductor Industry. In Strategic Trade Policy and the New International Economics, ed. Paul R. Krugman. Cambridge, MA: MIT Press. Branstetter, Lee, and Mariko Sakakibara. 2002. When Do Research Consortia Work Well and Why? Evidence from Japanese Panel Data. American Economic Review 92, no. 1: 143–59. Calder, Kent E. 1993. Strategic Capitalism: Private Business and Public Purpose in Japanese Industrial Finance. Princeton, NJ: Princeton University Press. Campos, E. J., and Hilton Root. 1996. The Key to the Asian Miracle: Making Shared Growth Credible. Washington: Brookings Institution. Chenery, Hollis B. 1959. The Interdependence of Investment Decisions. In The Allocation of Economics Resources, ed. Moses Abramovitz. Stanford: Stanford University Press. Cho, Soon. 1994. The Dynamics of Korean Economic Development. Washington: Institute for International Economics. Chung, Yong-Iob. 2002. Korea Under Seige, 1876–1945: Capital Formation and Economic Transformation. Eastern Michigan University, Ypsilanti, MI. Photocopy. Clark, Cal. 1989. Taiwan’s Development: Implications for Contending Political Economy Paradigms. Westport, CT: Greenwood Press. Clerides, Sofronis, Saul Lach, and James R. Tybout. 1998. Is Learning by Exporting Important? Microdynamic Evidence from Colombia, Mexico, and Morocco. Quarterly Journal of Economics 113, no. 3: 903–47. Coe, David T., and Elhanan Helpman. 1995. International R&D Spillovers. European Economic Review 39, no. 5: 859–87. Coe, David T., Elhanan Helpman, and Alexander Hoffmaister. 1997. North-South R&D Spillovers. Economic Journal 107, no. 440 (Sept): 134–49. Collins, Susan M., and Barry P. Bosworth. 1996. Economic Growth in East Asia: Accumulation Versus Assimilation. Brookings Papers on Economic Activity, 1996, vol. 2, 135–91. Corden, W. Max. 1991. Strategic Trade Policy. In Companion to Contemporary Economic Thought, eds. David Greenaway, Michael Bleaney, and Ian Stewart. London: Routledge. Dahlman, Carl, and Ousa Sananikone. 1997. Taiwan, China: Policies and Institutions for Rapid Growth. In Lessons from East Asia, ed. Danny M. Leipziger. Ann Arbor: University of Michigan Press. Deaton, Angus, and Christina Paxson. 1994. The Effects of Economics and Population Growth on National Saving. Demography 34, no. 1: 97–114. De Melo, Jaime. 1985. Sources of Growth and Structural Change in the Republic of Korea and Taiwan: Some Comparisons. In Export-Oriented Development Strategies, eds. V. Corbo, A. O. Krueger, and F. Ossa. Boulder, CO: Westview Press. Diamond, Peter, Daniel McFadden, and Miguel Rodriguez. 1972. Identification of the Elasticity of Substitution and the Bias of Technical Change. In An Econometric Approach to Production Theory, eds. Daniel McFadden and Peter Diamond. Amsterdam: North Holland. Dixit, Avinash, and Gene M. Grossman. 1986. Targeted Export Subsidies with Several Oligopolistic Industries. Journal of International Economics 21, no. 3/4: 233–50. Fields, Gary. 1984. Employment, Income Distribution, and Economic Growth in Seven Small Economies. Economic Journal 94, no. 373: 74–83. 104

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

06--References--103-110

3/14/03

10:16 AM

Page 105

Flamm, Kenneth. 1996. Mismanaged Trade? Strategic Policy and the Semiconductor Industry. Washington: Brookings Institution. Furman, Jason, and Joseph E. Stiglitz. 1998. Economic Crises: Evidence and Insights from East Asia. Brookings Papers on Economic Activity 30, no. 2: 1–114 Gold, Thomas. 1986. State and Society in the Taiwan Miracle. Armonk, NY: M. E. Sharpe. Goldstein, Morris. 2000. IMF Structural Programs. http://www.iie.com/papers/goldstein1200. pdf Goto, Akira, and Ryuhei Wakasugi. 1988. Technology Policy. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Grossman, Gene M. 1990. Explaining Japan’s Innovation and Growth: A Model of Quality Competition and Dynamic Comparative Advantage. Bank of Japan Monetary and Fiscal Studies 8, no. 2: 75–100. Grossman, Gene M., and Elhanan Helpman. 1991. Innovation and Growth in the World Economy. Cambridge, MA: MIT Press. Hall, Robert E. 1990. Invariance Properties of Solow’s Productivity Residual. In Growth/ Productivity/Unemployment: Essays to Celebrate Bob Solow’s Birthday, ed. Peter Diamond. Cambridge, MA: MIT Press. Haggard, Stephan. 1990. Pathways from the Periphery: The Politics of Growth in the Newly Industrializing Countries. Ithaca, NY: Cornell University Press. Haggard, Stephan, and Tun-jen Cheng. 1987. Newly Industrializing Asia in Transition: Policy Reform and American Response. Policy Papers in International Affairs 31. Berkeley: University of California, Institute of International Studies. Helpman, Elhanan, and Paul R. Krugman. 1989. Trade Policy and Market Structure. Cambridge, MA: MIT Press. Hobday, Mike. 1995. Innovation in East Asia: The Challenge to Japan. London: Edward Elgar. Horioka, Charles Yuji, and Wako Watanabe. 1997. Why Do People Save? A Micro-analysis of Motives for Household Saving in Japan. Economic Journal 107, no. 402: 537–52. Hou, Chi-Ming, and San Gee. 1993. National Systems Supporting Technical Advance in Industry: The Case of Taiwan. In National Innovation Systems: A Comparative Analysis, ed. Richard R. Nelson. New York: Oxford University Press. Inkeles, Alex, and David H. Smith. 1974. Becoming Modern: Individual Change in Six Developing Countries. Cambridge, MA: Harvard University Press. Itoh, Motoshige, and Kazuhara Kiyono. 1987. Welfare-Enhancing Export Subsidies. Journal of Political Economy 95, no. 1: 115–37. Johnson, Chalmers. 1982. MITI and the Japanese Miracle. Stanford: Stanford University Press. Johnson, Chalmers. 1984. The Industrial Policy Debate. San Francisco: Institute for Contemporary Studies. Johnson, Harry G. 1953–54. Optimal Tariffs and Retaliation. Review of Economic Studies 21, no. 55: 152–53. Jones, Leroy, and Il Sakong. 1980. Government, Business, and Entrepreneurship in Economic Development: The Korean Case. Cambridge, MA: Harvard University Press. Kang, David C. 2002a. Bad Loans to Good Friends: Money, Politics and the Developmental State in South Korea. International Organization 56, no. 1: 177–207. Kang, David C. 2002b. Crony Capitalism: Corruption and Development in South Korea and the Philippines. Cambridge, England: Cambridge University Press. Kang, Moonsung. 2000a. Patent Protection and Strategic Trade Policy. KIEP Working Paper 00-07. Seoul: Korea Institute for International Economic Policy. Kang, Moonsung. 2000b. Patent Infringement and Strategic Trade Policies: R&D and Export Subsidies. KIEP Working Paper 00-09. Seoul: Korea Institute for International Economic Policy. Khor, Martin. 1996. Globalisation: Implications for Development Policy. http://www.twnside. org.sg/title/eadi-cn.htm Kim, Ji Hong. 1990. Korean Industrial Policy in the 1970s: The Heavy and Chemical Industry Drive. KDI Working Paper 9015. Seoul: Korea Development Institute. REFERENCES

105

06--References--103-110

3/14/03

10:16 AM

Page 106

Kim, Jong-il, and Lawrence J. Lau. 1994. The Sources of Economic Growth of the East Asian Newly Industrialized Countries. Journal of the Japanese and International Economies 8, no. 3 (September): 235–71. Kim, Linsu. 1997. From Imitation to Innovation: Dynamics of Korea’s Technological Learning. Boston: Harvard Business School Press. Kim, Sun G., and Wankeun Oh. 1999. The Relations Between Government R&D and Private R&D Expenditure in the APEC Economies: A Time-Series Analysis. KIEP Working Paper 99-22. Seoul: Korea Institute for International Economic Policy. Komiya, Ryutaro, Masahiro Okuno, and Kotaro Suzumura. 1984. Nihon no Sangyo Seisaku. Tokyo: Diamond Press. English-language version published in 1988 as Industrial Policy of Japan, San Diego: Academic Press. Krueger, Anne O. 1978. Liberalization Attempts and Consequences. Cambridge, MA: Ballinger Publishing. Kuznets, Paul. 1977. Economic Growth and Structure in the Republic of Korea. New Haven: Yale University Press. Kwack, Sung Yeung. 2000. Total Factor Productivity Growth and the Source of Growth in Korean Manufacturing Industries, 1971–1993. Journal of the Korean Economy 1, no. 2: 229–65. Kwon, Jene K., and Hoon Paik. 1995. Factor Price Distortions, Resource Allocation, and Growth: A Computable General Equilibrium Analysis. Review of Economics and Statistics 77, no. 4: 664–76. Lawrence, Robert Z., and David E. Weinstein. 2001. Trade and Growth: Import Led or Export Led? Evidence from Japan and Korea. In Rethinking the East Asian Miracle, ed. Joseph E. Stiglitz and Shahid Yusuf. Oxford, England: Oxford University Press. Leamer, Edward E. 1987. Paths of Development in the Three-Factor, n-Good General Equilibrium Model. Journal of Political Economy 95, no. 5: 961–99. Lee, Hiro. 1993. General Equilibrium Evaluation of Industrial Policy in Japan. Journal of Asian Economics 4, no. 1: 25–40. Lee, Jong-wha. 1996. Government Interventions and Productivity Growth in Korean Manufacturing Industries. Journal of Economic Growth 1, no. 3: 391–414. Lee, T. H., and Kuo-shu Liang. 1982. Taiwan. In Development Strategies in Semi-industrial Economies, ed. Bela Balassa et al. Baltimore: Johns Hopkins University Press. Leff, Nathaniel H. 1979. Entrepreneurship and Economic Development: The Problem Revisited. Journal of Economic Literature 17, no. 1: 46–64. Lincoln, Edward. 1999. Troubled times: US-Japan trade relations in the 1990s. Washington: Brookings Institution Press. Little, Ian M. D. 1979. An Economic Reconnaissance. In Economic Growth and Structural Change in Taiwan, ed. Walter Galenson. Ithaca, NY: Cornell University Press. Little, Ian M. D. 1981. The Experience and Causes of Rapid Labour-Intensive Development in Korea, Taiwan Province, Hong Kong, Singapore and the Possibilities of Emulation. In Export-Led Industrialization and Development, ed. E. Lee. Geneva: International Labor Organization. Little, Ian M. D. 1994. Trade and Industrialization Revisited. Pakistan Development Review 33 (Winter). Little, Ian M. D., Tibor Scitovsky, and Maurice Scott. 1970. Industry and Trade in Some Developing Countries. New York: Oxford University Press. Magaziner, Ira, and Robert Reich. 1982. Minding America’s Business: The Decline and Rise of the American Economy. New York: Harcourt Brace Jovanovich. Mason, Edward S., Mahn Je Kim, Dwight H. Perkins, Kwang Suk Kim, and David C. Cole. 1980. The Economic and Social Modernization of the Republic of Korea. Cambridge, MA: Harvard University Press. McClelland, David C. 1961. The Achieving Society. New York: Free Press. McKinsey Global Institute. 1998. Productivity-Led Growth in Korea. Washington and Seoul: McKinsey & Co. 106

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

06--References--103-110

3/14/03

10:16 AM

Page 107

Morawetz, David. 1980. Twenty-five Years of Development. Baltimore: Johns Hopkins University Press. Morawetz, David. 1982. Why the Emperor’s New Clothes Are Not Made in Colombia. Baltimore: Johns Hopkins University Press. Morishima, Michio. 1982. Why Has Japan Succeeded? Western Technology and the Japanese Ethos. Cambridge, England: Cambridge University Press. Murphy, Kevin M., Andrei Shleifer, and Robert W. Vishny. 1989. Industrialization and the Big Push. Journal of Political Economy 97, vol. 5: 1003–26. Mutoh, Hiromichi. 1988. The Automotive Industry. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Nagaoka, Sadao. 1989. Overview of Japanese Industrial Technology Development. Industry Series Paper 6. Washington: World Bank, Industry and Energy Department. Nariai, Osamu. 2002. The Modern Japanese Economy. Tokyo: Foreign Press Center/Japan. Nehru, Vikram, and Ashok Dhareshwar. 1993. A New Database on Physical Capital Stock: Sources, Methodology, and Results. Rivista de Analisis Economico 8, no. 1: 37–59. Nehru, Vikram, and Ashok Dhareshwar. 1994. New Estimates of Total Factor Productivity Growth for Developing and Industrial Countries. Policy Research Working Paper 1313. Washington: World Bank. Nelson, Richard R., and Edmund Phelps. 1966. Investment in Humans, Technological Diffusion, and Economic Growth. American Economic Review 56 (May): 69–75. Nelson, Richard R., and Howard Pack. 1999. The Asian Growth Miracle and Modern Growth Theory. Economic Journal 109, no. 457: 416–36. Noguchi, Yukio. 1995. 1940 Nen-taisei [The 1940 System]. Tokyo: Toyo Keizai Shimposha. (In Japanese.) Noland, Marcus. 1993a. Industrial Policy and Japan’s Pattern of Trade. Review of Economics and Statistics 75, no. 2: 241–48. Noland, Marcus. 1993b. Selective Intervention and Growth: The Case of Korea. Washington: Institute for International Economics. Photocopy. Noland, Marcus. 1996. Research and Development Activities and Trade Specialization in Japan. Journal of the Japanese and International Economies 10, no. 2: 150–68. Noland, Marcus. 1997a. Public Policy, Private Preferences, and the Japan Trade Pattern. Review of Economics and Statistics 79, no. 2: 259–66. Noland, Marcus. 1997b. Has Asian Export Behavior Been Unique? Journal of International Economics 43: 79–101. Noland, Marcus. 2000a. Avoiding the Apocalypse: The Future of the Two Koreas. Washington: Institute for International Economics. Noland, Marcus. 2000b. The Philippines in the Asian Crisis: How the Sick Man Avoided Pneumonia. Asian Survey 40, no. 3: 401–12. Noland, Marcus. 2001. Economic Reform in Korea: Achievements and Future Prospects. http://www.iie.com/papers/noland0201-3.htm Ogura, Seiritsu, and Naoyuki Yoshino. 1988. The Tax System and the Fiscal Investment and Loan Program. In Industrial Policy of Japan, ed. Ryotaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Okazaki, Tetsuji. 1993. Corporate Systems. In Source of Current Japanese Economic Systems, ed. Tetsuji Okazaki and Masahiro Okuno-Fujiwara. Tokyo: Nihon Keizai Shimbun Press. (In Japanese.) Okazaki, Tetsuji. 2001. The Government-Firm Relationship in Post-war Japan: The Success and Failure of Bureau Pluralism. In Rethinking the East Asian Miracle, ed. Joseph E. Stiglitz and Shahid Yusuf. Oxford, England: Oxford University Press. Okita, Saburo. 1981. Tohon Seiso—Watakushi no Rirekisho. Tokyo: Nihon Keizai Shimbunsha. Okita, Saburo. 1985. Japan’s Challenging Years. Sydney: George Allen & Unwin. Okuno-Fujiwara, Masahiro. 1988. Interdependence of Industries, Coordination Failure, and Strategic Promotion of an Industry. Journal of International Economics 25, no. 1/2: 25–43. REFERENCES

107

06--References--103-110

3/14/03

10:16 AM

Page 108

Organization for Economic Cooperation and Development. 1998. Economic Surveys: Korea. Paris: Organization for Economic Cooperation and Development. Pack, Howard. 1992a. New Perspectives on Industrial Growth in Taiwan. In Taiwan: From Developing to Mature Economy, ed. Gustav Ranis. Boulder, CO: Westview Press. Pack, Howard. 1992b. Technology Gaps Between Industrial and Developing Countries: Are There Dividends for Latecomers? In Proceedings of the World Bank Annual Conference on Development Economics. Washington: World Bank. Pack, Howard. 1997. The Role of Exports in Asian Economic Development and Lessons for Latin America. In Pathways to Growth, ed. Nancy Birdsall and Fred Jaspersen. Baltimore: Johns Hopkins University Press. Pack, Howard. 2000. Industrial Policy: Growth Elixir or Poison? World Bank Research Observer 15, no. 1: 47–68. Pack, Howard. 2001a. Technological Change and Growth in East Asia: Macro and Micro Perspectives. In Rethinking the East Asian Miracle, ed. Joseph E. Stiglitz and Shahid Yusuf. Oxford, England: Oxford University Press. Pack, Howard. 2001b. The Role of Foreign Technology Acquisition in Taiwanese Growth. Industrial and Corporate Change 10, no. 3: 713–33. Pack, Howard, and Christina Paxson. 1999. Inter-industry Labor Mobility and Knowledge Spillovers in Taiwan. Discussion Paper 2102. Washington: World Bank. Pack, Howard, and David Lin. 2001. The Role of Industrial Policy in Taiwan’s Development. Photocopy. University of Pennsylvania, Philadelphia. Pack, Howard, and Kamal Saggi. 2000. Vertical Technology Transfer, Diffusion, and Competition. Journal of Development Economics (August): 389–416. Pack, Howard, and Larry E. Westphal. 1986. Industrial Strategy and Technological Change. Journal of Development Economics 22, no. 1: 87–128. Park, Seung-rok, and Jene K. Kwon. 1995. Rapid Economic Growth with Increasing Returns to Scale and Little or No Productivity Growth. Review of Economics and Statistics 77, no. 2: 332–51. Patrick, Hugh. 1986. Japan’s High Technology Industries: Lessons and Limitations of Industrial Policy. Seattle: University of Washington Press. Pilat, Dirk. 1994. The Economics of Rapid Growth: The Experience of Japan and Korea. Brookfield, VT: Edward Elgar. Prestowitz, Clyde V., Jr. 1988. Trading Places. New York: Basic Books. Psacharopoulos, George. 1973. Returns to Education. San Francisco: Jossey-Bass. Pyo, Hak-kil. 1989. Export-led Growth, Domestic Distortions, and Trade Liberalization. Paper presented at the United States-Korea Financial Policy Discussions. Washington, December. Radelet, Steven, and Jeffrey D. Sachs. 1998. The East Asian Financial Crisis: Diagnosis, Remedies, Prospects. Brookings Papers on Economic Activity 30, no.1: 1–74. Ranis, Gustav. 1979. Industrial Development. In Economic Growth and Structural Change in Taiwan, ed. Walter Galenson. Ithaca, NY: Cornell University Press. Ray, Debraj. 1998. Development Economics. Princeton, NJ: Princeton University Press. Rhee, Yung W., B. Ross-Larson, and G. Pursell. 1984. Korea’s Competitive Edge: Managing Entry into World Markets. Baltimore: Johns Hopkins University Press. Rhee, Yung-whee. 1994. Managing Entry into International Markets: Lessons from the East Asian Experience. In Manufactured Exports of East Asian Industrializing Economies, ed. Shu-chin Yang. Armonk, NY: M. E. Sharpe. Rivera-Batiz, L., and P. M. Romer. 1991. Economic Integration and Endogenous Growth. Quarterly Journal of Economics 106, no. 2: 531–56. Rob, Rafael. 1991. Learning and Capacity Expansion Under Demand Uncertainty. Review of Economic Studies 58, no. 4: 655–75. Rodrik, Dani. 1994. King Kong Meets Godzilla: The World Bank and the East Asian Miracle. In Miracle or Design? Lessons from the East Asian Experience. Policy Essay 11. Washington: Overseas Development Council. 108

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

06--References--103-110

3/14/03

10:16 AM

Page 109

Rodrik, Dani. 1995. Getting Interventions Right: How South Korea and Taiwan Grew Rich. Economic Policy 20: 55–107. Rodrik, Dani. 1999. The New Global Economy and Developing Countries: Making Openness Work. Policy Essay 24. Washington: Overseas Development Council. Rodrik, Dani. 2001. The Global Governance of Trade As If Development Really Mattered. New York: United Nations Development Programme, October. Rosenstein-Rodan, Paul N. 1943. Problems of Industrialization of Eastern and Southeastern Europe. Economic Journal 53: 202–11. Rosenzweig, Mark, and Kenneth Wolpin. 1993. Credit Market Constraints, Consumption Smoothing, and the Accumulation of Durable Production Assets in Low-Income Countries: Investment in Bullocks in India. Journal of Political Economy 101 (April): 223–44. Sakakibara, Mariko. 1997. Evaluating Government-Sponsored R&D Consortia in Japan: Who Benefits and How? CIPER Working Paper 97–4. Los Angeles: University of California, Los Angeles, John E. Anderson Graduate School of Management. Sakakibara, Mariko, and Michael E. Porter. 2001. Competing at Home to Win Abroad: Evidence from Japanese Industry. Review of Economics and Statistics 83, no. 3: 310–22. SaKong, Il. 1993. Korea in the World Economy. Washington: Institute for International Economics. Saxonhouse, Gary R. 1983. What Is All This About “Industrial Targeting” in Japan? World Economy 6, no. 3: 253–73. Sazanami, Yoko, Shujiro Urata, and Hiroki Kawai. 1995. Measuring the Costs of Protection in Japan. Washington: Institute for International Economics. Schott, Jeffrey J. 1994. The Uruguay Round: An Assessment. Washington: Institute for International Economics. Schultz, Theodore W. 1964. Transforming Traditional Agriculture. New Haven: Yale University Press. Schultz, Theodore W. 1975. The Value of the Ability to Deal with Disequilibria. Journal of Economic Literature 13 (Sept.): 827–46. Scitovsky, Tibor. 1954. Two Concepts of External Economies. Journal of Political Economy 62, no. 2: 143–51. Shinohara, Miyohei. 1982. Industrial Growth, Trade, and Dynamic Patterns in the Japanese Economy. Tokyo: Tokyo University Press. Shleifer, Andrei, and Robert W. Vishny. 1993. Corruption. Quarterly Journal of Economics 108, no. 3: 599–618. Shouda, Yasutoyo. 1982. Effective Rates of Protection in Japan. Nihon Keizai Kenkyu 11 (March): 68–70. Smith, Heather. 1997. Taiwan’s Industrial Policy in the 1980s: An Appraisal. Asian Economic Journal 11, no. 1: 1–34. Smith, Heather. 2000. Industry Policy in Taiwan and Korea in the 1980s. Cheltenham: Edward Elgar. Spencer, Barbara, and James Brander. 1983. International R&D Rivalry and Industrial Strategy. Review of Economic Studies 50, no. 4: 707–22. Stiglitz, Joseph E. 1993. The Role of the State in Financial Markets. Paper presented at the Annual Conference on Development Economics, World Bank, Washington, DC. Stiglitz, Joseph E. 1994. Whither Socialism? Cambridge, MA: MIT Press. Stiglitz, Joseph E. 2002. A Fair Deal for the World. New York Review of Books. May 23. Sturgeon, Timothy J., and Richard K. Lester. 2002. Upgrading East Asian Industries: New Challenges for Local Suppliers. Cambridge, MA: Massachusetts Institute of Technology, Industrial Performance Center. Tanaka, Naoki. 1988. Aluminum Refining Industry. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Thorbecke, Eric, and Henry Wan, eds. 1999. Taiwan’s Development Experience: Lessons on Roles of Government and Market. Boston: Kluwer Academic Publishers. Vestal, James E. 1989. Evidence on the Determinants and Factor Content Characteristics of Japanese Technology Trade. Review of Economics and Statistics 71, no. 4: 565–71. REFERENCES

109

06--References--103-110

3/14/03

10:16 AM

Page 110

Vogel, Ezra. 1979. Japan as Number One: Lessons for America. New York: Harper & Row. Wade, Robert. 1990. Governing the Market: Economic Theory and Taiwan’s Industrial Policies. Princeton, NJ: Princeton University Press. Wang, Jiann-Chyuan. 1998. The Industrial Policy of Taiwan, ROC. In Industrial Policies in East Asia, ed. Seiichi Masuyama, Donna Vandenbrink, and Chia Siow Yue. Singapore: Institute of Southeast Asian Studies. Weinstein, David E. 1995. Evaluating Administrative Guidance and Cartels in Japan (1957– 88). Journal of the Japanese and International Economies 9, no. 2: 200–23. Westphal, Larry E., Yung W. Rhee, and Gary Pursell. 1981. Korean Industrial Competence: Where It Came From. World Bank Staff Working Paper no. 469. Washington: World Bank. Westphal, Larry E., and Kwang Suk Kim. 1982. Korea. In Development Strategies for SemiIndustrial Economies, ed. Bela Balassa et al. Baltimore: Johns Hopkins University Press. Westphal, Larry E. 1990. Industrial Policy in an Export-Propelled Economy: Lessons from South Korea’s Experience. Journal of Economic Perspectives 4, no. 3: 41–59. Westphal, Larry E. 2002. Technology Strategies for Economic Development in a Fast Global Economy. Economics of Innovation and New Technology 11, no. 4–5 (August/October). Whalley, John, and Colleen Hamilton. 1996. The Trading System After the Uruguay Round. Washington: Institute for International Economics. Williamson, John. 1990. What Washington Means by Policy Reform. In Latin American Adjustment: How Much Has Happened? ed. John Williamson. Washington: Institute for International Economics. Winckler, Edwin A., and Susan Greenhalgh, eds. 1988. Contending Approaches to the Political Economy of Taiwan. Armonk, NY: M. E. Sharpe. World Bank. 1993. The East Asian Miracle. New York: Oxford University Press. World Bank. 1998. East Asia: The Road to Recovery. Washington, DC: World Bank. Yamawaki, Hideki. 1988. The Steel Industry. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Yamazawa, Ippei. 1988. The Textile Industry. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Yano, Makoto. 2001. Trade Imbalance and Domestic Market Competition Policy. International Economic Review 42, no. 3: 729–50. Yonezawa, Yoshie. 1988. The Shipbuilding Industry. In Industrial Policy of Japan, ed. Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura. San Diego: Academic Press. Yoo, Jung-ho. 1990. The Industrial Policy of the 1970s and the Evolution of the Manufacturing Sector. KDI Working Paper 9017. Seoul: Korea Development Institute. Yoo, Jung-ho. 1993. The Political Economy of Protection Structures in Korea. In Trade and Protectionism, ed. Anne O. Krueger and Takatoshi Ito. Chicago: University of Chicago Press. Yoo, Jung-ho. 1994. South Korea’s Manufactured Exports and Industrial Targeting Policy. In Manufactured Exports of East Asian Industrializing Economies, ed. Shu-Chin Yang. Armonk, NY: M. E. Sharpe. Yoo, Seong Min. 1999. Corporate Restructuring in Korea. Joint U.S.-Korea Academic Studies 9: 131–99. Zeile, William. 1990. Industrial Policy and Organizational Efficiency: The Korean Chaebol Examined. Program in East Asian Culture and Development Research Working Paper Series 30. Davis, CA: University of California, Davis, Institute for Government Affairs. Zeile, William. 1991. Industrial Targeting Through Government Rationing: The Korean Experience of the 1970s. Program in East Asian Culture and Development Research Working Paper Series 41. Davis, CA: University of California, Davis, Institute for Government Affairs.

110

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 111

Index

acquisition of technology, 65 Africa, Asian model, 2 agriculture Japan, 25, 26 effective rates of protection, 31t Taiwan, 53 Anglo-American model, 7 Argentina capital stock growth, 87t distributional indicators, 83t human capital, science and engineering students, 40t human capital accumulation, 39f Human Capital Index and per capita income, mid-1950s, 25t Asian financial crisis disappointing results, 1 IMF micromanagement, 90–91 Philippines, 85n second-round effects, 100 Asian miracle, 95 attributes, 100 Asian model, 2 Asian tigers, unorthodox policies, 2 Asia-Pacific Economic Cooperation, intellectual property rights, 56 atmospheric factors, 94, 95 input-output approach, 96n Austria, export similarity, 84t authoritarian regimes, analysis disregarded, 3 auto industry, Japan, 22n

banking sector early Japanese, 23 South Korea, 39, 41 loan evaluations, 71 start-ups, 18 suppression, policy of, 15 basic metals, Japan, normalized sectoral tax rates, 37t beverages, Taiwan, 55 beverages sector, South Korea, intersectoral purchases, 50t big push, pecuniary externalities, 17 book, purpose of, 8 brain drain, 94n, 95 Brander-Spencer model, 19 with retaliation, 20n Brazil capital stock growth, 87t distributional indicators, 83t export similarity, 84t human capital accumulation, 39f bureaucratic competence, 74 cautions to model, 100 political stability, 78 Canada, export similarity, 84t capital, sustaining returns, 97–98 capital accumulation, 97 capital assets, unsupervised, 71 capital channeling policies, 41 capital goods, imported, 98, 98n

111

07--Index--111-121

3/14/03

10:17 AM

Page 112

capital resource channeling, South Korea, 46 capital stock growth, 87t capital subsidies Japan, 26, 28 by sector, 27f cartels impacts studied, 36 “recession” or “rationalization,” 32 cement products, Taiwan, 55 ceramics/stone/glass, Japan, normalized sectoral tax rates, 37t chaebol balance with state, 68–69 private coordination, 46 use of directed credit, 70 chemical products, Taiwan, 55 chemicals Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t South Korea, intersectoral purchases, 50, 50t Taiwan, subsidized credit, 53 Chile Asian model, 15 human capital, science and engineering students, 40t industrialization, 11 more complex industries, 23 China local entrepreneurs, 14 sectoral development, 3 clothing industry, Colombia, 13n coal mining, Japan, 25 Cobb-Douglas, 36 Colombia, clothing industry, 13n comparative advantage anticipating change, 22n early Japanese, 23–24 initial industrialization, 99 Japan, R&D-intensive activities, 33–34 competition cartel impact, 36 export growth, 99 size of industry, 16–17 South Korea, measuring, 40 competition policy effects of, 6 FSX fighter dispute, 32 Japan anticompetitive behavior, 32 imports, 35, 35n lax enforcement, 32 competitiveness improving, strategies for, 16–18 learning-by-doing, 18 Comprehensive Stabilization Plan, 42 computer and robotics industries, Japan, 28 conditionality, 90

112

construction industry, 74–75 consumer surplus, 9 consumption levels, Japan-US compared, 9 corruption, 72–75 avoiding, 97 corruption surveys, 72, 72n, 73f credit access, 70 relative lack, 20 Costa Rica, more complex industries, 23 countervailing duty (CVD) cases, 89n Cournot competition, 18n Daewoo, 70 “deep” industrial structures, 10 democratization, South Korea, 69 Denmark, export similarity, 84t depreciation tax breaks Japan, 26, 26n specificity test, 91 WTO rules, 91 depressed regions, development of, 89 developing countries financial institutions’ constraints, 91 growth record, 2 development economics, 13–15 dispute settlement, WTO, 90 East Asia model debated, 15 replicability, 15, 77–78 East Asian Miracle (World Bank 1993), 7–8 East Asian tigers, unorthodox policies, 2 economic fundamentals, neoclassical view, 6 economic growth early industrial policy, 97 identifying broad sectors, 101 industrial policy related, 93–94 infrastructure facilitating, 95 policies advocated, 101 political legitimacy, 68 psychological factors, 95 quantitative or “atmospheric” factors, 94 economies of scope, 17 education brain drain, 94n, 95 chemical engineering, 4 comparing university levels, 79 improving, 3 investment in, 6 literacy rates, 79 marginal productivity of investment, 98 new technology, 98 other nations compared, 94 out-migration, 94n, 95 productivity, 98 effective rates of protection (ERPs), 30–31, 30n, 31t effects studied, 33 Egypt distributional indicators, 83t

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 113

human capital, science and engineering students, 40t human capital accumulation, 39f electrical machinery Japan capital subsidy-investment ratio, 27t domestic and imported, 51t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t Korea, domestic and imported, 51t electronics industry, 65 Japan, 68, 68n entrepreneurship, encouraging, 13–15 environmental requirements, 89 exchange rates oil shocks, 63 South Korea, exports, 47–48 undervalued, 5n executive branch, power centralized, 68 export growth Asian miracle, 100 benefits of industry specific, 63 national, 63 competitiveness, 5n, 99 explaining, 96 externalities, 64–65 macroeconomic management, 62–63 Export-Import Bank, Japan, 25 export-led development, competitiveness, 4 exports Asian experience, 94 corruption related, 74 export similarity, by country, 84t monitoring performance, 85–86 rebates, 78 role macroeconomic management, 62–63 as model of intervention, 63 productivity, 63–66 South Korea, 40–41 increase, 47 real exchange rates, 47–48 Taiwan, boom causes debated, 54n wastage allowances on duty-free imports, 41, 41n export subsidies, 89 WTO ban, 91 export targets, 91 external economies, 16 externalities by first entrant, 17–18, 18n explaining export growth, 96 export growth, 64–65 government intervention, 95–96 Japan, 9–10 potential interindustry, measuring, 48–49 South Korea, 46 TFP growth, 96

fabricated metal, Japan, normalized sectoral tax rates, 37t FDI (foreign direct investment) Singapore, 13 to start industrialization, 13 FILP. See Fiscal Investment and Loan Program financial institutions, corruption, 74 financial repression corruption, 74 credit access, 70 effects, 70 financial sector Japan, cleanup costs, 71 South Korea cleanup costs, 71 vulnerability, 70–71 Taiwan, cleanup costs, 71 vulnerability, 70 South Korea, 70–71 Finland, export similarity, 84t Fiscal Investment and Loan Program (FILP), 25–26 uses by sector, 27f fishery, Japan, effective rates of protection, 31t fishing, Japan, 25 food processing Japan capital subsidies-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t food products, Taiwan, 55 food sector, South Korea, intersectoral purchases, 50t forestry Japan, 25 effective rates of protection, 31t France capital stock growth, 87t export similarity, 84t FSX fighter dispute, 32 GDP Japan, decline (1990s), 71 South Korea, decline (1997–98), 71 General Dynamics, 32 general equilibrium model Japanese industrial policy, 33 South Korea, resource misallocation, 44 general machinery Japan capital subsidy-investment ratio, 27t domestic and imported, 51t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t South Korea, domestic and imported, 51t generic infrastructure, 3–4 Germany capital stock growth, 87t export similarity, 84t

INDEX

113

07--Index--111-121

3/14/03

10:17 AM

Page 114

Ghana, capital stock growth, 87t globalization connotations, 1 doubts about, 7 meanings of, 1 governance rankings, 73f, 74 government failure, 20 political instability and policy change, 22–23 government intervention, 4 distributional impact, 67 export productivity, 63–66 externalities, 95–96 initial industrialization, 99 success of, 63–66 TFP growth, 96 government official competence, 74 government procurement, 32 government role Japan, 30 monopsonistic power, 32 R&D positive impact, 34 South Korea, 44 as central agent, 46 Taiwan political economy considerations, 56 R&D, 55, 56f sectoral evolution, 51–52 Greece export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t growth accounting, 58–59 growth enhancement Japan, 36–37 South Korea, 42 industry-specific, 47 three criteria, 47 growth rates conflation with welfare enhancement, 16n coordination of planned investment, 17 hypotheses, 6 industrial policy related, 2–3 labor-intensive industries, 12 trade specialization, 19 guidance administrative, Japan, 36 informal, Japan, 36 informal administrative, 24 Harrod-neutral (labor-augmenting) technological advance, 59 heavy and chemical industry (HCI) assessed, 42, 43–44 chaebol oligopoly, 44 promoted versus other sectors, 60 South Korea, 41–42 heavy industries South Korea, 49, 50t intersectoral purchases, 50, 50t Hicks-neutral technical change, 59

114

high-technology industries, Taiwan, subsidized credit, 53 high-technology sector Japan, tax incentives, 28, 28n licensing technology, 65, 66 royalty payments, 31 Taiwan, 52, 55 Hitachi, 65 Hong Kong distributional indicators, 83t early development era, 12 export similarity, 84t industrial depth, 10, 10n laissez-faire, 12–13 skill base debated, 12 household level, 13–14 Housing Loan Corporation (Japan), 25 Hsinchu Science Park (Taiwan), 55 human capital, 7 economic growth, 94 Japan, 24 scatterplots, 79, 80f–81f South Korea, 38–40 accumulation compared, 39f science and engineering compared, 40t US aid, 39n Taiwan, 55 uncodified knowledge to other firms, 48, 49–50 Human Capital Index defined, 25n per capita income, mid-1950s, 25t Hyundai, 70 IMF constraints on industrial policy, 88, 90–91 orthodoxy enforced, 2 import diversification program, 88 import-substituting industrialization (ISI) economic growth, 94 investment rates, 94 Philippines, 85 policy latitude, 63 questioned, 3 strategy debated, 100–01 Taiwan, 53 incentives, 4 income distributional indicators, 83t Human Capital Index and, 25t Japan as middle-income, 22 Latin America, 2–3 per capita growth, 12 postwar Japan, 78–79 See also middle-income countries India capital stock growth, 87t cross-sector externalities, 96, 96n distributional indicators, 83t human capital, science and engineering students, 40t

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 115

human capital accumulation, 39f industrial structure, 10 technology imports, 99 TPF growth, 96 Indonesia capital stock growth, 87t distributional indicators, 83t export similarity, 84t WTO complaint, 90 industrial fertilizers, Taiwan, 55 industrialization initiation, 11–15 policy options, 12–13 industrial policy assessing impact, 6 caveats, 19–20 defining, 10 international system constraints, 88–92 pitfalls, 83, 85 scarce resources, 20 sectoral discrimination, 3 sectoral favoritism costs, 20, 20n sectoral TFP growth, 34–36, 35t as source of growth, 2 South Korea assessed, 42–51 effects on nonpromoted sectors, 49–51 promoted and neglected sectors, 48–49, 50t types of instruments, 10 World Bank-IMF constraints, 88, 90–91 industrial promotion, Japan, 28n Industrial Technology Research Institute (ITRI) (Taiwan), 55 industry councils (shingikai), 36 infant industry promotion, 23 influence peddling, 75 information technology sector, Taiwan, 57 infrastructure economic growth related, 95 generic, 3–4 provision of, 79 initial industrialization, 11–15 comparative advantage, 99 conflicts, 92 government intervention, 99 irreproducible conditions?, 78–83 Japan, needs, 21–22 initiation of industrialization, 11–15 input markets, distortions, 60 input-output approach atmospherics, 96n Taiwan, 57 intellectual property rights strategic effect, 19n Taiwan, 55–56 interdependence of two industries, model of, 18n interest groups, Japan, 68 intermediate goods domestic production, 48

imports, 98, 98n International Monetary Fund. See IMF international supply networks, 13, 13n international system constraints, 88–92 intervention East Asian “miracle,” 7–8 Taiwan, 8n investment economic growth, 94 South Korea boom, 47, 48 coordination, 46, 50 sustaining returns, 97–98 investment rates, 5 infrastructure versus, 95 iron and steel Japan capital subsidies-investment ratio, 27t effective rates of protection, 31t R&D subsidy share, 30t South Korea, intersectoral purchases, 50, 50t Israel export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t Italy export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t Japan auto industry, 22n capital stock growth, 87t construction industry, 74 corruption rankings, 72, 73f distributional indicators, 83t electronics industry, 65, 68, 68n exports, early, 62 export similarity, 84t externalities, 9–10 factor markets distortions, 60 general and electrical machinery, 51t governance rankings, 73f, 74 government role, 30 as growth model, 3 Hitachi, 65 human capital, science and engineering students, 40t Human Capital Index and per capita income, mid-1950s, 25t imported technology, 99 industrial policy, history, 23–24, 23n interest groups, 68 land reform, 82, 82n long-run costs, 5 paeans of 1980s, 8–9 political economy, 67–69 poster nation, 2 postwar economic reconstruction, 24

INDEX

115

07--Index--111-121

3/14/03

10:17 AM

Page 116

Japan—continued income levels, 78–79 needs of, 21–22 postwar policies, 11 product bans in South Korea, 88 purchasing power parity, 9 steel industry, 22n subsidies R&D, 93 sectoral impact, 93 success story, 5 telecommunications, 68, 68n World War II effects, 24 WTO complaint, 90 Japan Development Bank, 25, 28 Japanese model, explaining success, 22 Japan Inc., 8, 11 J. C. Penney Co., 64 keiretsu, government’s coordinating role, 46n Kenya capital stock growth, 87t distributional indicators, 83t Kim Dae-jung, 75 Kim Young-sam, 75 knowledge spillovers, 16 knowledge transmission, worker movement, 48, 49–50 Korean model, questioning, 15 Korean War (1950–53), 38 procurement boom, 21n Korea. See South Korea Korea Trade Promotion Corporation (KOTRA), 40–41 Kuomintang, 68, 69 labor flexibility, 98–99 reallocations among sectors, 63–64 scatterplots, 79, 80f–81f South Korea, deadweight loss, 78 labor-intensive industries, 12 labor organizations resistance to change, 70 South Korea, bans, 41n suppression, 60 wage repression, 60 labor productivity by sector, 63 South Korea, 41n protected sectors, 45 TFP growth rates, 97 laissez-faire Asian success not, 21 for starting industrialization, 12–13 land distributional indicators, 83t East Asia, 82 scatterplots, 79, 80f–81f

116

land reform, 68 Japan, 82, 82n South Korea, 82, 82n Taiwan, 82, 82n large-scale industrial firms, 14 Latin America disappointing results, 1 education rates, 95 income declining, 2–3 technology imports, 99 TFP growth rates, 86t least developed countries, models for, 22 leather and rubber, Japan, effective rates of protection, 31t leather sector, South Korea, intersectoral purchases, 50t Lee Teng-hui, 52 legitimacy (political) issue, 67–68 Leontief input-output coefficients, 49 LG. See Lucky Goldstar liberalization policies South Korea, 41 Taiwan, 54 licensing for technology, 65, 66 lifetime employment, 9 light industry South Korea, 42 rates of return, 43f List, Friedrich, 23 literacy rates, 79 loans, low-interest, 28, 28n capital channeling policies, 41 Japan, R&D, 29 liberalization policy, 41 South Korea, 41, 45t Taiwan, export subsidy loans, 53, 53f Lucky Goldstar (LG), 65 machine tool industry, local conditions/design, 18 machine tool manufacturers, 16–17, 17n macroeconomic management, export growth, 62–63 macroeconomic policies, Asian success stories, 5 Malaysia capital stock growth, 87t distributional indicators, 83t export similarity, 84t human capital, science and engineering students, 40t Human Capital Index and per capita income, mid-1950s, 25t international institutions, 90 manufacturing, Japan, effective rates of protection, 31t manufacturing sector South Korea, 42 rates of return, 43f market equilibrium, South Korea, 42

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 117

market failures methods of intervention, 22 need to identify, 15–16 market fundamentalism, 90 Marxist economists, Japan, 23n medicines, Taiwan, 55 Meiji Restoration, 23 metal products Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t R&D subsidy share, 30t South Korea, intersectoral purchases, 50, 50t Mexico capital stock growth, 87t distributional indicators, 83t export similarity, 84t human capital, science and engineering students, 40t human capital accumulation, 39f Human Capital Index and per capita income, mid-1950s, 25t Microsoft, 16, 17n microwave oven production, 64 middle-income countries Japan as model, 22 selective intervention, 90 World Bank-IMF constraints, 90–91 military production, 24 mining Japan capital subsidies-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidies share, 30t subsidies, 27f, 28 R&D, 29 Ministry of Finance (Japan), orthodoxy on competitiveness, 35n Ministry of International Trade and Industry (MITI), 9 “miracles”, economic, 62–63 Asian miracle, 95 miscellaneous manufacturing sector, South Korea, intersectoral purchases, 50t miscellaneous products, Japan, effective rates of protection, 31t modernization literature, 14 monitoring of firms international institutions’ constraints on, 91 protectionism, 85 monopsonistic power, government role, 32 moral hazard, 33 motor vehicles, Taiwan, 55 multinational corporations, Singapore, 13 Museveni, Yoweri, 2 natural resources, scatterplots interpreted, 82 natural resources, lack of, 79

neoclassical economics, Japan, 23, 23n neoclassical view, 6 “neoliberalism,” 7 neomercantilist orientation, 23n Nigeria, capital stock growth, 87t Nippon Telephone and Telegraph (NTT), 29 nonelectrical machinery, South Korea, intersectoral purchases, 50, 50t nonferrous metals as intervention target, 47 Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t R&D subsidy share, 30t nonmetallic mineral products, Japan, effective rates of protection, 31t nonmetallic minerals sector, South Korea, intersectoral purchases, 50t nonmetallic products as intervention target, 47 Japan capital subsidy-investment ratio, 27t R&D subsidy share, 30t Norway, export similarity, 84t OECD, non-US, TFP growth rates, 86t off-budget finance, Japan, indirect subsidies, 25 oil price increases, 63 O’Neill, Paul H., 2 Pakistan capital stock growth, 87t export similarity, 84t human capital, science and engineering students, 40t human capital accumulation, 39f paper, as intervention target, 47 paper sector, South Korea, intersectoral purchases, 50t Park Chung-hee, centralization of power, 68 pecuniary economies, 17 per capita income, growth, 12 Peru, export similarity, 84t petroleum and coal products as intervention target, 47 Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t pharmaceutical firms, Japan, 34 Philippines advantages, 83, 85 Asian financial crisis, 85n distributional indicators, 83t export similarity, 84t human capital, science and engineering students, 40t

INDEX

117

07--Index--111-121

3/14/03

10:17 AM

Page 118

Philippines—continued Human Capital Index and per capita income, mid-1950s, 25t reforms, 85n physical capital economic growth, 94 scatterplots, 79, 80f–81f interpreted, 82 political capture, 67–68 political economy factors, 20, 67–69 political stability, 77 POSCO (Korean iron and steel producer), 44, 99 postal savings system, 25 “post hoc, ergo propter hoc” fallacy, 5 precision instruments Japan capital subsidies-investment ratio, 27t normalized sectoral tax rates, 37t R&D subsidy share, 30t precision machinery, Japan, effective rates of protection, 31t present discounted value (PDV), 16n primary traded goods, Japan, effective rates of protection, 31t printing and publishing sector, South Korea, intersectoral purchases, 50t procurement, public, 32 production function estimates, 59–61 productivity, 4 acquisition of technology, 65–66 education and new technology, 98 export growth, success of government intervention, 63–66 imports impact, 34–35 improvements resulting from activities of other firms, 16 Samsung, 64 South Korea, export monitoring, 86 proprietary technologies, 65 “pro-producer” bias, 6 protectionism, 4 monitoring of firms, 85 sectoral TFP growth, 34 South Korea, 41, 42, 44 cross-sectional, 45–46 studies of Japan’s economy, 33 Taiwan debated, 51, 52 industrial policy tool, 53–54 US, 9 WTO rules, 88–89 Prussia, 23 public financial institutions, Japan, 25, 26 public utilities, Taiwan, subsidized credit, 53 publishing, Japan, effective rates of protection, 31t publishing sector, Japan, 30 pulp and paper Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t

118

normalized sectoral tax rates, 37t R&D subsidy share, 30t purchasing power parity, Japan-US compared, 9 R&D (research and development) benefits to other firms, 18 export growth, 65 Japan comparative advantage, 33–34 research consortia, 34, 34n subsidies, 93 outcomes possible, 19 sectoral patterns, 29, 34 subsidies, government share, 29, 30t Taiwan, 52, 55, 56f WTO, 89, 91–92 noncompliance, 92n Radio Shack, 65 real external economies, 16 regional development, 89 regulatory institutions, 98, 98n rents, capture of, 19 rent-seeking behavior, 20 replicability issues, 15, 77–78 research analyzed, 8 research and development. See R&D retaliation, possibility of, 20, 20n return rates, South Korea, 43f revisionist view, 6–7 risk taking, household level, 14 royalty payments, 31 rubber products sector, South Korea, intersectoral purchases, 50t Saburo Okita, 37n Samsung, 64, 65 SAP, 16 savings rates, 6 captive saving, 6n economic growth, 94 South Korea, 41n scale economies, productivity and export growth, 64 Schultz, Theodore, 14 second-generation reforms, 1, 6n second-round effects, 100 sectoral industrial policy, Japan, postwar, 24–32 sectoral neutrality, justifying, 7 sectoral productivity, South Korea, 44–45, 45t selective industrial policy case for, 15–20 defining, 10 identifying successful aspects, 4–5 internal determinants, 92 selective intervention corruption burden, 74 declining sectors, 67 distributional impact, 67 initial conditions, 78–83 middle-income countries, 90 success measured, 97

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 119

Seoul, 38 Singapore distributional indicators, 83t export similarity, 84t FDI (foreign direct investment), 13 reliance on FDI, 4 Small Business Finance Corporation (Japan), 28 small farmers, 13–14 small-scale firms, official support for, 14, 14n social security payments, 25 software industry, Japan, tax incentives, 28, 28n Sony, 34 South Asia, TFP growth rates, 86t Southeast Asia, education rates, 95 South Korea capital stock growth, 87t corruption rankings, 72, 73f scandals, 75 democratization, 69 dependence on US economic aid, 39, 39n distributional indicators, 83t earlier history, 38 early development era, 11–12 education levels, 22 export expansion and manufacturing, 63–64 exports, early, 62 export similarity, 84t factor markets distortions, 60 financial sector vulnerability, 70–71 general and electrical machinery, 51t governance rankings, 73f, 74 human capital, science and engineering students, 40t Human Capital Index and per capita income, mid-1950s, 25t industrial policy, 37–51 as Japanese colony, 38 Korean War (1950–53), recovery, 47 land reform, 82, 82n long-run costs, 5 Lucky Goldstar, 65 military coup (Park Chung-hee), 40 monitoring export performance, 85 political economy, 67–69 poster nation, 2 postwar recovery, 47 Samsung, 64, 65 successes interpreted, 8 success story, 5 Taiwan compared, OECD imports, 44 TFP growth, 86t, 96–97 trade policies, 78 transparency, 75 US military occupation, 38 WTO complaint, 90 Spain export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t state, opposition to, 77

static comparative advantage, industrial policy versus, 10 statist policies, reconsidered, 2 Statute for the Encouragement of Investment (Taiwan), 52 Statute for the Upgrading of Industry (Taiwan), 52 steel industry Japan, 22n, 24 South Korea, 44, 99 steel manufacturing, 17 Stiglitz, Joseph, 90 strategic trade policy, 18–19, 19n retaliation possibility, 20, 20n sub-Saharan Africa income declining, 2 industrialization, 11 subsidies capital, Japan, by sectors, 27t competitiveness, 5n credit and R&D policy, 29 declining sectors, 67 direct, R&D, 29 early Japanese, 23n export, 89 no longer possible, 100 output subsidization schemes, 19 PDV of producer surplus versus social cost of, 16n questioned, 19, 19n sectoral composition, Japan, 26f, 27f selective incentives, 91 Taiwan, effects hypothesized, 57 WTO, 89 subsidies, direct, Japan, postwar, 25 subsidies, indirect Japan FILP, 25 off-budget finance, 25 postwar, 25 tax system, 25 supercomputer procurement, 32, 32n Sweden, export similarity, 84t Symantec, 16, 17n Taiwan capital stock growth, 87t corruption rankings, 72, 73f construction and public works, 74–75 distributional indicators, 83t early development era, 11–12 education levels, 22 export growth and manufacturing, 63–64 exports, early, 62 export similarity, 84t governance rankings, 73f, 74 industrial policy, 51–56 major tools, 52–54 neoclassical interpretation, 51 success analyzed, 51–52 intervention, 8n

INDEX

119

07--Index--111-121

3/14/03

10:17 AM

Page 120

Taiwan—continued labor flexibility, 98–99 labor-intensive, 12 land reform, 82, 82n OECD import shares compared, 44 political economy, 67–69 poster nation, 2 success story, 5 TFP growth rates, 86t, 96–97 target setting, Japan, 36 tariff cuts Japan, 30–31 WTO bound tariff lines, 88 tariff rates, South Korea, 45t tariffs justifying, 7 Taiwan, 52 collections, 54, 54f tax system depreciation, Taiwan, 52 Japan, 26, 26n, 28, 28n accelerated depreciation, 26 indirect subsidies, 25 normalized, 36, 37t R&D, 29 sectoral TFP growth, 34 South Korea, 41–42, 45t Taiwan, 52, 54 See also depreciation tax breaks technical change, growth accounting exercises, 59 technology transfer, source of growth, 15 telecommunications, Japan, 68, 68n terms of trade effects, 19 textiles Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t South Korea, 42, 50t spinning, 31t weaving, 31t textiles and clothing sector, South Korea, intersectoral purchases, 50t Thailand capital stock growth, 87t distributional indicators, 83t export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t tobacco sector, South Korea, intersectoral purchases, 50t Tokyo Round, ERPs, 30–31 Tornqvist index, 58–59 Toshiba, 34 total factor productivity (TFP), 6 calculations, 36, 58–61 debate in Asia, 36 government intervention, 96 growth rates, 7, 86, 86t, 97

120

calculating, 58–61 calculating with growth accounting, 58–59 calculating with production function estimates, 58, 59–60 compared, 87 debated, 86 pitfalls, 61 reliability, 93 sectoral growth and industrial policy, 34–36, 35t South Korea, 43 interindustry externalities, 48 resource misallocation, 44, 44n Taiwan industrial policy effect, 57 neglected and favored sectors, 57 traded goods, Japan, effective rates of protection, 31t trade policies, South Korea, 78 trade specialization, growth rates, 19 transitional economies, disappointing results, 1 transparency, South Korea, 75 transparency risk premium, 75 transportation machinery Japan capital subsidy-investment ratio, 27t effective rates of protection, 31t normalized sectoral tax rates, 37t R&D subsidy share, 30t South Korea, intersectoral purchases, 50, 50t Tunisia, export similarity, 84t Turkey distributional indicators, 83t export similarity, 84t human capital, science and engineering students, 40t human capital accumulation, 39f Human Capital Index and per capita income, mid-1950s, 25t Uganda development process, 12 prerequisites, 12n United Kingdom, export similarity, 84t United States export similarity, 84t Human Capital Index and per capita income, mid-1950s, 25t purchasing power parity, 9 R&D expenditures, 29 WTO complaint, 90 upgrading, industrial, South Korea, 43–44 Uruguay Round ban on VERs, 88 effects on industry, 88 Japanese tariff rates, 31 venture capital funds, Taiwan, 55 VERs (voluntary export restraints), 88 Very Large Scale Integration Research Association, 29

INDUSTRIAL POLICY IN AN ERA OF GLOBALIZATION

07--Index--111-121

3/14/03

10:17 AM

Page 121

Washington consensus, 1 Asian successes, 3 government role, 3 pillars undermined, 7 questioned, 7 wastage allowances, South Korea, 48, 78 welfare consequences Japan, 9 physical and human investment impacts, 7 welfare enhancement conflated with growth-accelerating, 16n as exception, 33 Japan, 36–37 South Korea, 42, 46 welfare-enhancing policy, market failures, 15–18 wooden products, Japan, effective rates of protection, 31t wood products, as intervention target, 47

wood and wood products sector, South Korea, intersectoral purchases, 50t wool and worsted fabrics, Taiwan, 55 working hours, 9 World Bank Asian miracle, 95 constraints on industrial policy, 88, 90–91 orthodoxy enforced, 2 World Trade Organization (WTO) constraints on industrial policy, 88–90, 91–92 impact on industrial policy, 100 intellectual property rights, 56 orthodoxy enforced, 2 World War II effects on Japan, 24 postwar era, independence movements, 4 zaibatsu, 24

INDEX

121