Japan in East Asia: Trading and Investment Strategies 9789814379250

Citation preview

JAPAN IN EAST ASIA Trading and Investment Strategies

The Institute of Southeast Asian Studies (ISEAS) was established as an autonomous organization in 1968. It is a regional research centre for scholars and other specialists concerned with modern Southeast Asia, particularly the many-faceted problems of stability and security, economic development, and political and social change. The Institute is governed by a twenty-two-member Board of Trustees comprising nominees from the Singapore Government, the National University of Singapore, the various Chambers of Commerce, and professional and civic organizations. A ten-man Executive Committee oversees day-to-day operations; it is chaired by the Director, the Institute's chief academic and administrative officer.

!SEAS Series on Japan and the Asia-Pacific

JAPAN IN EAST ASIA Trading and Investment Strategies Wendy Dobson University of Toronto

I5IR5 INSTITUTE OF SOUTHEAST ASIAN STUDIES

Published by Institute of Southeast Asian Studies Heng Mui Keng Terrace Pasir Panjang Singapore 0511 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Institute of Southeast Asian Studies.

© 1993 Institute of Southeast Asian Studies Cataloguing in Publication Data Dobson, Wendy. Japan in East Asia : trading and investment strategies. (!SEAS series on Japan and the Asia-Pacific) 1. Japan-Commerce-East Asia.

2. East Asia-Commerce-Japan. 3. Japan-Commerce-ASEAN countries. 4. ASEAN countries-Commerce-Japan. 5. Investments, Japanese-East Asia. 6. Investments, Japanese-ASEAN countries. I. Title. II. Series. HF3828 A8D63 1993 sls93-28183 ISBN 981-3016-57-4 ISSN 0218-5474 The responsibility for facts and optmons expressed in this publication rests exclusively with the author, and her interpretations do not necessarily reflect the views or the policy of the Institute or its supporters.

Typeset by Letraprint Printed in Singapore by Prime Packaging Industries Pte Ltd

Dedicated to the memory of Professor K.S. Sandhu Director of !SEAS, 1972-1992

CONTENTS

List of Tables

zx

List of Charts

X

Acknowledgements

I II

INTRODUCTION ECONOMIC ENVIRONMENTS AND FIRM BEHAVIOUR

IV

v

1

Japan United States

9 9 9 12

Firm Behaviour

14

Economic Environments

III

xz

THE STRATEGIC BEHAVIOUR OF GOVERNMENTS Home Government Policies Host Government Policies ASEAN Policies Summary

19 20 21 23 24.

PATTERNS OF TRADE IN EAST ASIA

27

FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWORKS Production Networks Japanese and U.S. Affiliates in Asia Compared Japanese Affiliates' Behaviour Compared among the Regions Survey and Interview Findings Asymmetry in the Auto Industry

37 40 45 47 51 58

VI

VII

TESTING THE DETERMINANTS OF FOREIGN DIRECT INVESTMENT

61

CONCLUSIONS

67

Notes

73

Appendices

75

References

81

About the Author

87

LIST OF TABLES

1 Regional Shares of Stocks of U.S. and Japanese Outward Investment, 1990

1

2 Shares of Product Groups in World Merchandise Trade, 1990

5

3 Gravity Coefficients: The Strength of Trade Ties in Manufacturing among the East Asian Economies, Japan and the United States, 1988

28

4 Structure of Manufactured Exports in East Asian Economies, 1979 and 1988

29

5 Intra-Industry Trade in Manufacturing, 1980, 1985-89

32

6 Comparison of Procurement and Sales Behaviour of Japanese and U.S. Affiliates, 1980 and 1990

46

7 Sales, Inputs and Intra-Group Transactions: A Comparison of Japanese Affiliates in USA, EC, Asia and NIEs, 1990

48

8 MNE Relationships with Local Suppliers in Singapore: A Comparison

54

9 Determinants of Japanese FDI: Results of the Regression Analysis

65

IX

LIST OF CHARTS

1 Direction of Trade, ASEAN and NIEs with the USA, Japan and the Rest of the World, 1980 and 1990 2 Indices of Intra-Industry Trade, Japan and Eight Asian Economies, 1985 and 1989 3 Stocks of Outward FDI from Japan, and USA: Distribution by Economy, 1980 and 1990 4 Industrial Distribution of Stocks of Japanese and U.S. FDI, 1980 and 1990 5 Interdependency Chart for Electrical Apparatus in NIEs and ASEAN Countries

3 33 38 39 42

ACKNOWLEDGEMENTS

The economic dynamism of East Asia is one of the great transforming features of the post-Cold War world. Spontaneous and often unilateral reductions of barriers to both trade and investment by governments in the region have been important determinants of this dynamism, along with inflows since the mid-1980s, of large amounts of Japanese investment relative to the sizes of the host economies. This study examines the growth of Japanese trade with and investment in the region in the light of particular concerns voiced by some North Americans. They recognize the economic potential of the transformation and wish to contribute to, and benefit from, the opportunities afforded by rapid economic growth. But concerns abound about potential implications of differences in business systems and the role of governments in economic development. The purpose of this study is to cast light on, rather than add heat to, these concerns. In carrying out this study, I benefited greatly from the contributions of many people though the observations and conclusions contained herein are my own responsibility. The late Professor Kernial Sandhu, then Director of the Institute for Southeast Asian Studies, offered to host a seminar with scholars from the region, which took place in Singapore on 29-30 September 1992. ISEAS was also my host during the month spent revising the manuscript in response to the suggestions and discussion at that seminar. I am particularly indebted to Kernial Sandhu for his encouragement and enthusiasm about this project, which is dedicated to his memory. His untimely death means the region has lost a great intellectual entrepreneur who will be greatly missed by people all over the world. I am grateful, too, to Dr Lee Tsao Yuan, Deputy Director of the Institute for Policy Studies and Dr Chia Siow Yue, National University of Singapore, for assistance during that visit, and to the staff of the Institute. Financial support for this research was received from the Pacific 2000 Fund of the Canadian Department of External Affairs and the Ontario Centre for International Business. The Canadian International Development Agency made possible the 1992 seminar, as did the Canada-ASEAN Centre in Singapore. Xl

Last, but by no means least, I wish to acknowledge unfailing, high quality research assistance from Mr Robert Tran and Mr Walid Hejazi. Kazuo Noguchi provided access to Japanese-language publications.

I

INTRODUCTION

Amid concerns about slow growth and lagging competitiveness in many western industrial economies, the remarkable dynamism of the East Asian economies has attracted interest in the reasons for, and implications of, their economic success. Because of the diversity of economic structures and relationships in the region there are many different definitions of East Asia. This study applies the term to the four newly industrializing economies (NIEs) - Hong Kong, Korea, Singapore and Taiwan - and the ASEAN-4 - Indonesia, Malaysia, Philippines and Thailand. Although China is fast emerging as a major force, it is not included because of data problems. Recent inflows of Japanese investment to these economies, evident in the regional distributions of accumulated investment in Table 1, have drawn particular attention. Political initiatives to promote closer regional economic ties have evoked concerns that growing Japanese trade and investment implies the formation of a self-sufficient bloc. TABLE 1 Regional Shares of Stocks of U.S. and Japanese Outward Investment, 1990 U.S. Outward FDJ•

Canada North America Europe NIEs and ASEAN-4 Japan ROW T OTAL

16 48 6

Japanese Outward FDJb 2

44 19 14

5 25

21

100

100

See Appendix A for a compariso n of data sources. a SOURCE: U.S. Department of Commerce. Survey of Current Business, Aug. 1991, pp. 104 and 105. b Accumulated outflows, 1951-1990. SOURCE: Ministry of Finance, Dec. 1991. Zaisei Kinyu Geppo, pp. 31-35.

2

JAPAN IN EAST ASIA

What validity is there to these concerns? The region's continued dependence on western export markets is evident in Chart 1 and implies the region is some distance from self-sustaining economic growth. Implicit in such concerns are differences in frameworks, assumptions and interpretations. For example, some analysts assert that characteristics of the Japanese market, which make it difficult for foreigners to penetrate, also play a role in the possible creation of exclusionary production networks in East Asia. 1 Others observe that U.S.-Japanese economic rivalry is being carried into the region. 2 Such assertions build on the U.S. literature on Japanese trade and investment that accentuates its asymmetrical characteristics. Japanese foreign direct investment (FDI) outflows have grown rapidly. By 1989 the stock of outward investment stood at US$ 156 billion and accounted for 11 per cent of the world totaP Yet Japan has not been a major recipient of FDI; its stock of inward FDI in 1989 was US$28 billion. 4 While it has been a large exporter of manufactures, it has not been a large importer of these products. 5 Many non-Japanese complain about the difficulties of selling to Japan. These perceptions are then generalized beyond the Japanese market, particularly to the East Asian economies, as recipients of the recent wave of Japanese Investment. Explanations of these asymmetries, especially by North Americans, generate considerable disagreement. Some analysts draw on comparisons with other western economies at comparable stages of development and on recognition of Japan's lack of natural resources and relative geographic isolation - until its Asian neighbours began to develop. 6 Indeed, Japan's share of overseas production in manufacturing is still only around 6 per cent compared to 17 and 20 per cent for U.S. and German firms, respectively Qulius 1991, p. 8). Clearly, the rapid surge in outward investment since the mid-1980s is a stock adjustment which has a long way yet to go. Others accentuate the uniqueness and "differentness" of Japanese institutions, particularly the role of government and the organizational features of keiretsu - networks of firms linked together by cross-shareholdings and by long-term repetitive relationships between suppliers and customers. There are also the nonarm's-length relationships, and the idea that Japan pursues a different form of capitalism. 7 If keiretsu relationships are being replicated abroad in key industries in w hich firms from many nations are now com-

CHART 1 Direction of Trade, ASEAN and NIEs with USA, Japan and Rest of the World, 1980 and 1990 (Shares, per cent)

Destination of Exports 100

90 80

IExports, NIEs to: I

IExports, ASEAN-4 to: I

69

70

60

•

1980

share (o/o) 50

01990 40

30 20 10 0

USA

Japan

ROW

USA

Japan

ROW

Origin of Imports 100 90 80

IImports to NIEs from: I IImports to ASEAN-4 from: I

70

60

•

1980

share (o/o) 50

01 990 40

30 20 10 0

USA

Japan

ROW

USA

Japan

N OTE: NIEs does not incl ude Taiwan. SOURCE: IMF, Direction of Trade Statistics Yearbook, var ious years.

ROW

4

JAPAN IN EAST ASIA

peting, then U.S. firms stand to lose out on potential growth opportunities, particularly in a rapidly-growing region like East Asia. The view one takes with respect to these questions is important, particularly for third countries in North America and Europe. If one believes Japan, or Japanese firms, are following an exclusionary strategy in East Asia, then seeking to invest in, and trade with, these economies would entail more risk than if economic developments were largely accounted for by the operation of market forces. In effect, one would conclude that being efficient would be insufficient to ensure success. Similarly, the relevance of these debates to the economist is not whether Japan is different or unique, but the economic efficiency of Japanese agents; the impact of their behaviour on global welfare; and the long-run sustainability of their behaviour. Thus, the questions posed in this book are the following: What are the reasons for the recent growth of Japanese trade with, and investment in, the East Asian economies? How do these reasons compare with prevailing theories of trade and investment? With U.S. behaviour? What are the implications of Japanese patterns for third countries and for the future of the systems of international trade and finance? The central analytical issue is: Why and how do firms produce abroad? The essence of FDI is the transfer within the firm of firmand industry-specific assets such as technology, managerial, marketing and organizational knowhow that are necessary to compete with local producers in host countries. Dunning's (1980) "eclectic" paradigm of international production is perhaps the best model to capture these features of FDI behaviour. Dunning identifies three factors which influence a firm's decision to carry out FDI: 1. ownership (0) advantages, such as economies of scale, other technological advantages, or management skills which enable the firm to recover the costs of investing abroad; 2. location (L) factors which contribute to the decision to employ ownership advantages to produce abroad; these factors may be risks and barriers in export markets or the availability of lower-cost labour or natural resources; and 3. internalization (I) factors which determine that foreign production occurs through FDI, that is, within the firm. Firms decide to create

5

INTRODUCTION

an internal market among parent and affiliates or subsidiaries in order to control key sources of competitiveness or to reduce the risks that the firm might lose control of such intangible factors as knowledge and technology. Dunning's work is part of the transactions cost literature which originated with Hymer (1976), Kindleberger (1969), Caves (1982), Rugman (1980), among others. Firms in oligopolistic industries, characterized by knowledge and financial advantages, produce abroad rather than export their products or license their technologies. As we see in Table 2, more than 40 per cent of global merchandise trade takes place in industries with these characteristics. Yet as Komiya (1991, p. 144) argues, between a fifth and a third of Japanese global FDI flows have been accounted for by independent small and medium-sized firms (SMEs) without oligopolistic characteristics. Similarly, much Taiwanese investment in neighbouring economies is undertaken by small and medium-sized enterprises (SMEs) in highly-competitive industries. These realities suggest the need for a less restrictive transactions framework, such as Dunning's, which acknowledges firm- and industry-specific assets associated with economies of scale, as well as locational advantages associated with relative factor scarcities, and home and host government policies. TABLE 2 Shares of Product Groups in World Merchandise Exports, 1990 Product group

Value ($)

Share(%}

Machinery and transport equipment (nes) (of which office machines and telecommunications equipment) (Automotive products)

1,237

35.2

Chemicals Iron & steel

8.5 9.0

298

8.5

109

3.0 46.7

TOTAL SOU RCE:

303 322

GATT, International Trade 1990-91 , Volume II.

6

JAPAN IN EAST ASIA

The role of home and host governments as locational influences on FDI are a major focus of this study. Conceptually, incorporating them into the analysis is motivated by two approaches. The first is the concept of long-run dynamic comparative advantage and Schumpeterian innovation (Yoshitomi 1991), where technical change results from the deliberate efforts of firms. Innovating firms can reap at least temporary monopoly gains from such innovation until imitation by rivals, resulting from intense competition, drains away excess profits. The second approach is that of strategic trade theory, which in contrast, applies industrial organization models of imperfect competition to trade. In this framework, oligopolistic rents accrue to manufacturing firms due to barriers to entry and externalities arising from innovation. Governments have incentives to intervene in ways that shift rents to home firms, or in instances of FDI, to intervene to shift benefits from foreign-owned firms to the domestic economy. Thus, emerging manufacturing industries using advanced technology become strategic sectors in which capital and labour can earn significantly higher returns. The Japanese Government recognized this, says Krugman ( 1986), and used policy tools to nurture industry and shield it from foreign competition. The distinctions between dynamic comparative advantage and the new trade theory are part of an international policy debate about strategic industries - industries which governments identify and promote on the grounds of serving national interests - but for which there are various identifying criteria. One criterion for identifying strategic industries is those in which the long-term availability of products and technologies is considered crucial to the national interest; another criterion is those industries characterized by dynamic increasing returns; a third criterion is those in which there are regional or national externalities which have a significant impact on growth through backward and forward linkages in material and knowledge (Stevens 1991, p. 98). The organization of this study reflects the importance of such strategic factors as determinants of trade and investment. Chapter 2 focuses on the home environments of Japanese firms. Economic performance and policy have been important influences on the development of ownership and internalization advantages of Japanese firms.

INTRODUCTION

7

This chapter illustrates (with some comparisons drawn with the U.S. economy) how, as the economy has restructured and modernized in the post-war period, firm behaviour has evolved too. Chapter 3 is an assessment of the concept of strategic behaviour of governments and firms, what it might look like and how it matters to trade and investment. Chapter 4 focuses on an examination of East Asian trade patterns with Japan and their evolution; while Chapter 5 focuses through firm-level comparisons and surveys - on patterns of FDI, in particular on production networks in the region. Regression analysis is used in Chapter 6 to test for the significance of "strategic" behaviour by governments as a determinant of FDL Conclusions of the study, in the final chapter, can be summarized as follows: • While Japanese public policy is increasingly neutral towards export promotion, it continues to encourage outward FDI; Japan's growing labour shortage and rising incomes in the East Asian economies suggest why this emphasis is unlikely to diminish in future. • Production relationships in manufacturing affiliates differ by industry and home country of foreign investors. U.S. investors in the electrical industry in the region are well-established low-cost suppliers to their home market while Japanese affiliates are recent arrivals and are more heavily oriented to local and export markets. In the transportation industry, production is oriented to local markets and is dominated by the Japanese - more by default than design. • Japanese affiliates in the electrical industry tend to internalize production in networks that are less open than those of their American counterparts; yet in the long term, behaviour of production networks will converge. Fierce global competition in the industry is forcing non-Japanese firms into network patterns to achieve efficiencies comparable to Japanese producers, while local governments are beginning to require more open procurement behaviour which will open up the Japanese networks. • Host governments' policy liberalization towards foreign investment creates comparative advantages in manufacturing activities, but at what cost? Inducements can lead to inefficient operations behind tariff walls; resources allocated to inducement of investment already

8

JAPAN IN EAST ASIA

headed for the region will reduce welfare. Differential inducements create the basis for co-operation and disciplines which governments in the region should now develop.

II

ECONOMIC ENVIRONMENTS AND FIRM BEHAVIOUR

To evaluate and apply the distinctions implied by the introductory discussion, it is important to be aware of certain characteristics of the post-war Japanese economy and to make some comparisons with those in the U.S. economy. For purposes of this study, the comparison is organized around Dunning's framework to address differences in ownership factors including country- and firm-specific characteristics and advantages; differences in internalization and locational advantages, such as the evolution of relative factor endowments, relative costs, barriers to trade and inducements to invest. The areas of contrast include differences in resource endowments and economic frameworks; policy goals and government and industrial responses to rapid economic change and to outward FDI.

Economic Environments

Japan Japan is a small crowded island, poor in natural resources, and - until less than two decades ago - preoccupied with creating employment for a growing labour force and achieving rapid economic modernization and growth in per capita incomes. In the post-war period, in contrast, the United States has been the world's hegemonic leader, supported in no small measure by the power, dynamism and natural resource riches of its huge economy. Where Japan's economic development has involved protection of infant industries, economic planning and close business-government co-operation, economic development in the United States has, like in other Anglo-Saxon countries, emphasized laissezfaire behaviour in an independent private sector, and the private sector as the engine of growth. The public sector provides public goods, corrects market failure and fights concentration of economic power. 9

10

JAPAN IN EAST ASIA

Japan's economic growth in the immediate post-war period was fuelled by procurement requirements for the Korean and Vietnam wars. Its industrial base was provided by labour-intensive light manufacturing and heavy and chemical industries that produced substitutes for imports. 8 Heavy dependence on energy and natural resource imports changed rapidly after the first oil shock, world recession and collapse of the Bretton Woods system of fixed exchange rates in the 1973-74 period. In response to the growt!:! slow.:ao~ and escalation in production costs, both the public and private sectors invested heavily in R&D. Productivity growth was rapid and exports began to grow rapidly, mainly to Asia. In addition, the industrial structure began to evolve towards a heavier emphasis on electrical and other machinery and information technologies. The public sector played an important role in this transformation by co-ordinating national and corporate interests and by encouraging an export orientation. Wage restraint; export support policies including capital controls and credit allocation to exporters; low inflation and an undervalued yen all contributed to export success. It was only when Asian growth slowed down in the early 1980s that Japanese exporters became dependent on U.S. markets and the current account surplus began to pile up. Around the same time, these policies that had contributed to export success became successive targets for policy liberalization and deregulation. These also were factors that contributed to outward direct investment by Japanese firms. The history of Japanese foreign direct investment is much shorter than that of U.S. or European firms. Japanese FDI in the 1950s and 1960s was undertaken by labour-intensive industries, followed by the heavy and chemical industries and resource seeking multinationals. The magnitude of FDI was negligible, however, because of capital controls (Komiya 1990, p. 117). 9 Ninety seventy-two was the watershed year for outward FDI, triggered by the large yen appreciation and its negative impact on domestic production costs, and positive impact on the relative prices of foreign assets (Komiya 1991, p. 118) that occurred at the time of the collapse of the Bretton Woods fixed exchange rate system. This FDI was largely subcontracting-dependent, assemblybased as multinationals transplanted operations to serve the Asian market (Ozawa 1991). The oil shock and resulting slow-down quickly moderated this surge, as did the undervalued yen.

ECONOMIC ENVIRO NMENTS AND FIRM BEHAVIOUR

11

The second major surge of outward direct investment was triggered by yen appreciation in the 1985-88 period and another industrial restructuring urged in the 1986 Maekawa Report which advocated a switch from export-led to domestic-demand-led growth. 10 At that time, the rapidly-growing current account surplus was becoming an increasing focus of U.S.-Japan political friction. After peaking at 4.4 per cent of GDP in 1986, the surplus dropped to 2 per cent by 1989 (before rising again in 1991-92), facilitated by the 30 per cent yen appreciation in real effective terms 11 between 1985 and 1988. Political frictions and the impact on production costs of the largely-unanticipated yen shock on production costs caused corporations to rethink their strategies. Many rapidly globalized their operations, locating production through FDI in the Unites States and Europe to secure market share behind trade barriers, and in the Asian economies to reduce production costs of exports that had climbed at home as a result of the exchange rate realignment. An additional, longer term, influence on corporate production decisions was Japan's growing labour shortage caused by the ageing of its population. Options to respond to this constraint included exporting labour-intensive jobs by investing abroad; substituting capital for labour at home through increased automation; or more intensive use of the existing labour force by extending the retirement age and attracting more women into the lifetime labour force. Japanese industry chose both to automate and to invest abroad. Between 1986 and 1991 Japanese companies invested US$3300 billion in new plants and equipment in Japan; the Japan Development Bank estimates that 30 per cent of private investment in Japan during this period was allocated to increasing production capacity; 15 per cent was allocated to new product development; and R&D absorbed another 10 per cent. Nearly 20 per cent of capital spending is estimated to have been devoted to rationalization and labour saving in the face of rising labour and services shortages (Rowley 1991). This restructuring also led to a vertical division of labour in production in East Asia, illustrated by the W'hite Paper on International Trade 1992 (MITI 1992): Through the 1980s in the work-sharing structure between Japan and Asia, the work-sharing in manufacturing process mainly in

12

JAPAN IN EAST ASIA

machinery sector has developed rapidly with foreign companies (mainly Japanese companies) advancing into the area. Capital goods supplied from Japan are much used in manufacturing process in Asia and Japanese companies support adjusting the foundation of industries and elevating the level of technology through technology transfer and import. On the other side, the work-sharing by vertical product differentiation between Japan and Asia has developed in recent years in the cause of increasing demand for high-value added products by rising income in Asian NIEs. Moreover at the present, parts imported from Asia are being used in the manufacturing process in Japan, and the work-sharing between Japan and Asia has been rapidly intensified. This view of the world is not entirely shared by others in Asia; while the division of labour is accepted as necessary to industrialization (UNESCAP 1991), some analysts (Narongchai and Suthiphand 1989) have emphasized both the desire and the determination to move beyond the complementarity implicit in the vertical division of labour: ... Actually trade relations between Asian NICs and ASEAN are still limited. But it is expected that intra-industry specialization led by Japan will be able to enhance trade among them as the degree of interdependence will become larger. The issue is that all parties concerned should find ways and different alternatives in order to co-operate to the maximum benefit.

United States U.S. economic circumstances and environment have been rather different. For much of the post-war period, the conduct of macroeconomic policy tended to give greater weight to the goal of high employment than to price stability. The 1970s oil shocks were resisted rather than accepted; lower-than-world domestic prices were adopted for a period. Accelerating inflation and balance of payment crises contributed to the collapse of the Bretton-Woods system in the early 1970s and caused a dollar crisis again in 1978. 12 In 1979, the conduct of U.S. monetary policy shifted abruptly to fight inflation with high interest rates. The Reagan Administration, elected in 1980, then cut taxes and increased defence spending without an accompany ing cut in other government expenditures. The economy

ECONOMIC ENVtRONMENTS AND FIRM BEHAV IOU R

13

subsequently boomed but ballooning twin deficits became a policy problem that has persisted into the 1990s. One deficit was a growing fiscal deficit which, because of low U.S. domestic savings rates, had to be financed by attracting foreign savings with high real interest rates. Its twin was a large current account deficit. The inflow of foreign capital pushed up the U.S. dollar to a peak in early 1985. During the preceding years, because imports were relatively cheaper and exports expensive, trade and current account deficits opened up. Japan, in contrast, maintained high savings rates during this period, and pursued a course of medium-term fiscal consolidation. Accumulating current account surpluses was a predictable consequence of the divergence in economic performance. Despite the obvious role in the external deficit played by macroeconomic divergence, concerns were also growing about the microeconomic causes of U.S. competitiveness problems. A number of attempts were made to diagnose these causes. One of the most widelyrespected analyses was an investigation of U.S. industrial performance undertaken by social scientists at the Massachusetts Institute of Technology (Dertouzos et al. 1989). This study defined six problems that helped to account for declining productivity and prescribed corrective actions to be taken by government, the private sector and educational institutions. The list of problems illustrates some of the key areas of difference between Japanese and U.S. decision-making and decisionmaking environments, including the use of outdated strategies by U.S. firms, short-time horizons in corporate decision-making, lagging investment in new technology, neglect of human resources, lack of co-operative attitudes and poor business-government relations. The history of U.S. FDI also differs from Japan's. While U.S. firms invested abroad in the 1960s, the destination was Europe, in order to be present in the large market anticipated as a result of initiatives to form the European Community. At that time the main sending countries - the United States, the UK, and Japan - attempted to bring some coherence to their policies towards outward FDI. Their main concerns related to the home country's policy environment towards outward investment, that is, whether tax policies favoured or inhibited outward investment; the balance of payments and employment consequences of foreign production; and concerns about possible unintended transfer from the home country of some of the technological

14

JAPAN IN EAST ASIA

benefits. Dunning (1991, p. 232) points out that, with the exception of the Japanese Government which perceived the benefits of outward investment for upgrading the domestic economy, rarely was outward direct investment perceived to be complementary to domestic investment. In the 1980s, internationalization of economic transactions and development of MNE production strategies that integrated both foreign and domestic operations, forced governments to change their attitudes towards FDI and to reconsider how their macroeconomic policies affect, and are affected by, FDI.

Firm Behaviour The discussion to this point has focused on country-specific factors that have affected ownership and internalization advantages of Japanese and American firms. Domestic policies, cultural and historical factors are bound to determine firms' behaviour both at home and when production is transferred abroad. As competition between Japanese and American firms intensified in major industries, such as autos and electronics, international debate has increasingly focused on differences in firm-level characteristics as well. Of particular relevance to this study are three relationships - between companies and shareholders, among companies, and between companies and their employees - which characterize Japanese firms. 13 Many major Japanese firms are part of keiretsu. First, Japanese firms are afforded the advantages of patient equity capital that is not widely traded, since their main shareholders are other companies in the same group. In contrast, American firms' shares are widely-held, often by large institutional investors. Share prices are strongly determined by quarterly profits reports. The heaviest pressures, therefore, are felt by American managers during downturns when they must cut costs to stay in the black. Investment plans are one of the victims of cost-cutting pressures, as are employees. In contrast, many Japanese firms do not face the same sources of short-term pressures. Second, keiretsu production relationships, which apply to most segments of the value chain, mean that suppliers and customers become very familiar with each other's requirements and are an important factor in achieving high levels of quality required to reduce inventories,

ECONOMIC ENV IRO NMENTS AND FIRM BEHAVIOU R

15

thereby reducing transactions and production costs. American firms, in contrast, are independent companies which forge and sever ties, including the selection of suppliers, in a flexible fashion on an arm's length basis of open bidding. These differences contribute to criticisms that the Japanese system is exclusionary. In addition, when loyalty and familiarity of close ties dominate price and quality as business criteria, the system is also seen to be inefficient. Some analysts see other strengths of the Japanese system as critical to firms' competitive advantages. Porter ( 1990, p. 407) believes keiretsu are overrated as the source of competitive advantage. Instead, he stresses the importance of supporting industries, of emphasis on backward and forward integration and networks of small- and medium-sized suppliers and sub-contractors, known as shita-uke, through which Japanese firms work closely together. He argues the unique aspect of Japanese behaviour is to work co-operatively in groups such as these, and to use trade associations and trading houses as information collectors and disseminators as important sources of competitive advantage. Womack et al. ( 1990) found that these forms of inter-company relationships in the auto industry created major ownership advantages for firms, regardless of nationality, and play important roles in the development of process innovations such as economies of scope and "lean" (as opposed to mass) production which are sources of future competitive advantage in the industry. Third, the system of lifetime employment in large companies means that employees expect guaranteed employment until retirement, and regular salary increments as they gain experience. This system allows for investment in human capital development through regular training on the job. These investments, and involvement of employees in decision-making, contribute to low employee turnover and oftenintense identification with the firm's business objectives. These characteristics have also been achieved in Japanese affiliates located in the United States where employees are Americans. But the system can lead to slow career advancement which non-Japanese employees often find unacceptable. In contrast, the American employment system tends to compensate employees by the type of job; employees wishing to advance quickly expect to change jobs several times during their lifetime. Some large companies, such as IBM, have committed themselves to

16

JAPAN IN EAST ASIA

lifetime employment systems, but such a commitment has been the exception rather than the rule. One attribute of the Japanese market that is often misunderstood, is that, despite the emphasis on group relationships and information sharing, domestic competition is usually fierce. The consequences for firm behaviour are described in Morita (1992): Since the end of the war ... Japanese manufacturers concentrated their resources on developing technologies and products, and increasing productivity and quality control with the objective of catching up to, and surpassing European and American industries . . . . Japanese companies were successful in overseas markets because they could supply consumers with products that fit their needs exactly .... While Japan has a free economic system, as do Europe and America, it seems that the price of a Japanese-made product tends to be lower than the prices of comparable products manufactured by European and American companies. Why ... ? ... fierce competition in the Japanese market has given rise to a unique priceoriented approach by Japanese companies, which is clearly not employed by European and American companies. A Japanese company first sets a price that aims to gain market share, and then it tries to cut costs and profits according to the price ... . To European and American companies that are forced to confront these unfamiliar Japanese practices in their home markets, Japanese corporate behaviour seems like an invasion or a strangulation. This is where the problem lies.

This discussion highlights a number of characteristics on which Japanese and American firms differ and which should be expected to characterize their production abroad. With their managerial advantages in key industries such as autos and electronics, it is not surprising that, as relative costs have changed, instead of vacating those segments in which they have lost comparative advantage, Japanese firms have moved offshore those market segments. Outward FDI from Japan averaged as much as 4 per cent of private gross fixed capital formation during the 1983-89 period, rising to 6.3 per cent in 1990 (OECD 1991, p. 64). By 1992, MIT! was urging offshore production by electronics firms of all low-technology production "to avoid excessive increases in exports when demand in Japan slumps" (Nikkei Weekly 1992).

ECONOMIC ENVIRONMENTS AND FIRM BEHAVIOUR

17

This discussion has also brought out some of the now well-known differences between corporate governance and business systems that are likely to influence how these firms behave as investors in the East Asian economies. Government policies can also be an important determinant of FDI; these are factors that have received little attention so far. They are the subject of Chapter 3.

III

STRATEGIC BEHAVIOUR OF GOVERNMENTS

In this chapter some of the major characteristics of home and host government behaviour that are considered to be "strategic" are examined. It is to be expected that such policies will influence FDI through the impact of home country policies on ownership and internalization advantages and the impact of host government policies on location advantages. Japanese and host government FDI-related policies are then summarized. The resource and welfare implications of such policies are also examined. Strategic behaviour refers to policies to promote economic activity in certain industries because of the impact of those industries on longterm growth of incomes and employment. By fostering backward linkages to the local economy and forward linkages to markets for goods and services, and by promoting knowlege and skills accumulation, governments can "create" comparative advantages independent of traditional factor endowments. Home government policies include trade protection and support policies that target the growth of certain industries; host government policies induce foreign investment in certain industries to catch up to industrialized economies and to create comparative advantage. Teece ( 1991) summarizes what is known about the benefits and costs of home government support policies: i.e., subsidies and protective measures targeted at certain industries. From the available evidence, he makes several observations. First, targeting has a better chance of success when it is used for the purpose of "catching up" than for advancing the technological frontier, because of the greater efficiency with which resources can be used when the way forward is known. Second, targeting works better when governments can co-ordinate policies to ensure all are supportive of the economic objective. He notes that, with the possible exception of Japan, government guidance is less and less used in industralized countries today. Firm-level studies suggest that firms' ownership advantages, such as organization and management, are more important to competitiveness than are public sector 19

20

JAPAN IN EAST ASIA

support policies. Linkages, for example, are better achieved using the market than government guidance. These observations raise the question of the opportunity cost of allocating public resources to such uses. The benefits may be visible, but the alternative uses are usually invisible. Third, measures that have been effective have tended to be those that support basic and generic research. There is a better case for supporting R&D and other forms of information and information collection because they are public goods.

Home Government Policies Has Japanese Government behaviour been strategic? In the broad sense of promoting the overall economic objective of modernizing to catch up with the west, many authors, such as those addressing the strategic trade policy issue in Krugman ( 1986), answer the question in the affirmative. More recently, however, as policies have been liberalized and deregulation has occurred, the industrial response has been more consistent with dynamic comparative advantage. In 1986, the Maekawa Report advocated what amounted to a massive reorientation of economic policy necessitating industrial restructuring; even so, exporters were unprepared for the magnitude of the rise in the yen in the 1985-87. Before the yen subsequently stabilized in the range of ¥125- 150 per U.S. dollar, exporters strongly protested the abrupt changes in relative prices that occurred. Industry did adapt, however, aided by easy monetary policy and low real interest rates in the late 1980s. By 1991, it was becoming apparent that the domestic investment boom had over-reached as evidence began to appear in some areas of manufacturing, that returns to invesment had fallen below financing costs (Rowley 1991). The area where the Japanese Government is widely regarded to behave strategically, however, is in the geographic targeting and commercial orientation of its official development assistance (ODA). Cronin (1992, p. 39) asserts that Japanese ODA plays a key role in Asia-Pacific economic development and that Japanese public agencies have plans to rationalize and co-ordinate economic development in the region. East Asian economies have historically been major recipients of Japanese ODA, receiving a 30-40 per cent share for much of the 1980s. The New AID (Asian Industrial Development) Plan,

STRATEGIC BEHAVIOUR OF GOVERNMENTS

21

introduced in 1989, established aid criteria such as the identification of possibilities for horizontal industrial diversification and infrastructure development such as roads, ports, power plants and assistance with upgrading technical and managerial standards. Japanese firms tended to regard the existence of ODA in a country as the home government's "seal of approval" and invested in ways that took advantage of the availability of aid money (Orr 1991). Bilateral ODA flows to Southeast Asia rose 200 per cent between 1980 and 1990, to US$2.3 billion from US$0.8 billion, with over 70 per cent in the form of yen loans (Ministry of Foreign Affairs, various issues). Governmentbusiness partnerships are an important part of the aid programme; the Overseas Economic Cooperation Fund (OECF), for example, established a joint venture with business in 1981 to create the JapanASEAN Investment Company. In 1989 it established the Japan International Development Organization in co-operation with the keidanren (federation of Japanese employers, representing the interests of large businesses in Japan) to provide seed money for development proJects. Large numbers of organizations are involved in what appears to be a co-ordinated and commercially-oriented effort. They include the OECF, which provides yen loans; the Export Import Bank, which provides loan to private firms and enterprises; the Japan International Cooperation Agency QICA) which provides technical assistance; the Japan External Trade Organization QETRO) which promotes trade and investment; and the Association for Overseas Technical Scholarships which send trainees and students to Japan. As many as 20,000 East Asian students currently study in Japan under the auspices of this Association. While there are pressures to decommercialize Japanese ODA, Cronin believes rivalries among these agencies will prevent such a change in behaviour. Host Government Policies Around the time of the yen shock and the Maekawa Report in Japan, major changes were also underway in policies, particularly in the ASEAN economies, in response to recession and dropping commodity prices. Wide-ranging liberalization occurred, to ease restrictions on

22

JAPAN IN EAST ASIA

FDI, attract the necessary capital to diversify their industrial structures, and promote rapid industrialization and employment growth. 14 As the new wave of Japanese investment began to materialize, host governments adopted two kinds of policies: investment-related trade policies and strategic investment policies (UNCTC 1992a, p. 257) . Trade policies had long been used as instruments to induce direct investment as a substitute for exports and they continue to be used in the auto industry. For example, in Thailand, import duties on autos averaged 180-300 per cent until they were slashed in July 1991 to 60-100 per cent (Oishi 1992). Indonesia also imposes prohibitive duties on foreign auto imports. Increasingly, however, free trade zones (FTZs) have become widely used, in conjunction with incentives and performance requirements, to influence the composition of direct investment and to capture jobs and incomes benefits in the local economy. Investment-related trade measures that have been introduced in Malaysia, Thailand, Taiwan and the Philippines have emphasized export processing zones (EPZ) and free trade zones (FTZ) where foreign firms receive duty free imports and financial incentives if they export all or most of their production. Hong Kong and Singapore, of course, have been duty free ports for many years. More recently, there has been a tendency to link trade measures and performance requirements in order to achieve certain policy targets. Malaysia, for example, has developed a complex set of inducements whereby affiliates in FTZs who meet export requirements are exempted from local content or joint venture requirements. Thailand has plans to develop a data processing zone (DPZ) to permit free passage of materials and data, as well as liberal rules for employing foreign personnel by multinationals which locate in Bangkok (Lim and Pang 1991). Strategic investment policies have also been widely used. Foreign investors in Taiwan are elegible for tax breaks if they export most of their production and/or meet local content requirements (Lim and Pang 1991 ). Malaysia has targeted such industries as agriculture and forestry and their associated processing industries, chemicals and pharmaceuticals, textiles, electrical parts and components where trade-offs can be made between local ownership and content requirements and export targets (Lim and Pang 1991 ). Singapore uses financial incentives to encourage multinationals to use the island as their regional headquarters for their management, information and R&D activities.

STRATEGIC BEHAVIOUR O F GOVERNMENTS

23

Although leading American multinationals have based regional headquarters in Singapore for some time, Singapore's 1986 legislative changes to induce such activities have caught the _attention of Japanese multinationals. Singapore has developed financial incentives to encourage multinationals to base their international procurement operations (IPO) there, and through its Local Industry Upgrading Program (LIUP), provides organizational support and financial incentives to encourage the electronics industry to work with local suppliers in upgrading their technological capabilities and product quality. In summary, trade measures and strategic investment policies have been liberalized by host economies during the 1980s and further liberalization can be expected in future. In some cases, such as in Thailand, the government allows relatively few tax incentives and questions their cost effectiveness. In Indonesia, investment incentives were abolished in 1985 and low interest export credits and other measures substituted. Korea targets its incentives to firms which contribute significantly to balance of payments and technology transfer (Healey 1991). Even these measures can be expected to be scrutinized in future as both trade and strategic policies become targets for elimination in negotiations at the GATT multilateral trade talks.

ASEAN Policies In addition to host government measures, a number of ASEAN-wide policies have been developed. ASEAN industrial co-operation, pursued during the past two decades, is based on several premises: individually, ASEAN economies (i.e., Brunei, Indonesia, Malaysia, the Philippines, Singapore, and Thailand) suffer from small market size, whereas the combined market was 325 million people by mid-1991 (Asian Development Bank 1992, p. 313) with rapidly increasing purchasing power; industrial co-operation will increase the region's economic potential and enable each economy to specialize in production of components of manufactured goods, particularly autos. Industrial co-operation fostered by ASEAN has taken several forms: ASEAN Industrial Projects (AlP) initiated by governments which ran into implementation problems early on; ASEAN Industrial Complementation (AIC), which aimed to allocate production of selected

24

JAPAN IN EAST ASIA

auto components that could be assembled into an "ASEAN car" which also has not been implemented; and ASEAN Industrial Joint Ventures (AIJV) which is more flexible and can claim to have given life to six joint ventures (Chee 1988). Obstacles to co-operation have included problems of identifying projects, approval problems and reluctance of member countries to share their markets (Chee 1988). Even so, the AIC concept has interested Japanese auto manufacturers. Mitsubishi Motor Corporation (MMC), for example, proposed in 1987 a scheme to produce parts of a Mitsubishi car in various countries on condition import duties were relaxed. Other assemblers followed suit with what has come to be known as Brand to Brand Complementation. Under this scheme, affiliates agree to observe rules of origin in return for which governments grant tariff preferences. The private sector then allocates production among the participating economies. Originally the scheme was intended to fully implemented by 1995, in order to meet projected demand of 600 thousand units per year (Chiasakul and Yoshida 1990), but Indonesia has not yet become a participant (Pangestu 1992) which effectively denies many of the potential benefits of economies of scale.

Summary The combination of home and host government policies outlined in this chapter suggests that Japanese FDI is subject to strategic influences, by both home and host governments. Supports for FDI such as information gathering or manpower training - measures that are generally available - can improve efficiency with which FDI is allocated. Evidence with respect to Japanese ODA indicates it has been targeted geographically and to commercial support activities that reduce the private cost to firms of locating production abroad. One implication is that such assistance makes it possible for Japanese firms to overinvest in production capacity in order to disadvantage their competitors. On the other hand, perceptions of bias in the procurement of Japanese goods may themselves be biased, since if Japanese products are price competitive, they should be purchased anyway, assuming perfect information about alternative sources of supply. It is the latter assumption, however, that may not be valid.

STRATEGIC BEHAVIOUR OF GOVERNMENTS

25

Host government policies towards FDI have evolved in their orientation from import substitution to export promotion. Extensive use is made of investment-related trade measures, particularly export processing zones, as well as such strategic measures as incentives and performance requirements. The evidence suggests these incentives have reduced the costs to firms that have decided to produce in the region, no matter what their home country. But they are not without social cost since they have beggar-thy-neighbour effects of luring investment to one economy at the expense of its neighbour. When economic growth is rapid, however, these negative effects are not strong enough to create political frictions. In the short term, these measures accelerate job creation, income growth, and technology transfer. In the longer term, however, they can be expected to have some perverse effects. Restrictive local content requirements, for example, can undermine the market by forcing the use of low quality inefficient locally-produced inputs. These measures induce MNEs to regionalize, rather than localize, operations in ways that avoid local content requirements and achieve economies of scale and required quality. Furthermore, production will have to be rationalized as trade barriers are reduced in the brand complementation scheme or ASEAN Free Trade Area (AFTA). When this rationalization occurs, leading to increased specialization in particular products in different economies, it will cause less disruption if it occurs in a fast- rather than slow-growth environment.

IV

PATTERNS OF TRADE IN EAST ASIA

The policy orientations of East Asian governments, discussed in the previous chapter, are based on assumptions that recognize important linkages between trade and FDI. By encouraging foreign firms' investment in key industries such as autos and electronics, comparative advantages in industrial production have been created. With these industries, opportunities for product differentiation have also grown, contributing, in turn, to the growth of intra-industry trade and the horizontal division of labour. In this chapter, patterns of inter-industry and intra-industry trade are first examined and the reasons for intraindustry trade (IIT) are then studied. Determinants of IIT include converging income differentials, economic structures and geographic proximity. FDI should also be a significant explanatory variable because of its role in the evolution of domestic industrial structures beyond reliance on natural resource and labour endowments. Beginning with inter-industry trade, also referred to as the "vertical division of labour", trade patterns are established by differences in endowments of land, labour, capital and technology, and by countries specializing in things they do well. A simple gravity analysis in Table 3 of the intensity of bilateral trade ties in 1988 suggests that Japanese and U.S. trade with the East Asian economies is still characterized by inter-industry trade. Gravity coefficients compare the bilateral export shares of each pair of economies with the export shares they have with all the others in order to measure the relative strength of each bilateral trade tie. 15 The close ties between Singapore and Malaysia are evident in coefficients of greater than 5; ties between Singapore and Hong Kong and their immediate neighbours come next with values greater than 2. Japan's strongest ties are with Indonesia (because of raw materials) and the United States; while U.S. ties are strongest with Taiwan and Korea (because of exports by the latter of capital and labour-intensive consumer goods, respectively). Beyond that, coefficients around 1 indicate bilateral trade ties no stronger than the average, while coefficients less than 1 indicate low levels of integration. 27

TABLE 3 Gravity Coefficients: The Strength of Trade Ties in Manufacturing among the East Asian Economies, Japan and the United States, 1988 Japan Japan Hong Kong Indonesia Korea Malaysia Philippines Singapore Thailand Taiwan USA

0.78 2.30 1.74 0.50 0.94 0.55 1.23 1.00 2.54

Hong Kong

Indonesia

Korea

Malaysia

Philippines

Singapore

Thailand

Taiwan

USA

0.91

1.25 1.92

1.21 0.79 0.94

0.75 0.61 1.35 0.40

0.73 1.69 1.53 0.81

0.90 0.98 1.73 0.56 5.05 0.00

1.38 0.77 0.82 0.52 0.94 2.33 2.56

1.23 0.97 0.66 0.28 0.30 0.09 0.82 0.00

1.25 1.27 0.49 1.17 0.92 0.00 0.99 1.00 1.24

1.54 1.10 1.11 1.02 1.07 1.21 1.63 0.97

0.68 0.72 0.45 0.00 0.47 0.83 1.02

SOURCE: Author's calculations from UN trade tapes.

0.33 0.35 0.68 0.27 0.55 2.35

0.49 1.09 5.84 1.46 0.63 1.37

1.87 0.65 1.05 1.38

2.75 0.73 1.60

0.00 1.11

2.26

TABLE 4 Structure of Manufactured Exports in East Asian Economies, 1979 and 1988 (Shares, percentage, of SITC categories 5-8) A. Japan

36 51

1988 2 6 50 42

TOTAL

100

100

B. N/Es Natural resource-intensive products Unskilled labour-intensive products Technology-intensive products Human capital-intensive products

Hong Kong 1979 1988 1 1 65 51 10 24 24 23

South Korea 1979 1988 6 2 51 41 16 24 27 33

Singapore• 1979 1988 8 5 20 12 49 63 23 20

Taiwan 1979 1988 7 5 51 43 26 33 17 20

TOTAL

100

100

100

100

C. ASEAN Natural resource-intensive products Unskilled labour-intensive products Technology-intensive products Human capital-intensive products

Indonesia 1979 1988 56 51 15 31 20 6 9 12

Malaysia 1979 1988 43 10 11 17 39 56 7 17

Philippines 1979 1988 28 21 47 39 13 34 11 5

Thailand 1979 1988 41 12 38 45 13 28 8 15

TOTAL

100

100

100

100

Natural resource-intensive products Unskilled labour-intensive products Technology-intensive products Human capital-intensive products

a

1979 2 11

100

100

100

100

100

100

100

100

Re-exports are included.

SOURCE: The data in this table are drawn from UN trade data, SITC (Rev. 2) at the 3-digit level. The four product groups listed here are classified by manufactured commodities based on their factor intensity following Krause (1987). See Fukusaku (1992), pp. 18 and 48-52.

30

JAPAN IN EAST ASIA

Table 4, which summarizes an analysis of the factor intensity of exports in the region, illustrates the operation of dynamic comparative advantage during the 1979-88 period. Only ASEAN-4 exports were natural resource-intensive in 1979, and these shares dropped sharply during the 1980s as secondary and tertiary activities expanded. The NIEs emphasized unskilled labour-intensive exports (such as textiles, clothing and furniture) though, again, these shares dropped sharply during the 1980s. Japan's exports increasingly concentrated into the technology-intensive products such as chemicals, machinery, electronics, office and other sophisticated kinds of equipment, and autos. By 1988, the NIEs were beginning to evolve according to this pattern. Hidden within these aggregates are important linkages between FDI and FDI-related trade. Multinationals use their affiliates to distribute their products in host economies and to purchase capital and intermediate goods required to produce for export markets or for the local economy. Chee ( 1992) surveyed several empirical studies that also conclude that multinationals in South East Asian economies export a higher proportion of their output than do their local counterparts. But this behaviour is not unique to Asian affiliates: Julius {1990) traced FDI-related trade among multinationals (foreign-owned firms - FOFs) in the major industrialized economies and found they are more active international traders than their domestic counterparts. In addition, the modern production technologies in autos and electronics that use extensive supplier relationships make it advantageous for affiliates to specialize in differentiated products that are exchanged in intra-industry trade. Gains from trade, therefore, can accrue to countries with similar factor endowments and industrial structures. Fukasaku {1992) identifies two types of liT. The first is between producers at different stages of the product cycle; firms in home countries have the more advanced technology associated with early stages of the product cycle. They invest in affiliates and joint ventures in host economies at the later stage of more mature products and standard technologies. Chi (1991) traced a surge in Taiwanese exports of capital and intermediate goods that succeeded outflows of Taiwanese FDI in the 1980s that could be accounted for by firms exploiting the product cycle in this way. The second type occurs among globalizing manufacturers, where parent firms invest in affiliates to produce components

PATTERNS OF TRADE IN EAST ASIA

31

and parts at the most cost-efficient location, which are then exchaged among affiliates in other economies and with the parent. 16 liT is the dominant feature of trade in major industries in Europe and North America. Japan, in comparison, has engaged much less in liT than the other industrialized economies. Some analysts such as Lincoln (1990) have argued the reason is that obstacles exist in Japan's economic structure. Yamazawa and Hirata (forthcoming) have documented a temporary increase in Japan's manufactured imports from the Asian economies since the mid-1980s. However, the downturn in Japanese domestic demand in the early 1990s and fierce price competition in the Japanese market have since driven many of these suppliers out of business. While liT is an important conceptual possibility, measuring it is subject to debate. Trade data are imprecise with respect to the actual comparability of products measured in SITC classifications. For example, the more precise and disaggregated the product category, the smaller will be the liT index; indeed, at the extreme, it is possible to envisage distinctive products, or even brands, which if they were measured at a sufficiently-disaggregated SITC level, say 7-digit, would bring the liT index close to zero. This issue is discussed further in Appendix B. One would expect liT patterns in East Asia to have two characteristics: first, liT indices will begin at low levels (with the exceptions of Singapore and Hong Kong which have large shares of entrepot trade) reflecting the vertical division of labour arising from differing endowments and, in manufacturing, different stages in product cycles. With the growth of American and Japanese FDI in manufacturing in the 1970s and 1980s, industrial structures began to diversify, but also to converge, leading to rising liT indices. Second, one would expect to see liT in finished products such as, for example, in the electrical industry, reflecting the exchange of capital and intermediate goods from Japan for lower-end products such as electric fans, refrigerators, air conditioners produced in the East Asian economies. Thus, liT indices should be higher in finished products (which are differentiated) and low in parts (which are distinct intermediate products with few close substitutes); in other words, liT at two-digit SITC levels should be higher than at higher levels of disaggregation. Aggregate liT indices at two-digit levels for selected years in the

32

JAPAN IN EAST ASIA

1980-89 interval appear in Table 5. These indices measure the liT index through time between each economy and all the others on the list. Japan engaged in much lower levels of liT with the others in the table than the other industrialized countries or Hong Kong and Singapore, where entrepot trade has a distorting influence. Japan's liT has been increasing through time, but more slowly than Taiwan's liT, for example. This finding suggests that geographic proximity to Hong Kong and the fast-growing Southeast Asian economies plays an important role in the growth of liT in the East Asian economies. In comparison, however, rates of increase in liT have been much faster in the least industrialized economies, such as Malaysia, Indonesia and the Philippines, possibly reflecting the role of FDI in development of their manufacturing sectors. An industry-by-industry comparison is made in Chart 2. Here, liT indices measure bilateral trade in these SITC categories between Japan and each of the other Asian economies in 1985 and 1989. As the peaks in the charts in the left-hand panel confirm, high levels of liT already existed in 1985 in the four NIEs. In these economies and in Malaysia and Thailand, liT indices increased dramatically in this period in TABLE 5 Intra-Industry Trade in Manufacturing, 1980, 1985-89 Economy

1989

1988

1987

1986

1985

1980

Japan Hong Kong Indonesia Korea Malaysia Philippines Singapore Thailand Taiwan

40.9 56.2 27.3 49.3 64.0 46.6 72.1 45.1 54.1 64.1 63.5 71.4

43.1 54.8 24.9 40.8 64.4 33.4 70.5 43.0 51.8 62.0 65.3 67.7

33.8 55.7 18.0 48.7 67.6 50.0 69.8 38.8 44.6 57.3 66.2 66.9

29.6 53.5 20.0 49.0 65.9 33.9 72.2 42.5 43.0 56.2 67.6 68.1

32.1 53.2 19.2 48.5 60.1 30.9 73.0 37.5 43.0 60.4 69.9 68.0

36.9 56.7 11.1 46.5 34.4 18.3 63.2 32.1 38.6 65.1 64.1 65.3

USA Canada Europe*

* France, Germany, Italy, Britain and the N etherlands only. Author's calculations from UN trade tapes.

SOURCE :

CHART 2 Indices of Intra-Industry Trade, Japan and Eight Asian Economies, 1985 and 1989 SITC 75: OFFICE MACHINES

SITC

.

j~

.800

I

.800 .400

Produc t

/ ~1

1.000

~·.-.

\

.

1

\

75 01nce mach. & ADP equip. 751 01nce machines

\

\

.;

I ·.. 1./ ·. 1

752 ADP machines 759 Parts

·

~ II . ·762. : . ..-75 g .000 '-=--'----"..:..-='---~~ .200

.

2

5

4

6

7

SITC 76: TELECOM & RADIO EQUIPMENT

1.000 .800 76 Telecom & recording

.600

'

.400

762 Radio receivers

.200

.000

'--'----~-'1-----

3

4

6

7

SITC 77: ELECTRICAL MACHINES

1.000

1.000

.800

.600

n

.600

Elecuical mach. app & appt.

712 SwitChes, relays, !uses.... n22 Printed circl.its 776 cathOde valves

.400 .200 .000

2

4

5

7

6

SITC 7B: ROAD VEHICLES

1.000

1.000

.BOO

.800

.600

.600

.400

.400

78 Road vehicles 78" Paris & accessories 7849 Other par1s & accessories

.200

.200

.000

.000 =~:......_-~~=""'--'-~ 2 4 6

2

4

6

1.000

f'-1

.BOO

.600

.600

.600

.400

.400

.200

.200

.ooo

I

' - \

I

\

so Cherricals 51 Organic

5" Pharmaceuticals

.I . -54

\..---

.000 3

Hong Kong

Korae

lrdonesla

7 Philippines Malaysia

2

Thailand

Slroapore

Taiwan

Hong Kong

Korea

InDOnesia

Philippines Malaysll

Thailand

Singapore

Taiwan

34

JAPAN IN EAST ASIA

office machinery (SITC 75) and chemicals (50); smaller increases occurred in telecommunications (76) and electrical machinery (77). No change occurred in road vehicles (78). Indonesia and the Philippines stand out as the two economies in which liT with Japan grew very little. In Indonesia liT in auto parts (SITC 784) declined during this period. Why these changes? Increasing liT in SITC 75-77 categories reflects increased exchange of products between Japan and the Asian economies. Greater disaggregation would be necessary to confirm exchange of intermediate and capital goods for finished products, and to cast light on how much of this activity is carried out within firms. The low liT indices for road vehicles reflect the import substitution stage of the development of that industry in most economies, with only Korea achieving significant levels of trade with Japan by 1989. In chemicals (50) the high liT indices reflect growing exchange of products between Japan and the Asian economies; growing similarity of products as, for example, the NIEs develop pharmaceutical industries; and possibly the activity of foreign-owned multinationals in all economies. Hasen {1992) points out in his study ofJapanese liT that by 1989, chemicals accounted for 27 per cent of cumulative investment in manufacturing in Japan by foreigners and 13 per cent of Japan's cumulative investment abroad. Intra-firm trade among multinationals which carried out this investment probably contributed to high levels of liT in chemicals that he found. An examination of determinants of liT (the theory is summarized in Appendix B) follows Balassa (1986) and Fukasaku (1992). Differences in levels of economic development are measured by relative per capita incomes, YPC, reflecting differential demand. The coefficient should be positive in sign; the greater the similarity in per capita income between economies, the more liT should occur. Relative factor endowments, RFE, measures similarities between economic structures. The sign of this coefficient should also be positive; the greater the similarities in factor endowments (and therefore production structures) the more product differentiation and liT could be expected. A third explanatory variable is geographic proximity, DIS, between major ports (measured in logs); this coefficient should have a negative sign. Several dummy variables are also used in this analysis to reflect special trade relationships, such as the existence of the ASEAN (D1)

PATTERNS OF TRADE IN EAST ASIA

35

and Canada-U.S. (D2) preferential trading arrangements, entrepot trade in Hong Kong (D3) and Singapore (D4). This study of determinants of liT is confined to liT in manufacturing, SITC (Rev.2). The dependent variable is the bilateral liT ratio between each economy and each other economy for each 2-digit SITC category during the years 1985-89, inclusive. Using OLS methodology, these data are pooled in a cross-section time-series of 550 observations for the eight Asian economies, Japan, the United States, and Canada. The regression model is described in Appendix B. The results of the regression are as follows: liT

=

0.874** - 0.182 YPC** - 0.004 RFE** - 0.062 LDIS** 11.47 - 6.63 - 4.72 - 6.06

+ 0.034 D1 * + 0.045 D2* + 0.064 D3** + 0.092 D4** 1.92 1.81 2.52 3.15 Adjusted R-square Number of observations

= =

0.245; 550;

** coefficient is significant at 1 per cent level of confidence; * significant at the 5 per cent level. These results confirm most of the hypothesized relationships. liT is more likely to occur between economies with convergent living standards, as measured by per capita income differentials, and with convergent patterns of factor endowments. liT is also more likely to occur among economies situated close to each other, among economies with high levels of entrepot trade, and among economies linked by trade agreements, although the Canada-U.S. and ASEAN dummy variables are not highly significant. The significance of D2, the Canada-U.S. dummy, suggests two possibilities: the first is that the globalization of North American firms plays a role in trade within the East Asian region; the second possibility is that this dummy picks up the activity of Japanese firms between the two regionsY These findings are significant because they imply convergence of industrial structures, growth of liT and related economic integration. A study of this kind should also test the significance of FDI flows as an explanatory variable having a positive influence on the growth of

36

JAPAN IN EAST ASIA

liT through the product lif~cycle and globalization channels outlined above. Yet the lack of comparable data for the economies in the region precluded testing such a hypothesis directly. Patterns and determinants of FDI are examined next.

v FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWORKS

This chapter focuses on foreign direct investment. While it begins with a comparative summary of the distributions of accumulated U.S. and Japanese FDI across economies and industries in the region, the main focus in this chapter is on the interactions among firms, and between firms and governments, to create production networks. Such networks increasingly characterize the behaviour of globalizing multinational enterprises, but some empirical analyses of firm behaviour in East Asia have observed that Japanese firms are creating de facto exclusive production networks, replicating keiretsu and other domestic structures abroad (Borrus 1992; UNCTC 1991a; Doner 1991). Between 1980 and 1990, the stock of world FDI tripled to $1,500 billion from $524 billion. (UNCTC 1992c, p. v). Japan's share of this stock grew to 11 from 4 per cent in the same period (UNCTC 1991a, p. 32). The distribution of stocks of Japanese and U.S. investment among the major regions in 1990 was compared in Table 1. Europe and North America have been the main destinations for both Japanese and American investors. While the East Asian economies have small shares of the totals, the region's share of Japanese FDI was roughly twice that of U.S. FDI by 1990. Charts 3 and 4, comparing the distributions of U.S. and Japanese stocks of FDI in East Asia in 1980 and 1990, show similar distributions among economies and industries. Chart 3 contains a comparison of distributions of FDI among host economies. Both Japanese and U.S. FDI was concentrated in both years in Indonesia, Hong Kong, and Singapore. Japanese FDI in 1980 was concentrated in Korea while U.S. FDI was concentrated in the Philippines, reflecting in both cases the density of contacts, knowledge and commitments each had built up in these economies during and after World War II. Between 1980 and 1990 Japanese stocks built up rapidly, with a focus on Hong Kong, Singapore and Thailand, and to a lesser extent Taiwan, while in Korea, the Philippines and Indonesia, shares had dropped. 37

CHART 3 Stocks of Outward FDI from Japan, and USA: Distribution by Economy, 1980 and 1990

Distribution of Japanese FDI 100

90 80 70 60 %50 40

30 20 10 0 Thailand

South

""'"

Malaysia

\ndonMI.a

Philppl.-..

I ASEAN-4 1

Distribution of U.S. FDI 100

90 80 70 60 %

50 40

30 20 10

South

"""'

Singapore

Thailand

Mallyt.ia

lr'ldonMI.I

Ptlilfppi,...

IASEAN-4 1

SOURCE: MOF, Japan, "Zaisei Kinyu Geppo", December 1991 , pp. 30 and 31; U.S. Department of Commerce, Bureau of Economic Analysis, Survey of Current Business (Aug. 1981 , Table 12, p. 32 and Aug. 1991, Table 17, p. 104).

CHART 4 Industrial Distribution of Stocks of Japanese and U.S. FDI, 1980 and 1990

Japanese FDI 100 90 80 70 60

~

%50

990

40 30 20 10

Tolol

U.S. FDI 100 90 80 70 60 ., 50

~ 0

40 30 20 10

NOTE: Industry shares for total Asia were used because the 8 Asian economies were not

broken out by industry in "Zaisei Kinyu Tokei Geppo". These shares are close approximations since the 8 Asian economies accounted for 98 per cent of Japan's FDI to the region in 1980, and 93 per cent in 1990. SOURCE: MOF, Japan, "Zaisei Kinyu Tokei Geppo", December 1991, pp. 56 and 57; U.S. Department of Commerce, Bureau of Economic Anal ysis, Survey of Current Business (Aug. 1981, Table 12, p. 32 and Aug. 1991 , Table 11.3, p. 88).

40

JAPAN IN EAST ASIA

Stocks of U.S. FDI in Korea, Taiwan, Thailand and Malaysia were distributed roughly equally in both years, although by 1990, Korea's share had increased, while Malaysia's had declined. These differences may reflect differences in corporate strategies and strategic decisions by hosts (Korea's attitude towards Japan and Malaysia's towards the United States) more than deep differences in the behaviour of the sending economies. The main conclusion suggested by the comparison in this chart is that U.S. investment in the region is stagnant and being overtaken by Japanese investment. However, it is important to keep in mind the relative maturity of U.S. investment compared to Japanese; the latter is going through a rapid stock adjustment, building from a small base. The industrial comparison in Chart 4 provides further insights into this issue. In Chart 4, comparisons of industrial shares of U.S. and Japanese FDI brings out further similarities. In both cases, FDI was heavily concentrated in non-manufacturing, particularly natural resources {petroleum for the United States, mining for Japan). Other similarities appear on the non-manufacturing side as service activities, such as finance and distribution, have been placed abroad in the 1980s. Japanese shares also reflect the high level of activity of trading houses. In manufacturing, investment from both countries increased in electronics. The chemicals share was about the same for both in 1990 but Japanese investment dropped off while American stayed stable. The major difference can be found in transportation equipment where U.S. investment was almost non-existent in the 1980s while Japanese investment increased. This contrast adds some strength to the observation that U.S. investment in the region is stagnating, at least in the auto industry. Subsequent chapters will examine this difference.

Production Networks The logic of the arguments is roughly as follows: the world is breaking into economic regions; these regions are inward looking; Japan is the dominant economic player in Asia, investing more than the United States in the Asian economies; the yen is supplanting the U.S. dollar as the reserve currency. Leaving aside the accuracy of these observations, a number of microeconomic assumptions are also made: the

FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWORKS

41

supply architecture in Asia is Japanese-controlled; Japanese companies now think about Asia as part of a single (domestic) market from which to pursue a global strategy, particularly in high tech industries. Such a strategy has two elements: first, to spread sub-systems assembly throughout Asia and, second, to keep tight control of underlying technologies (Barrus 1992). UNCTC asserts that some Japanese affiliates in Asia (and Europe) are part of self-contained regional networks in which affiliates in the Asian economies are linked operationally to Japanese parents and to other affiliates in North America and Europe. They perform the following functions: serve local markets; sell finished goods in the United States and Europe; and act as low cost suppliers to other affiliates in the three major economies (UNCTC 1991a, pp. 46-47). These assumptions and assertions are frequently supported by newspaper articles and some references to other studies (Barrus 1992), selective use of data from the MITI 1989 general survey of affiliates' procurement and sales behaviour, and periodic publication of charts tracing the interdependence of their operations such as in Chart 5. Other evidence points to networks as increasingly universal behaviour of multinationals, particularly in the electronics industry. UNCTC (1992a, p. 253) observes that the emergence of integrated world-wide or regional networks of producing subsidiaries is a response to three changes in the global competitive environment: the lowering of tariffs in successive GATT rounds, the emergence of dynamic new competitors in the Asian economies, and the increased international dispersion of technological capability that has broken the monopoly on competitive technologies held for many years by U.S. and European firms. It further implies that increased use of networks is a universal phenomenon (UNCTC 1992a, p. 252): . .. The increased competition felt by established transnational corporations throughout their global operations has led to the pursuit of increased effectiveness at all stages of the value-added chain. Thus, a technological lead, and the ability to innovate effectively from this in a range of markets, remain central to the operations of transnational corporations, but an increased emphasis has also emerged on making optimal effective use of each overseas producing unit. An implication of the Dunning framework is that Japanese firms, as the creators of ownership advantages of production and distribution

CHART 5 Interdependency Chart for Electrical Apparatus in NIEs and ASEAN Countries Audio Equipment [PARTS AND COMPONENTS)

• Transistor, IC-MECS (Singapore) • Switch, Speaker-SINCOM (Singapore) • Variable Resistance Capacitor, Tuner, Electric Capacltor-MECOM< MEDEM (Malaysia) • Electric Capacitor-PEe (Philippines) • Motor--MEM (Singapore) • Speaker--NAEBEL (Indonesia)

[ASSEMBLER)

[DESTINATION) 'Other Asia Japan, America Europe, Middle East

~

TAMACO (Taiwan)

~

~

MESA (Singapore)

~ ~Japan, America

~

INTC (Thailand) PEC (Philippines) NABEL (Indonesia)

Europe, Middle East

I~

Local Market

TV •Transistor, IC-MECS (Singapore) • Switch, Speaker--SINCOM (Singapore) • Variable Resistance Capacitor, Electric Capacitor--MECOM, MEDEM (Malaysia) • Flyback, Yoke--MPI (Malaysia) • Cathode Ray Tube--MEC (Malaysia)

~ ~

MTV (Malaysia) MELCOM (Malaysia) NTC (Thailand) NABEL (Indonesia) PEC (Philippines) MELCOA (Australia) TAMACO (Taiwan)

~

~Japan, Europe

~

Local Market

Middle East

COMPRESSOR FOR REFRIGERATOR MARIS (Singapore)

MELCOM (Malaysia) APN (Thailand) TAMACO (Taiwan) PEC (Philippines) NABEL (Indonesia)

~ Local Market

~I Japan China

* TAMACO (Audio Equipment), MESA, MTV - production base for re-export. SOURCE: Matsushita Electr ic Co. as reported in H aj ime Ohta (199 1).

FOREIGN DIRECT INVESTMENT AND PRODUCTIO N NETWORKS

43

networks in autos and electronics, should internalize these advantages through FDI. Host government incentives would provide an added locational advantage. Although these government incentives are generally available to all foreign investors, Japanese firms may have "strategic" first mover advantages. Those who penetrate a new market first are likely to reap greater gains until imitators catch up. Thus, investing in a new market may not be profit maximizing in the short run; but being there, in markets in which incomes and market opportunities are likely to materialize in the longer term, is still rational long-run maximizing behaviour. What might production networks look like? How would they work? One would expect they would have characteristics similar to networks in Japan. They would involve clusters of Japanese firms; exclusive use of Japanese suppliers regardless of price and quality; heavy reliance on intra-firm channels for the movement of goods to minimize transactions costs; and to retain shares of world markets they would be expected to be export oriented. Following Morita's observations above, firms might also engage in fierce competition with third country producers by over-investing. Furthermore, one would not expect to find American corporations behaving the same way as Japanese firms (assuming U.S. firms' behaviour is largely market-determined and

laissezfaire). While available price and profit data are not very specific, it is possible to compare sales and procurement data of Japanese affiliates, using the MITI general surveys, and of U.S. affiliates, using Department of Commerce data. It is also possible to compare sales and procurement behaviour of Japanese affiliates in different locales. Beyond that, one must rely on firm level interviews to study the allocation of production and reliance on backward linkages in procurement. Available studies on these subjects are analysed, and the results of firm interviews are also reported in this chapter First, because of the manufacturing focus of this study, characteristics of key industries in Japan are briefly summarized. In autos and electronics, Japanese firms dominate both domestic and international markets, while in chemicals, foreign firms dominate, both in Japan and in international markets. Both the autos and electronics industries are characterized by closely-related supply-assembly networks. The

44

JAPAN IN EAST ASIA

industrial structure m the chemical industry, m contrast, 1s more heterogeneous. Electrical industry: The electrical industry accounted for nearly 50 per cent of Japan's merchandise exports in 1990; computers and telecommunications exports alone accounted for 22 per cent of world merchandise trade in the same year (GATT 1992). The 6 major Japanese computer manufacturers started out in the electrical fabrication business and subsequently organized computer divisions. Each manufacturer is a member of one or more keiretsu. Since each finished product has many intermediate components, parent companies can create networks of suppliers, either as foreign subsidiaries to take advantage of lower labour costs offshore, or as sub-contractors in Japan. Skills to manage the complex set of relationships implied are essential. Japanese Government measures have also greatly shaped the industry over the past two decades through laws giving MITI legal power to influence key factors on the supply side of the industry, such as promoting R&D, provision of financial assistance, influencing industry organization; and to protect software development against external competition, support basic research in computer development, artificial intelligence and optoelectronics. Government procurement has also played a key role on the demand side. Barrus (1992) distinguishes quite different supply "architectures" in the electronics industries in America, Europe and Asia, with the Asian architecture influenced by the highly-strategic leverage inherent in the organization of the Japanese industry. He compares this with the European structure which relied more on foreign supply (mainly from the United States historically) and with the U.S. structure which has also been local supply based, but in an industry characterized by many firms and intense competition among them, and one that is wide open to foreign competition. Auto industry: Automotive products accounted for 23 per cent of Japan's merchandise exports in 1990 and for 20.5 per cent of world merchandise exports (GATT 1992). The Japanese auto industry consists of 11 vehicle OEMs (compared to three in the United States), locked in intense competition in both home and external markets. There are about 1,400 parts suppliers and more than 10,000 suppliers of materials and subcontractors of processes supporting the vehicle manufacturing industry.

FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWO RKS

45

Sourcing of parts is similar among the major manufacturers. On average, OEMs produce about 25 per cent of their parts inhouse, and source the rest from outside suppliers. There are two tiers to the supplier network. The first tier supplies about 85 per cent of the parts procured externally and is typically involved with the OEMs in joint product development. Second tier companies generally operate as secondary sources to the first tier producers or produce items designed by the OEM. In 1988, the top five Japanese auto makers in Japan controlled 90 per cent of the Japanese domestic passenger car market. Hence the opportunity to engage in extensive division of labour; keeping sophisticated components at home and taking advantage of production opportunities in economies with more plentiful labour supplies. Chemicals: Chemicals accounted for only 5.5 per cent of Japanese merchandise exports in 1990 (little changed from 1980) and those exports accounted for 5.5 per cent of world merchandise trade (GATT 1992). This industry is more fragmented in Japan than the electrical and auto industries with the leading competitors being foreigners. Six of Japan's top nine chemical producers are associated with one of the eight major keiretsu. Their combined 1989 sales were roughly only 75 per cent of Toyota's entire sales that year (Dodwell 1989). By 1988, organic chemicals represented the single largest sector within the chemical industry, accounting for slightly more than a third of the industry total. Since Japan is highly reliant on foreign supplies of raw materials, its chemical industry is sensitive to fluctuations in oil prices and international currency markets. Labour and transportation costs have also placed new pressures on the industry in recent years. To offset these effects, chemical companies have been diversifying into pharmaceuticals, biotechnology and other new business areas.

Japanese and U.S. Affiliates in Asia Compared Table 6 compares breakdowns of inputs and revenues for Japanese and U.S. affiliates in the Asian economies for manufacturing overall and for these industries. 18 Overall, U.S. firms in manufacturing are more closely tied into the production structure of the U.S. parent; they

TABLE 6 Comparison of Procurement and Sales Behaviour of Japanese and U.S. Affiliates, 1980 and 1990 (Per cent, shares) All Manufacturing USA

Transport

Electrical

Japan

USA

USA

Japan

Chemicals USA

Japan

Japan

1988

1980

1990

1982

1988

1980

1990

1982

1988

1980

1990

1982

1988

1980

1990

{NA) {NA)

{37)

{29)

{NA) {NA)

{40)

{35)

{NA) (NA)

{4)

{22)

{NA) {NA)

{22)

{34)

1982 A. Inputs 1. Total intra-firm

2. Home-country Intra-firm

NA {NA)

42 {67)

50 (63)

NA

45 (78)

42 (65)

(45)

NA (NA)

51 (56)

58 (49)

NA

(97)

NA (NA)

NA

{91)

(92)

NA (NA)

31 (24)

56 (84)

(80)

(36)

(41)

(21)

(8)

(10)

(12)

(37)

(0)

(NA)

(4)

(7)

(4)

(7)

(4)

{12)

48

41

10 {77)

16 {59)

65

60

16 (73)

27 (60)

2 {36)

1

(NA) (NA)

2 (46)

2

(NA) (NA)

(NA) (NA)

4 (84)

10 (40)

(NA)

26 (24)

20 (37)

(NA) (NA)

32 (32)

36 (44)

(NA) (NA)

6 (62)

6 (9)

(NA) (NA)

5 (2)

12 (35)

NA 36 {NA) {NA)

64 {9)

64 {7)

{NA) {NA)

52 {10)

{13)

NA 100 {NA) {NA)

89 {9)

92 {6)

{NA) {NA)

35 {3)

78 {4)

NA

B. Sales 1. Total intra-firm

2. Home-country Intra-firm 3. Third-country Intra-firm 4. Local Intra-firm

{NA) {NA) 22 (66)

22

23 19

27 13

37

65 33

0 0

NOTE: The figures in parenthesis () represent intra-firm shares. SOURCES: U .S. Department of Commerce, U.S. Direct Investment Abroad; and MITI, General Survey, various issues.

10 92

23 76

FOREIGN DIRECT INVESTMENT AND PRODUCTIO N NETWORKS

47

rely more heavily than Japanese affiliates on the home market for both inputs and revenues. This pattern is borne out in electrical machinery, where sales of U.S. affiliates to the home market account for more than 60 per cent of their output. For Japanese firms, less than a third of their sales are to Japan. As a corollary, U.S. firms source and sell less to local markets, suggesting lower levels of backward and forward linkages than among Japanese affiliates in the region. In the electrical/electronics industries, Japanese firms have better internalization between parents and affiliates, and their superior 0 advantages have enabled them to develop better forward linkages in Asian markets (a greater share of sales are to local markets in line B.4). Beyond these differences the sketchy data show that in the transportation equipment industry, American producers are largely absent suggesting that U.S. auto firms lack the ownership and internalization advantages of Japanese firms. In contrast, in chemicals, Japanese firms appear to lack ownership characteristics that would distinguish them from U.S. firms. 19

Japanese Mfiliates' Behaviour Compared among the Regions Table 7 breaks down 1990 data on inputs and revenues of Japanese affiliates in the three major regions in overall manufacturing, as well as in the electrical, transport and chemical industries. Overall, as much as 60 per cent of inputs are sourced locally in the NIEs, compared to only 30-50 per cent in Europe and the United States. This marked contrast reflects NIE governments' local content requirements. Industry specific characteristics vary considerably. Local sources of inputs are highest in chemicals and transportation and lowest in electronics, where key high-priced technical components must be imported from Japan. On the sales side, in contrast, local sales account for a much smaller share of revenues in the Asian economies than in the United States and Europe, especially in electronics. But sales of this industry are also distributed quite evenly across local, home and third country markets. Following the exclusive networks hypothesis, one should expect heavy emphasis on internalization (I) advantages among the Asian

TABLE 7 Sales, Inputs and Intra-group Transactions: A Comparison of Japanese Affiliates in USA, EC, Asia and NIEs, 1990 (Per cent, shares) All Manufacturing

Electrical

Transport

USA EC Asia N!Es

USA

100 100 100 48 55 27 2 4 2

100 100 100 76 77 36 2 8 2

100 40 1

EC Asia N!Es

USA

Chemicals

EC Asia N!Es

USA EC Asia N!Es

A. Input Shares 1. Total

Intra-group Local intra-group

100 26 3

100 100 100 38 44 21 4 1 0

100 16 4

100 100 100 5 16 34 0 2

100 48 2

2. Local Sources Intra-group

47 5

34

55 4

61 5

13

22 36

42 5

44 3

53 7

47 0

59 2

70 6

95

11

79 0

56 3

49 4

3. Japanese imports Intra-group

50 88

42 91

35 63

28 71

73 97

50 96

46 65

48 74

47 72

52 84

41 49

29 41

5 80

15 84

35 84

47 93

4. Third-country imports Intra-group

4 35

23 57

11

12 27

9 32

28 75

12 30

8 41

0 8

2 0

0

24

1 0

0 100

6 52

9 34

5 49

19

TABLE 7 (continued) All Manufacturing

Electrical

Transport

Chemicals

USA

EC Asia N!Es

USA EC Asia N!Es

USA

EC Asia N!Es

USA EC Asia N/Es

100 10 7

100 35 13

100 19 5

100 15 3

100 6 4

100 38 18

100 38 5

100 32 2

100 15 13

100 46

100 7 6

100 8 6

100 5 1

100 17 0

100 11 3

100 10 1

2. Local Sales Intra-group

95 7

66 20

70 7

74 5

97 4

73 24

37 13

37 5

95 14

51

92 6

81 7

96 1

52 0

78 4

74

3. Sales to Japan Intra-group

4 66

2 59

14 59

11 56

2 98

14

28 60

30 54

4 30

100

2 36

4 14

4 100

0 90

10 40

12 32

4. Third-country sales Intra-group

1 38

32 63

16 37

15 36

1 48

26 79

36 44

33 41

1 8

48 92

7 9

15 12

0 57

48 36

12 35

14 39

B. Sales Shares 1. Total Intra-group Local intra-group

NOTE: Total local intra-group shares were calculated by taking the product of local sales or local sources (whichever is the case) and its associated intragroup share. Similarly, total intra-group shares were calculated by summing the products of intra-group shares by each of three categories: local, Japan and third-country. MIT! Research Institute data, while based on the original questionnaire data obtained in the MIT! General Survey, have been corrected to ensure greater consistency. The sum of A2, A3 and A4, and the sum of B2, B3 and B4 may not equal 100 per cent because of rounding errors. SOURCES: All data for Asia and NIEs excluding intra-group shares were provided by MIT! Research Institute. The U.S. EC, Asia, and NIEs intra-group data were taken from the 1991, MIT!, General Survey [Kaigai Toshi Tokei Soran ] (Tokyo), pp. 200-30.

so

JAPAN IN EAST ASIA

affiliates, in the form of intra-group transactions - that is, that local inputs and revenues would be within the group. This is not the case in Table 7. Instead, the electrical industry in Europe has the highest levels of internalization of any of the affiliates, with 36 per cent of local inputs and 24 per cent of local sales being within the group. There is also interesting evidence in Table 7 of regional differentials in backward and forward linkages. These linkages are most highlydeveloped in the United States, where local sources and sales account for between 50 and 90 per cent shares. Local sourcing in the Asian electrical industries is higher (over 40 per cent) than in Europe or North America (around 20 per cent), but less occurs within the group. Local sales account for a much lower share of the total in Asia (less than 40 per cent compared to over 70 per cent) than among the other two regions, and again, less within the group. These differentials show that Japanese components suppliers are active in Asia in providing inputs, but internalization within the group is not very mature; instead, a supplier sells to any buyer. On the sales side, local linkages among Asian affiliates are more mature than those in Europe and the United States - emphasis on local markets is roughly equal to that for Japanese and third markets. In transportation, local sourcing is roughly the same in Asia as in Europe and North America (between 50 and 70 per cent of inputs are locally supplied) and the heavy emphasis on supplying the local market is similar as well. Intra-group supply and sales channels play little role in any of the regions in this industry. In chemicals, the high degree of local linkages is apparent and is quite similar among the regions (although Asian affiliates rely more heavily on supplies from Japan). The other dimension that can be probed using Table 7 is export behaviour of affiliates. The networks hypothesis would predict that Asian affiliates should be highly export-oriented. Yet, on average, Asian manufacturing affiliates rely less on exports than their European counterparts. They export 30 per cent of sales ( 1 - B.2 share of local sales), compared to 5 per cent in the United States and 34 per cent in Europe. This comparison undoubtedly reflects the huge size of the U.S. market and the integration of the European economies. Industrial comparisons are important here: Asians export 63 per cent of sales in the electronics industry, compared to 27 per cent in Europe and 3 per cent in the United States. These figures have given rise to many

FOREIGN DIRECT INVESTMENT AND PRODUCTIO N NETWORKS

51

generalizations, however, which are not borne out by examination of the auto and chemicals industries: in the auto industry exports account for only 8 per cent of sales; and for 22 per cent in the chemicals industry. In summary, regional differentials in the sourcing and sales characteristics of Japanese affiliates in key manufacturing industries tend to distinguish Asian affiliates as more export-oriented. In addition, Asian affiliates' transactions are more internalized than are the transactions of their U.S. counterparts in Asia, as the previous table showed, but not as highly internalized as their Japanese counterparts in Europe and the United States. Operations of Japanese affiliates in Asia are most highly internalized in the electrical industry, which suggests the possibility of exclusionary behaviour. Yet in the absence of readily available local suppliers use of locally-based Japanese suppliers may be the only alternative to reliance on costly imports of parts and components. This issue is examined in greater detail in the next section.

Survey and Interview Findings While the industry level analysis in the previous section provides some insights into firm behaviour, the supplier-linkage issue must also be pursued through firm level surveys and interviews. The issue in East Asia has two dimensions: the first dimension is the allocation of production among host economies. The small size of some economies and the existence of host government local content requirements means that it can be difficult to achieve economies of scale unless production is aimed at regional or global market demand. Some analysts (Barrus 1992; Urata 1991) are concerned that production networks which are vertically integrated across national boundaries will be exclusionary closed to local or non-Japanese suppliers and not amenable to pressures or incentives from host governments. The second issue is backward linkages to suppliers, either local- or foreign-controlled, in the host economy. Lim and Pang {1991, p. 96) note that this kind of vertical integration reduces the possibilities within an economy to develop indigenous technical capabilities. Pasuk (1990) reports a 1987 JETRO survey of 400 Japanese firms in five ASEAN countries, of which 25 per cent were wholly owned and 2/3 were joint

52

JAPAN IN EAST ASIA

ventures with local capital. Over half the firms reported using less than 50 per cent local content by volume, yet more than two-thirds sourced from locally-owned firms. Those with the highest ratios of local content were resource based industries (Pasuk 1990, p. 57). Manufacturing firms reported a marked increase in local content after the yen appreciated, but they also indicated improved quality of local products was an important factor. Wong (1991) has surveyed the sub-contracting practices of Japanese, U.S. and European MNEs in the electrical machinery sector in Singapore, Malaysia, Korea and Japan using case studies. Like the UNCTC (UNCTC 1992a), he noted the impact of globalization on subcontracting: clustering in a few regional centres of production activities related to product lines; many MNEs integrating stages of processes using just-in-time and local subcontractors. In a detailed study of suppliers in Singapore, where local suppliers were distinguished from foreign-controlled, 77.5 per cent of the 338 suppliers surveyed were local-controlled. But they accounted for only a third of the value added in total production by the group (Wong 1991, pp. 28-30). Since the purpose of his work was to study technological transfer from MNEs to SMEs, he distinguished the MNEs by nationality and with respect to their knowledge of the supplier technology and their commitment to work with suppliers in long-term relationships. Use of such longterm relationships did not vary by nationality. Factors such as product quality, the cyclicality of the industry and technological sophistication of the products (suppliers often have greater technological sophistication than their customers) were more important influences on longterm relationships. Company philosophy was another factor. Some Japanese companies are more "internationalized" than others and more aware of their role in the local economy; just as some American and European MNEs are more "civic minded" than others. 20 Anecdotal evidence also stressed the distinctiveness of multinationals: each has its own strategy. Openness to local suppliers varies but perception of potential transactions costs is an important determinant. U.S. MNEs are more open to local suppliers, for example, because there are very few U.S. alternatives in the region. They also tend to be strict in applying price- and quality-criteria to transactions. Japanese MNEs are open to local suppliers, it was reported, but are difficult to do business with. Suppliers reported they had to better

FOREIGN DIRECT INVESTMENT AND PRO DUCTION NETWORKS

53

the price competition by 30 to 40 per cent to obtain business. This differential may well reflect the added transactions costs perceived by Japanese MNEs who have access to more familiar Qapanese) alternatives. Other findings in survey data indicate that subcontracting arrangements are evolving and dynamic. Chi (1990) found in a study in Taiwan that well-established firms buy more local products than do newcomers. After a time, newly-arrived firms may be attracted by demonstrated capabilities of local suppliers. Wong found the same thing in Singapore. Chi also found that firm specific advantages like ownership structure, export propensity, status as final or intermediate goods producer, proximity to local producers and scale of operation all affect the frequency of backward linkages. Wong (1991, p. 52) noted that Japanese subcontractors are also experiencing increasing competitive pressures from NIE producers which has strained long established relationships with Japanese MNEs. In addition, some Japanese subcontractors have developed sufficiently-advanced technological capabilities that they have some market power and are less willing to confine their growth strategies to the plans of their traditional customers. Despite some instances of the use of market power by the customers, the traditional subcontracting system in Japan is beginning to become more open. In this study, procurement practices of four MNEs in the electrical/electronics sector in Singapore were studied (two Japanese and two American MNEs). Three of the subjects had been in the Asian region since the early 1970s; one had arrived since 1985. For the earlyarriving Japanese firm, the original intent had been to supply local markets beginning in Taiwan, Thailand and the Philippines. The recently-arrived Japanese firm had created Asian affiliates as part of its globalization strategy. For one American, the original intent was to reduce production costs to supply the home market by locating in Singapore where there was cultural familiarity, an educated labour force, and positive government attitudes. Three MNEs have numerous manufacturing and sales sites scattered through the region (see Table 8). One has a large manufacturing presence in Japan, and uses its regional operations to procure parts and components. Where the Japanese MNEs had allocated specific products to each affiliate, however, the American manufacturing affiliates tended

TABLE 8 MNE Relationships with Local Suppliers in Singapore: A Comparison Firm A ljapan) 1. Industry

consumer electronics

Firm B ljapan)

Firm C (USA)

Firm D (USA)

consumer electronics

industrial electronics

industrial electronics • first FDI in 1950s in Japan; since then, most FDI in Asia has been to organize procurement

2. History of local presence

• first FDI in early 1970s in Taiwan, Thailand and Philippines; • objective: to serve the local market

• first FDI in 1987 as part of globalization strategy

• first FDI in early 1970s beginning in Singapore; • objective: to reduce production costs in serving home market

3. Geographic allocation of production

• OHQ in Singapore 1 • 24 manufacturing affiliates in region • 7 sales offices in: Malaysia, Thailand, Hong Kong and Taiwan • 3 manufacturing and sales affiliates in: China, Thailand and Philippines • 2 research affiliates in: Malaysia and Taiwan • 1 support office in: Taiwan

• RHQ in Singapore 2 • 15 manufacturing affiliates in: Singapore, Korea, Taiwan, Thailand, Indonesia and Malaysia • 4 sales offices in: Singapore, Hong Kong, Thailand and Malaysia • 1 engineering and design affiliate in: Korea • 4 support offices in: Singapore, Thailand and Malaysia • 1 rep office in: Taiwan

• Design Centre in Singapore • 6 manufacturing affiliates in: China, Korea, Malaysia and Singapore • 13 sales offices in: China, Hong Kong, Korea, Malaysia, Taiwan, Singapore and Thailand • 2 distributors in: P hilippines and Indonesia

• manufacturing affiliates in Japan and Australia. • R&D joint venture in Singapore

4. Rationale for locations

• globalization • to reduce production costs • host government incentives

• globalization

• to reduce production costs to serve global markets; • to serve local markets

• to obtain costcompetitive parts and components

TABLE 8 (continued)

Firm A (japan)

Firm B (japan)

Firm C (USA)

Firm D (USA)

5. Use of Local Suppliers

• own production highlyautomated • 24% parts obtained from local suppliers • 20% parts are key components from Japanese parent; • 56% parts obtained from local Japanese suppliers

• 80% parts from local Japanese suppliers; • 20% parts from Japan

• for new products use 100% US components for first 12 months; • after 12 months production: use 30-50% parts from suppliers in reg10n; • after 24 months: 50-60% parts from local and regional suppliers

• 20% from locallycontrolled companies; • 20% from Japanesecontrolled local suppliers; • 60% from other US MNEs

6. Products supplied by locally-controlled firms

• paper and plastic packaging, plastic parts, metal parts

• same

• same

• keyboards • mechanical assemblies

7. Selection Criteria Applied

• quality standards met • ·priced competitively • timely

• timely • quality standards met • price competitive • information on product readily available • reliability as supplier

• quality • proximity

• quality • proximity

1 2

Operational Headquarters Regional Headquarters

56

JAPAN IN EAST ASIA

to be fewer in number and to have more comprehensive production and development mandates. All firms located production to take advantage of local skills and lower relative production costs. These costs were affected by host government incentives and, for the Japanese, by the availability of Japanese ODA. All made heavy use of free trade zones that allow them to ship parts and components about the region in bond, thereby avoiding local duties. Attitudes towards, and use of, local suppliers varied by firm rather than by home country. None relied on locally-controlled suppliers for more than 20 per cent of parts or components by volume. In all cases, the products supplied were packaging materials, moulds, plastic and metal parts, and mechanical assembles used in computers. In nearly all cases, key components were sourced in Japan or the United States; the remainder being supplied by locally-based Japanese firms. One Japanese affiliate stated it uses 80 per cent local content; 20 per cent of its components came from Japan, and tended to be high technology components such as compact disc mechanisms and semiconductors. Local inputs, however, originated mainly from Japanese suppliers (70 per cent by volume). Only 30 per cent came from locallycontrolled suppliers. An American affiliate, in comparison, reported its use of local suppliers depended on the length of time it had been producing the product locally. For start-up production, all the parts would be imported, using the original suppliers for the first 12 months. By the end of 12 months, it would be likely that 25-30 per cent of the inputs would be obtained "within the region" (plastic parts and packaging, moulds, metal parts obtained locally; semiconductors from U.S., Korean, Taiwanese or Japanese suppliers). By the end of 24 months, it was estimated that 50-60 per cent of components would come from within the region; sophisticated and costly parts, like engines, might still account for 30 per cent coming from the United States. In all cases, while attitudes to local suppliers varied, these suppliers had to meet some universally-applied criteria. These criteria, stated explicitly by one Japanese firm, included: timeliness of supply; ability to satisfy quality standards; price competitiveness; availability of product information; and reliability. What these surveys of subcontracting practices reveal is that, above all, intense cost competition, rigid quality standards, and short

FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWORKS

57

product cycles are the driving factors in procurement arrangements. In this environment, internalization within the group, if not within the firm, minimizes material and transactions costs. For these reasons, Americans rely just as heavily on local Japanese suppliers as Japanese MNEs do. This preference is explained by the fact that mature Japanese suppliers' standards are well-established and well-known. Like all large firms, MNEs are conservative and slow to experiment with outsiders and with innovations - an additional factor working against local firms trying to establish themselves. In addition, Japanese firms are distinguished by their belief in relationships; by their reliance on Japanese managers, which reduces contact with participants in the local market; and by a factor that one Japanese firm noted in an interview. Until recently his firm had a sixth product supply criterion: that the producer be a Japanese firm. On balance, it seems fair to conclude that Japanese networks in the East Asian economies function in more exclusive ways than American firms, in part because of reliance on Japanese managers who lack local networks, and in part because of diffidence towards nonJapanese products. This conclusion is not the same as one that states keiretsu structures are replicated abroad. All interviewees indicated that they supply each other at the same time as competing in final product markets. If local suppliers meet quality standards, they are used in MNEs. This survey also suggests that market forces, if allowed to function, should bring about convergence in networking behaviour of MNEs. On the one side, Japanese firms are losing their distinctive ownership advantages as competitors imitate those advantages. On the other side, Japanese networks are being pried open by host government performance requirements and incentives in East Asia, and by increasingly independent and sophisticated Japanese intermediate goods suppliers. The implication of these observations is that there is considerable potential for host governments to influence MNE behaviour through market-oriented incentives and performance requirements. If these requirements vary across economies, however, they can create distortions and misallocation of resources. Thus, the preferred, but longer-term solution is to develop common standards and rules by which performance of investors can be monitored and assessed. In closing this discussion on the behaviour of firms, it is worth noting that contrary to popular perception, taking advantage of inter-

58

JAPAN IN EAST ASIA

nalization opportunities in Japanese MNEs is neither universal nor problem-free. One of the affiliates interviewed observed how the rationalization of operations from multiple products to streamlined product mandates had not yet been accomplished in certain economies because of difficulties reaching consensus in the Japanese parent about how such rationalization should occur. Aoki and Tachiki (1992, p. 37) have studied regional headquarters policies of Japanese MNEs and note that some entities are still underdeveloped in East Asia because of head office rivalries and continued head office decision-making which can undercut decentalized authority. Similarly, head office decisions can undercut regional headquarters: engineers, for example, who have the final say whether components meet quality standards, may not support decisions made by regional procurement authorities who are trying to be responsive to local content issues. Asymmetry in the Auto Industry As a brief postscript to this section on investment, a few comments on the virtual absence of American auto producers in East Asia are in order. One possible explanation for this striking fact is that American firms, confronted by relatively small markets in most of the NIEs and ASEAN-4, by differing driving practices and by performance requirements, have considered Asian investments to be inefficient, at least in the short run. An important corollary of this explanation would be that Japanese firms acted strategically; they were willing to take short run losses in order to be positioned to expand their market share as Asian incomes rose. An alternative explanation is that the U.S. Big Three have ignored the Asian market. Their ownership advantages may not have extended to meeting different consumer demands such as right-hand drive (but which is supplied in Europe), or to bridging the distance from the U.S. base - especially if primary concerns were to meet foreign competition in the huge U.S. market and to penetrate the Japanese market. The latter explanation seems to account for the absence of U.S. producers.21 Local performance requirements in the Asian economies are a minor irritant. The first priority for FDI has been to penetrate the pace-setting Japanese market, both to expand sales and to acquire new technology.

FOREIGN DIRECT INVESTMENT AND PRODUCTION NETWORKS

59

Yet Ford {1991a, p. 16) observed that 60 per cent of market growth in auto demand in the next twenty years is likely to be in markets in which Ford currently has no manufacturing presence. In 1991 North America accounted for 66 per cent of total vehicle sales; Japan and Taiwan accounted for 4 per cent of the total (Ford 1991b, p. 41). The World Bank (1992a, p. 2) projects that real per capita incomes in East Asia will rise at an average annual pace of nearly 6 per cent a year in the 1990-2000 period. At this rate, incomes will double in 12 years; which means Hong Kong and Singapore incomes will approach current U.S. and Canadian levels of US$ 20,000 by the beginning of the next century, and Taiwan and Korea will reach current levels of those two economies by then. More significantly, Indonesia's and the Philippines' 240 million people, as well as those in South China, will have crossed at the US$1,000 threshold by then.22 The U.S. Big Three are absent except in joint ventures like Chrysler's with MMC in Bangkok (MMC/Sittipol) and Ford with Mazda in Taiwan and Malaysia. Japanese production capacity is spread among the eight economies, often in the form of small-scale "branch plants"; some produce older-vintage models and may be inefficient. As the costs of flexible manufacturing technologies drop, however, such plants may become efficient at much smaller scales of production than has been the case in the past. In the meantime, both ASEAN governments and Japanese producers have expended considerable efforts to promote scale economies by allocating production among different economies in return for reduced protective tariffs, so far with limited success. Progress towards the ASEAN Free Trade Area will also create opportunities for more efficient local production in the auto industry. Despite the barriers, Japanese producers find Asian production profitable. The 1991 Export Import Bank of Japan (Tejima 1992, p. 37) survey of Japanese firms engaged in FDI reports that 70.5 per cent of auto firms in ASEAN and 48.3 per cent of firms in the NIEs reported surpluses in fiscal year 1990. In contrast, 66 per cent of North American affiliates reported deficits.

VI TESTING THE DETERMINANTS OF FOREIGN DIRECT INVESTMENT

The purpose of this chapter is to apply the conceptual discussion and qualitative data analysis in preceding chapters to econometric tests of the determinants of Japanese FDI in the East Asian economies. The existing econometrics literature consists almost exclusively of studies of European and U.S. FDI. It is surveyed in UNCTC {1992b) which summarizes results from studies using cross-sectional and time series data. Very few econometric tests exist of Japanese or other Asian outward FDI because of the paucity of comparable data. As will be seen, such problems arise in this study as well. Cross-sectional studies reported by UNCTC test the significance of factors such as the intensity of a firm's use of R&D, skills and advertising. Time series studies, most of which are focused on FDI in industrialized countries, examine the impact of dynamic variables such as market size in the host country (as a proxy for potential economies of scale), tariffs {to reflect barriers to trade), capital controls (as influence on FDI), financial variables such as long bond yields (as a measure of the opportunity cost of capital); trade balance; measures of institutional arrangements such a free trade areas and common markets {the latter as proxies for trade barriers); and currency changes. One of the few attempts to study determinants of Japanese outward FDI to East Asia was carried out by Urata {1992, p. 188) on data for the 1977-86 period. 23 He analysed flows of Japanese FDI to Asia across 8 manufacturing sectors. Using Dunning's eclectic framework, he tested the influence on FDI of establishment size, product differentiation, and R&D-intensity as well as measures of expQrt dependence and import penetration. Urata found that neither product differentiation nor technological superiority were important determinants of outward FDI in Asia during the study period. Trade variables both provided significant explanatory power, however: both export dependence and import penetration were positively associated with FDI. He observed that these findings point to FDI as a sub61

62

JAPAN IN EAST ASIA

division of production processes into subprocesses performed in that host economy where they were carried out at least cost; the findings also suggest a close association to intra-industry trade. In this study the dependent variable is outward flows of Japanese FDI, using notification data from the Ministry of Finance, to the manufacturing industries in the East Asian economies by host economy and year, during the 1980-90 period. The equations are estimated by ordinary least squares, using pooled cross-section data. Independent variables include the impact of host country market size and other variables consistent with the Dunning framework. The ownership variable is a measure of host economy manufactures exports as a share to total exports. In economies that do well in manufacturing, the share is large. Thus, investing firms seeking sites to make the most of their ownership advantages would be less likely to choose economies that already have well-developed manufacturing sectors since production costs are likely to be rising relative to those in Japan. The internalization effect can be measured by tracking accumulated Japanese FDI in manufacturing in host economies in each year of the study period. Here the rationale would be that Japanese FDI would be likely to flow, others things equal, to sites where other FDI has preceded it, in order to take advantage of familiar suppliers and other internalization advantages. l.ocational variables include real wages in host economies relative to those in Japan; the ratio of manufactured imports as a share of total imports; and host government incentives and performance requirements. The ratio of manufactured imports in total host country imports is a proxy for low relative production costs and attractiveness of the economy as an offshore production site. Measurement of the impact of host country policies that influence or regulate FDI inflows follows UNCTC (1991b) which provides a valuable source of data and definitions of government policy variables in 46 countries over 11 years and econometric tests of their influence. In this study, seven types of policy towards foreign investment are studied, including ownership policies, taxes and subsidies, currency regulation, price controls, performance requirements, sector-specific incentives and application and entry procedures. Changes in these policies are scored each year, creating a kind of dummy variable. The scores consist of the following: - 1 for a restrictive measure; 0 for no change in policies; and + 1 for a

TESTING THE DETERMINANTS OF FOREIGN DIRECT INVESTMENT

63

liberalizing change. Assuming these scores have a cumulative effect, the total for a particular economy in a particular year is included at that value in subsequent years, and further changes are added or subtracted, as the case may be. Host government policies are also a measure of the impact of strategic variables as determinants of FDI. ODA is the second strategic policy variable used in this study. ODA flows are a relevant measure of Japanese Government behaviour since these reduce the private cost to firms of FDI by creating the necessary infrastructure such as power, transportation and communications and roads. Measures of the annual flows of Japanese overseas development assistance are the most readily available measure published annually by the Ministry of Foreign Affairs (MOFA, various issues). The variables in the regression analysis are listed below, along with the expected sign of coefficients for independent variables: FDI is the dependent variable which is a measure of annual outward Japanese FDI in manufacturing to each economy studied during the 1980-90 period. Independent variables include: GNP ( + ): a proxy for market size and opportunities to realize economies of scale in production; Manexp (- ): ratio of manufactured exports to total host economy exports in each year; this coefficient is expected to be negative since the higher the ratio of manufactured exports in total exports, the less likely Japanese firms can exercise their ownership advantages effectively; Manimp ( + ): ratio of manufactured imports to total host economy imports in each year; a positive relationship would be consistent with the locational attraction of lower production costs in manufacturing in the host economy relative to Japan; ODA ( + ): level of Japanese ODA inflow to the host economy in each year; a positive relationship is expected since ODA flows should lower the private cost of FDI; POLICY ( + ): this is a dummy variable reflecting the degree of liberalization in host economy policies towards inward FDI in each year; a positive relationship is expected resulting from policy liberalization encouraging FDI; RelW (- ): host economy real wage in manufacturing as a ratio of Japanese real wage in manufacturing in each year is expected to be

64

JAPAN IN EAST ASIA

negatively related to FDI; the higher the host economy wage, the less attractive it is as a site for labour-intensive FDI. Regression results are reported in Table 9. In Equation 1, the sample includes the NIEs - Korea, Hong Kong, and Singapore for which relative wage data were available. Canada and the United States are also included to increase sample size. Independent variables are lagged two years (with the exception of relative wages which have been lagged only one year). All coefficients are of the expected signs and are significant at the 1 per cent level of confidence, with the exception of the strategic variables. The ODA coefficient has the expected sign, but is insignificant. POLICY, which signifies degree of liberalization of host economy policies towards FDI, is significant at the 5 per cent level. One reason this statistical relationship is not as strong is because of the inclusion of Canada. The liberalizing effects of NIE policies are offset by restrictive Canadian policies in the early 1980s in the form of the National Energy Program and the Foreign Investment Review Agency. These findings provide a positive test of the hypothesized relationships included in the qualitative discussion in this book. They are based in part, however, on levels of trade and FDI variables among which one would expect to find auto-correlation and two-way relationships, given the close relationship between trade and FDI. The varying sizes of host economies can also be expected to contribute to heteroskedasticity. One way to reduce some of these statistical effects is to scale the level variables by host economy GNP. The results are re-ported in equations 2 and 3. The dependent variable, FDI/GNP, is a ratio of inflows of Japanese FDI in each year to host economy GNP. Independent variables, X and M, are ratios of host economy exports (X) and imports (M) to host economy GNP. ODA is similarly scaled. POLICY retains its dummy variable characteristics. In addition, CUMFDI is added as an internalization variable which measures accumulated Japanese FDI in manufacturing in the host economy using a 3-year moving average. The rationale for this variable, as indicated above, is that Japanese FDI would be likely to flow, other things equal, to sites where other FDI has preceded it, in order to take advantage of familiar suppliers and other internalization advantages. Lagging and/or cumulating the independent variables, however, reduces the number of observations.

TABLE 9 Determinants of Japanese FDI: Results of the Regression Analysis Dependent Variable Independent Variable

N!Esa 1 FDI

Constant

-7.88 ( -0.80)

GNP

132.37** (13.14)

Manexp

-44.63**c ( -3.30)

Manimp

69.35 '~*c

All Economies b

2 FDIIGNP

3 FDIIGNP

-30.10 ( -1.29)

-28.43** ( - 2.41)

-9.09 ( (-1.21)

-12.11 ** ( -2.28)

(3.91)

X

M ODA POLICY ReiW

15.25'~ C

4.65C (1.39) 1.01 *c (1.83) -14.67*'~d

( -4.94)

CUMFDI

(2.09) -1510.00( ( -0.88) 3.82* (1.83)

18.42** (3.28) 335.57* (2.07) 5.01 ** (3.35)

14.82 (0.99) 0.53** (2.87)

Adjusted R 2 DW Statistic F Value

0.80

0.66

1.47

0.83

1.48

15.00

Sample Size

50

45

0.44** (3.03) 0.60 DW "h"• 16.85 54

Includes H ong Kong, Korea, Singapore, USA and Canada. Includes Hong Kong, Indonesia, Korea, Malaysia, Singapore and Thailand; data for Taiwan and the Philippines were not available. c Lagged 2 years. d Lagged I year. e See footnote 23. a

b

** Significant at 1 per cent level of confidence. * Significant at 5 per cent level of confidence.

66

JAPA N IN EAST ASIA

Equation 2 reports results for the NIEs for which relative wage data are available. Coefficients for the location variables, M and POLICY, are of the expected sign, but POLICY is still significant only at the 5 per cent level of confidence. The internalization variable, CUMFDI, is highly significant and of the expected sign. Equation 3 reports results for all East Asian economies for which comparable data were available. In this equation, the United States and Canada have been removed. Since wage data were not available for ASEAN economies, RelW is not included in this regression. All coefficients are of the expected signs and are statistically significant.24 These regression results support the hypothesized determinants of Japanese FDI in the East Asian economies and confirm the influence of strategic policies of both home and host governments. Evidence was found, in the ownership variables, Manexp and X, for the thesis that Japanese investors seek out those economies with smaller shares of manufacturing in their exports. Locational variables such as relative wages (RelW) and ratios of imports of manufactured goods, (Manimp and M), suggest support for the thesis that Japanese investors seek lower cost environments. The significance in equations 2 and 3 of CUMFDI as a proxy for internalization factors also supports the thesis that established Japanese investor presence is a determinant of subsequent flows of investment. The other significant finding is the statistical support for the strategic policy variable, POLICY, which implies host government policies are significant influences on direct investment flows. The other policy variable, ODA, is not very strongly supported by the regression results. There are alternative explanations for this result. One is that ODA is no longer a significant influence on investment flows. The other explanation is that the data are too highly aggregated to pick up the effect. All the findings in this study must be interpreted with caution because of the aggregated nature of the data and the small sample size. Since Japanese ODA is no longer tied, but was targeted both to the region and to commercial purposes during the study period, the latter explanation is a more valid one.

VII CONCLUSIONS

This study has examined why and how Japanese manufacturing firms have located production in the East Asian economies; it has also assessed the strategic role of government policies in this investment and the implications for global welfare. The conclusions from this study are organized into two parts: conclusions about public policies of home and host governments as factors affecting firms' decisions to invest abroad; and conclusions about firms' behaviour and the implications of this behaviour. First, Japanese public policy, although it has become more neutral towards exporting with such changes as deregulation of Japanese capital markets since the late 1970s, still encourages outward FDI. ODA activity in Asia has lowered the cost to MNEs of investment by providing essential infrastructure in host economies. One reason for this positive stance is Japan's labour shortage; while automation is one possible response, moving labour-intensive manufacturing abroad has also been a way to release the labour constraint. A second reason for continued Japanese policy activism towards FDI is the perceived need to deflect pressures from the U.S. congress and administration for reduction of the persistent large bilateral trade surplus. By moving goods production offshore, Japanese firms can maintain export market shares. Another major policy change that influenced outward FDI, though not in an overtly intentional way, was the decision in 1985 (as in 1971) to allow yen/dollar realignment. Although firms failed to anticipate the realignment, once exchange rate expectations had adjusted to a stronger yen and related increases in production costs, FDI became a channel of real-side adjustment. Location by international businesses of production in economies on both sides of major exchange rate relationships is increasingly employed as a way to manage around exchange rate uncertainties as well as trade barriers. At the same time, host governments have liberalized their policies towards foreign investors to create comparative advantage in manufacturing activities. For host governments, one of the desirable outcomes of these policies is the introduction of knowledge and creation of skills 67

68

JAPAN IN EAST ASIA

and employment that would be longer coming in the absence of access to foreign savings and know-how. Because of these social objectives, host governments have sought to accelerate flows beyond what the operation of markets might have produced. Second, at the firm level, Japanese firms have developed over many years significant ownership advantages in the form of management and technological skills in key industries like autos and electronics and have pressed these advantages abroad. Host economy locational advantages in the form of low-wage policies attracted production; policy inducements by governments seeking to diversify and industrialize their economies augmented these locational advantages. This study has cast light on the question of whether the resulting production relationships among Japanese affiliates in the East Asian economies are unique and exclusionary. Although the stock of U.S. FDI in manufacturing in the region is of much longer standing than Japanese FDI, dating back to the late 1960s and 1970s, the distribution of FDI across economies and industries, illustrated in Charts 3 and 4, is similar with the exception of the auto industry where American investment is absent. Comparison of affiliates' sales and procurement behaviour, in Tables 6 and 7, showed that U.S. manufacturing affiliates are more closely tied into the production structures of the home country than are Japanese affiliates. In the electrical and electronics industries, Japanese affiliates' activities in the region are more heavily oriented to the local market than are their American counterparts, and are more internalized, indicating the successful application of their ownership advantages. Yet comparisons across regions, among Japanese affiliates, show even greater internalization of local procurement and sales among their counterparts located in Europe. Generalizations from patterns in the electrical industry can be misleading, however. Japanese firms by design dominate certain segments of that industry, but in autos they dominate by default. During the past two decades, American firms have been largely absent from the auto industry in the region. Japanese auto firms have applied their superior ownership and internalization capabilities to investment in the region, but these investments also reflect the impact of host government performance requirements. As consumer incomes increase at the rapid rate projected for the next ten years, another major Japan-U.S.

CONCLUSIONS

69

trade friction could develop over the Asian, as well as the Japanese, auto market if U.S. producers seek to penetrate the region. Firm-level surveys and interviews confirm that Japanese firms are less open than American firms to outside suppliers for several reasons. One is that they are more recent investors; a second factor is that, with familiar alternatives to fall back on in the form of Japanese suppliers, the relative transactions costs of using non-Japanese suppliers are higher than they are to non-Japanese firms. A third factor is a tendency, because of commitments to lifetime employment, to rely heavily on Japanese managers. This practice reduces their exposure to alternative supplier opportunities. Even so, these surveys and interviews suggest that a convergence process is underway by which affiliates in the electrical/electronics industries will more closely resemble each other over time. Two factors contribute to this process. On one side, fierce global competition is forcing non-Japanese firms to imitate Japanese ownership advantages. While American firms do not imitate Japanese supplier networks (they are more open to outside suppliers including those used by the Japanese), increasingly they are investing in long-term relationships in order to achieve exacting quality standards. Anecdotal evidence indicates the American practice of recruiting local executives is helpful in developing such linkages. These recruits tend to have local knowlege and contacts that are of considerable assistance in creating supplier relationships. The other forces for convergence are host government pressures for localization and the maturing of Japanese suppliers in the home market. Incentives, such as those provided by Singapore's Local Industry Upgrading Program, provide encouragement and pressures to open up Japanese internalization practices to use local suppliers. The other factor is the gradual opening of supplier relationships in Japan as SME suppliers, whose technological capabilities are sometimes superior to those of their customers, behave in more independent ways. These firms are taking advantage of production opportunities in the East Asian economies, to supply both Japanese and western firms, but also to export their own products. Is Japanese FDI in East Asia global welfare-enhancing? Many of the North American studies of outward investment in the 1970s were preoccupied with the costs to the home economy of losses of jobs and

70

JAPAN IN EAST ASIA

loss of control over technology diffusion. In marked contrast, Japan's outward FDI to Asia has promoted the export of jobs as part of the solution to the labour shortage at home. Technology diffusion has been carefully controlled by Japanese producers, no doubt applying lessons of their long experience as absorbers and adaptors of external technologies (much to the chagrin of some of their more-advanced neighbours, particularly in Korea and Taiwan). By shifting production according to comparative advantage, Japanese MNEs have enhanced global welfare. They have enhanced the efficiency of the Japanese economy. Evidence also exists (although not the focus of this study) that host economies have received foreign savings and transfers of managerial skills and technology and the social benefits of higher employment and incomes than would otherwise have been the case. In addition, as Table 5 and Chart 3 illustrate, liT in manufacturing has increased rapidly in the 1990s among many of the Asian economies, and to some extent with Japan. liT allows for product specialization within certain industries which has contributed to higher volumes of two-way trade among countries in Europe and North America. Studies cited earlier have begun to trace these links between trade and FDI in the Asian economies in more detail. Much remains to be done, but one of the implications is that liT with Japan will be limited by the lack of FDI inflows to Japan from the region. Anecdotal evidence also suggests that liT in East Asia is intra-firm trade, which is confirmed in the electrical industries by the sales and procurement data in Table 7. If efficient locally-owned suppliers or suppliers from third countries are excluded by these intra-firm transactions, however, the positive welfare impact will be offset. In addition, host government inducements can lead to inefficient operations behind tariff walls as has occurred in the auto industry. Host government policies can also reduce welfare by misallocating resources to bid for FDI, already headed for the region, to locate in one economy rather than in one of its neighbours. Thus, host government performance requirements and incentives for FDI should be reconsidered. One objective of the Uruguay Round of multilateral trade negotiations is to discipline the use of performance requirements (known as trade-related investment measures). Some host governments themselves are recognizing the costly and distorting impact of incen-

CONCLUSIONS

71

tives, local content and export requirements, and are moving to phase them out. The emergence of differentials among host government policies provides the opportunity to begin the journey towards co-operation and disciplines in this area. The first step is to gather accurate information on such practices to provide a basis for comparisons. The next step is for governments to agree to report changes in regulations, legislation and practices to a central body so that monitoring can become a part of ongoing inter-governmental co-operation. The third step is to develop agreed benchmarks or norms against which comparisons can be made in monitoring. All of these steps are needed to provide the factual basis for co-operation, and eventually for bargaining, that leads to agreed common disciplines. Without this objective basis, calls for the creation of an investment agreement in the region, such as that of Guisinger (1991) which would establish common standards for home and host governments and limit investment incentives, would be premature. The preferred, but longer-term, process to head off emergent investment frictions requires the creation of understanding and of objective measures and benchmarks as the basis for any bargain. 25 Finally, in looking to the future, the recent recessions in Japan and the other G-7 economies raises questions about the sustainability of investment flows into the region. Examination of determinants in this study suggests why Japanese investment will continue, despite the slowdown. First, although home country conditions and government policies affect the decision-making environment, FDI is the result of production decisions by multinationals. They have the ownership and internalization capabilities and face a slow-down in home country d~ mand, rising production costs and labour shortages. Location variables in the East Asian economies continue to be attractive. Abundant lowcost labour continues to become available in mainland China and Indochina; host government incentives and development of more sophisticated factor supplies in the NIEs and ASEAN allow for higher valu~added activities; rising incomes will expand domestic demand in the region. For all these industrial organization reasons, continued Japanese FDI flows into the region can be expected despite temporary adverse financial conditions such as high real interest rates. Japanese production abroad is going through a stock adjustment; comparisons

72

JAPAN IN EAST ASIA

to other major investors such as the United States (and Germany) imply there is much room yet for additional FDI before an equilibrium is reached. While the speed of the recent adjustment may moderate, the direction of the future trend is unmistakable.

NafES 1. For example, Doner (1991) and Borrus (1992). 2. Thurow (1992) develops a confrontational theme implying a zero sum rivalry. Encarnation (1992) in examining why Japanese sell more in the United States than Americans in Japan, examines this rivalry theme in the Southeast Asian regwn. 3. UNCTC (1992a, p. 20). In comparison, UNCTC estimated that U.S. and EC stocks each totalled US$370 billion and US$376 billion, respectively, with each accounting for about 27 per cent of the world stock (54 per cent taken together). 4. The source of these estimates is UNCTC (1992a, p. 20). The stock of inward FDI in Japan is based on data on outward investment from the six major home countries for foreign investment. This figure contrasts with Japanese sources, such as the Ministry of Finance and JETRO (1992), which report accumulated inward direct investment at around US$20 billion in 1991. 5. See, for example, Lawrence (1987) and Takeuchi (1989). 6. Saxonhouse (1986) provides an excellent example of such a reasoned argument. 7. Johnson (1982) and (1992). 8. The discussion in this paragraph is drawn from Pasuk (1990) Chapter 3. 9. For the same reason, inward FDI to Japan was also restricted; no earnings could be repatriated. 10. Some authors identify the 1980s as the "third" wave of outward investment, the second and much smaller wave having taken place after yen appreciation in 1978. 11. Morgan Guaranty Trust Company (1989, p. 12). 12. Volcker and Gyohten ( 1992, chapters 2 and 3) provide a detailed account of U.S. and Japanese policy-making during this period. 13. The discussion in this chapter draws on Irimajiri (1992). 14. For a concise summary of such policy reforms, see Asian Development Bank (1991, pp. 20-35). 15. The term "gravity coefficient" is used here to measure the intensity of trade linkages among the group of East Asian economies which are the focus of this study. The coefficients measure an economy's export share of the market relative to its share in total trade among the group. Thus, A(i,j)

=

x(i,j)/x(t,j) x(i,t)/x(t,t)

where A(i,j) is the coefficient; the numerator is the exports from country i to country j relative to all exports, (x(t,j)), to country j from all other countries in the group. The denominator is the exports from country i to all countries in the group, (x(i,t)), relative to total exports within the group, (x(t,t)). 16. Fukusaku (1992, pp. 25-26). 17. Fukasaku (1992) also picks up such a significant relationship but interprets it in terms of the first of these two possibilities. 73

74

JAPAN IN EAST ASIA

18. Because of the paucity of data on U.S. affiliates' procurement behaviour, local and third country procurement patterns cannot be compared and are omitted from the table. 19. Encarnation (1992), in an examination of Japanese-U.S. rivalry in trade and FDI, also found the behaviour of both affiliates to be quite similar in the Asian market. 20. Interview with the author, October 1992. 21. Personal interviews with industry officials and independent analysts in Tokyo in 1992. 22. GNP per capita estimates from World Bank ( 1992b) statistical tables. 23. Thorough and innovative studies of the determinants of inward FDI into Taiwan have also been reported by Tu (1990) and Chi and Tu (1992, pp. 142-71). 24. In this equation, the use of the lagged cumulated dependent variable as an independent variable might be expected to create a problem of auto-correlation. To test for this problem, Durbin's "h" statistic was calculated by regressing the OLS residual on the lagged residual and all the explanatory variables. No autocorrelation was found by this test. See Kennedy (1985) for further explanation of this test. 25. For a conceptual framework along these lines, applied in the area of macroeconomic policy, see Dobson (1991).

APPENDICES

APPENDIX A Sources of Data on Foreign Direct Investment There are two national sources of data on Japanese FDI: the balance of payment statistics of the Bank of Japan (BOJ) and notification data collected by the Ministry of Finance (MOF). BOJ statistics are compiled on actual investment transactions by Japanese residents, on a calendar year basis, in overseas branches, subsidiaries or associated companies in which Japanese parents' ownership exceeds 10 per cent. These data are available only in aggregate form, with no national or industrial breakdowns. MOF statistics, which do provide such breakdowns, are anticipatory, compiled at the time the Ministry is notified by firms of their intentions to invest, subject to approval of the host government. Because these data are collected before the transaction occurs, they overstate actual investment. U.S. Department of Commerce (U.S. DOC) data measures actual capital flows by U.S. firms that own at least 10 per cent of the voting equity of a foreign enterprise. This measure of FDI includes retained earnings. Comparisons between Japanese and American foreign direct investment must, therefore, be made with caution and are best confined to general ratios and trends rather than to levels. Data from host governments also vary considerably because of the variety of methods used to track investment inflows. They are, therefore, of limited assistance in reconciling differences in home country data.

75

76

JAPAN IN EAST ASIA

APPENDIX B Intra-industry Trade in East Asia 1. Indices of liT Intra-industry trade ratios have been computed following the GubelLloyd index, using U.N. trade data calculated at SITC (Rev.2) at 2- and 3-digit levels. For each industry (i), the intra-industry trade ratio, IITi, is calculated according to the Grubel-Lloyd measure (Grubel and Lloyd 1975): liTi

=

{

1- abs(xi - mi)} x 100

(xi + mi) exports of industry i, imports of industry i.

where xi

mi

An aggregate liT index (liT) is also constructed for each economy included in the study according to the following aggregation technique: liT

=

E liTi x (xi + m)

X+M where X

M

total exports, total imports.

The analytical relevance of this measure has been subject to criticisms based on the argument that arbitrary statistical classifications in the trade data do a poor job of reflecting actual 2-way exhanges. But as Fukasaku (1992, p. 27) has pointed out, studies show that simultaneous exports and imports remain quantitatively significant, even at 7-digit SITC levels.

2. Determinants of liT The theory of the determinants of liT follows contributions of Linder (1961) and Lancaster (1980) who suggested, respectively, that intraindustry trade will increase with levels of economic development (because of increasing product differentiation) and with market size

APPENDICES

77

(owing to economies of scale). Krugman (1980) pointed out that transportation costs will reduce the volume of such trade. The analysis here follows Balassa (1986) who tested these and other theoretical propositions econometrically on cross-country data, and Fukusaku ( 1992) who analysed the determinants of liT in manufactures among Pacific-Asian economies. Fukusaku's study is a crosssectional cross-country analysis of determinants of liT ratios in the years 1979 and 1988, using OLS and weighted least squares techniques.

3. The Regression

A. Dependent Variable The dependent variable is a set of bilateral liT ratios among the economies studied (including the United States and Canada), for SITC (Rev.2) 2-digit categories 500-1000 for the years 1985-89, inclusive. Thus,

liT''"

=

{

1- abs(xij - mi)} x 100

and

iitij

1000 E liT''" x (x 1ij + m 1i) 1 = 500

(xij + mi)

where

xij represents exports from country i to country j, mij represents imports to country i from country j, and 1 from 500 to 1000 represents SITC industries.

B. Independent Variables Following Balassa and Fukusaku, independent variables were chosen that allow for empirical tests of the significance for liT of relative differences in levels of economic development (measured by relative differences in per capita incomes between economies); differences in relative factor endowments (measured by share of primary commodities in primary exports); distance (measured by shipping distances

78

JAPAN IN EAST ASIA

between major ports); and by a number of trading arrangements and peculiarities of trade. These independent variables are defined as follows: YPC

=

relative per capita income differences, where ypcij =

abs(gnppci - gnppcj) (gnppci + gnppcj)/2

and gnppc RFE

=

=

gross national product per capita in countries i and j;

relative factor endowments, where rfe lj

=

abs(l/s l - lis)J 1/sJ

and rfeij

=

factor endowments of country i relative to country j;

s

=

the share of primary commodities in total exports of country t or ;;

DIS

shipping distance from country i to country j, measured in logs;

DUMMY VARIABLES Dl D2 D3 D4

ASEAN preferential trade arrangement Canada-U.S. preferential trade arrangement Entrepot trade in Hong Kong Entrepot trade in Singapore

C. Expected Signs of Coefficients

Coefficient YPC RFE DIS Dl D2 D3 D4

Sign

+ + + +

APPENDICES

79

D. Methodology The sample includes 5 years of data and 11 countries. Ordinary least squares (OLS) are used. Further methodological limitations are that, while our dependent variable is an aggregate measure of liT in manufacturing industries across economies and through time, the independent variables are country-level variables, and therefore serve only as rough indicators of factors determining liT in particular industries.

REFERENCES

Aoki, Akira and Dennis S. Tachiki. "Overseas Japanese Business Operations: The Emerging role of Regional Headquarters". RIM: Pacific Business and Industries I, no. 15 (1992): 28-39. Asian Development Bank (ADB). Asian Development Outlook. ADB: Manila, vanous years. Balassa, Bela. "Intra-industry Specialization: A Cross-Country Analysis". European Economic Review 30 (1986): 27-42. Barrus, Michael. Reorganizing Asia: Japan's New Development Trajectory and the Regional Division of lAbor. BRIE Working Paper no. 53, March 1992. Caves, Richard. Multinational Enterprises and Economic Analysis. Cambridge University Press: Cambridge, 1982. Chee, P.L. Industrial Development, Policies and Cooperation Among A SEAN Countries and their Implications on Trade Patterns in the Region. CAPS Discussion Paper Series 15. Tokyo: Seikei University, 1988. _ _ ."Asia-Pacific Economic Development and Transnational Corporations". Paper presented at the Kyushu University International Symposium, 17 September 1992. Chi Schive. "Linkages: Do Foreign Firms Buy Locally in Taiwan?" Asian Economic Journal IV (1990): 1-15. _ _ ."How Did Taiwan Solve Its Dutch Disease Problem?" Paper presented at the Fifth Biennial Conference on U.S.-Asia Economic Relations, sponsored by the American Committee on Asian Economic Studies, 20-22 June 1991. _ _ and Jenn-Hwa Tu. "Foreign Firms and Structural change in Taiwan". In Eric D. Ramstetter, ed. Boulder, Colorado: Westview Press, 1992. Chiasakul, Samart and Mikimasa Yoshida. Thai Economy in the Changing Decade and Industrial Promotion Policy. ASEDP no. 6. Tokyo: Institute for Developing Economies, 1990. Cronin, Richard. Japan, the United States, and prospects for the Asia-Pacific Century. Singapore: ISEAS, 1990. Dertouzos, Michael L., Richard K. Lester and Robert M. Solow. Made in America. Cambridge, Mass.: MIT Press, 1989. 81

82

JAPAN IN EAST ASIA

Dobson, Wendy. Economic Policy Coordination: Requiem or Prologue? Washington: Institute for International Economics, 1991. Dodwell Marketing Consultants. Industrial Groupings in Japan. 8th ed. Tokyo: Dodwell, 1989.

_ _ . 1he Structure of the Japanese Auto Parts Industry. Tokyo: Dodwell, 1990. Doner, Rick. "Japanese Foreign Investment and the Creation of a PacificAsian Region". Mimeographed, 1991. Dunning, J.H . "Towards an Eclectic Theory of International Production: Some Empirical Tests". Journal ofInternational Business Studies 11 (1980): 9-31. _ _ . "Governments and Multinational Enterprises: From Confrontation to Cooperation?" Millennium Journal of International Studies 20, no. 2 (1991): 225-44. Encarnation, Dennis]. Rivals Beyond Trade: America versus Japan in Global Competition. Ithaca and London: Cornell University Press, 1992. Ford Motor Company Qapan) Ltd. Company Profile. Tokyo: Ford, 1991a. Ford Motor Company. Annual Report 1991. Dearborn, Michigan: Ford Motor Company, 1991b. Fukusaku, Kiichiro. Economic Regionalization and Intra-industry Trade: Pacific-Asian Perspectives. Technical Paper no. 53. Paris: OECD, 1992. General Agreement on Tariffs and Trade. International Trade 1990-91. GATT: Geneva, 1992. Grubel, Herbert G. and Lloyd, P.J. Intra-Industry Trade: 1he Theory and Measurement of International Trade in Differential Products. London: The MacMillan Press Ltd, 1975. Guisinger, Stephen. "Foreign Direct Investment Flows in East and Southeast Asia: Policy Issues". ASEAN Economic Bulletin 8 no. 1 (1991): 29 Healey, Derek. Japanese Capital Exports and Asian Economic Development. Paris: OECD, 1991. Hosen, Mitsuo. "Recent Trends in Intra-industry Trade in Major Developed Countries". Mimeographed. Tokyo, 1992. Hymer, S.H. 1he International Operations of National Firms: A Study of Direct Investment. Cambridge, Mass: MIT Press, 1976. International Monetary Fund. Direction of Trade Statistics Yearbook. Washington: IMF, various years.

REFERENCES

83

Irimajiri, Shoichiro. "Highlighting the Contrasts in Japanese and American Business". Economic Eye 13 (1992): 22-25. Johnson, Chalmers. MIT! and the Japanese Miracle. Tokyo: Charles E. Tuttle Co., 1982. _ _ . "Japan in Search of a 'Normal' Role". Mimeographed. University of California, San Diego, 1992. Julius, DeAnne. Global Companies and Public Policy. London: Royal Institute of International Affairs, 1990. _ _ .Foreign Direct Investment: The Neglected Twin of Trade. Washington, D.C.: Group of Thirty, 1991.

Kennedy, Peter. A Guide to Econometrics. Cambridge, Mass.: MIT Press, 1985. Kindleberger, C.P. American Business Abroad: Six Lectures on Direct Invest· ment. New Haven: Yale University Press, 1969. Komiya, Ryutaro. The Japanese Economy: Trade, Industry, and Government. Tokyo: University of Tokyo Press, 1990. Krause, Lawrence B. "The Structure of Trade in Manufactured Goods in the East and Southeast Asian Region". In Trade and Structural Change in Pacific Asia, ed. Colin I. Bradford, Jr., and William H. Branson, pp. 205-260. Chicago and London: University of Chicago Press, 1987. Krugman, Paul R. "Scale Economies, Product Differentiation, and the Pattern of Trade". American Economic Review 70 (1980): 950-59. _ _ , ed. Strategic Trade Policy and the New International Economics. Cambridge, Mass. and London: MIT Press, 1986. Lancaster, Kelvin. "Intra-industry Trade under Perfect Monopolistic Competition". Journal of International Economics 10 (1980): 151- 75. Lawrence, Robert Z. "Imports in Japan: Closed Markets or Minds?" Brookings Papers on Economic Activity 2 (1987): 517-54. Lim, Linda Y.C. and Pang, Eng Fong. Foreign Direct Investment and Industrialization in Malaysia, Taiwan and Thailand. Paris: OECD Development Centre, 1991. Lincoln, Edward]. Japan 's Unequal Trade. Washington, D.C.: The Brookings Institution, 1990. Linder, Staffan Burenstam. An Essay on Trade and Transformation. New York: Wiley, 1961 .

84

JAPAN IN EAST ASIA

Ministry of Finance (MOF). Zaisei Kinyu Geppo. Tokyo: various years. Ministry of Foreign Affairs (MOFA). Japan's ODA: Official Development Assistance Annual Report. Tokyo: Association for Promotion of International Cooperation, various issues. Ministry of International Trade and Industry (MITI). General Survey of Foreign Investment [Kaigai Toshi Tokei Saran: Kaigai Katsudo Kihon Chosa}. Tokyo: Keibun Publishing, 1989. _ _ . White Paper on International Trade 1990. Tokyo: MITI, 1991. _ _ . White Paper on International Trade 1992 (Summary). Tokyo: Ministry of International Trade and Industry, 1992.

Morgan Guaranty Trust Company. World Financial Markets. Issue 3, 12 October 1989. Morita, Akio. "A Turning Point for Japanese Managers?" International Economic Insights May/June 1992, pp. 2-10. Narongchai, Akrasanee and Suthiphand Chirathivat. "Trade Issues among ASEAN, Asian NICs and Japan". In Global Adjustment and the Future AsianPacific Economy, ed. by Miyohei Shinohara and Lo Fu-chen. Tokyo: Asian and Pacific Development Centre and IDE, 1989. Nikkei Weekly. "MITI urges electronics firms to produce abroad". 13 June 1992, pp. 1, 23.

Ohta, Hajime. "Prospects for Japanese FDI in Asia and the Pacific". Roundtable on Foreign Direct Investment in Asia and the Pacific in the 1990s, East West Center, Honolulu, March 1991. Oishi, Noboyuki. "Auto industry seen paring operations". Nikkei Weekly, 15 February 1992. Organization for Economic Cooperation and Development. Japan Survey. Paris: OECD, 1991. Orr, Robert M. "Japan's Foreign Aid: More Than a Good Housekeeping Seal-of-Approval". Japan Economic Survey. Washington, D.C.: Japan Economic Institute, 1991. Ozawa, Teratumo. "Japan in a New Phase of Multinationalism and Industrial Upgrading: Functional Integration of Trade, Growth and FDI". Journal of World Trade 25, no. 1 (1991): 43- 60. Pangestu, Mari. "The Asean Free Trade Area: Going Forward the Asean Way". Mimeographed, 1992.

REFERENCES

85

Pasuk, Phongpaichit. The New "Wave of Japanese Investment in ASEAN. Singapore: ISEAS, 1990. Porter, Michael E. The Competitive Advantage of Nations. New York: The Free Press, 1990. Ramstetter, Eric D. Direct Foreign Investment in Asia's Developing Economies and Structural Change in the Asia· Pacific Region. Boulder, Colorado: Westview Press, 1992. Rowley, Anthony. "Price of Expansion". FEER, 14 November, 1991. Rugman, Alan. "A New Theory of the Multinational Enterprise: Internationalization versus Internalization". Columbia Journal of World Business 15 (1980): 23-29. Saxonhouse, Gary R. "Japan's Intractable Trade Surplus in a New Era". World Economy 9 (1986): 239-57. Sekiguchi, Sueo, ed. Japanese Direct Investment. Montclair: Allanheld, Osmun and Co. for the Atlantic Institute of Foreign Affairs, 1979. Stevens, Barrie. "Support Policies for Strategic Industries: An Assessment and Some Policy Recommendations". In Strategic Industries in a Global Economy. Paris: OECD International Futures Programme, 1991. Takeuchi, Kenji. "Does Japan Import Less Than It Should? A Review of the Economic Literature". Asian Economic Journal, September 1991, pp. 138-69. Teece, David J. "Support Policies for Strategic Industries". In Strategic Indus· tries in a Global Economy. Paris: OECD, 1991. Tejima, Shigeki. "Japanese Foreign Direct Investment in the 1980s and Its Prospect for the 1990s". EXIM Review 11, no. 2 (1992): 25. Thurow, Lester. Head to Head: The Coming Economic Battle Among Japan, Europe, and America. New York: William Morrow and Company, 1992. Tu, Jenn-Hwa. Direct Foreign Investment and Economic Growth: A Case Study of Taiwan. Monograph Series no. 48. Taiwan: Institute of Economics Academia Sinica, 1990. UNCTC. World Investment Report: The Triad in Foreign Direct Investment. New York: United Nations, 1991a. _ _ . Government Policies and Foreign Direct Investment. UNCTC Current Studies Series A no. 17. New York: United Nations, 1992b. _ _ .World Investment Report 1992. New York: UNCTC, 1992a.

86

JAPAN IN EAST ASIA

_ _ . The Determinants ofForeign Direct Investment. New York: UNCTC, 1992b. _ _ . World Investment Directory: Asia-PAcific. New York: UNCTC, 1992c.

UNESCAP. Industrial Restructuring in Asia and the Pacific. Bangkok: United Nations, 1991. Urata, Shujiro. "Japanese Foreign Direct Investment and its Impact on Foreign Trade in Asia". Mimeographed, 1991. _ _ . "The Rapid Increase of Direct Investment Abroad and Structural Adjustment in Japan". In Eric D. Ramstetter, ed. (1992). U.S. Department of Commerce. Survey of Current Business. Washington, D.C.: Bureau of Economic Analysis, Aug. 1981, p. 32 and Aug. 1991, pp. 88 and 104. _ _ .U.S. Direct Investment Abroad. Washington: Bureau of Economic Analysis, various years.

Volcker, Paul and Toyoo Gyohten. Changing Fortunes. New York: Random House, 1992. Womack, James et al. The Machine That Changed The World. Cambridge, Mass.: MIT Press, 1990. Wong Poh Kam. Technological Development Throught Subcontracting Linkages. Tokyo: Asian Productivity Organization, 1991. World Bank. Global Economic Prospects and the Developing Countries. Washington, D.C.: World Bank, 1992a. _ _ . World Development Report 1992. Washington, D.C.: World Bank, 1992b.

Yamazawa, Ippei and Akira Hirata. Trade Policies Toward Developing Countries. Macmillan: New York, forthcoming. Yoshitomi, Masaru. "New Trends of Oligopolistic Competition in the Globalisation of High1'ech Industries". In Strategic Industries in a Global Economy. Paris: OECD, 1991.

ABOUT THE AUTHOR

Wendy Dobson is Director, Centre for International Business, and Professor at the University of Torontds Faculty of Management where her research focuses on comparative studies of trade and foreign direct investment in East Asia. Before joining the University in 1990, she was Associate Deputy Minister for Finance in the Canadian Government in 1987-89. Prior to that, she headed the C. D. Howe Institute, Canada's leading non-partisan independent economic policy research organization in 1981-87. She is author of Economic Policy Coordination: Requiem or Prologue-? (1991), co-editor of Shaping Comparative Advantage (1987), and editor of Canada-Japanese Relations in a Triangular Perspective ( 1987).